INTERNAL - RACP · Australian and New Zealand Journal of Medicine, is the official journal of the Adult Medicine Division of The Royal Australasian College of Physicians (RACP). Its

Volu

me 48, Issu

e 2, Febru

ary 2018, Pages 107–234IN

TE

RN

AL

ME

DIC

INE

JOU

RN

AL

Climate change allergens and allergic diseases

Documenting patient body weight

EVOLVE: low value practices in rheumatology

Non-alcoholic fatty liver disease

Hospital-associated venous thromboembolism

Adrenal incidentalomas

Management of neurosyphilis

INTERNALMEDICINEJOURNAL

Volume 48

Issue 2

February 2018

ISSN 1444-0903

Total-pages: Spine-width: Paper-type:previous Total-pages:128 Spine-width: 4.2mm Paper-type: UPM Classic 64Current

Editorial 113 What we know and do not know about women and kidney

diseases: questions unanswered and answers unquestioned: reflection on World Kidney Day and International Women’s Day

Giorgina B. Piccoli, Mona Alrukhaimi, Zhi-Hong Liu, Elena Zakharova, and Adeera Levin, on behalf of the World Kidney Day Steering Committee

Review

124 Recording patient bodyweight in hospitals: are we doing well enough?

Kate M. Flentje, Colin L. Knight, Ingrid Stromfeldt, Anindita Chakrabarti and N. Deborah Friedman

Clinical Perspectives 129 Climate change: allergens and allergic diseases

Constance H. Katelaris and Paul J. Beggs

Original Articles 135 EVOLVE: The Australian Rheumatology Association’s ‘top five’

list of investigations and interventions doctors and patients should question

Kathleen Morrisroe, Ayano Nakayama, Jason Soon, Mark Arnold, Les Barnsley, Claire Barrett, Peter M. Brooks, Stephen Hall, Patrick Hanrahan, Pravin Hissaria, Graeme Jones, Veera S. Katikireddi, Helen Keen, Rodger Laurent, Mandana Nikpour, Katherine Poulsen, Philip Robinson, Muriel Soden, Nigel Wood, Nicola Cook, Catherine Hill and Rachelle Buchbinder

144 Underappreciation of non-alcoholic fatty liver disease by primary care clinicians: limited awareness of surrogate markers of fibrosis

Preya J. Patel, Xuan Banh, Leigh U. Horsfall, Kelly L. Hayward, Fabrina Hossain, Tracey Johnson, Katherine A. Stuart, Nigel N. Brown, Nivene Saad, Andrew Clouston, Katharine M. Irvine, Anthony W. Russell, Patricia C. Valery, Suzanne Williams and Elizabeth E. Powell

151 Acute oxygen therapy: an audit of prescribing and delivery practices in a tertiary hospital in Perth, Western Australia

Anam Kamran, Elisa Chia and Claire Tobin

157 Incidence of in-hospital and post-discharge diagnosed hospital-associated venous thromboembolism using linked administrative data

Joanne M. Stubbs, Hassan Assareh, Jennifer Curnow, Kerry Hitos and Helen M. Achat

165 Off-label use of rituximab in autoimmune disease in the Top End of the Northern Territory, 2008–2016

Sarah Wongseelashote, Vipin Tayal and Peter Francis Bourke

173 Characteristics of adrenal incidentalomas in a New Zealand centre

Ziwei Goh, Ian Phillips, Penny J. Hunt, Steven Soule and Tom J. Cawood

178 United we stand, divided we conquer: pilot study of multidisciplinary General Medicine Heart Failure Care Program

Omar Wahbi-Izzettin, Ingrid Hopper, Edward Ritchie, Vathy Nagalingam and Ar Kar Aung

184 Delays in presentation and diagnosis of pulmonary tuberculosis: a retrospective study of a tertiary health service in Western Melbourne, 2011–2014

Eloise Williams, Allen C. Cheng, Garry P. Lane and Stephen D. Guy

Brief Communications 194 Metformin-induced encephalopathy: the role of thiamine

Caoimhe McGarvey, Catherine Franconi, David Prentice and Michael Bynevelt

198 Assessment of potential opioid toxicity and response to naloxone by rapid response teams at an urban Melbourne hospital

Bharathy Gunasekaran, Jennifer Weil, Tom Whelan, John Santamaria and Mark Boughey

201 Lipid-lowering therapy use and achievement of cholesterol targets in an Australian diabetes clinic

Katerina V. Kiburg, Glenn M. Ward, David N. O’Neal and Richard J. MacIsaac

204 Management of neurosyphilis: time for a new approach?

Olivia C. Smibert, Adam W. J. Jenney and Denis W. Spelman

207 High urinary interleukin-8 levels is associated with poor prognosis in idiopathic membranous nephropathy

Jilin Chen, Xiaoye Fu, Yanling Sun, Shengkun Zhang, Hua Xie and Hongli Lin

Position Paper 210 Considerations for pre-transfusion immunohaematology

testing in patients receiving the anti-CD38 monoclonal antibody daratumumab for the treatment of multiple myeloma

Hang Quach, Simon Benson, Helen Haysom, Anne-Marie Wilkes, Nicole Zacher, Merrole Cole-Sinclair, Henry Miles Prince, Peter Mollee, Andrew Spencer, Phoebe Joy Ho, Simon J. Harrison, Cindy Lee, Bradley Augustson and James Daly

Letters to the Editor

Clinical-scientific notes

221 Benefit of routine testicular examination: hypogonadism in a person with 47XYY

Eman Negm and Bernard E. Tuch

222 Boerhaave syndrome: a common manifestation of a rare disease

Adel Ekladious and Mark K. Tiong

223 Role of radiotherapy in management of gingival infiltration of chronic myelomonocytic leukaemia

Bahare Moradi, Stephen Thompson and Timothy Brighton

224 Encephaloclastic cyst: a rare complication of a malfunctioning methotrexate Ommaya reservoir

Michal Lubomski, Malcolm Pell, Alistair Lochhead and Martin Jude

General correspondence

226 Time of administration of treatment for hypertension in renal patients

Martin S. Knapp

227 Starting beta-blockers during exacerbations of chronic obstructive pulmonary disease

Robert J. Hancox, Catherina L. Chang, Richard Beasley and Christine Jenkins

228 Author reply

Pieter A. Neef, Louise M. Burrell, Christine F. McDonald, Louis B. Irving, Douglas F. Johnson and Daniel P. Steinfort

229 Iron polymaltose infusion therapy during pregnancy

Raj Ramakrishna and Arumugam Manoharan

230 Clinical need for standardised multidisciplinary meeting assessment processes

Wing H. Yau, Bianca Devitt, Soe Y. Aung, Phillip Parente, Sharad Sharma and Carmel Pezaro

231 Author reply

Claire E. Johnson, Neli Slavova-Azmanova and Christobel Saunders

233 Corrigendum

VOLUME 48 issue 2 February 2018

IMJ_48_2Cover.indd 1IMJ_48_2Cover.indd 1 02/02/18 1:21 PM02/02/18 1:21 PM

IMJ.JEBJan16

Subspecialty Editors

Cardiology (General)Paul Bridgman, Christchurch

Cardiology (Arrhythmias)Andrew McGavigan, Adelaide

Clinical GeneticsLes Sheffield, Melbourne

Clinical PharmacologyJenny Martin, NewcastleYvonne Bonomo, Melbourne (Addiction Medicine)

Continuing Education (Clinical Perspectives)

Christopher Pokorny, Sydney

Emergency MedicineJoseph Ting, Brisbane

EndocrinologyMorton Burt, AdelaideAnthony Russell, Brisbane

Ethics Paul Komesaroff, Melbourne

GastroenterologyDavid M. Russell, Melbourne

Geriatric MedicineLeon Flicker, Perth

Haematology (General)Peter Browett, Auckland

Haemostasis/ThrombosisClaire McLintock, Auckland

Immunology and AllergyMarianne Empson, Auckland

Infectious DiseasesIan Woolley, Melbourne

Intensive CareMichael O’Leary, Sydney

Internal MedicineIan Scott, Brisbane

NephrologyZoltan Endre, SydneyKarin Jandeleit-Dahm, Melbourne

NeurologyDavid Blacker, Perth

Nuclear MedicineSze Ting Lee, Melbourne

Occupational and Environmental Medicine; Health Economics; Editorials Editor

Des Gorman, Auckland

OncologyDamien Thomson, Brisbane

Palliative MedicineJanet Hardy, Brisbane

Public Health MedicineMark Ferson, Sydney

Respiratory MedicineMatthew Naughton, Melbourne

RheumatologyPeter Youssef, Sydney

Sexual Health MedicineDarren Russell, Cairns

Honorary Advisory Board

Peter Doherty, MelbourneKar Neng Lai, Hong KongRichard Larkins, MelbourneSir Gustav Nossal, MelbourneLawrie W. Powell, BrisbaneNicholas Saunders, NewcastleJohn Shine, SydneyChorh Chuan Tan, SingaporeSir David Weatherall, OxfordJudith Whitworth, Canberra

Editorial Ombudsman

Graham Macdonald, Sydney

Manager

Virginia Savickis, Sydney

Editorial Assistant

Aparna Avasarala, Sydney

Previous Editors-in-Chief

Internal Medicine JournalEdward Byrne (1999–2004)

The Australian and New Zealand Journal of MedicineGraham Macdonald (1989–1999)Michael O'Rourke (1981–1989)Akos Z. Gyory (1975–1981)Charles Kerr (1970–1975)

The Australasian Annals of MedicineRonald Winton (1957–1970)Mervyn Archdall (1952–1956)

Editor-in-ChiefJeff Szer, Melbourne

Continuing EducationDeputy Editor-in-Chief

Zoltan Endre, SydneyDeputy Editor-in-Chief

Paul Bridgman, Christchurch

bs_bs_banner

View this journal online at wileyonlinelibrary.com/journal/imj

aims and scopeThe Internal Medicine Journal, formerly known as the Australian and New Zealand Journal of Medicine, is the official journal of the Adult Medicine Division of The Royal Australasian College of Physicians (RACP). Its purpose is to publish high-quality internationally competitive peer-reviewed original medical research, both laboratory and clinical, relating to the study and research of human disease. Papers will be considered from all areas of medical practice and science. The Journal also has a major role in continuing medical education and publishes review articles relevant to physician education. Except where otherwise stated, articles are peer reviewed.

abstracting and indexingThis journal is indexed by Abstracts on Hygiene and Communicable Diseases, AgBiotech News and Information, AIDS Abstracts, Australian Medical Index, BIOBASE, Biological Abstracts (BIOSIS), Biomedical Reference (EBSCO), Cambridge Scientifi c Abstracts, Chemical Abstracts Service, Current Contents/Clinical Medicine (an ISI product), Derwent Biotechnology Abstracts, EMBASE/Excerpta Medica, Environmental Sciences and Pollution Management, Health and Safety Science Abstracts (Online version), Helminthological Abstracts, InPharma Weekly, International Pharmaceutical Abstracts (IPA), Journals @ Ovid, MEDLINE, Nutrition Abstracts and Reviews, Pharmacoeoconomics and Outcomes News, Reactions Weekly, Science Citation Index, SCOPUS, Tropical

Diseases Bulletin, Vitis-Viticulture and Oenology Abstracts (Online Edition), World Agricultural Economics and Rural Sociology Abstracts, and CINAHL.

address for editorial correspondenceEditor-in-Chief, Internal Medicine Journal, The Royal Australasian College of Physicians, 145 Macquarie Street, Sydney, NSW 2000, Australia (tel: +61 2 9256 5431; fax: +61 2 9252 3310). For enquiries regarding ScholarOne Manuscripts (formerly known as ManuscriptCentral) submissions please email [email protected] (e.g. IMJ-0000-2017).General enquiries should be directed to Virginia Savickis, the Editorial Office, Internal Medicine Journal, using [email protected] on published papers are welcomed. Authors are offered right of reply (no more than 500 words) at the discretion of the Editor and discussion will not be entered into. Given the current pressures on editorial space, however, invited comments are restricted to one reply.

disclaimerThe Publisher, RACP and Editors cannot be held responsible for errors or any consequences arising from the use of information contained in this journal; the views and opinions expressed do not necessarily reflect those of the Publisher, RACP and Editors, neither does the publication of advertisements constitute any endorsement by the Publisher, RACP and Editors of the products advertised.

Copyright © 2018 Royal Australasian College of Physicians.

For submission instructions, subscription and all other information visit

wileyonlinelibrary.com/journal/imj

Wiley’s Corporate Citizenship initiative seeks to address the environmental, social, economic, and ethical challenges faced in our business and which are important to our diverse stakeholder groups. Since launching the initiative, we have focused on sharing our content with those in need, enhancing community philanthropy, reducing our carbon impact, creating global guidelines and best practices for paper use, establishing a vendor code of ethics, and engaging our colleagues and other stakeholders in our efforts. Follow our progress at www.wiley.com/go/citizenship

Wiley is a founding member of the UN-backed HINARI, AGORA, and OARE initiatives. They are now collectively known as Research4Life, making online scientific content available free or at nominal cost to researchers in developing countries. Please visit Wiley’s Content Access – Corporate Citizenship site: http://www.wiley.com/WileyCDA/Section/id-390082.html

ISSN 1444-0903 (Print)ISSN 1445-5994 (Online)

IMJ.JEBJan16

bs_bs_banner

Editorial

113 What we know and do not know about women and kidney diseases: questions unanswered and answers unquestioned: reflection on World Kidney Day and International Women’s Day


Review



Clinical Perspectives

129 Climate change: allergens and allergic diseases


Original Articles

135 EVOLVE: The Australian Rheumatology Association’s ‘top five’ list of investigations and interventions doctors and patients should question
















Brief Communications

194 Metformin-induced encephalopathy: the role of thiamine




bs_bs_banner

February 2018, Volume 48, Issue 2







Position Paper

210 Considerations for pre-transfusion immunohaematology testing in patients receiving the anti-CD38 monoclonal antibody daratumumab for the treatment of multiple myeloma














Martin S. Knapp



228 Author reply






231 Author reply


233 Corrigendum

bs_bs_banner

February 2018, Volume 48, Issue 2

Consolidate your testing with a growing assay menu. Make the most of space and time with single-platform consolidation.

Big possibilities. Small footprint.

SS-00504-AUS-EN Rev. 001 ©2017 Hologic, Inc. All rights reserved. Hologic, The Science of Sure, Panther Fusion, Open Access and associated logos are trademarks and/or registered trademarks of Hologic, Inc. and/or its subsidiaries in the United States and/or other countries. This information is intended for medical professionals in the U.S. and other markets and is not intended as a product solicitation or promotion where such activities are prohibited. Because Hologic materials are distributed through websites, eBroadcasts and tradeshows, it is not always possible to control where such materials appear. For specific information on what products are available for sale in a particular country, please contact your local Hologic representative or write to [email protected]. Hologic (Australia) Pty Ltd, Level 4, 2-4 Lyon Park Rd, Macquarie Park NSW 2113. Tel. +61 2 9888 8000. ABN 95 079 821 275.

Panther is evolving, adding PCR capabilities to the proven TMA technology on our fully automated, sample-to-result system.

To see how the Panther Fusion system can optimise workflow and consolidate your menu, please visit PantherFusion.com

P718.indd 1P718.indd 1 15/06/17 5:37 PM15/06/17 5:37 PM

EDITORIAL

What we know and do not know about women and kidneydiseases: questions unanswered and answers unquestioned:reflection on World Kidney Day and InternationalWomen’s Day

Introduction

Chronic kidney disease (CKD) affects approximately

10% of the world’s adult population: it is within the

top 20 causes of death worldwide,1 and its impact on

patients and their families can be devastating. World

Kidney Day and International Women’s Day in 2018

coincide, thus offering an opportunity to reflect on the

importance of women’s health and specifically their

kidney health, on the community, and the next gener-

ations; as well as to strive to be more curious about

the unique aspects of kidney disease in women, so that

we may apply those learnings more broadly.Girls and women, who make up approximately 50%

of the world’s population, are important contributors to

the society and their families. Besides childbearing,

women are essential in childrearing and contribute to

sustaining family and community health. Women in the

21st century continue to strive for equality in business,

commerce and professional endeavours, while recognis-

ing that in many situations, equality does not exist. In

various locations around the world, access to education

and medical care is not equitable among men and

women; women remain under-represented in many

clinical research studies, thus limiting the evidence base

on which to make recommendations to ensure best out-

comes (Fig. 1).In this editorial, we focus on what we know and do

not know about women’s kidney health and kidney dis-

ease, and what we might learn in the future to improve

outcomes for all.

What we know and do not know

Pregnancy is a unique challenge and is a major cause

of acute kidney injury (AKI) in women of childbearing

age; AKI and pre-eclampsia (PE) may lead to subse-

quent CKD, but the entity of the risk is not completely

known.2–5 CKD has a negative effect on pregnancy

even at very early stages.6,7 The risks increase with

CKD progression thus posing potentially challenging

ethical issues around conception and maintaining

pregnancies.6–8 We do know that PE increases the

probability of hypertension and CKD in later years, but

we have not evaluated a surveillance or reno-protective

strategy to determine if progressive loss of kidney func-

tion can be attenuated.9–12

Specific systemic conditions, such as systemic lupus

erythematosus (SLE), rheumatoid arthritis (RA) and

systemic scleroderma (SS), are more likely to affect

women than men. We do not know the relative con-

tribution of these acute and chronic conditions on

progression to end-stage renal disease (ESRD) in

women.In CKD cohorts, the prevalence in women is always

less than in men, and they have slower progression to

ESRD.13–15 We do not know why and how much of this

is due to differences in identification of kidney impair-

ment, different access to care, or true difference in dis-

ease severity and prevalence.Women with CKD have a higher cardiovascular risk

than women without CKD16; but their risk is still lower

than that of men with similar degrees of kidney impair-

ment. In haemodialysis cohorts, there are differences in

vascular access types in women versus men, which may

be due to biological or systemic factors. In some loca-

tions, there is differential use of peritoneal and haemo-

dialysis in women and men.Women are more likely to donate kidneys for

transplantation than to receive them. We do not

know if this is because of the differential incidence of

CKD in men versus women, cultural factors or other

reasons.There remain gender differences in access to care in

different regions of the world, and we do not have data

to evaluate directly the extent of these differences, in the

poorest parts of the world in particular.

As a reflection of the worldwide support for this internationalevent, the WKD 2018 editorial is being published in a numberof international journals.

Internal Medicine Journal 48 (2018) 113–123© 2018 Royal Australasian College of Physicians

113

doi:10.1111/imj.13694

Pregnancy, PE, pregnancy-inducedhypertensive disorders and foetalhealth: the importance of women’shealth and kidney health of presentand future generations

What we know

PE is the principal cause of AKI and maternal death, par-ticularly in developing countries.2,17 Pregnancy is themost common cause of AKI in women of childbearingage.10,18,19 Several diseases and conditions, besides PE,hypertensive disorders of pregnancy and CKD, can leadto pregnancy-related AKI. Causes vary in differentregions. Septic abortion after an illegal procedure is theleading cause of early AKI in countries where legal abor-tions are not available, while PE after assisted fertilisa-tion is becoming a leading cause in developedcountries.12,20–22

PE and hypertensive disorders of pregnancy occur in3–10% of all pregnancies;2,3,18 in these disorders, the kid-ney is the main target of an unbalanced pro-angiogenicand anti-angiogenic derangement, leading to hyperten-sion, proteinuria and widespread endothelial damage.The incidence of PE, higher in low-middle income coun-tries (possibly reflecting undiagnosed predisposing dis-eases), peaks at the extremes of reproductive age forreasons mentioned above.12,20–22

The relationship between kidney and placenta is biu-

nivocal, and the presence of CKD is a risk factor for PE

and hypertensive disorders of pregnancy (Fig. 2). Besides

CKD, other conditions cited as risk factors for PE (diabe-

tes, immunologic diseases, baseline hypertension, obesity

and metabolic syndrome) are also risk factors for CKD.

Given that even minor alterations of kidney function are

present in many of these disorders, the importance of

kidney function is indirectly recognised in the develop-

ment of PE. Newer definitions of PE recognise differ-

ences between ‘placental’ and ‘maternal’ causes of PE,

based on novel angiogenic–antiangiogenic markers,23,24

which may be important for management during and

after pregnancy.There are long-term effects of PE on both maternal

and foetal health, but this remains an area of active

research with many unknowns.PE is a risk factor for the future development of CKD

and ESRD in the mother.3–5 The reasons are not fullyunderstood; podocyte loss is a hallmark of PE, suggestingpermanent glomerular damage.25 Endotheliosis, associ-ated with PE, but also found in normal pregnancies, mayherald glomerulosclerosis; tubular and vascular damagemay co-exist.26,27

Besides maternal risks, PE is associated with intrauter-ine and perinatal death, preterm delivery and restrictedintrauterine growth; the latter two are linked to ‘small

Figure 1 Sex differences throughout the continuum of CKD care. AI, autoimmune; AKI, acute kidney injury; AVF, arteriovenous fistula; CKD, chronic

kidney disease; HD, haemodialysis; KT, kidney transplant; RA, rheumatoid arthritis; SLE, systemic lupus erythematosus; SS, systemic scleroderma.

Editorial


114

babies’.2,3,5 Small babies and preterm babies have highlyincreased risks of neurological deficits and postnatalcomplications, especially sepsis.28–32 The risks may behigher in low-income countries, since survival anddeficit-free survival depend on the provision of postnatalintensive care.20,21 In the long term, small babies are atrisk for the development of diabetes, metabolic syn-drome, cardiovascular diseases (CVD) and CKD inadulthood.33–37 Since kidney development is completedin the last phases of pregnancy, delayed, insufficient kid-ney growth, resulting in low nephron number is proba-bly the basis of the increased risk of CKD andhypertension in small for gestational age and pretermbabies.33–37

Pregnancy in CKD, dialysis andtransplantation

What we know

Chronic kidney disease

CKD is a risk factor for adverse pregnancy outcomesfrom its early stages (Table S1, Supporting informa-tion).6,38,39 The risks increase from CKD stage 1 to CKDstage 5, and may be higher in glomerular nephropathies,autoimmune diseases and diabetic nephropathy.6,7,38–41

Results of pregnancy after kidney donation suggest thatreduction of kidney parenchyma may be associated witha higher risk of PE and hypertensive disorders ofpregnancy.42,43

Hypertension and proteinuria at baseline are impor-tant modulators of pregnancy-related risks; among therisks, we know that malformations are not increasedwith respect to the overall population (out of the contextof inherited diseases, such as reflux nephropathy, poly-cystic kidney disease or congenital anomalies of the kid-ney and urinary tract), maternal death is unusual(in highly resourced countries), while the incidence ofpreterm delivery and of small for gestational age babies,intrinsically linked, is increased in stage 1 CKD patientsand rises with the worsening of kidney function. Simi-larly, the effect of pregnancy on CKD progression is notfully understood because of different study designs,obstetric policies and duration of follow-up. Overall,short- and long-term decrease in kidney function isunusual in early CKD, but the risk increases as CKDseverity increases.6,7,38–41,44–48

Pregnancy is a potential occasion for the initial diagno-sis of CKD. In poorly or unevenly resourced countries,advanced CKD may be discovered only during preg-nancy. The implications of dialysis initiation may presentimportant clinical and ethical issues; in highly resourcedcountries with established prenatal care, the diagnosis ofearlier stages of CKD may lead to more intensive therapyand surveillance.49–51

Dialysis and transplantation

Fertility is reduced in ESRD; Australian and Europeandata suggest a 1:10 ratio from general population totransplantation and from transplantation to dialysis(1:100 probability as compared to the general popula-tion).52,53 The first sporadic cases of successful pregnancy

Placenta and kidney are highly vascularized; filter blood; divide

compartments; are sophisticated metabolic machines

Pregnancy is a precious occasion

to diagnose CKD.

From the placenta to the kidneyPE may induce permanent kidney damage, via

AKI, tubular damage, podocyte loss

From the kidney to the placentaCKD may induce placental dysfunction

with an increased risk of pre-term delivery, hypertensive disorders of

pregnancy and PE

Figure 2 Pregnancy and kidney

function: complex interactions

between two organs, the kidney

and placenta. AKI, acute kidney

injury; CKD, chronic kidney dis-

ease; PE, pre-eclampsia.

Editorial


115

on dialysis were described in the 1970s, but in the newmillennium this became an acknowledged real clinicalpossibility.8,54,55

More than 1000 pregnancies have been reported indialysis patients.55 The most important advance has beenthe demonstration of a strong relationship between theintensity (frequency and duration) of the dialysis ses-sions and positive pregnancy results: thus, intensifyingdialysis up to daily, is the current standard of care.8,54

Changing attitudes towards counselling women withadvanced CKD may be impacted, with the knowledge ofpositive outcomes on dialysis for women and theiroffspring.

Fertility is partly restored after kidney transplantation.56–60

However, even in an ideal situation (normal kidney func-tion, no hypertension or proteinuria, at least 2 years aftertransplantation, without recent rejection episodes), the riskof complications is higher in women with transplanted kid-neys than in the general population. However, if teratogendrugs are avoided (mycophenolic acid and rapamycin), theoutcomes of pregnancy after kidney transplantation sharethe same risk factors as CKD (kidney function, hypertensionand proteinuria).59

Experience with pregnancy in patients with a reducedrenal function or failing kidney graft is limited andcounselling is still forcibly based on personal experienceor indirect evidence.61,62 Assisted fertilisation techniquesare increasingly popular in some settings, but dedicatedstudies in CKD patients are few; multiple pregnanciesmay bear an added risk in CKD patients, with bothnative and transplanted kidneys.

Autoimmune diseases, women andkidney disease

What we know

Autoimmune diseases, such as SLE, RA and SS, prefer-entially affect women and are characterised by systemicinflammation leading to target organ dysfunction,including the kidneys. Sex differences in the incidenceand severity of these diseases result from a complexinteraction of hormonal, genetic and epigenetic factors

(Table 1). The public health burden of autoimmune dis-eases, which collectively represent a leading cause ofmorbidity and mortality among women throughoutadulthood, is substantial.63–65

SLE is an autoimmune disease with multiple organinvolvement, affecting approximately five million peopleworldwide; disproportionately predominant in women(9:1 female-to-male ratio) and individuals ofnon-European ancestry. The highest female predomi-nance (up to 15:1) is in peak reproductive years. Thebiology of these differences has been explored: oneexplanation is the number of X chromosomes andgenetic variants on the X chromosome;66–68 anotherimportant aetiological explanation is the role of oestro-gen in SLE. Oestrogen’s primary effects are mediated bytranscription activity of the intracellular oestrogen recep-tors, whose profile is altered in T-cells from female SLEpatients.69,70 Cathepsin S protein has recently been iden-tified as a potential cause of lupus, triggering theimmune system to attack healthy cells, particularly infemales.71 Numerous non-human leukocyte antigengenetic markers may predispose individuals ofEuropean, Hispanic and Afro-American ancestry tolupus.72 Susceptibility to SLE during pregnancy is alsomultifactorial; one factor being upregulation of IFN-α.Elevated IFN-α, expressed by the placenta, plays a patho-genic role in SLE, contributing both to the success of pla-cental reproduction and to increased susceptibility toSLE.73 Regulatory T-cells (which may be the key to cellmodulating foeto-maternal tolerance) have abnormali-ties of structure and function, and may contribute topregnancy pathology in women with SLE and to chal-lenges of managing them during pregnancy.74 SLEaffects kidneys in about 50% of patients, including glo-merular, interstitial and vascular lesions. Lupus nephritisis a major risk factor for overall morbidity and mortalityin SLE, and despite potent therapies still leads to signifi-cant impairment of kidney function for many patients.75

Kidney disease is a critical concern in counsellingwomen with lupus considering pregnancy, with previouskidney involvement and lower C4 levels conferring highrisk of active nephritis occurring in pregnancy.76 Socio-economic disparities are also linked to the health of

Table 1 Sex differences in the incidence and severity of autoimmune diseases

SLE RA SS

Peak incidence Reproductive age Perimenopausal After 50–60 yearsFemale/male ratio Peak: 15:1 Peak: 4:1 Peak: 14:1

Total: 9:1 After 60 years: 1:1 Total: 3:1Influence of oestrogenHigh levels Negative Positive ?Low levels ? Negative Negative

RA, rheumatoid arthritis; SLE, systemic lupus erythematosus; SS, systemic scleroderma.

Editorial


116

patients with lupus. Poverty is associated with anincreased long-term level of accumulated disease-associated damage and a 1.67 times increased likelihoodof experiencing a clinically meaningful increase in dam-age. Frequency of adverse pregnancy outcomes inwomen with lupus is twofold higher in black and His-panic women than in white women. In blacks, socioeco-nomic status was a determinant of pregnancy outcomesand a key contributor to adverse pregnancyoutcomes.77,78

RA also preferentially affects women (4:1 ratio tomen) with the peak incidence at age 45–55, coincidingwith the perimenopausal years. This suggests a possibleassociation between oestrogen deficiency and diseaseonset. Female-to-male incidence ratio after the age of60 years is approximately 1:1, potentially implicatingchanges in sex hormones in the development of RA, anda pattern of RA symptom improvement or even remis-sion during pregnancy is well recognised.79–81 Renalinvolvement in RA is relatively common and multifacto-rial and is a predictor of mortality in RA patients. Therisk of CKD is significantly higher in patients with RAthan in the general population. The development ofCKD may result from several ongoing processes,including specific renal involvement associated withRA (e.g. glomerulonephritis, interstitial nephritis),chronic inflammation, comorbidities and nephrotoxicanti-rheumatic drugs. The strong association betweenRA activity and AA amyloidosis increases morbidity andis the main cause of ESRD with RA and nephropathy.Importantly, some of the life-long and combined RApharmacotherapy can lead to various renal sideeffects.82–84

SS predominantly affects women (female-to-maleratios ranging from 3:1 to 14:1), with the peak inci-dence in the fifth and sixth decades. Oestrogen mayplay a role in scleroderma pathogenesis through itsstimulatory effect on transforming growth factor-beta1 receptor and platelet-derived growth factor recep-tor.85 Vasculopathy is an important disease-relatedmanifestation in SS, and the low oestrogenic state asso-ciated with menopause has been suggested to aggravatevascular manifestations in affected women.86 SS canalso be complicated by several different forms of kidneydisease, including scleroderma renal crisis, which repre-sents a form of malignant hypertension with acuterenal failure; or more commonly ischaemic nephropa-thy leading to slowly progressive CKD, accompanied byhypertension and albuminuria.78 Normotensive acuterenal failure in patients with SS may be caused byinterstitial nephritis or anti-neutrophil cytoplasmic anti-bodies (ANCA) vasculitis, a separate entity in sclero-derma with poor outcome.87–89

Women, CKD and access to renalreplacement therapies

What we know

Although renal replacement therapy (RRT), includingdialysis and transplantation is life sustaining, not allpatients receive RRT. The rate of ESRD treated by RRTdiffers greatly between countries and regions, and intri-cately depends on the economy of a country and health-care system.90,91 Worldwide, only 50% of patientsrequiring RRT receive treatment,92 and in low- andmiddle-income countries and regions, even less; in largeparts of sub-Saharan Africa, less than 2% of ESRD aretreated by RRT.93 The equality of access to RRT forwomen and girls is of particular concern because, inmany societies, they are disadvantaged by discriminationrooted in sociocultural factors.94,95

Sex differences in access to dialysis

At least 2.284 million people may have died prematurelydue to lack of access to RRT with treatment gaps beingmuch larger in low-income countries, with conservativeestimates in Asia and Africa of 1.907 million and432 000 people not receiving RRT. By 2030, the esti-mated number of RRT should be more than double to5.439 million (3.899–7.640 million), with the mostgrowth in Asia (0.968 million to a projected 2.162 mil-lion (1.71–3.14 million)).92 These numbers are derivedfrom an extensive systematic review.There are few data to compare the gender difference

for the treatment gaps. Studies in Africa show that menwere more likely to receive RRT than women.96,97 InJapan, the incidence of treated ESRD in females was lessthan half of that in males (3287 in males vs 1764 womenper million population treated):91 no explanations aregiven for this finding. One US study reports women hav-ing significantly higher odds ratio of 1.70 for late initia-tion of dialysis compared to men.98 Awareness levels ofprevious kidney disease in women were reported muchlower than in men (2.9 � 1.6% in women vs 17.9 �5.9% in men), which may contribute to later initiationof RRT.99

Mortality rates are similar in men and women on dial-ysis, but the incident rates of some dialysis-associatedcomplications and morbidity are higher in women. A USreport of hospitalisations in 111 653 patients undergoingmaintenance haemodialysis describes higher hospitalisa-tion rates in women and higher risk for 30-dayreadmissions.100

In addition, the prevalent use of arteriovenous fistula,which is associated with reduced mortality, complication

Editorial


117

and costs, is lower among female than male haemodialy-sis patients.101 This may be due to several different fac-tors, including anatomical/surgical issues related tovessel size, timing of referral and attitudinal differences.This has not been systematically studied.

Dialysis dose which is evaluated by Kt/V may result inunder-dialysis in women who have an average smallervolume of urea distribution or total body water thanmen.102 Women receiving dialysis have also beenreported to have worse clinical parameters, includinganaemia, nutrition and quality of life.103 Reasons areuncertain.

Sex differences in access to kidneytransplantation

Transplantation represents the best form of RRT inpatients without contraindications. Worldwide datadescribe that women are less likely than men to be kidneytransplant patients, either from a cadaveric or livingdonor, but are more likely to serve as living donors forkidney transplantation.104 Data from different countries,including the United States, France, China and India, con-firm differential kidney transplant rates (lower in womenthan men), less likelihood of women being registered onnational transplant waiting lists and longer time from dial-ysis initiation to listing. Mothers are more likely to bedonors, as are female spouses.91,105–108 Sex inequality alsoexists in the paediatric population. A survey from 35 coun-tries participating in the European Society for PediatricNephrology/European Renal Association-EuropeanDialysis and Transplant Association Registry, reportedgirls had a lower access to renal transplantation thanboys.109

Socioeconomic factors undoubtedly play a role in theinequality of transplantation between sexes, especially inthe low- and middle-income countries and regions. Gen-erally, men provide the major income for their familywhich may discourage them to donate kidneys. Differentemployment status and incomes between genders maycontribute to sex differences in transplantation becauseemployment and income status is usually associated withbetter healthcare insurance which cover the costs fortransplantation. Psychosocial factors and education ofwomen have been suggested as a contribution to sex dis-parity. US data found black women were less likely towant living donor kidney transplantation compared withmen, despite being twice as likely as men to receiveunsolicited offers for kidneys. They were also less likelyto have been evaluated for a kidney transplant.110 Otherreports describe disparities in age and sex in access tokidney transplantation which originate at the time ofpre-referral discussions about kidney transplantation;

irrespective of age, women were more likely not to havehad discussions with medical professionals. This resultmay imply that there is a need for better clinical guide-lines and education for women, their social network andtheir providers.111

Present and future what we do not know

Given the data presented above with respect to preg-nancy, AKI, autoimmune diseases, CKD, dialysis andtransplantation, there are many unanswered questions.In high-income countries with increasing maternal ageand assisted fertilisation, there may be an increase in PEwhich may impact future generations if associated withadverse foetal outcomes. The increase in in vitro fertilisa-tion techniques for those of advanced maternal age maylead to multiple pregnancies, which may predispose toPE, intrauterine growth restriction or both. Will this leadto an increase in CKD and CVD for women in thefuture?

Due to the high heterogeneity of CKD, we do notknow if and how pregnancy outcomes are modulated bythe different nephropathies, as besides the most commonones, such as IgA or lupus nephropathy, diabeticnephropathy and reflux nephropathy, evidence isscant.44,45,112–114 How should we define preconceptionrisks of pregnancy with respect to current proteinuriacut-offs? Neither indications on when to start dialysis inpregnancy are well established nor is the specific role offrequency and duration. In those with kidney trans-plants, given the changing expanded donor policies,higher age at transplantation and reduced fertility inolder women, there may be changes in attitudes towardspregnancy with less than optimal kidney function.56,60

How this will impact short- and long-term outcomes ofmothers and their babies is not clear.

Teen pregnancies are very common in some parts ofthe world, and they are often associated with lowincome and cultural levels. The uneven legal rules forassisted fertilisation and the lack of systematic assess-ment of the kidney function point to the need for furtherresearch.

Despite elegant demonstrations for the role of sex hor-mones in vascular health and immunoregulation, thestriking predominance in females of SLE, RA and SSremains unexplained relative to other systemic diseases,such as ANCA vasculitis and haemolytic-uraemic syn-drome. Note that thrombotic thrombocytopenic purpurahas a higher incidence in women, though this is likelydue to the association with other conditions more com-mon in women. The incidence of kidney involvement inSLE during pregnancy and similarities/differences inthose with PE has not been well studied. The role of

Editorial


118

different medications and responses to medications forautoimmune diseases relative to sex has also not beenwell studied.More attention to similarities between conditions, the

importance of sex hormones in inflammation, immunemodulation and vascular health, may lead to importantinsights and clinical breakthroughs over time. If womenare more likely to be living donors, at differential ages,does this impact both CVD risk, and risk for ESKD: havewe studied this well enough, in the current era, withmodern diagnostic criteria for CKD and sophisticatedtools to understand renal reserve? Are the additionalexposures that women have after living donation com-pounded by hormonal changes on vasculature as theyage? And are the risks of CKD and PE increased in theyounger female kidney living donor?In the context of specific therapies for the treatment or

delay of CKD progression, do we know if there are sexdifferences in therapeutic responses to ACEi/ARB?Should we look at dose finding/adjustments by sex? Ifvascular and immune biology is impacted by sex hor-mones as described earlier, do we know the impact ofvarious therapies by level or ratio of sex hormones? Inlow-middle income countries how does changing eco-nomic and social cultures impact women’s health, andwhat is the nutritional impact on CKD of increasing pre-dominance of obesity, diabetes and hypertension?

Summary

Women have unique risks for kidney diseases: kidneydiseases, as well as issues related to access to care, have aprofound impact on both the current and next genera-tions. Advocating for improved access to care for womenis critical to maintain the health of families, communitiesand populations.Focused studies on the unique contribution of sex hor-

mones, or the interaction of sex hormones and other physi-ology, are important to improve our understanding of theprogression of kidney diseases. Immunological conditions,such as pregnancy (viewed as a state of tolerance to non-

self ) as well as SLE and other autoimmune and systemicconditions common in women, better studied may also leadto breakthroughs in understanding and care paradigms.There is a clear need for higher awareness, timely

diagnosis and proper follow-up of CKD in pregnancy. Inturn, pregnancy may also be a valuable occasion forearly diagnosis of CKD, allowing planning of therapeuticinterventions.On this occasion, World Kidney Day and the Interna-

tional Women’s Day 2018 are commemorated on the sameday, offering us the opportunity to highlight the importanceof women’s health and particularly their kidney health. Onits 13th anniversary, World Kidney Day promotes afford-able and equitable access to health education, healthcareand prevention for all women and girls in the world.The coincidence of World Kidney Day and Interna-

tional Women’s Day offers an opportunity to developand define best practices and future research agendas,and ultimately, to optimise the outcomes of all peopleliving with or at risk for kidney disease.

Received 2 November 2017; accepted 20 November 2017.

Giorgina B. Piccoli,1,2 Mona Alrukhaimi,3 Zhi-Hong Liu,4

Elena Zakharova5,6,7 and Adeera Levin,8 on behalf of theWorld Kidney Day Steering Committee†

1Department of Clinical and Biological Sciences, University of

Torino, Torino, Italy, 2Department of Nephrology, Centre

Hospitalier Le Mans, Le Mans, France, 3Department of Medicine,

Dubai Medical College, Dubai, United Arab Emirates, 4National

Clinical Research Center of Kidney Diseases, Jinling Hospital,

Nanjing University School of Medicine, Nanjing, China,5Department of Nephrology, Moscow City Hospital S.P. Botkin,

6Department of Nephrology, Moscow State University of Medicine

and Dentistry, and 7Department of Nephrology, Russian Medical

Academy of Continuous Professional Education, Moscow, Russia,

and 8Department of Medicine, Division of Nephrology, University of

British Columbia, Vancouver, British Columbia, Canada

†Members of the World Kidney Day Steering Committee are: Philip

Kam Tao Li, Guillermo Garcia-Garcia, Mohammed Benghanem-

Gharbi, Kamyar Kalantar-Zadeh, Charles Kernahan, Latha

Kumaraswami, Giorgina Barbara Piccoli, Gamal Saadi, Louise Fox,

Elena Zakharova and Sharon Andreoli.

References

1 GBD 2015 Disease and Injury

Incidence and Prevalence

Collaborators. Global, regional, and

national incidence, prevalence, and

years lived with disability for

310 diseases and injuries, 1990–2015:

a systematic analysis for the Global

Burden of Disease Study 2015. Lancet

2016; 388: 1545–602.

2 von Dadelszen P, Payne B, Li J,

Ansermino JM, Broughton Pipkin F,

Côté A-M et al. Prediction of adverse

maternal outcomes in pre-eclampsia:

development and validation of the

fullPIERS model. Lancet 2011; 377:

219–27.

3 Mol BWJ, Roberts CT,

Thangaratinam S, Magee LA, de

Groot CJM, Hofmeyr GJ. Pre-

eclampsia. Lancet 2016; 387: 999–1011.

4 Vikse BE, Irgens LM, Leivestad T,

Skjaerven R, Iversen BM.

Preeclampsia and the risk of end-stage

renal disease. N Engl J Med 2008; 359:

800–9.

5 Theilen LH, Fraser A,

Hollingshaus MS, Schliep KC,

Varner MW, Smith KR et al. All-cause

and cause-specific mortality after

hypertensive disease of pregnancy.

Obstet Gynecol 2016; 128: 238–44.

Editorial


119

6 Piccoli GB, Cabiddu G, Attini R,

Vigotti FN, Maxia S, Lepori N et al. Risk

of adverse pregnancy outcomes in

women with CKD. J Am Soc Nephrol

2015; 26: 2011–22.

7 Zhang J-J, Ma X-X, Hao L, Liu L-J,

Lv J-C, Zhang H. A systematic review

and meta-analysis of outcomes of

pregnancy in CKD and CKD outcomes

in pregnancy. Clin J Am Soc Nephrol

2015; 10: 1964–78.

8 Alkhunaizi A, Melamed N,

Hladunewich MA. Pregnancy in

advanced chronic kidney disease and

end-stage renal disease. Curr Opin

Nephrol Hypertens 2015; 24: 252–9.

9 Piccoli GB, Cabiddu G, Castellino S,

Gernone G, Santoro D, Moroni G et al.

A best practice position statement on

the role of the nephrologist in the

prevention and follow-up of

preeclampsia: the Italian study group

on kidney and pregnancy. J Nephrol

2017; 30: 307–17.

10 Liu Y, Ma X, Zheng J, Liu X, Yan T.

Pregnancy outcomes in patients with

acute kidney injury during pregnancy:

a systematic review and meta-analysis.

BMC Pregnancy Childbirth 2017;

17: 235.

11 Jim B, Garovic VD. Acute kidney

injury in pregnancy. Semin Nephrol

2017; 37: 378–85.

12 Acharya A. Management of acute

kidney injury in pregnancy for the

obstetrician. Obstet Gynecol Clin North

Am 2016; 43: 747–65.

13 Iseki K. Gender differences in chronic

kidney disease. Kidney Int 2008; 74:

415–7.

14 Nitsch D, Grams M, Sang Y, Black C,

Cirillo M, Djurdjev O et al. Associations

of estimated glomerular filtration rate

and albuminuria with mortality and

renal failure by sex: a meta-analysis.

BMJ 2013; 346: f324.

15 Levin A, Djurdjev O, Beaulieu M, Er L.

Variability and risk factors for kidney

disease progression and death

following attainment of stage 4 CKD in

a referred cohort. Am J Kidney Dis

2008; 52: 661–71.

16 Weiner DE, Tighiouart H, Elsayed EF,

Griffith JL, Salem DN, Levey AS et al.

The Framingham predictive

instrument in chronic kidney disease.

J Am Coll Cardiol 2007; 50: 217–24.

17 Oladapo OT, Adetoro OO, Ekele BA,

Chama C, Etuk SJ, Aboyeji AP et al.

When getting there is not enough: a

nationwide cross-sectional study of

998 maternal deaths and 1451 near-

misses in public tertiary hospitals in a

low-income country. BJOG 2016; 123:

928–38.

18 Tranquilli AL, Dekker G, Magee L,

Roberts J, Sibai BM, Steyn W et al. The

classification, diagnosis and

management of the hypertensive

disorders of pregnancy: a revised

statement from the ISSHP. Pregnancy

Hypertens 2014; 4: 97–104.

19 Liu Y, Bao H, Jiang Z, Huang Y,

Wang N. Pregnancy-related acute

kidney injury and a review of the

literature in China. Intern Med 2015;

54: 1695–703.

20 Prakash J, Pant P, Prakash S,

Sivasankar M, Vohra R, Doley PK et al.

Changing picture of acute kidney

injury in pregnancy: study of 259 cases

over a period of 33 years. Indian J

Nephrol 2016 Aug; 26: 262–7.

21 Ibarra-Hernández M, Orozco-

Guillén OA, De la Alcantar-Vallín ML,

Garrido-Roldan R, Jiménez-

Alvarado MP, Castro KB et al. Acute

kidney injury in pregnancy and the

role of underlying CKD: a point of

view from México. J Nephrol 2017; 30:

773–80.

22 Blázquez A, García D, Rodríguez A,

Vassena R, Figueras F, Vernaeve V. Is

oocyte donation a risk factor for

preeclampsia? A systematic review and

meta-analysis. J Assist Reprod Genet

2016; 33: 855–63.

23 O’Gorman N, Wright D, Poon LC,

Rolnik DL, Syngelaki A, de Alvarado M

et al. Multicenter screening for pre-

eclampsia by maternal factors and

biomarkers at 11–13 weeks’ gestation:

comparison with NICE guidelines and

ACOG recommendations. Ultrasound

Obstet Gynecol 2017; 49: 756–60.

24 Zeisler H, Llurba E, Chantraine F,

Vatish M, Staff AC, Sennström M et al.

Predictive value of the sFlt-1:PlGF ratio

in women with suspected

preeclampsia. N Engl J Med 2016; 374:

13–22.

25 Garovic VD. The role of the podocyte

in preeclampsia. Clin J Am Soc Nephrol

2014; 9: 1337–40.

26 Wide-Swensson D, Strevens H,

Willner J. Antepartum percutaneous

renal biopsy. Int J Gynaecol Obstet 2007;

98: 88–92.

27 Shiiki H, Dohi K, Hanatani M, Fujii Y,

Sanai H, Ichijo M et al. Focal and

segmental glomerulosclerosis in

preeclamptic patients with nephrotic

syndrome. Am J Nephrol 1990; 10:

205–12.

28 Linsell L, Malouf R, Morris J,

Kurinczuk JJ, Marlow N. Risk factor

models for neurodevelopmental

outcomes in children born very

preterm or with very low birth weight:

a systematic review of methodology

and reporting. Am J Epidemiol 2017;

185: 601–12.

29 Guellec I, Lapillonne A, Marret S,

Picaud J-C, Mitanchez D,

Charkaluk M-L et al. Effect of intra-

and extrauterine growth on long-term

neurologic outcomes of very preterm

infants. J Pediatr 2016; 175: 93–99.e1.

30 Moore T, Hennessy EM, Myles J,

Johnson SJ, Draper ES, Costeloe KL

et al. Neurological and developmental

outcome in extremely preterm

children born in England in 1995 and

2006: the EPICure studies. BMJ 2012;

345: e7961.

31 Guillén Ú, DeMauro S, Ma L,

Zupancic J, Roberts R, Schmidt B et al.

Relationship between attrition and

neurodevelopmental impairment rates

in extremely preterm infants at 18 to

24 months: a systematic review. Arch

Pediatr Adolesc Med 2012; 166: 178–84.

32 Ranke MB, Schweizer R,

Rodemann SM, Bevot A, Martin DD,

Goelz R et al. Schoolchildren born

VLBW or VLGA show height-related

changes in body composition and

muscle function but no evidence of

metabolic syndrome risk factors.

Results from the NEOLONG Study.

J Pediatr Endocrinol Metab 2015; 29:

163–72.

33 Castanys-Muñoz E, Kennedy K,

Castañeda-Gutiérrez E, Forsyth S,

Godfrey KM, Koletzko B et al.

Systematic review indicates postnatal

growth in term infants born small-for-

gestational-age being associated with

later neurocognitive and metabolic

outcomes. Acta Paediatr 2017; 106:

1230–8.

34 Ong KK, Kennedy K, Castañeda-

Gutiérrez E, Forsyth S, Godfrey KM,

Koletzko B et al. Postnatal growth in

preterm infants and later health

outcomes: a systematic review. Acta

Paediatr 2015 Oct; 104: 974–86.

35 Low Birth Weight and Nephron

Number Working Group. The impact

of kidney development on the life

Editorial


120

course: a consensus document for

action. Nephron 2017; 136: 3–49.

36 Luyckx VA, Bertram JF, Brenner BM,

Fall C, Hoy WE, Ozanne SE et al. Effect

of fetal and child health on kidney

development and long-term risk of

hypertension and kidney disease.

Lancet 2013; 382: 273–83.

37 Luyckx VA, Brenner BM. Birth

weight, malnutrition and kidney-

associated outcomes – a global

concern. Nat Rev Nephrol 2015; 11:

135–49.

38 Davison JM, Lindheimer MD.

Pregnancy and chronic kidney disease.

Semin Nephrol 2011; 31: 86–99.

39 Hall M. Pregnancy in women with

CKD: a success story. Am J Kidney Dis

2016; 68: 633–9.

40 Nevis IF, Reitsma A, Dominic A,

McDonald S, Thabane L, Akl EA et al.

Pregnancy outcomes in women with

chronic kidney disease: a systematic

review. Clin J Am Soc Nephrol 2011; 6:

2587–98.

41 Cabiddu G, Castellino S, Gernone G,

Santoro D, Moroni G, Giannattasio M

et al. A best practice position statement

on pregnancy in chronic kidney

disease: the Italian Study Group on

Kidney and Pregnancy. J Nephrol 2016;

29: 277–303.

42 Garg AX, Nevis IF, McArthur E,

Sontrop JM, Koval JJ, Lam NN et al.

Gestational hypertension and

preeclampsia in living kidney donors.

N Engl J Med 2015; 372: 124–33.

43 Josephson MA. Transplantation:

pregnancy after kidney donation:

more questions than answers. Nat Rev

Nephrol 2009; 5: 495–7.

44 Gianfreda D, Quaglini S, Frontini G,

Raffiotta F, Messa P, Moroni G. Does

pregnancy have any impact on long

term damage accrual and on the

outcome of lupus nephritis?

J Autoimmun 2017; 84: 46–54.

45 Blom K, Odutayo A, Bramham K,

Hladunewich MA. Pregnancy and

glomerular disease: a systematic

review of the literature with

management guidelines. Clin J Am Soc

Nephrol 2017; 12: 1862–72.

46 Imbasciati E, Gregorini G, Cabiddu G,

Gammaro L, Ambroso G, Del Giudice A

et al. Pregnancy in CKD stages 3 to 5:

fetal and maternal outcomes.

Am J Kidney Dis 2007; 49: 753–62.

47 Fischer MJ. Chronic kidney disease

and pregnancy: maternal and fetal

outcomes. Adv Chronic Kidney Dis 2007;

14: 132–45.

48 Bramham K. Diabetic nephropathy

and pregnancy. Semin Nephrol 2017;

37: 362–9.

49 Eswarappa M, Rakesh M, Sonika P,

Snigdha K, Midhun M, Kaushik K

et al. Spectrum of renal injury in

pregnancy-induced hypertension:

experience from a single center in

India. Saudi J Kidney Dis Transpl 2017;

28: 279–84.

50 Prakash J. The kidney in pregnancy: a

journey of three decades. Indian J

Nephrol 2012; 22: 159–67.

51 Piccoli GB, Fassio F, Attini R, Parisi S,

Biolcati M, Ferraresi M et al. Pregnancy

in CKD: whom should we follow and

why? Nephrol Dial Transplant 2012; 27

(Suppl 3): iii111–18.

52 Piccoli GB, Cabiddu G, Daidone G,

Guzzo G, Maxia S, Ciniglio I et al. The

children of dialysis: live-born babies

from on-dialysis mothers in Italy – an

epidemiological perspective comparing

dialysis, kidney transplantation and

the overall population. Nephrol Dial

Transplant 2014; 29: 1578–86.

53 Jesudason S, Grace BS, McDonald SP.

Pregnancy outcomes according to

dialysis commencing before or after

conception in women with ESRD.

Clin J Am Soc Nephrol 20147;

9: 143–9.

54 Hladunewich MA, Hou S, Odutayo A,

Cornelis T, Pierratos A, Goldstein M

et al. Intensive hemodialysis associates

with improved pregnancy outcomes: a

Canadian and United States cohort

comparison. J Am Soc Nephrol 2014;

25: 1103–9.

55 Piccoli GB, Minelli F, Versino E,

Cabiddu G, Attini R, Vigotti FN et al.

Pregnancy in dialysis patients in the

new millennium: a systematic review

and meta-regression analysis

correlating dialysis schedules and

pregnancy outcomes. Nephrol Dial

Transplant 2016; 31: 1915–34.

56 Deshpande NA, James NT,

Kucirka LM, Boyarsky BJ, Garonzik-

Wang JM, Montgomery RA et al.

Pregnancy outcomes in kidney

transplant recipients: a systematic

review and meta-analysis.

Am J Transplant 2011; 11: 2388–404.

57 Deshpande NA, Coscia LA, Gomez-

Lobo V, Moritz MJ, Armenti VT.

Pregnancy after solid organ

transplantation: a guide for obstetric

management. Rev Obstet Gynecol 2013;

6: 116–25.

58 Bramham K, Nelson-Piercy C, Gao H,

Pierce M, Bush N, Spark P et al.

Pregnancy in renal transplant

recipients: a UK national cohort study.

Clin J Am Soc Nephrol 2013; 8: 290–8.

59 Piccoli GB, Cabiddu G, Attini R,

Gerbino M, Todeschini P, Perrino ML

et al. Outcomes of pregnancies after

kidney transplantation: lessons learned

from CKD. A comparison of

transplanted, nontransplanted chronic

kidney disease patients and low-risk

pregnancies: a multicenter nationwide

analysis. Transplantation 2017; 101:

2536–44.

60 Webster P, Lightstone L, McKay DB,

Josephson MA. Pregnancy in chronic

kidney disease and kidney

transplantation. Kidney Int 2017; 91:

1047–56.

61 Pietrzak B, Mazanowska N,

Kociszewska-Najman B, Szymusik I,

Grzechoci�nska B, Pazik J et al.

Successful pregnancy outcome after

in vitro fertilization in a kidney graft

recipient: a case report and literature

review. Ann Transplant 2015; 20:

338–41.

62 Norrman E, Bergh C, Wennerholm U-

B. Pregnancy outcome and long-term

follow-up after in vitro fertilization in

women with renal transplantation.

Hum Reprod 2015; 30: 205–13.

63 Tedeschi SK, Bermas B,

Costenbader KH. Sexual disparities in

the incidence and course of SLE and

RA. Clin Immunol 2013; 149: 211–8.

64 Marder W, Vinet É, Somers EC.

Rheumatic autoimmune diseases in

women and midlife health. Womens

Midlife Health 2015; 1: 11.

65 Ortona E, Pierdominici M, Maselli A,

Veroni C, Aloisi F, Shoenfeld Y. Sex-

based differences in autoimmune

diseases. Ann Ist Super Sanita 2016; 52:

205–12.

66 Petri M. Epidemiology of systemic

lupus erythematosus. Best Pract Res Clin

Rheumatol 2002; 16: 847–58.

67 Weckerle CE, Niewold TB. The

unexplained female predominance of

systemic lupus erythematosus: clues

from genetic and cytokine studies. Clin

Rev Allergy Immunol 2011; 40: 42–9.

68 Scofield RH, Bruner GR, Namjou B,

Kimberly RP, Ramsey-Goldman R,

Petri M et al. Klinefelter’s syndrome

(47,XXY) in male systemic lupus

Editorial


121

erythematosus patients: support for

the notion of a gene-dose effect from

the X chromosome. Arthritis Rheum

2008; 58: 2511–7.

69 Pierdominici M, Ortona E. Estrogen

impact on autoimmunity onset and

progression: the paradigm of systemic

lupus erythematosus. Int Trends Immun

2013; 1: 24–34.

70 Maselli A, Conti F, Alessandri C,

Colasanti T, Barbati C, Vomero M et al.

Low expression of estrogen receptor βin T lymphocytes and high serum

levels of anti-estrogen receptor αantibodies impact disease activity in

female patients with systemic lupus

erythematosus. Biol Sex Differ 2016;

7: 3.

71 Kim SJ, Schätzle S, Ahmed SS,

Haap W, Jang SH, Gregersen PK et al.

Increased cathepsin S in Prdm1(−/−)

dendritic cells alters the TFH cell

repertoire and contributes to lupus.

Nat Immunol 2017; 18: 1016–24.

72 Langefeld CD, Ainsworth HC,

Cunninghame Graham DS, Kelly JA,

Comeau ME, Marion MC et al.

Transancestral mapping and genetic

load in systemic lupus erythematosus.

Nat Commun 2017; 17: 16021.

73 Niewold TB, Hua J, Lehman TJA,

Harley JB, Crow MK. High serum IFN-

alpha activity is a heritable risk factor

for systemic lupus erythematosus.

Genes Immun 2007; 8: 492–502.

74 Tower C, Mathen S, Crocker I,

Bruce IN, Regulatory T. Cells in

systemic lupus erythematosus and

pregnancy. Am J Reprod Immunol 2013;

69: 588–95.

75 Almaani S, Meara A, Rovin BH.

Update on lupus nephritis. Clin J Am

Soc Nephrol 2017; 12: 825–35.

76 Buyon JP, Kim MY, Guerra MM, Lu S,

Reeves E, Petri M et al. Kidney

outcomes and risk factors for nephritis

(flare/de novo) in a multiethnic cohort

of pregnant patients with lupus. Clin J

Am Soc Nephrol 2017; 12: 940–6.

77 Yelin E, Yazdany J, Trupin L.

Relationship between process of care

and a subsequent increase in damage

in systemic lupus erythematosus.

Arthritis Care Res 2017; 69: 927–32.

78 Kaplowitz ET, Ferguson S, Guerra M,

Laskin CA, Buyon JP, Petri M et al.

Socioeconomic status contributes to

racial/ethnic disparities in adverse

pregnancy outcomes among women

with systemic lupus erythematosus.

Arthritis Care Res (Hoboken).

doi:10.1002/acr.23263

79 Myasoedova E, Crowson CS,

Kremers HM, Therneau TM,

Gabriel SEI. The incidence of

rheumatoid arthritis rising?: results

from Olmsted County, Minnesota,

1955–2007. Arthritis Rheum 2010; 62:

1576–82.

80 Goemaere S, Ackerman C, Goethals K,

De Keyser F, Van der Straeten C,

Verbruggen G et al. Onset of symptoms

of rheumatoid arthritis in relation to

age, sex and menopausal transition.

J Rheumatol 1990; 17: 1620–2.

81 de Man YA, Dolhain RJEM, van de

Geijn FE, Willemsen SP, Hazes JMW.

Disease activity of rheumatoid arthritis

during pregnancy: results from a

nationwide prospective study. Arthritis

Rheum 2008; 59: 1241–8.

82 Icardi A, Araghi P, Ciabattoni M,

Romano U, Lazzarini P, Bianchi G.

Kidney involvement in rheumatoid

arthritis. Reumatismo 2003; 55:

76–85.

83 Anders H-J, Vielhauer V. Renal co-

morbidity in patients with rheumatic

diseases. Arthritis Res Ther 2011;

13: 222.

84 Chiu H-Y, Huang H-L, Li C-H, Chen H-

A, Yeh C-L, Chiu S-H et al. Increased

risk of chronic kidney disease in

rheumatoid arthritis associated with

cardiovascular complications – a

National Population-Based Cohort

Study. PLos One 2015; 10: e0136508.

85 Vinet É, Bernatsky S, Hudson M,

Pineau CA, Baron M. Effect of

menopause on the modified Rodnan

skin score in systemic sclerosis.

Arthritis Res Ther 2014; 16: R130.

86 Sammaritano LR. Menopause in

patients with autoimmune diseases.

Autoimmun Rev 2012; 11: A430–6.

87 Penn H, Denton CP. Diagnosis,

management and prevention of

scleroderma renal disease. Curr Opin

Rheumatol 2008; 20: 692–6.

88 Anders HJ, Wiebecke B, Haedecke C,

Sanden S, Combe C, Schlöndorff D.

MPO-ANCA-positive crescentic

glomerulonephritis: a distinct entity of

scleroderma renal disease? Am J Kidney

Dis 1999; 33: e3.

89 Zakharova EV, Makarova TA,

Stolyarevich ES. ANCA-associated

vasculitis in patient with CREST-

syndrome – case report [Internet].

[cited 2017 Oct 19]. Available from

URL: https://www.peertechz.com/

Clinical-Nephrology/ACN-2-115.php

90 Eckardt K-U, Coresh J, Devuyst O,

Johnson RJ, Köttgen A, Levey AS et al.

Evolving importance of kidney disease:

from subspecialty to global health

burden. Lancet 2013; 382: 158–69.

91 Saran R, Robinson B, Abbott KC,

Agodoa LYC, Albertus P, Ayanian J

et al. US renal data system 2016

annual data report: epidemiology of

kidney disease in the United States.

Am J Kidney Dis 2017; 69(3 Suppl

1): A7–8.

92 Liyanage T, Ninomiya T, Jha V, Neal B,

Patrice HM, Okpechi I et al. Worldwide

access to treatment for end-stage

kidney disease: a systematic review.

Lancet 2015; 385: 1975–82.

93 Ojo A. Addressing the global burden of

chronic kidney disease through clinical

and translational research. Trans Am

Clin Climatol Assoc 2014; 125: 229–43

discussion 243-246.

94 World Health Organization.

Addressing gender within primary

health care reforms. In: World Health

Organization editor. Gender, Women

and Primary Health Care Renewal: A

Discussion Paper [Internet] [cited 2017

Oct 19]. Available from URL: http://

apps.who.int/iris/bitstream/10665/

44430/1/9789241564038_eng.pdf

95 Eguavoen ANT, Odiagbe SO,

Obetoh GI. The status of women, sex

preference, decision-making and

fertility control in Ekpoma Community

of Nigeria. J Soc Sci 2007; 15: 43–9.

96 Halle MP, Takongue C, Kengne AP,

Kaze FF, Ngu KB. Epidemiological

profile of patients with end stage renal

disease in a referral hospital in

Cameroon. BMC Nephrol 2015; 16: 59.

97 Ajayi S, Raji Y, Bello T, Jinadu L,

Salako B. Unaffordability of renal

replacement therapy in Nigeria. Hong

Kong J Nephrol 2016; 18(Suppl

C): 15–9.

98 Kausz AT, Obrador GT, Arora P,

Ruthazer R, Levey AS, Pereira BJ. Late

initiation of dialysis among women

and ethnic minorities in the United

States. J Am Soc Nephrol 2000; 11:

2351–7.

99 Coresh J, Byrd-Holt D, Astor BC,

Briggs JP, Eggers PW, Lacher DA et al.

Chronic kidney disease awareness,

prevalence, and trends among

U.S. adults, 1999 to 2000. J Am Soc

Nephrol 2005; 16: 180–8.

Editorial


122

info:doi/10.1002/acr.23263

https://www.peertechz.com/Clinical-Nephrology/ACN-2-115.php

https://www.peertechz.com/Clinical-Nephrology/ACN-2-115.php

http://apps.who.int/iris/bitstream/10665/44430/1/9789241564038_eng.pdf



100 Adams SV, Rivara M, Streja E,

Cheung AK, Arah OA, Kalantar-

Zadeh K et al. Sex differences in

hospitalizations with maintenance

hemodialysis. J Am Soc Nephrol 2017;

28: 2721–8.

101 Ethier J, Mendelssohn DC, Elder SJ,

Hasegawa T, Akizawa T, Akiba T et al.

Vascular access use and outcomes: an

international perspective from the

dialysis outcomes and practice patterns

study. Nephrol Dial Transplant 2008;

23: 3219–26.

102 Depner TA. Prescribing hemodialysis:

the role of gender. Adv Ren Replace Ther

2003; 10: 71–7.

103 Sehgal AR. Outcomes of renal

replacement therapy among blacks

and women. Am J Kidney Dis 2000; 35:

S148–52.

104 Jindal RM, Ryan JJ, Sajjad I,

Murthy MH, Baines LS. Kidney

transplantation and gender disparity.

Am J Nephrol 2005; 25: 474–83.

105 Couchoud C, Bayat S, Villar E,

Jacquelinet C, Ecochard R, REIN

Registry. A new approach for

measuring gender disparity in access to

renal transplantation waiting lists.

Transplantation 2012; 94: 513–9.

106 Liu G, Li X, Liu T, Zhao X, Zhang S,

Wang J et al. Gender disparity of

living donor renal transplantation in

East China. Clin Transplant 2013; 27:

98–103.

107 Naghibi O, Naghibi M, Nazemian F.

Gender disparity in kidney

transplantation. Saudi J Kidney Dis

Transpl 2008; 19: 545–50.

108 Bal MM, Saikia B. Gender bias in renal

transplantation: are women alone

donating kidneys in India? Transplant

Proc 2007; 39: 2961–3.

109 Hogan J, Couchoud C, Bonthuis M,

Groothoff JW, Jager KJ, Schaefer F

et al. Gender disparities in access to

pediatric renal transplantation in

Europe: data from the ESPN/ERA-

EDTA registry. Am J Transplant 2016;

16: 2097–105.

110 Gillespie A, Hammer H, Kolenikov S,

Polychronopoulou A, Ouzienko V,

Obradovic Z et al. Sex differences and

attitudes toward living donor kidney

transplantation among urban black

patients on hemodialysis. Clin J Am Soc

Nephrol 2014; 9: 1764–72.

111 Salter ML, McAdams-Demarco MA,

Law A, Kamil RJ, Meoni LA, Jaar BG

et al. Age and sex disparities in

discussions about kidney

transplantation in adults undergoing

dialysis. J Am Geriatr Soc 2014; 62:

843–9.

112 Piccoli GB, Attini R, Cabiddu G,

Kooij I, Fassio F, Gerbino M et al.

Maternal-foetal outcomes in pregnant

women with glomerulonephritides.

Are all glomerulonephritides alike in

pregnancy? J Autoimmun 2017;

79: 91–8.

113 Seeger H, Salfeld P, Eisel R,

Wagner CA, Mohebbi N. Complicated

pregnancies in inherited distal renal

tubular acidosis: importance of acid-

base balance. J Nephrol 2017; 30:

455–60.

114 Yefet E, Tovbin D, Nachum Z.

Pregnancy outcomes in patients with

Alport syndrome. Arch Gynecol Obstet

2016; 293: 739–47.

Supporting Information

Additional supporting information may be found in the online version of this article at the publisher’s web-site:

Table S1. Adverse pregnancy outcomes in patients with chronic kidney disease (CKD) and in their offspring.

Editorial


123

doi:10.1111/imj.13519

REVIEW

Recording patient bodyweight in hospitals: are we doing wellenough?Kate M. Flentje , Colin L. Knight, Ingrid Stromfeldt, Anindita Chakrabarti and N. Deborah Friedman

Department of General Medicine, University Hospital Geelong, Geelong, Victoria, Australia

Key wordspatient weight, hospital admission, medicationprescribing.

CorrespondenceKate Flentje, Department of General Medicine,University Hospital Geelong, Corner Bellarineand Ryrie Street, Geelong, Vic. 3220, Australia.Email: [email protected]

Received 8 February 2017; accepted28 May 2017.

Abstract

Recording patient weight is a standard practice for all hospital admissions, with this

measurement influencing other daily practices that rely on the delivery of safe and

effective patient care. Patient weight is important in the areas of medication prescribing,

fluid balance and assessment of nutrition. In particular, prescribing narrow therapeutic

index medications may result in significant harm as a potential consequence of inaccu-

rate dosing. Despite its importance, it is evident that bodyweight measurements are

recorded in only 13.5–55% of hospital patients, in a variety of settings including the

emergency department, intensive care unit, medical and surgical wards. Barriers to

compliance of healthcare staff include additional workload, patient handling and avail-

ability of appropriate weighing equipment. Hospitals and patients would benefit from

enhancing compliance with the systematic weighing of patients, staff training and

removing barriers to performing this task.

Introduction

Recording accurate patient weight is a fundamental partof the initial patient assessment as it potentially influ-ences a variety of clinical tasks during hospital admission(Table 1). This includes accurate prescribing, fluid assess-ment, nutritional and obesity screening and safe patientlifting practices.1–7 Most research regarding weighing ofhospitalised patients is from the United States (US) andthe United Kingdom (UK)1,2,4–6,8–13 with only limiteddata from Australia.5,14 Data reveal that in a variety ofhospital settings, including the emergency department(ED), intensive care unit (ICU), medical and surgicalwards, staff compliance with best practice of weighingpatients ranges between 13.5 and 55%.2,5–8

Prescribing medications

Errors in medication dose administration have the poten-tial to lead to adverse outcomes.1,2,5–8,15 Doses are regu-larly calculated based on patient weight, particularly formedications with a narrow therapeutic index, in patientswith renal dysfunction and in the paediatric popula-tion.1,2,5,15,16 There is a large discrepancy in the practice

of weighing paediatric and adult patients, with the com-pliance of paediatric weight recording near 100%.7

Despite the importance, recording of adult bodyweightis universally done poorly. A multi-centre study in theUS revealed that only 65.7% of patients were weighedwithin the first 36 h of admission.17 Of those notweighed, 67% of patients were asked about their weight,and these estimations were more likely to vary from theactual weight by >5 lbs (2.27 kg). This inaccurate weightrecording could potentially lead to adverse outcomesrelated to medication prescribing.

Narrow therapeutic index medications

Patient weight is often required to guide initial dosing ofnarrow therapeutic index medication such as thrombo-lytic agents, anticoagulants and antibiotics such as theaminoglycosides and vancomycin. A study across threehospitals in London tested the prevalence of weightrecording in patients who were prescribed narrow thera-peutic index antibiotics. Of the 8.8% of patients pre-scribed these medications, 34% did not have a recordedweight.2 In another UK study, only 64.5% of patientswho had received gentamicin or low molecular weightheparin (LMWH) had their weight recorded.7 In anAustralian setting, patients receiving heparin, enoxaparinor gentamicin were weighed approximately 25% of the

Funding: None.Conflict of interest: None.


124

time.5 Incorrect dosing of antibiotics could lead to under-dosing with lack of efficacy, antibiotic treatment failure,long-term antimicrobial resistance development or over-dosing with toxic side-effects.18

In a UK study of stroke patients administered alte-plase, 19.7% had been administered a dose outside theacceptable range due to the use of patient weight esti-mation for dosing rather than objective weight mea-surement.1 A study in an emergency department in theUS found that among the 958 patients whose weightswere estimated by hospital staff but not measured,55.5% were found to be outside the thrombolytic toler-ance range, with 34.8% dosed too low, and 21% dosedtoo high.9 In a study of cardiac patients in Scotland pre-scribed LMWH, weights were either not recorded orestimated incorrectly by staff or patients, leading to dos-ing errors in 49% of cases. Forty-three per cent of doseswere found to be excessive after the drug had alreadybeen administered.6

A clinically significant consequence of inaccurateweight recording is bleeding caused by overdosing ofthrombolytic or anticoagulant medication.6 In anAustralian study, 43.7% of patients administered thera-peutic anticoagulation who were not weighed experi-enced haemorrhagic complications, compared to nohaemorrhagic complications in patients who wereweighed (P = 0.03).5 Among the affected group, 71.4%were classified as major haemorrhagic complications,where there was a haemoglobin drop of >20 points orintervention required. In a US study of patients receivingtherapeutic enoxaparin for non-ST elevation myocardialinfarction, 18.7% received an excessive dose (>10 mg)even after being weighed.15 An excessive dose was foundto be associated with significant bleeding (OR 1.43, 95%CI 1.18–1.75) and death (OR 1.35, 95% CI 1.03–1.77).This supports that excessive dosing from any cause canlead to complications including death.

Obesity risk

Obesity poses a significant risk to patient outcomes in anumber of areas. Individuals that are overweight are at

increased risk of disease, medical complications anddeath.3 The recognition of obesity allows clinicians todiscuss bodyweight with patients with a view to encour-aging weight reduction, lifestyle changes and betterhealth outcomes.The intra-operative and post-operative complication

risk in obese patients is higher than non-obese patients,suggesting that identification in the pre-operative settingmay be a prognostic factor.19,20 Following moderate ormajor non-cardiac surgery, obese patients are at signifi-cantly higher risk of myocardial infarction (P = 0.001),wound infection (P = 0.001), nerve injury (P = 0.03)and urinary tract infection (P = 0.004).19 Weighingpatients would provide an objective measurement toscreen for these at-risk patients and allow for appropriateperi-operative preparation. Identification of obesepatients is also important from a logistical perspective,because they may require specialised care in terms oftechnique, staffing and use of equipment such as bariat-ric beds and hoist transfers.

Nutritional assessment

Malnutrition is associated with multiple adverse patientoutcomes including depression of the immune system,impaired wound healing, muscle wasting, extended hos-pital stay, higher treatment costs and increased mortal-ity.21 Accordingly, malnutrition is an important riskfactor to identify in patients admitted to hospital.In the acute hospital setting, malnutrition is a highly

prevalent condition, with Australian and internationalstudies reporting rates of approximately 40% of admis-sions.21 There were similar findings in a study of residen-tial aged care facilities in Australia, which found 43.1%of residents to be moderately malnourished and 6.4%severely malnourished.22 Patient weight has been vali-dated as a part of many effective malnutrition screeningtools.23 There is a correlation between more consistentweighing practices and a reduction in the prevalence ofmalnutrition in hospitalised patients.24

Volume status and fluid management

Fluid management is important in critically ill patients,with numerous studies showing an association betweenpositive fluid balance and increased mortality.16,25,26 In spe-cific medical conditions such as cardiac, liver and kidneyfailure, strict fluid balance during hospital admission aids inthe prevention of clinical deterioration.16,25,26 The accurateand timely measurement of patient weights is essential inguiding fluid management in this select cohort wheremaintaining optimal volume status is critical.

Table 1 Reasons for weighing patients in hospital

• Dosing medications with a narrow therapeutic window• Dosing medications safely in the very young, elderly and morbidly

obese• Monitoring patients with heart, liver or renal failure• Managing patients on haemodialysis• Assessing peri-operative risk• Nutritional assessment and malnutrition prevention• Improving management of patients at extremes of weight ranges• Accurate fluid and volume assessment

Recording bodyweight in hospitals


125

Increasing weight in a haemodialysis patient can indi-cate fluid accumulation. A ‘dry weight’ reflects the low-est post-dialysis weight, and therefore the fluid balanceeach patient can tolerate without developing hypoten-sion.27 Adhering to tight control of ‘dry weight’ overtime in this cohort is thought to be associated with a sig-nificant reduction in the risk of cardiovascular mortal-ity.28 Other studies have found an association betweenchronic fluid overload and increased mortality risk inhaemodialysis patients.29,30

One method of fluid balance monitoring is throughinput/output balance charts, however these charts aretime consuming and the accuracy of information isdebatable with data often incomplete.31 Daily, and insome cases twice daily, measurement of patient weight isanother marker of volume status, as most changes inweight over a short period of time are associated withchange in body fluids.25 Use of daily weight measure-ments has been assessed in a variety of patient popula-tions to aid clinical decision-making.32

Weight estimation in place ofmeasurement

Several hospital-based studies across the UK and UShave investigated the efficacy of weight estimation tech-niques compared with objectively measured weights andfound that clinicians are generally poor at estimatingweight.1,6,9–11

Hospital staff inaccurately estimate weight approxi-mately 50% of the time, particularly in patients with lowand high body mass index.4,6,9,10 Staff overestimate theweight of lighter patients and underestimate the weightof heavier patients, particularly among female patients.11

In contrast, patients more accurately estimate their ownweight compared with hospital staff.4,6,11 In one US study,patients were nine times more likely than medical andnursing staff to estimate accurately their own weight.11 Inanother study in the UK, patients were 80% accurate attheir own weight estimation compared to 39% of staff(P < 0.001).6 These data indicate that where possible apatient’s weight should be measured and not estimated.

Barriers to weighing patients inhospital

Additional workload

Of the barriers preventing staff from regularly recordingpatient weights during hospital admissions, the burdenof additional workload, inadequate staffing and interrup-tion to workflow are the most common reasons given(Table 2).2,12,14,33 A trial of different weight scales used in

a Melbourne study drew complaints of extra workloadto busy nursing staff, and concerns that the weight onthe scales did not match personal visual estimates.14 Thetiming of patient weighing may also interfere with othernursing tasks, and is worth considering when imple-menting daily weighing into routine practice.33

Difficulty handling patients

Nurses in an Australian study ranked lifting and weigh-ing patients among their eight most stressful handlingtasks, and that cognitively impaired and aggressivepatients were the most difficult to weigh regularly.5

Other studies have found the burden of locating equip-ment too time consuming, and that asking the patientfor their own weight estimate is more convenient.4,6 Inother survey responses, staff have reported bed boundpatients and very unwell patients as barriers to measur-ing weight.6

It is reasonable to acknowledge that not every patientcan stand on scales and be easily weighed in every situa-tion, such as patients in the intensive care unit (ICU),the elderly and those in post-operative care. The use ofalternative weighing methods, such as bed scales andchair scales, may be used among patients who cannotindependently stand or follow direction safely. In a Mel-bourne ICU study, the most efficient and economicalmethod was using a bed weighing scale, where thepatient does not need to move from the supine/semi-recumbent position. This could be an alternate methodfor hospitals with patients too unwell to move frombed.14

Variations in weighing equipment

Variation in the accuracy of weighing equipment mayadversely confound clinical decision-making. A US studycomparing 223 different scales for accuracy used acrossmultiple healthcare centres found that scales calibratedwithin the previous year, scales used on carpeted floorsand scales with less wear and tear were more precise.13

Other studies have found faulty equipment to be a com-mon issue.2,4,5 A minimum standard for patient scale

Table 2 Barriers to weighing patients in hospital

• Burden of additional workload• Stress of manually handling patients• Lack of effective weighing equipment• Poor understanding of the importance of recording accurate weight• Confusion due to multiple areas in the patient’s medical records to

document weight• Perceived patient distress

Flentje et al.


126

accuracy in the UK has been implemented by theEuropean Parliament. This legislation imposes a require-ment in every hospital for a robust program of equip-ment testing and calibration, as well as basic stafftraining on how to weigh accurately and correctly differ-ent cohorts of patients.34

Inconsistent documentation

Recording of patient weights in multiple areas of docu-mentation is commonplace in hospital admissions.Although perhaps not a direct barrier to the act ofweighing patients, inconsistent documentation may cre-ate confusion among treating teams, leading to barriersin accurate weight recording and interpretation, andpoor communication during patient admissions.2,5 Ifweight assessment and documentation are to beimproved, hospital policies must be updated and stream-lined with timely staff education to enable consistentand coherent weight documentation.

Opinion divergence: perceived lack of utilityand patient distress

Some healthcare workers believe that weighing patientsis either unnecessary or should be omitted depending onthe state of the patient’s health. In an audit in 2009,nursing staff on a medical ward in the UK thought thatweighing was unnecessarily invasive for the patient.12

Others have proposed that weighing patients may be dis-tressing for the patient and would not change manage-ment. In a qualitative study in a UK hospice withterminally ill patients, staff did not want patients think-ing about their weight during palliative care, with 55%declaring a desire not to weigh patients.8 This indicatesthat in areas of health such as end of life care, recordingpatient weight may be considered futile or unimportant.However, 98% of the patients in this study felt that get-ting weighed was not upsetting, 89% wanted to know iftheir weight was changing and 84% of patients wanted

to be weighed at future hospital appointments. This maydemonstrate a divergence between perceived inconve-nience of weighing by staff and patients.

Limitations in existing data

Most of the available literature fails to identify direct cor-relation between actual adverse clinical outcomes as aresult of failure to weigh patients as the sole causalityfactor.2,5,7,10,11,13,34 There are no blinded, randomisedcontrol studies in this area, and available studies sufferfrom biases that commonly occur in retrospective andnon-blinded observational cohort studies.1,2,9,10,15,35 Forexample, patients may not have been weighed becausethey had more pre-existing co-morbidities confiningthem to bed, creating practical difficulties in facilitatingthe weighing process.6,8 Studies describing consequencesof excess thrombolytic administration such as bleedingwere underpowered and limited to single or few centresin one region, making it potentially difficult to generaliseresults to other populations.5

Conclusion

While there is limited direct evidence of harm from fail-ure to weigh patients, there is a greater overall benefit topatients if they are weighed when they access health-care. These benefits are most pronounced in the areas offluid balance, assessment of nutrition and medicationprescribing. Excessive dosing of anticoagulation has beenshown almost to double the risk of major bleeding anddeath.15 The choice of weighing equipment needs to beconsidered depending on patient case mix, as uniquepatient cohort characteristics likely require a modifiedapproach equipment to selection. Availability of thisequipment may help to remove the barriers to weighingpatients in hospital. Institutions would benefit from stafftraining to improve compliance with weighing patients,and regular equipment calibration to maintain consistentand accurate measurements.

References

1 Barrow T, Khan MS, Halse O,

Bentley P, Sharma P. Estimating weight

of patients with acute stroke when

dosing for thrombolysis. Stroke 2016; 47:

228–31.

2 Charani E, Gharbi M, Hickson M,

Othman S, Alfituri A, Frost G et al. Lack

of weight recording in patients being

administered narrow therapeutic index

antibiotics: a prospective cross-sectional

study. BMJ Open 2015; 5: e006092.

3 Hahler B. Morbid obesity: a nursing care

challenge. Dermatol Nurs 2002; 14:

249–52, 255–6.

4 Hendershot KM, Robinson L, Roland J,

Vaziri K, Rizzo AG, Fakhry SM.

Estimated height, weight, and body

mass index: implications for research

and patient safety. J Am Coll Surg 2006;

203: 887–93.

5 Hilmer SN, Rangiah C, Bajorek BV,

Shenfield GM. Failure to weigh patients

in hospital: a medication safety risk.

Intern Med J 2007; 37: 647–50.

6 Spicer K, Gibson P, Bloe C, Cross SJ,

Leslie SJ. Weight assessment in cardiac

patients: implications for prescription of

low molecular weight heparin. Postgrad

Med J 2009; 85: 124–7.

7 Willson JS, Thomas SJ, Mackenzie SC,

Armitage AJ, Sen S, Ratliff WD et al.

Recording inpatient weight:

implications for medicines

administration. Br J Nurs 2016; 25:

16–18, 20–1.

8 Watson M, Coulter S, McLoughlin C,

Kelt S, Wilkinson P, McPherson A et al.

Recording bodyweight in hospitals


127

Attitudes towards weight and weight

assessment in oncology patients: survey

of hospice staff and patients with

advanced cancer. Palliat Med 2010; 24:

623–9.

9 Kahn CA, Oman JA, Rudkin SE,

Anderson CL, Sultani D. Can ED staff

accurately estimate the weight of adult

patients? Am J Emerg Med 2007; 25:

307–12.

10 Bloomfield R, Steel E, MacLennan G,

Noble DW. Accuracy of weight and

height estimation in an intensive care

unit: implications for clinical practice

and research. Crit Care Med 2006; 34:

2153–7.

11 Hall WL 2nd, Larkin GL, Trujillo MJ,

Hinds JL, Delaney KA. Errors in weight

estimation in the emergency

department: comparing performance by

providers and patients. J Emerg Med

2004; 27: 219–24.

12 Lees L, Allen-Mills G. Auditing the

nursing standard for weighing patients

on an acute medical unit. Nurs Times

2009; 105: 12–13.

13 Stein RJ, Haddock CK, Poston WS,

Catanese D, Spertus JA. Precision

in weighing: a comparison of scales

found in physician offices, fitness

centers, and weight loss centers.

Public Health Rep 2005; 120:

266–70.

14 Freitag E, Edgecombe G, Baldwin I,

Cottier B, Heland M. Determination of

body weight and height measurement

for critically ill patients admitted to the

intensive care unit: a quality

improvement project. Aust Crit Care

2010; 23: 197–207.

15 LaPointe NM, Chen AY, Alexander KP,

Roe MT, Pollack CV Jr, Lytle BL et al.

Enoxaparin dosing and associated risk

of in-hospital bleeding and death in

patients with non ST-segment elevation

acute coronary syndromes. Arch Intern

Med 2007; 167: 1539–44.

16 Payen D, de Pont AC, Sakr Y, Spies C,

Reinhart K, Vincent JL. A positive fluid

balance is associated with a worse

outcome in patients with acute renal

failure. Crit Care 2008; 12: R74.

17 Jensen GL, Friedmann JM, Henry DK,

Skipper A, Beiler E, Porter C et al.

Noncompliance with body weight

measurement in tertiary care teaching

hospitals. JPEN J Parenter Enteral Nutr

2003; 27: 89–90.

18 Roberts JA, Kruger P, Paterson DL,

Lipman J. Antibiotic resistance – what’s

dosing got to do with it? Crit Care Med

2008; 36: 2433–40.

19 Bamgbade OA, Rutter TW, Nafiu OO,

Dorje P. Postoperative complications in

obese and nonobese patients. World J

Surg 2007; 31: 556–60; discussion 61.

20 Friedman ND, Sexton DJ, Connelly SM,

Kaye KS. Risk factors for surgical site

infection complicating laminectomy.

Infect Control Hosp Epidemiol 2007; 28:

1060–5.

21 Barker LA, Gout BS, Crowe TC.

Hospital malnutrition: prevalence,

identification and impact on patients

and the healthcare system. Int J Environ

Res Public Health 2011; 8: 514–27.

22 Gaskill D, Black LJ, Isenring EA,

Hassall S, Sanders F, Bauer JD.

Malnutrition prevalence and nutrition

issues in residential aged care facilities.

Australas J Ageing 2008; 27: 189–94.

23 Guigoz Y. The mini nutritional

assessment (MNA) review of the

literature – what does it tell us? J Nutr

Health Aging 2006; 10: 466–85;

discussion 85–7.

24 O’Flynn J, Peake H, Hickson M,

Foster D, Frost G. The prevalence of

malnutrition in hospitals can be

reduced: results from three consecutive

cross-sectional studies. Clin Nutr 2005;

24: 1078–88.

25 Bouchard J, Mehta RL. Fluid

accumulation and acute kidney injury:

consequence or cause. Curr Opin Crit

Care 2009; 15: 509–13.

26 Sakr Y, Vincent JL, Reinhart K,

Groeneveld J, Michalopoulos A,

Sprung CL et al. High tidal volume and

positive fluid balance are associated

with worse outcome in acute lung

injury. Chest 2005; 128: 3098–108.

27 Canaud B, Lertdumrongluk P. Probing

‘dry weight’ in haemodialysis patients:

‘back to the future’. Nephrol Dial

Transplant 2012; 27: 2140–3.

28 Charra B, Calemard E, Ruffet M,

Chazot C, Terrat JC, Vanel T et al.

Survival as an index of adequacy of

dialysis. Kidney Int 1992; 41: 1286–91.

29 Kalantar-Zadeh K, Regidor DL,

Kovesdy CP, Van Wyck D,

Bunnapradist S, Horwich TB et al. Fluid

retention is associated with

cardiovascular mortality in patients

undergoing long-term hemodialysis.

Circulation 2009; 119: 671–9.

30 Wizemann V, Wabel P, Chamney P,

Zaluska W, Moissl U, Rode C et al. The

mortality risk of overhydration in

haemodialysis patients. Nephrol Dial

Transplant 2009; 24: 1574–9.

31 Eastwood GM. Evaluating the reliability

of recorded fluid balance to

approximate body weight change in

patients undergoing cardiac surgery.

Heart Lung 2006; 35: 27–33.

32 Guenter PA, Moore K, Crosby LO,

Buzby GP, Mullen JL. Body weight

measurement of patients receiving

nutritional support. JPEN J Parenter

Enteral Nutr 1982; 6: 441–3.

33 Pan A. The significance of timing of

patient daily weights and the barriers.

Paper 85. Vol. Master’s Projects. San

Francisco, CA: Department of Nursing,

University of San Francisco; 2014; 1–22.

34 Clarkson DM. Patient weighing:

standardisation and measurement. Nurs

Stand 2012; 26: 33–7.

35 Determann RM, Wolthuis EK,

Spronk PE, Kuiper MA, Korevaar JC,

Vroom MB et al. Reliability of height

and weight estimates in patients

acutely admitted to intensive care units.

Crit Care Nurse 2007; 27: 48–55;

quiz 56.

Flentje et al.


128

doi:10.1111/imj.13699

CL INICAL PERSPECT IVES

Climate change: allergens and allergic diseasesConstance H. Katelaris 1 and Paul J. Beggs2

1Campbelltown Hospital, Clinical Immunology and Allergy, and 2Department of Environmental Sciences, Faculty of Science and Engineering, MacquarieUniversity, Sydney, New South Wales, Australia

Key wordspollen, spores, fungal, allergy, asthma, foodhypersensitivity.

CorrespondenceConstance H. Katelaris, Clinical Immunologyand Allergy Unit, Campbelltown Hospital,Sydney, NSW 2560, Australia.Email: [email protected]

Received 9 March 2017; accepted14 August 2017.

Abstract

Climate change has been described as the biggest global health threat of the 21st cen-

tury. The atmospheric concentrations of greenhouse gases, such as carbon dioxide,

methane and nitrous oxide, have increased significantly since the start of the Industrial

Era around 1750, with much of this increase occurring over just the last 50 years or

so. This is resulting in warming of the climate system as well as changes in precipitation

and weather and climate extremes. These changes in climate are having wide-ranging

impacts on the Earth’s physical, biological and human systems, including human

health. It is these impacts of climate change on human health that are the focus of this

paper, particularly the impacts on allergens and allergic diseases. Such impacts are par-

ticularly significant in many countries where the prevalence of such diseases is high

and/or increasing. There is now compelling evidence that rising air temperatures and

carbon dioxide concentrations are, in some plant species, resulting in increased pollen

production and allergenicity and advancement and lengthening of the pollen season.

Changes in extreme events, such as thunderstorms and tropical cyclones, will also have

impacts on allergic diseases, with, for example, the flooding associated with tropical

cyclones leading to proliferation of mould growth in damp homes. The article also con-

siders a range of responses to these health threats, including greenhouse gas mitigation,

and adaptation strategies, such as enhanced environmental monitoring and health sur-

veillance and adequate planning for the future medical workforce.

Introduction

The Lancet recently stated that ‘Climate change is the big-gest global health threat of the 21st century’,1 and withthis in mind, it is particularly significant that 2016 wasthe warmest year on record – warmer than 2015, whichwas much warmer than 2014, which was warmer thanall previous years in the modern global temperaturerecord. The year 2016 was, of course, significant formany reasons, but is also of particular significance to thisarticle; 21 and 22 November 2016 witnessed by far the

most severe thunderstorm asthma2 event ever recorded,overwhelming emergency services and hospitals in Mel-bourne, Australia, with an estimated 3365 more publichospital emergency department respiratory-related pre-sentations than expected and as many as nine deaths.3

This review explores the implications of climatechange for human health, particularly for allergens andallergic diseases, and how this might be managed in clin-ical practice. We first outline the nature of climatechange, particularly those aspects of it most relevant toallergens and allergic diseases. This is followed by a briefoverview of the impacts of climate change on humanhealth. We then describe the impacts of climate changeon allergens and allergic diseases, focusing on allergicrespiratory diseases and food allergy. We then outlineseveral responses to these health threats.

What is climate change?

While climate change is perhaps most commonly consid-ered to involve warming of the climate, and this is

Funding: None.Conflict of interest: None.Disclosure: C. H. Katelaris and P. J. Beggs are Chief Investigatorson National Health and Medical Research Council (NHMRC)Partnership Project APP1116107. Industry partners on thisproject are Australasian Society of Clinical Immunology andAllergy, Asthma Australia, Bureau of Meteorology,Commonwealth Scientific and Industrial Research Organisation(CSIRO), Stallergenes and MeteoSwiss. P. J. Beggs is a ChiefInvestigator on Australian Research Council (ARC) DiscoveryProject DP170101630.


129

indeed a fundamental aspect of it, climate change actu-ally involves what is referred to as the climate system, ahighly complex system consisting of the atmosphere, thehydrosphere (oceans, seas, rivers, fresh water lakes,underground water, etc.), the cryosphere (sea ice, snowcover, glaciers and ice sheets, frozen ground, etc.), theland surface and the biosphere (all living organisms).Human influence on the climate system is clear, andwarming of the climate system is unequivocal.4

The largest driver of climate change is the increase inthe atmospheric carbon dioxide (CO2) concentrationsince the start of the Industrial Era around 1750. Theannual mean atmospheric CO2 concentration in 2016was 404.21 ppm, measured at the Mauna Loa Observa-tory in Hawaii.5 This is an increase of well over 100 ppmfrom the pre-industrial concentration of just 280 ppm.The increase in atmospheric CO2 concentration from1750 to the present has not been gradual, with twothirds of the increase occurring over just the last 50 years(since 1967). CO2 is not the only greenhouse gas to haveincreased since 1750 due to human activity. The atmo-spheric concentrations of methane and nitrous oxidehave also increased significantly, by about 150 and 20%respectively.4

These increases in greenhouse gases have resulted inan uptake of energy by the climate system, which hasled to warming. The average temperature of the Earth’ssurface is about 14 �C and has increased by almost 1 �Cover the past 100 years or so. Importantly, the extent ofwarming varies over the surface of the Earth, with someplaces warming less than this global average and otherswarming more. Projections of future climate changeindicate that the global mean surface temperature willincrease by a further 0.3–4.8 �C by the end of this cen-tury relative to the period 1986–2005, depending on thegreenhouse gas emissions scenario. Again, these futuretemperature increases will vary spatially, with meanwarming over land being higher than over the ocean.4

Many other aspects of climate have also changed andwill continue to do so into the future. The changes inprecipitation have been more complex than those intemperature, in that some regions have observed anincrease in precipitation while some have observed adecrease in precipitation and yet others have experi-enced no change at all. Projections for the future indicatethat the contrast in precipitation between wet and dryregions and between wet and dry seasons will increase.4

Climate change also involves changes in extremeweather and climate events. Hot days and nights will bewarmer and/or more frequent, and heat waves will bemore frequent and/or longer lasting over most landareas.4 There will be increases in the frequency, intensityand/or amount of heavy precipitation in some regions.4

There is also evidence that thunderstorms and intensetropical cyclone activity may increase into the future.

The impacts of these many and varied changes in cli-mate have been observed for some time now. Glaciersare shrinking, Arctic sea ice is thinning, the sea level isrising and oceans are becoming more acidic as theyabsorb some of the excess CO2 from the atmosphere.Many plant and animal species have shifted their geo-graphic ranges, seasonal activities, migration patterns,abundances and species interactions in response to cli-mate change.6 Impacts are also occurring on what arereferred to as human systems, of which human health isone. The following section outlines these impacts.

Health impacts of climate change

The health impacts of global climate change will bewidespread and will have variable effects depending ongeographic region, socioeconomic status of a populationand pre-existing population vulnerabilities. There aremany consequences, mostly negative, for public health,and these have been summarised in global terms by theWorld Health Organisation (WHO) as threats to safedrinking water, adequate shelter, stable food source andclean air.7

Threats to public health result from the major conse-quences of climate change:

a increases in natural, weather-related disasters ormajor weather events that create havoc, with destruc-tion of housing, dislocation of populations and threatof spreading disease;

b changes in rainfall – in some areas excessive, leadingto flooding, and in others a striking lack, resulting inprolonged drought and crop failure;

c heat waves are associated with documented excess mor-tality.1 Increasing heat will also have important impactson air quality and pollutant levels, and these are particu-larly threatening to children, the elderly and those withcardiovascular and respiratory disorders;

d warmer temperatures and increased CO2 have impactson allergens and allergic diseases, such as asthma andallergic rhinosinusitis. These will be among the mostimportant climate change influences on human healthand are the focus of the remainder of this article.

Aeroallergens and allergic diseases

A comprehensive review of the impacts of climatechange on allergens and allergic diseases has been pub-lished recently.8 Through the effects of key climatechange factors, we expect significant changes in expo-sure patterns to plants, pollen and fungi. In addition,

Katelaris & Beggs


130

complex interactions between pollutants, dust stormmaterial, thunderstorm events and allergens may beexpected to impact respiratory health negatively as well.Respiratory allergic diseases, such as asthma and allergicrhinoconjunctivitis, have become very prevalent inmany parts of the world over the last few decades.Therefore, there is a very large population at risk of sig-nificant allergic disease with any increase in pollen expo-sure and allergenicity. Furthermore, pollen allergy is alsoimplicated in certain forms of food-allergic disorders, sochanges in pollen distribution and allergenicity may beexpected to impact these conditions as well.Clinical evidence of the importance of pollen expo-

sure in causation of allergic respiratory diseases comesfrom a variety of approaches, ranging from exploringthe patterns of sensitisation to various pollen allergensto examining correlations between the pollen count andasthma exacerbations, usually measured by hospitalattendances or admissions.9,10 Asthma exacerbations arean important parameter as they carry significant eco-nomic implications.

Pollen production and allergenicity

There is now compelling evidence that rising tempera-tures and CO2 levels impact plant and pollen production.For many plants, rising CO2 levels represent an increasein a vital resource, and they respond accordingly withincreased growth and reproduction and greater pollenyields.11,12 Singer et al.,13 have shown that with increas-ing CO2 levels, the major allergen from ragweed, Amb a1, increases, although there is no change in total pollenprotein level.Ragweed, a native of North America, has been invad-

ing large areas of South America and Europe for the lastfew decades. It is a major cause of respiratory allergy.Hamaoui-Laguel et al.,14 have used modelling frame-works that account for various factors under high-endand moderate climate and land-use change scenarios topredict airborne ragweed pollen concentrations inEurope in the future. Their modelling shows that, by2050, airborne ragweed pollen concentrations will beabout four times higher than they are now, almost cer-tainly increasing the incidence and prevalence of rag-weed allergy. Indeed, in subsequent work,15 it has beenfound that sensitisation to ragweed will more than dou-ble in Europe, from 33 to 77 million people, by2041–2060; that the greatest proportional increases willoccur where sensitisation is uncommon (e.g. Germany,Poland, France) and that higher pollen concentrationsand a longer pollen season may also increase the severityof symptoms.15 While ragweed is not an important aller-gen in Australia at present, it has been introduced and is

likely to spread and become a problem if effective eradi-cation measures are not in place.

Pollen seasons

Phenology is the study of the influence of climate onperiodic plant and animal life-cycle events. There aresome long-term phenological records in many Europeancountries, and these demonstrate measurable changesoccurring in recent decades. Flowering is particularlysensitive to temperature over the preceding month. Fit-ter and Fitter16 have demonstrated an average 4.5-dayadvancement in the first flowering date for nearly400 British plant species over the 1990s compared withthe previous four decades. Menzel17 has examined datafrom a Europe-wide network, the International Pheno-logical Gardens, and reports that the average annualgrowing season has lengthened by approximately 11 dayssince the 1960s.In countries where there are long-term aerobiological

survey data, it is possible to examine trends in airbornepollen and fungal spore counts. The trends are by nomeans uniform but vary with geographical location andplant type. Analysis of these data makes it possible notonly to track changes in the pollen season but also toexamine changing patterns in flowering phenology, oneof the most valuable indicators of climate changeimpact.18

Ariano et al.,19 had a unique opportunity to study vari-ations in pollen levels and allergic sensitisation in West-ern Liguria, Italy, because of the existence of almostthree decades of pollen monitoring and meteorologicalvariable data and skin test and clinical data from resi-dents in the region. They describe a progressive increasein the duration of the pollen season for Parietaria

(+85 days), olive (+18 days) and cypress (+18 days). Allpollen monitored, except for grasses, showed an increasein total counts. They report an increase in the percent-ages of patients sensitised to pollen over these years,whereas sensitisation rates to the house dust miteremained stable.Consistent with the findings of Ariano et al.,19 a 30-

year (1982–2011) olive pollen record from Spain hasshown a trend towards increasing pollen production,most likely caused by longer flowering periods as a resultof higher temperatures.20

Fungal exposure and allergy

Fungi are a large and diverse group of eukaryotic organ-isms that have complex metabolisms, secreting numer-ous enzymes into their surroundings. Many of these arewell-described allergens.21 Other chemicals include

Climate change and allergies


131

ergosterol, which can be used as a measure of fungal bio-mass; constituents of cell walls, such as β-glucans thathave been shown to cause respiratory symptoms, itchingand fatigue in a dose-dependent manner and myco-toxins, which are low molecular weight organic com-pounds important in agriculture (e.g. aflatoxin). The roleof the latter in producing human disease in domesticenvironments is far more contentious. Volatile fungalmetabolites are responsible for the musty smell associ-ated with fungal growth.22

Adverse health effects from fungal exposure can occurby a variety of mechanisms, including infection, allergy,irritation and toxicity, depending on the nature and doseof the exposure. Infection may be seen in normal andimmune-compromised patients. Fungal components,such as β-glucan, may produce effects through activationof the innate immune system or through T cell and othermechanisms. Allergic sensitisation to fungi is an impor-tant risk factor for allergic asthma, and fungal exposurehas been linked to asthma exacerbations and hospitalpresentations23 as well as a described association withasthma mortality.24 In addition to allergic rhinitis andasthma, fungal exposure has been linked to conditionslike allergic broncho-pulmonary aspergillosis, hypersen-sitivity pneumonitis, allergic fungal sinusitis and atopicdermatitis.

Increased flooding in many areas of the world is seenas a consequence of climate change. Flooding leading tolong-term dampness in residential dwellings promotesfungal growth. In the aftermath of Hurricane Katrina inNew Orleans, USA, high indoor and outdoor fungalcounts were noted.25 Increased moisture along withhigher temperatures and CO2 levels encourage fungalgrowth.

Food allergy

Food allergy has become a very significant public healthconcern as 4–8% of children and 3–4% of adults in Wes-ternised countries have a food allergy.26,27 Aeroallergens,and sensitisation to them, are important in some expres-sions of food allergy.

IgE-mediated food allergy to common foods, such ascow’s milk, egg, soy, nuts, wheat and seafood (known asClass 1 food allergens), may result from sensitisationthrough the gastrointestinal tract. A less well-recognisedform of food allergy may occur as a result of primarysensitisation to homologous pollen allergens through therespiratory tract, causing reactivity to cross-reactive foodallergens (Class II allergens).18 Global differences in foodsensitisation patterns have been particularly observed forthese plant food allergens, whereby differences in aller-genic plant distribution, agriculture and dietary patterns

determine the predominant pattern of pollen and foodallergy.18 For instance, in Europe, prevalence of plantfood allergy is significantly influenced by sensitisation toparticular proteins in birch pollen, such as Bet v 1 andBet v 2, while in the Mediterranean region, there is ahigher sensitisation rate to profilins and non-specificlipid transfer proteins.28 A recent study has shown differ-ences in the pattern of allergen reactivity causing peanutsensitisation across different geographic regions, andthese differences are largely determined by aeroallergenexposure.29 Thus, it is likely that changes in climate thatresult in altered distribution of various allergenic plantsmay, in time, bring about a change in the pattern of foodallergy, especially that caused by plant food allergens.18

Furthermore, there is now limited experimental evi-dence that increasing CO2 concentrations may directlyalter the allergenicity of some plant-derived foods, witha recent study showing increased allergen concentrationin peanut when grown under such conditions.30

Eosinophilic oesophagitis has become a commonlyrecognised condition in children with dysphagia, withantigen exposure being the driving force for the eosino-philic inflammation seen in this condition.31 While foodallergens are the obvious major allergens involved in thisprocess, there are data to suggest that aeroallergen expo-sure is capable of triggering eosinophilic inflammationeither because of the swallowed fraction or because ofingestion of foods cross-reacting with pollen allergens.The potential role of aeroallergens in provoking paediat-ric eosinophilic oesophagitis has recently been studied byFahey et al.32

Response to climate change healththreats

For years, a reduction in greenhouse gases has been cen-tral to our approach to managing the effects of climatechange, and for the sake of planet and human health,these efforts must continue and grow much stronger.Immediate and sustained reduction of all pollutants,coal-burning technologies and transport-related pollu-tion must be enabled. However, there are many indica-tors that the rising greenhouse gases have alreadybrought about significant changes that must beaddressed and managed. This realisation has shifted thefocus from purely attempting to reduce greenhousegases to one of managing the many aspects of healthimpacts caused by the changing climate. The WHO hasmade adaptation a critical component of the UnitedNations Framework Convention on Climate Change(UNFCCC), with particular emphasis on planning strate-gies for the developing world where the impact on popu-lation health will be the greatest.33

Katelaris & Beggs


132

Central to endeavours for managing climate changeand associated health impacts is the establishment ofprecise, ongoing measurements of all those parametersdemonstrated to be important for their impacts onhuman health. For instance, in the case of aeroallergenimpact on respiratory health, the studies on influences ofclimate change on plant growth and distribution, andpollen production, have come predominantly from theNorthern Hemisphere. In Australia, until recently, therehas been no systematic aerobiological monitoring. This isbeginning to be addressed, with the establishment of anational pollen monitoring service within a partnershipknown as the AusPollen project.34 Long-term, longitudi-nal studies to map allergenic pollen and fungi distribu-tion are necessary to help understand the patterns ofrespiratory allergy and to plan for times of peak expo-sures and likely hospital presentations. The significanceand importance of this have been noted very recently inthe context of the November 2016 Melbourne thunder-storm asthma event.35

This paper has highlighted the important role of aero-allergens in driving inflammation in many allergic condi-tions, so the management of aeroallergen sensitisationand allergy will be an important aspect of planning spe-cific mitigation strategies given the exposure increaseswe can expect with climate change. Although allergicconditions are some of the most common afflictions inmedical practice, highly trained specialists in this fieldare few in number. Endeavours to upskill the generalmedical workforce in the recognition and managementof allergic disorders will be an important component inany mitigation strategy. Immunotherapy, as a treatmentstrategy for those expressing the clinical consequences ofinhalant allergy, has been in use for a century. Improve-ments in allergen characterisation, methods of delivery

and length of treatment programmes will enhance itsutility in addressing some of the morbidity produced byaeroallergen exposure.

Conclusion

The medical community has engaged in understanding

and managing risks within the health arena for many

decades. As such, all members of the medical community

have roles to play in managing responses to climate

change, from data collection and mitigation to adapta-

tion. We need to be advocates for rapid and sustained

reductions in greenhouse gas emissions, leading by

example in institutions like hospitals. Alongside these

efforts must be planning and adoption of strategies for

adaptation to the inevitable climate change factors that

so powerfully impact many facets of health.Adaptation strategy must incorporate improved and

effective monitoring of the many variables and conse-quences associated with climate change factors. Thisincludes accurate measures of particulates and pollut-ants; precise records of infection transmission and vectorpopulations and accurate, long-term aerobiological mon-itoring to map changing patterns of pollen and mouldspore distribution.Another vital aspect is appropriate forward planning

for workforce diversity that will be required to managespecific challenges, some of which have been outlined inthis paper. Finally, it is imperative that the more affluentcommunities assist those in developing countries toachieve these same goals as they are likely to bear theconsequences of climate change to an even greaterextent than those in affluent communities.

References

1 Costello A, Abbas M, Allen A, Ball S,

Bell S, Bellamy R et al. Managing the

health effects of climate change: Lancet

and University College London Institute

for Global Health Commission. Lancet

2009; 373: 1693–733.

2 Dabrera G, Murray V, Emberlin J,

Ayres JG, Collier C, Clewlow Y et al.

Thunderstorm asthma: an overview of

the evidence base and implications for

public health advice. QJM 2013; 106:

207–17.

3 Department of Health and Human

Services. The November 2016 Victorian

epidemic thunderstorm asthma event:

an assessment of the health impacts.

The Chief Health Officer’s Report,

27 April 2017. Melbourne: Victorian

Government; 2017.

4 IPCC. In: Stocker TF, Qin D, Plattner G-

K, MMB T, Allen SK, Boschung J, et al.,

eds. Climate Change 2013: The Physical

Science Basis. Contribution of Working

Group I to the Fifth Assessment Report of the

Intergovernmental Panel on Climate

Change. Cambridge, UK and New York:

Cambridge University Press; 2013.

5 Tans P, Keeling R. Trends in

Atmospheric Carbon Dioxide. US

Department of Commerce, National

Oceanic and Atmospheric

Administration (NOAA), Earth System

Research Laboratory (ESRL), Global

Monitoring Division, and Scripps

Institution of Oceanography. [updated

2017 Mar 6; cited 2017 Mar 7].

Available from URL: www.esrl.noaa.

gov/gmd/ccgg/trends/

6 IPCC. In: CWT, Pachauri RK and

Meyer LA, eds. Climate Change 2014:

Synthesis Report. Contribution of Working

Groups I, II and III to the Fifth Assessment

Report of the Intergovernmental Panel on

Climate Change. Geneva: IPCC; 2014.

7 Luber G, Knowlton K, Balbus J,

Frumkin H, Hayden M, Hess J et al.

Human health. In: Melillo JM,

Richmond TC and Yohe GW, eds.

Climate Change Impacts in the United

States: The Third National Climate

Assessment. Washington: U.S. Global

Change Research Program; 2014;

220–56.

8 Beggs PJ, ed. Impacts of Climate Change on

Allergens and Allergic Diseases.

Climate change and allergies


133

http://www.esrl.noaa.gov/gmd/ccgg/trends/

http://www.esrl.noaa.gov/gmd/ccgg/trends/

Cambridge: Cambridge University Press;

2016.

9 Erbas B, Chang J-H, Dharmage S,

Ong EK, Hyndman R, Newbigin E et al.

Do levels of airborne grass pollen

influence asthma hospital admissions?

Clin Exp Allergy 2007; 37: 1641–7.

10 Erbas B, Akram M, Dharmage SC,

Tham R, Dennekamp M, Newbigin E

et al. The role of seasonal grass pollen

on childhood asthma emergency

department presentations. Clin Exp

Allergy 2012; 42: 799–805.

11 Ziska LH. Rising atmospheric carbon

dioxide and plant biology: the

overlooked paradigm. DNA Cell Biol

2008; 27: 165–72.

12 Ziska LH, Bunce JA. Predicting the

impact of changing CO2 on crop yields:

some thoughts on food. New Phytol

2007; 175: 607–18.

13 Singer BD, Ziska LH, Frenz DA,

Gebhard DE, Straka JG. Increasing Amb

a 1 content in common ragweed

(Ambrosia artemisiifolia) pollen as a

function of rising atmospheric CO2

concentration. Funct Plant Biol 2005; 32:

667–70.

14 Hamaoui-Laguel L, Vautard R, Liu L,

Solmon F, Viovy N, Khvorostyanov D

et al. Effects of climate change and seed

dispersal on airborne ragweed pollen

loads in Europe. Nat Clim Chang 2015; 5:

766–71.

15 Lake IR, Jones NR, Agnew M,

Goodess CM, Giorgi F, Hamaoui-

Laguel L et al. Climate change and

future pollen allergy in Europe. Environ

Health Perspect 2017; 125: 385–91.

16 Fitter AH, Fitter RSR. Rapid changes in

flowering time in British plants. Science

2002; 296: 1689–91.

17 Menzel A. Trends in phenological

phases in Europe between 1951 and

1996. Int J Biometeorol 2000; 44: 76–81.

18 Katelaris CH. Impacts of climate change

on allergic diseases. In: Beggs PJ,

ed. Impacts of Climate Change on Allergens

and Allergic Diseases. Cambridge:

Cambridge University Press; 2016;

157–78.

19 Ariano R, Canonica GW, Passalacqua G.

Possible role of climate changes in

variations in pollen seasons and allergic

sensitizations during 27 years. Ann

Allergy Asthma Immunol 2010; 104:

215–22.

20 García-Mozo H, Yaezel L, Oteros J,

Galán C. Statistical approach to the

analysis of olive long-term pollen

season trends in southern Spain. Sci

Total Environ 2014; 473–474: 103–9.

21 Radauer C, Bublin M, Wagner S,

Mari A, Breiteneder H. Allergens are

distributed into few protein families and

possess a restricted number of

biochemical functions. J Allergy Clin

Immunol 2008; 121: 847–52.

22 Bräse S, Encinas A, Keck J, Nising CF.

Chemistry and biology of mycotoxins

and related fungal metabolites. Chem

Rev 2009; 109: 3903–90.

23 Tham R, Dharmage SC, Taylor PE,

Katelaris CH, Vicendese D,

Abramson MJ et al. Outdoor fungi and

child asthma health service attendances.

Pediatr Allergy Immunol 2014; 25:

439–49.

24 Targonski PV, Persky VW,

Ramekrishnan V. Effect of

environmental molds on risk of death

from asthma during the pollen season.

J Allergy Clin Immunol 1995;

95: 955–61.

25 Barbeau DN, Grimsley LF, White LE,

El-Dahr JM, Lichtveld M. Mold

exposure and health effects following

Hurricanes Katrina and Rita. Annu Rev

Public Health 2010; 31: 165–78.

26 Burks AW. The changing field of food

allergy. J Allergy Clin Immunol Pract

2015; 3: 39–41.

27 Koplin JJ, Osborne NJ, Wake M,

Martin PE, Gurrin LC, Robinson MN

et al. Can early introduction of egg

prevent egg allergy in infants? A

population-based study. J Allergy Clin

Immunol 2010; 126: 807–13.

28 Burney PGJ, Potts J, Kummeling I,

Mills ENC, Clausen M, Dubakiene R

et al. The prevalence and distribution of

food sensitization in European adults.

Allergy 2014; 69: 365–71.

29 Vereda A, van Hage M, Ahlstedt S,

Ibañez MD, Cuesta-Herranz J, van

Odijk J et al. Peanut allergy: clinical and

immunologic differences among

patients from 3 different geographic

regions. J Allergy Clin Immunol 2011;

127: 603–7.

30 Ziska LH, Yang J, Tomecek MB,

Beggs PJ. Cultivar-specific changes in

peanut yield, biomass, and allergenicity

in response to elevated atmospheric

carbon dioxide concentration. Crop Sci

2016; 56: 2766–74.

31 Atkins D. Aeroallergens in eosinophilic

esophagitis: significant triggers or noise

in the system? J Pediatr Gastroenterol

Nutr 2017; 64: 1–2.

32 Fahey L, Robinson G, Weinberger K,

Giambrone AE, Solomon AB.

Correlation between aeroallergen

levels and new diagnosis of

eosinophilic esophagitis in New York

City. J Pediatr Gastroenterol Nutr 2017;

64: 22–5.

33 Ebi KL, Prats EV. Health in national

climate change adaptation planning.

Ann Glob Health 2015; 81: 418–26.

34 AusPollen: The Australian Pollen

Allergen Partnership: Towards a

Standardized National Pollen Count

Network [homepage on the Internet].

[cited 2017 Mar 6]. Available from

URL: http://pollenforecast.com.au/

index.php

35 Inspector-General for Emergency

Management. Review of Response to

the Thunderstorm Asthma Event of

21–22 November 2016: Final Report.

Melbourne: Inspector-General for

Emergency Management, Victoria State

Government; 2017.

Katelaris & Beggs


134

http://pollenforecast.com.au/index.php

http://pollenforecast.com.au/index.php

doi:10.1111/imj.13654

ORIG INAL ARTICLES

EVOLVE: The Australian Rheumatology Association’s ‘top five’list of investigations and interventions doctors and patientsshould questionKathleen Morrisroe,1,2 Ayano Nakayama,3,4 Jason Soon,5,6 Mark Arnold,7 Les Barnsley,8 Claire Barrett,9

Peter M. Brooks ,10 Stephen Hall,11 Patrick Hanrahan,12 Pravin Hissaria,13 Graeme Jones,14

Veera S. Katikireddi,15 Helen Keen,12,16 Rodger Laurent,17 Mandana Nikpour,1,2 Katherine Poulsen,15

Philip Robinson,18 Muriel Soden,19,20 Nigel Wood,21 Nicola Cook,22 Catherine Hill23 and Rachelle Buchbinder 24,25

1Department of Rheumatology, St Vincent’s Hospital, 5Royal Australasian College of Physicians, 6Menzies Centre for Health Policy, Sydney School ofPublic Health, and 7School of Rural Health, University of Sydney, 8Department of Rheumatology, Concord Repatriation General Hospital, and17Department of Rheumatology, Royal North Shore Hospital, Sydney, New South Wales, 2Department of Medicine, and 10Centre for Health Policy,School of Population and Global Health, The University of Melbourne, 11Department of Medicine, Monash University and Cabrini Health, 24MonashDepartment of Clinical Epidemiology, Cabrini Institute, 25Department of Epidemiology and Preventive Medicine, School of Public Health and PreventiveMedicine, Monash University, Melbourne, and 21Department of Rheumatology, University Hospital Geelong, Geelong, Victoria, 3RheumatologyDepartment, Canberra Hospital, and 4College of Medicine, Biology and Environment, Australian National University, Canberra, Australian CapitalTerritory, 9Department of Rheumatology, Redcliffe Hospital, 15Department of Medicine, The Prince Charles Hospital, 18School of Medicine, RoyalBrisbane Hospital, University of Queensland, Brisbane, 19Department of Rheumatology, The Townsville Hospital, and 20College of Medicine andDentistry, James Cook University, Townsville, Queensland, 12Department of Medicine, University of Western Australia, 16Department of Rheumatology,Royal Perth Hospital and Fiona Stanley Hospital, and 22Department of Rheumatology, Royal Perth Hospital, Perth, Western Australia, 13Department ofImmunology, SA Pathology, Clinical Immunology Unit, Royal Adelaide Hospital and University of Adelaide, and 23Rheumatology Unit, The QueenElizabeth and Royal Adelaide Hospital and University of Adelaide, Adelaide, South Australia, and 14Menzies Institute for Medical Research, University ofTasmania, Hobart, Tasmania, Australia

Key wordsEVOLVE, evidence-based practice,implementation, low-value care, rheumatology.

CorrespondenceRachelle Buchbinder, Monash DepartmentClinical Epidemiology, Suite 41 Cabrini MedicalCentre, 183 Wattletree Road, Malvern, Vic.3144, Australia.Email: [email protected]

Received 18 September 2017; accepted25 October 2017.

Abstract

Background: The EVOLVE (evaluating evidence, enhancing efficiencies) initiative

aims to drive safer, higher-quality patient care through identifying and reducing low-

value practices.

Aims: To determine the Australian Rheumatology Association’s (ARA) ‘top five’ list of

low-value practices.

Methods: A working group comprising 19 rheumatologists and three trainees compiled

a preliminary list. Items were retained if there was strong evidence of low value and

there was high or increasing clinical use and/or increasing cost. All ARA members

(356 rheumatologists and 72 trainees) were invited to indicate their ‘top five’ list from a

list of 12-items through SurveyMonkey in December 2015 (reminder February 2016).

Results: A total of 179 rheumatologists (50.3%) and 19 trainees (26.4%) responded.

The top five list (percentage of rheumatologists, including item in their top five list) was:

Do not perform arthroscopy with lavage and/or debridement for symptomatic osteoar-

thritis of the knee nor partial meniscectomy for a degenerate meniscal tear (73.2%); Do

not order anti-nuclear antibody (ANA) testing without symptoms and/or signs suggestive

of a systemic rheumatic disease (56.4%); Do not undertake imaging for low back pain

for patients without indications of an underlying serious condition (50.8%); Do not use

ultrasound guidance to perform injections into the subacromial space as it provides no

additional benefit in comparison to landmark-guided injection (50.3%) and Do not order

anti-double-stranded DNA antibodies in ANA negative patients unless the clinical suspi-

cion of systemic lupus erythematosus remains high (45.3%).

Conclusions: This list is intended to increase awareness among rheumatologists, other cli-

nicians and patients about commonly used low-value practices that should be questioned.

Funding: R. Buchbinder is supported by an NHMRC Senior Principal Research Fellowship.Conflict of interest: None.


135

Introduction

The cost of healthcare in Australia is growing faster thanpopulation growth. For example, there was a near dou-bling of health expenditure over the decade 2001–2002to 2011–2012.1 This has placed an increased focus onhealthcare quality, affordability and value. The RoyalAustralasian College of Physicians’ (RACP) EVOLVE(evaluating evidence, enhancing efficiencies) initiative isa clinician-led partnership between the College and itsspecialty societies. It aims to drive safer, higher-qualitypatient care through identifying and reducing low-valuemedical care, defined as tests, treatments or proceduresthat are overused, inappropriate or of limited effective-ness and/or potentially harmful.2

Modelled on Choosing Wisely initiatives in the UnitedStates and other countries,3 and working in conjunctionwith Choosing Wisely Australia,4 specialist physiciansfrom over 20 medical specialities have completed or aredeveloping their EVOLVE ‘top five’ lists of low-valueclinical practices.2 The guiding principles of EVOLVE arethat the ‘top five’ list should be within or significantlyimpact the specialists’ domain of practice with the poten-tial to make a real impact in reducing low-value care;the practices should be either growing in use or currentlycommonly used and use of the Delphi consensusmethod,5 as the overarching methodology for identifyinga ‘top-five’ list.

In this article, we present the Australian Rheumatol-ogy Association’s (ARA) ‘top five’ list of low-valuepractices.

Methods

The EVOLVE ARA working group comprised 19 rheuma-tologists and 3 advanced rheumatology trainees formedafter a call for interested ARA members. At a face-to-facemeeting in 2015, the guiding EVOLVE principles werediscussed and it was agreed that items should beincluded if they were either primarily a rheumatologistissue or an issue that rheumatologists should advocatefor on behalf of their patients.

A preliminary list of low-value clinical practices wascreated based upon the working group’s clinical experi-ences, as well as consideration of potentially relevantitems identified from lists generated by others.6–10 Theworking group reduced the initial list to 12 items, notingthat some items included multiple components. Twoitems were excluded (Do not prescribe bisphosphonatesfor patients at low risk of fracture and Do not performwhole body bone scans for diagnostic screening forperipheral and axial arthritis in the adults), as these werenot considered relevant to the Australian context.

Small teams for each topic were formed to review theevidence and determine if the preliminary list of low-value practices met all of the following criteria:

1 Strong evidence of low-value clinical practice from aliterature review and2 Evidence of high or increasing clinical use and signifi-cant and/or increasing cost to the Australian communitybased upon publicly available Medicare Benefits Sched-ule (MBS) item usage and cost data relating to eachstatement from 2004 to 2015.11

Medicare Statistics provides data for MBS item num-

bers divided by the number of Medicare participants

enrolled at the end of each month. For this project, usage

data are expressed as number of services per financial

year and costs are expressed as total benefits paid out for

these services by financial year. The number of services

and costs included in the Medicare Statistics data only

relate to services that are performed by a registered pro-

vider, qualify for the Medicare benefit and for which a

claim has been processed by Medicare Australia. They do

not include services provided by hospital doctors to public

patients in public hospitals or services that qualify for a

benefit under the Department of Veterans’ Affairs

National Treatment Account. Another important caveat

of MBS data is that some single items can be used for

multiple indications and the specific indication for which

that item is used is not collected. For example, while there

are MBS item numbers for ultrasound-guided injections,

these do not differentiate between ultrasound-guided

injections for different body parts. In most, but not all

instances we excluded item numbers for diagnostic imag-

ing if the site being imaged was not specified.One item, ‘Do not order an HLA-B27 unless spondy-

loarthritis is suspected based on specific signs or symp-toms’, was removed from the list after the reviewrevealed it did not fulfil the criteria of high or increasingusage or high cost in Australia. Following review of theevidence, a new item was included: ‘Do not order anti-neutrophil cytoplasmic antibodies (ANCA) testing fordiagnosis of vasculitis unless one of the consensus guide-line indications is present’. We retained two items, ‘Donot use ultrasound guidance to perform injections intothe subacromial space (or trochanteric bursa), as itprovides no additional benefit in comparison tolandmark-guided injection’, because even though it wasnot possible to extract the exact number and cost ofthese subsidised ultrasound-guided injections, consensusamong the working group was that a large and increas-ing number of ultrasound-guided injections is being per-formed (inappropriately) into these sites.

Morrisroe et al.


136

The working group refined the ‘do-not-do’ statementsand wrote brief summaries of the evidence in support ofit being a low-value clinical practice using the NationalHealth and Medical Research Council (NHMRC) recom-mendations for summarising the level of evidence,strength of recommendation and quality.12

An anonymous survey was created in SurveyMon-key.13 All ordinary (356 rheumatologists) and associate(72 rheumatology trainees) ARA members were invitedto participate through email on 10 December 2015 with areminder sent 17 February 2016. The ARA Boardapproved the survey and ethical approval was not sought.Respondents were provided with the survey purpose

and background information about EVOLVE, presentedwith the 12 proposed recommendations for not under-taking a particular test, treatment or procedure and asummary of the evidence for each recommendation.They were asked to select the five recommendationsfor which they considered the evidence to be the stron-gest. They could also provide comments for any of thestatements in free text. Finally, they were asked to pro-vide some demographic and clinical practice details:gender, setting in which the majority of hours areworked (public, private, academic, retired and other),fellowship status (fellow for <10, 10–20, 21–30 or>30 years) and practice location (urban/metropolitan,large rural centre, small rural centre, remote). For thepurposes of our ‘top five’ list we excluded traineeresponses.

Results

Respondents included 179 rheumatologists (50.3%response rate) and 19 trainees (26.4% response rate).The majority of rheumatologists were male (n = 115,64.3%, 4 missing responses) and just over half workedprimarily in private practice (n = 95, 53.1%, 5 missingresponses).Table 1 presents the proportion of rheumatologists

who put each of the 12 statements into their ‘topfive’ list in order of ranking. Endorsement of individ-ual statements ranged from 20.7 to 73.2% of respon-dents. The highest endorsement was for notperforming arthroscopic treatments for knee osteoar-thritis and/or degenerative meniscal tears (73.2%),while over half endorsed not performing anti-nuclearantibody (ANA) testing for patients without rheu-matic symptoms (56.4%), imaging for low back painin those without specific indications (50.8%) andultrasound guidance for shoulder injections (50.3%).Nearly all of the comments indicated that respon-dents would have liked to endorse more than fivestatements. Trainee responses were similar with fourof the same recommendations chosen in the top fivealthough there was even stronger endorsement fornot performing arthroscopic treatments for knee oste-oarthritis and/or degenerative meniscal tears (84.2%)and not performing ANA testing for patients withoutrheumatic symptoms (73.7%).

Table 1 Support for inclusion of each recommendation to be in the top five list by all rheumatologists and by gender and public private practice andby trainees†

Recommendations Rheumatologists(n = 179) (%)

Male(n = 115) (%)

Female(n = 60) (%)

Public(n = 78) (%)

Private(n = 95) (%)

Trainees(n = 19) (%)

Arthroscopy for symptomatic knee osteoarthritis ordegenerate meniscal tear

73.2 80 63.3 66.7 82.1 84.2

ANA testing without rheumatic symptoms 56.4 56.5 53.3 57.7 52.6 73.7Imaging for low back pain without red flags 50.8 47.8 60.0 64.1 43.2 42.1Ultrasound-guided shoulder injections 50.3 52.2 46.7 46.2 52.6 36.8Anti-dsDNA testing in ANA negative patients 45.3 44.4 45.0 43.6 44.2 52.6DEXA scans more often than two-yearly 44.1 39.1 55.0 47.4 43.2 31.6Ultrasound guidance to perform trochanteric bursainjections

39.1 39.1 38.3 32.1 45.3 42.1

Shoulder ultrasound to diagnose non-specific shoulderpain

36.3 40.9 28.3 37.2 34.7 31.6

Ultrasound to investigate lateral hip pain 31.3 32.2 30.0 30.8 32.6 21.1ENA testing in patients with negative ANA 27.9 28.7 25.0 21.8 31.6 42.1ANCA testing for low probability of diagnosis ofvasculitis

24.6 22.6 25.0 24.4 22.1 26.3

Glucocorticoid injections for non-specific low back pain,facet joint arthritis or spinal canal stenosis

20.7 16.5 30.0 28.2 15.8 15.8

†The percentage in each cell relates to the proportion of respondents that rated the item in their ‘top five’. ANA, anti-nuclear antibody; ANCA, anti-neutrophil cytoplasmic antibodies; Anti-dsDNA, anti-double-stranded DNA; DEXA, dual energy X-ray absorptiometry; ENA, extractable nuclearantibodies.

ARA top five ‘do not do’ list


137

The top five recommendations together with a sum-mary of the evidence that they are a low-value test ortreatment and their current use/cost is summarisedbelow. The remaining seven recommendations aredescribed in Appendix I (Supporting information).

Recommendation 1: Do not performarthroscopy with lavage and/or debridementor partial meniscectomy for patients withsymptomatic osteoarthritis of the kneeand/or degenerate meniscal tear

Strength of recommendation: ANHMRC level of evidence: ICategory of evidence: IaThere is consistent evidence to indicate that arthro-

scopic lavage and/or debridement to treat people forsymptomatic knee osteoarthritis and/or partial menis-cectomy for patients with a degenerate meniscal tear(with or without underlying osteoarthritis), is no moreeffective than placebo surgery or non-operativealternatives.14–19 There appears to be a high rate of con-version from knee arthroscopy to total knee arthro-plasty, which rises with increased age, further suggestingarthroscopic surgery should be avoided in people overthe age of 50 years.20–22 Additionally, arthroscopy isassociated with peri -and post-operative risks and consid-erable cost.18,23

To determine the trend in performance of kneearthroscopic treatment for knee osteoarthritis over timewe considered five of nine MBS codes for knee arthro-scopic washout, debridement and/or partial meniscect-omy (Fig. 1). In total these item numbers, in peoplewith private health insurance, increased in usage from2004 to 2012 financial years, then appeared to plateau,and reduced by 5.9% between 2012 and 2015. Overthe entire period there was an almost 2%p.a. increase. The total benefit paid out for these ser-vices was $17.3 million in 2004 and almost $27.1 mil-lion in 2015, corresponding to an annual growth rateof 4.15%.

Recommendation 2: Do not order ANAtesting in patients without symptoms and/orsigns suggestive of a systemic rheumaticdisease

Strength of recommendation: BCategory of evidence: III-2ANAs are present in healthy individuals and ANA test-

ing is only useful in patients with symptoms and/or signsof a rheumatic disease where it can aid in the confirma-tion or exclusion of systemic connective tissues diseases.

ANA testing has a very high negative predictive value forexcluding connective tissue diseases. However a positiveANA does not have a high positive predictive value fordiagnosing these conditions in isolation and further sub-serology testing is needed to diagnose accurately andclassify these conditions.24,25

Despite guidelines and recommendations not to per-form an ANA test in patients without symptoms and/orsigns suggestive of a connective tissue disease,26–30 therehas been a steady increase over the last decade in thenumber of MBS-funded ANA tests ordered (Fig. 2). Thetotal benefits paid out for these services has increasedfrom $7.76 million in the 2004 financial year to $10.96

Figure 1 Medicare Benefits Schedule (MBS)-funded arthroscopic wash-

out, debridement and/or partial meniscectomy. (MBS item numbers

included in this figure. 49558: Knee, arthroscopic surgery of, involving

one or more of: debridement, osteoplasty or chondroplasty – not asso-

ciated with any other arthroscopic procedure of the knee region.

49559: Knee, arthroscopic surgery of, involving chondroplasty requiring

multiple drilling or carbon fibre (or similar) implant; including any associ-

ated debridement or osteoplasty – not associated with any other

arthroscopic procedure of the knee region. 49560: Knee, arthroscopic

surgery of, involving one or more of: partial or total meniscectomy,

removal of loose body or lateral release – not being a service associ-

ated with any other arthroscopic procedure of the knee region. 49561:

Knee, arthroscopic surgery of, involving one or more of: partial or total

meniscectomy, removal of loose body or lateral release; where the pro-

cedure includes associated debridement, osteoplasty or chondroplasty

– not associated with any other arthroscopic procedure of the knee

region. 49562: Knee, arthroscopic surgery of, involving one or more of:

partial or total meniscectomy, removal of loose body or lateral release;

where the procedure includes chondroplasty requiring multiple drilling

or carbon fibre (or similar) implant and associated debridement or oste-

oplasty – not associated with any other arthroscopic procedure of the

knee region.) ( ), 49558; ( ), 49559; ( ), 49560; ( ), 49561;

( ), 49562.

Morrisroe et al.


138

million in the 2015 financial year, corresponding to anannual growth rate of 3.2%.

Recommendation 3: Do not undertakeimaging for low back pain in patientswithout indications of a serious underlyingcondition

Strength of recommendation: ANHMRC level of evidence: ICategory of evidence: IaMost episodes of low back pain (~90%) do not require

imaging. Imaging may identify irrelevant incidental find-ings and increase the risk of exposure to unnecessaryand sometimes invasive treatment, in addition toincreasing costs.31–33 For patients with low back pain andno suggestion of serious underlying conditions there areno significant differences in pain or disability outcomesbetween immediate imaging as compared with usualcare without imaging.34,35

MBS-funded imaging for low back pain has beenincreasing consistently since 2004 primarily due toincreased numbers of computed tomography (CT) andmagnetic resonance imaging (MRI) scans (Fig. 3). Thetotal MBS benefit paid out for MRI imaging has grownfrom $14.76 million in 2004 to almost $27.96 millionin 2015, an annual growth rate of almost 6%. The totalbenefit paid out for the other imaging modalities of CTimaging and radiography has also grown from $58.4million in 2004 to $99.08 million in 2015, an annualgrowth rate of 4.9%.

Recommendation 4: Do not use ultrasoundguidance to perform injections into thesubacromial space, as it provides noadditional benefit in comparison to landmark-guided injection

Strength of recommendation: ANHMRC level of evidence: ICategory of evidence: IaCurrently there is no high quality evidence to support

the superiority of ultrasound-guided subacromial injec-tions compared with injections guided by landmarksalone. Based upon moderate evidence from five trials, aCochrane review was unable to find any advantage ofultrasound-guided injection over either landmark-guided or intramuscular injection.36 These results areconsistent with a more recent trial.37 In view of the cur-rently available data and the significant added cost, thereis little clinical justification in using ultrasound to guideinjections for shoulder pain.

Figure 3 Number of Medicare Benefits Schedule (MBS)-funded plain

radiographs and CT scans for low back pain and magnetic resonance

imaging (MRI) for sciatica and spinal stenosis in Australia, 2004–2015.

(MBS item numbers included in this figure. Radiography: 58106 and

58111: Spine, lumbosacral. Computed tomography (CT): 56223, 56229:

Computed tomography – scan of spine, lumbosacral region, without

intravenous contrast medium. 56226 and 56232: Computed tomogra-

phy – scan of spine, lumbosacral region, with intravenous contrast

medium and with any scans of the lumbosacral region of the spine prior

to intravenous contrast injection when undertaken. MRI: 63176, 63191,

63234, 63262: sciatica. 63179, 63192, 63237, 63263: spinal stenosis.)

( ), MRI sciatica; ( ), MRI spinal stenosis; ( ), CT; ( ),

radiography.

Figure 2 Number of Medicare Benefits Schedule (MBS)-funded antinu-

clear antibody tests in Australia, 2004–2015 testing. (MBS item num-

bers included in this figure. 71907: Antinuclear antibodies – detection in

serum or other body fluids, including quantitation if required). ( ),

71907.



139

The exact number and costs of subsidised ultrasound-guided injections into the subacromial space areunknown as there are two MBS item numbers thatinclude an ultrasound-guided intervention and neitherspecify a body site. We consider that a substantial num-ber of these procedures are likely to have been per-formed for shoulder pain. There has been an annualincrease of 26.8% in the number of ultrasound-guidedinjections for the period 2004–2015 (Fig. 4). In the2014/2015 financial year the total benefits paid throughthe MBS for ultrasound-guided injection was almost$27.5 million.

As a comparison the total benefits paid through theMBS for landmark-guided joint injections (MBS items50124 and 50125) in the 2008/2009 financial year was$12.8 million. These were removed from the MBS on 1November 2009 due to a Budget decision by the govern-ment that these services are minor and routine in natureand can be delivered as part of a standard consultation.While removal of this MBS item may have resulted in areduction in landmark-guided injection in primary andsecondary care, it may have also contributed to theobserved increase in more expensive image-guidedinjections. Several respondents made comments aboutthe lack of reimbursement for landmark-guided injec-tion, subsequent deskilling of GPs, long wait times forpublic rheumatology clinics, and radiologist-driven self-referrals as possible reasons for the increase in image-guided injection.

Recommendation 5: Do not order anti-double-stranded (ds) DNA antibodies in ANAnegative patients unless clinical suspicion ofsystemic lupus erythematosus (SLE)remains high

Strength of recommendation: BCategory of evidence: III-2International recommendations advise testing for anti-

dsDNA antibodies only after detecting a positive ANA inpatients with symptoms consistent with SLE.25 Inpatients who are ANA negative, anti-dsDNA should onlybe ordered in clinical situations where the pre-test prob-ability of SLE is very high.30 Where positive, repeatinganti-dsDNA antibodies titres is a useful test for monitor-ing disease activity, especially in lupus nephritis.38

The number of MBS funded anti-dsDNA tests performedover 2004–2015 has steadily increased (Fig. 5) and thetotal benefits paid out for these tests more than doubled inthe last decade from $2.1 million dollars in 2004 to $4.4million dollars in 2015. This amounts to an average per

annum growth of almost 7%. There are no epidemiologicaldata suggesting that the incidence of SLE is rising. Forexample over roughly the same time period for which hos-pital separations data are available (2004–2014), the num-ber of hospital separations with a principal diagnosis of SLEincreased by less than 2.8% p.a.39

Discussion

In this paper, we report the top five evidence- andconsensus-based recommendations for tests and

Figure 4 Medicare Benefits Schedule (MBS)-funded ultrasound-guided

injections for all musculoskeletal indications in Australia from 2004 to

2015. (MBS item numbers included in this figure. 55850 and 55851:

Musculoskeletal cross-sectional echography, in conjunction with a surgi-

cal procedure using interventional techniques, inclusive of a diagnostic

musculoskeletal ultrasound service, where the referring practitioner has

indicated on a referral for a musculoskeletal ultrasound that a ultra-

sound guided intervention be performed if clinically indicated.) ( ),

55850 and 55851.

Figure 5 Medicare Benefits Schedule (MBS)-funded dsDNA testing in

Australia from 2004 to 2015. (MBS item numbers included in this figure.

71099: Double-stranded DNA antibodies – quantitation by one or more

methods other than the Crithidia method.) ( ), 71099.

Morrisroe et al.


140

procedures that Australian rheumatologists consider tobe low-value care. An additional eight recommenda-tions, while not included in the top five, were alsoendorsed by a significant number of rheumatologists.The most endorsed recommendation regarding arthros-copy osteoarthritis of the knee and/or degenerate menis-cal tear is consistent with the recently launchedAustralian Clinical Care Standards for Osteoarthritis ofthe Knee,40 as well as a new clinical practice guidelinepublished in the BMJ.41

While we also include similar recommendationsregarding ANA, ENA, dsDNA testing and frequency ofBMD monitoring to some other countries, other recom-mendations were not transferrable to the Australian con-text. For example items, such as testing for Lyme diseaseand prescribing biologic agents prior to methotrexatewere not deemed applicable to Australia due to differ-ences in disease prevalence and mandated Medicare

restrictions. This highlights the importance of creatingrecommendations based on local clinical practices.

Conclusion

In order for our ‘top five’ recommendations to be imple-mented into daily practice, consideration of enablers andbarriers will be required. As a first step we intend to dis-seminate our recommendations widely to cliniciansthrough peer-review publication, news sites, conferencesand presentations; and to consumers through the use ofsocial media, such as twitter. Additionally, some of our rec-ommendations may be supported by other initiatives thatare already taking place, such as the MBS review,42 andnew models of care for back pain.43 While this means thatwe will not be able to determine the precise impact of theARA EVOLVE initiative, we plan to monitor the uptake ofour recommendations using Medicare Statistics data.

References

1 Australian Institute of Health and

Welfare. Australia’s Health 2014.

Australia’s Health Series Number 14. Cat.

No. AUS 178. Canberra: AIHW; 2014.

2 Soon J, Buchbinder R, Close J, Hill C,

Allan S, Turnour C. Identifying low-

value care: the Royal Australasian

College of Physicians’ EVOLVE

initiative. Med J Aust 2016; 204:

180–1.

3 Choosing Wisely Canada. Toronto:

University of Toronto, Canadian

Medical Association, St Michael’s, 2015

[cited 2017 Sep 18]. Available from

URL: http://choosingwiselycanada.org

4 Choosing Wisely Australia. Sydney: NPS

Medicine Wise, 2016 [cited 2017 Sep

18]. Available from URL: http://

choosingwisely.org.au

5 Linstone AT, Turoff M, eds. The Delphi

Method: Techniques and Applications.

Reading, MA: Addison-Wesley

Publishing Co.; 1975.

6 American College of Rheumatology: Five

Things Physicians and Patients Should

Question. Philadelphia, PA: ABIM

Foundation, 2017 [cited 2017 Sep 18].

Available from URL: http://www.

choosingwisely.org/societies/american-

college-of-rheumatology/

7 Canadian Rheumatology Association: Five

Things Physicians and Patients Should

Question. Toronto: Canadian

Rheumatology Association, 2017 [cited

2017 Sep 18]. Available from URL:

http://www.choosingwiselycanada.org/

rheumatology/

8 Endocrine Society: Five Things Physicians

and Patients Should Question.

Philadelphia, PA: ABIM Foundation,

2017 [cited 2017 Sep 18]. Available

from URL: http://www.choosingwisely.

org/societies/endocrine-society

9 American Academy of Orthopaedic

Surgeons: Five Things Physicians and

Patients Should Question. Philadelphia,

PA: ABIM Foundation, 2013 [cited 2017

Sep 18]. Available from URL: http://

www.choosingwisely.org/societies/

american-academy-of-orthopaedic-

surgeons

10 Elshaug AG. Over 150 potentially low-

value health care practices: an Australian

study. Med J Aust 2012; 197: 56–60.

11 Medical Item Reports. Canberra:

Australian Government Department of

Human Services, 2017 [cited 2017 May

12]. Available from URL: http://

medicarestatistics.humanservices.gov.au

12 NHMRC Additional Levels of Evidence and

Grades for Recommendations for Developers of

Guidelines. Canberra: National Health and

Medical Research Council, 2009 [cited

2017 Sep 18]. Available at URL: https://

www.nhmrc.gov.au/_files_nhmrc/file/

guidelines/developers/nhmrc_levels_

grades_evidence_120423.pdf

13 SurveyMonkey Inc. PA, California,

2017 [cited 2017 Sep 18]. Available at

URL: http://www.surveymonkey.com

14 Buchbinder R, Richards B, Harris I.

Knee osteoarthritis and role for surgical

intervention: lessons learned from

randomized clinical trials and

population-based cohorts. Curr

Opin Rheumatol 2014; 26: 138–44.

15 Khan M, Evaniew N, Bedi A, Ayeni OR,

Bhandari M. Arthroscopic surgery for

degenerative tears of the meniscus: a

systematic review and meta-analysis.

CMAJ 2014; 186: 1057–64.

16 Katz JN, Brophy RH, Chaisson CE, de

Chaves L, Cole BJ, Dahm DL et al.

Surgery versus physical therapy for a

meniscal tear and osteoarthritis. N Engl

J Med 2013; 368: 1675–84.

17 Sihvonen R, Paavola M, Malmivaara A,

Itala A, Joukainen A, Nurmi H et al.

Arthroscopic partial meniscectomy

versus sham surgery for a degenerative

meniscal tear. N Engl J Med 2013; 369:

2515–24.

18 Thorlund JB, Juhl CB, Roos EM,

Lohmander LS. Arthroscopic surgery for

degenerative knee: systematic review

and meta-analysis of benefits and

harms. BMJ 2015; 350: h2747.

19 Kise NJ, Risberg MA, Stensrud S,

Ranstam J, Engebretsen L, Roos EM.

Exercise therapy versus arthroscopic

partial meniscectomy for degenerative

meniscal tear in middle aged patients:

randomised controlled trial with

two year follow-up. BMJ 2016; 354:

i3740.

20 Fedorka CJ, Cerynik DL, Tauberg B,

Toossi N, Johanson NA. The

relationship between knee arthroscopy

and arthroplasty in patients under

65 years of age. J Arthroplasty 2014; 29:

335–8.

21 Wai EK, Kreder HJ, Williams JI.

Arthroscopic debridement of the knee

for osteoarthritis in patients fifty years

of age or older: utilization and



141

http://choosingwiselycanada.org

http://choosingwisely.org.au

http://choosingwisely.org.au

http://www.choosingwisely.org/societies/american-college-of-rheumatology



http://www.choosingwiselycanada.org/recommendations/rheumatology

http://www.choosingwiselycanada.org/recommendations/rheumatology

http://www.choosingwisely.org/societies/endocrine-society

http://www.choosingwisely.org/societies/endocrine-society

http://www.choosingwisely.org/societies/american-academy-of-orthopaedic-surgeons




http://medicarestatistics.humanservices.gov.au

http://medicarestatistics.humanservices.gov.au

https://www.nhmrc.gov.au/_files_nhmrc/file/guidelines/developers/nhmrc_levels_grades_evidence_120423.pdf




http://www.surveymonkey.com

outcomes in the Province of Ontario.

J Bone Joint Surg Am 2002; 84-a: 17–22.

22 Hawker G, Guan J, Judge A, Dieppe P.

Knee arthroscopy in England and

Ontario: patterns of use, changes over

time, and relationship to total knee

replacement. J Bone Joint Surg Am 2008;

90: 2337–45.

23 Bohensky MA, Ademi Z, deSteiger R,

Liew D, Sundararajan V, Bucknill A

et al. Quantifying the excess cost and

resource utilisation for patients with

complications associated with elective

knee arthroscopy: a retrospective cohort

study. Knee 2014; 21: 491–6.

24 Solomon DH, Kavanaugh AJ, Schur PH.

Evidence-based guidelines for the use of

immunologic tests: antinuclear antibody

testing. Arthritis Rheum 2002; 47:

434–44.

25 Agmon-Levin N, Damoiseaux J,

Kallenberg C, Sack U, Witte T,

Herold M et al. International

recommendations for the assessment of

autoantibodies to cellular antigens

referred to as anti-nuclear antibodies.

Ann Rheum Dis 2014; 73: 17–23.

26 Yazdany J, Schmajuk G, Robbins M,

Daikh D, Beall A, Yelin E et al. Choosing

wisely: the American College of

Rheumatology’s top 5 list of things

physicians and patients should

question. Arthritis Care Res 2013; 65:

329–39.

27 British Columbia Guidelines: Antinuclear

Antibody (ANA) Testing for Connective

Tissue Disease. British Columbia:

Ministry of Health, 2017 [cited 2017

Sep 18]. Available from URL: http://

www2.gov.bc.ca/gov/content/health/

practitioner-professional-resources/bc-

guidelines/ana-testing

28 Tozzoli R, Bizzaro N, Tonutti E,

Villalta D, Bassetti D, Manoni F et al.

Guidelines for the laboratory use of

autoantibody tests in the diagnosis and

monitoring of autoimmune rheumatic

diseases. Am J Clin Pathol 2002; 117:

316–24.

29 Qaseem A, Alguire P, Dallas P,

Feinberg LE, Fitzgerald FT, Horwitch C

et al. Appropriate use of screening and

diagnostic tests to foster high-value,

cost-conscious care. Ann Intern Med

2012; 156: 147–9.

30 Kavanaugh AF, Solomon DH.

Guidelines for immunologic laboratory

testing in the rheumatic diseases: anti-

DNA antibody tests. Arthritis Rheum

2002; 47: 546–55.

31 Suri P, Boyko EJ, Goldberg J,

Forsberg CW, Jarvik JG. Longitudinal

associations between incident lumbar

spine MRI findings and chronic low

back pain or radicular symptoms:

retrospective analysis of data from the

longitudinal assessment of imaging

and disability of the back

(LAIDBACK). BMC Musculoskelet Disord

2014; 15: 152.

32 Graves JM, Fulton-Kehoe D, Martin DP,

Jarvik JG, Franklin GM. Factors

associated with early magnetic

resonance imaging utilization for acute

occupational low back pain: a

population-based study from

Washington state workers’

compensation. Spine 2012; 37: 1708–18.

33 Webster BS, Bauer AZ, Choi Y,

Cifuentes M, Pransky GS. Iatrogenic

consequences of early magnetic

resonance imaging in acute, work-

related, disabling low back pain. Spine

2013; 38: 1939–46.

34 Chou R, Fu R, Carrino JA, Deyo RA.

Imaging strategies for low-back pain:


Lancet 2009; 373: 463–72.

35 Jarvik JG, Gold LS, Comstock BA,

Heagerty PJ, Rundell SD, Turner JA

et al. Association of early imaging for

back pain with clinical outcomes

in older adults. JAMA 2015; 313:

1143–53.

36 Bloom JE, Rischin A, Johnston RV,

Buchbinder R. Image-guided versus

blind glucocorticoid injection for

shoulder pain. Cochrane Database Syst Rev

2012; 8: CD009147.

37 Dogu B, Yucel SD, Sag SY,

Bankaoglu M, Kuran B. Blind or

ultrasound-guided corticosteroid

injections and short-term response in

subacromial impingement syndrome: a

randomized, double-blind, prospective

study. Am J Phys Med Rehabil 2012; 91:

658–65.

38 Linnik MD, JZ H, Heilbrunn KR,

Strand V, Hurley FL, Joh T.

Relationship between anti-double-

stranded DNA antibodies and

exacerbation of renal disease in

patients with systemic lupus

erythematosus. Arthritis Rheum 2005;

52: 1129–37.

39 National Quality Statement: National

Hospital Morbidity Database 2013–14.

Canberra: Australian Institute Health

and Welfare, 2017 [cited 2017 May 12].

Available from URL: http://meteor.

aihw.gov.au/content/index.phtml/

itemId/611030

40 Australian Commission on Safety and

Quality in Health Care. Osteoarthritis of

the Knee Clinical Care Standard. ACSQHC:

Sydney; 2017.

41 Siemieniuk RHI, Thomas A, Poolman R,

Brignardello-Petersen R, van de

Velde S, Buchbinder R et al.

Arthroscopic surgery for

degenerative knee disease including

arthritis and meniscal tears: a clinical

practice guideline. BMJ 2017; 357:

j1982.

42 Medicare Benefits Schedule Review.

Canberra: Australian Government

Department of Health, 2017 [cited

2017 May 12]. Available from URL:

http://www.health.gov.au/internet/

main/publishing.nsf/content/

mbsreviewtaskforce

43 NSW Agency for Clinical Innovation.

Management of People With Acute Low

Back Pain: Model of Care. Sydney: NSW

Health; 2016.

44 Hillier TA, Stone KL, Bauer DC,

Rizzo JH, Pedula KL, Cauley JA et al.

Evaluating the value of repeat bone

mineral density measurement and

prediction of fractures in older women:

the study of osteoporotic fractures. Arch

Intern Med 2007; 167: 155–60.

45 Screening for osteoporosis:

U.S. preventive services task force

recommendation statement. Ann Intern

Med 2011; 154: 356–64.

46 Screening for osteoporosis:

recommendation statement. Am Fem

Phys 2011; 83: 1197–200.

47 Osteoporosis: Fragility Fracture Risk:

Osteoporosis: Assessing the Risk

of Fragility Fracture. National

Clinical Guideline 2012 National

Institute for Health and Clinical

Excellence.

48 Grimaldi A, Mellor R, Hodges P,

Bennell K, Wajswelner H, Vicenzino B.

Gluteal tendinopathy: a review of

mechanisms, assessment and

management. Sports Med 2015; 45:

1107–19.

49 Cohen SP, Strassels SA, Foster L,

Marvel J, Williams K, Crooks M et al.

Comparison of fluoroscopically guided

and blind corticosteroid injections for

greater trochanteric pain syndrome:

multicentre randomised controlled trial.

BMJ 2009; 338: b1088.

50 Minagawa H, Yamamoto N, Abe H,

Fukuda M, Seki N, Kikuchi K et al.

Morrisroe et al.


142

http://www2.gov.bc.ca/gov/content/health/practitioner-professional-resources/bc-guidelines/ana-testing




http://meteor.aihw.gov.au/content/index.phtml/itemId/611030



http://www.health.gov.au/internet/main/publishing.nsf/content/mbsreviewtaskforce



Prevalence of symptomatic and

asymptomatic rotator cuff tears in the

general population: from mass-

screening in one village. J Orthop 2013;

10: 8–12.

51 Yamaguchi K, Tetro AM, Blam O,

Evanoff BA, Teefey SA, Middleton WD.

Natural history of asymptomatic rotator

cuff tears: a longitudinal analysis of

asymptomatic tears detected

sonographically. J Shoulder Elbow Surg

2001; 10: 199–203.

52 Yamamoto A, Takagishi K, Osawa T,

Yanagawa T, Nakajima D, Shitara H

et al. Prevalence and risk factors of a

rotator cuff tear in the general

population. J Shoulder Elbow Surg 2010;

19: 116–20.

53 Titchener AG, White JJ, Hinchliffe SR,

Tambe AA, Hubbard RB, Clark DI.

Comorbidities in rotator cuff disease: a

case-control study. J Shoulder Elbow

Surg 2014; 23: 1282–8.

54 Savige J, Gillis D, Benson E, Davies D,

Esnault V, Falk RJ et al. International

consensus statement on testing and

reporting of antineutrophil cytoplasmic

antibodies (ANCA). Am J Clin Pathol

1999; 111: 507–13.

55 Mandl LA, Solomon DH, Smith EL,

Lew RA, Katz JN, Shmerling RH. Using

antineutrophil cytoplasmic antibody

testing to diagnose vasculitis: can test-

ordering guidelines improve diagnostic

accuracy? Arch Intern Med 2002; 162:

1509–14.

56 Arnold DF, Timms A, Luqmani R,

Misbah SA. Does a gating policy for

ANCA overlook patients with ANCA

associated vasculitis? An audit of

263 patients. J Clin Pathol 2010; 63:

678–80.

57 Csernok E, Moosig F. Current and

emerging techniques for ANCA

detection in vasculitis. Nat Rev

Rheumatol 2014; 10: 494–501.

58 Watts RA, Mahr A, Mohammad AJ,

Gatenby P, Basu N, Flores-Suarez LF.

Classification, epidemiology and clinical

subgrouping of antineutrophil

cytoplasmic antibody (ANCA)-

associated vasculitis. Nephrol Dial

Transplant 2015; 30(Suppl 1): i14–22.

59 Friedman JH, Dighe G. Systematic

review of caudal epidural injections in

the management of chronic back pain.

Rhode Island Med J 2013; 96: 12–16.

60 Friedly JL, Comstock BA, Turner JA,

Heagerty PJ, Deyo RA, Sullivan SD et al.

A randomized trial of epidural

glucocorticoid injections for spinal

stenosis. N Engl J Med 2014; 371: 11–21.

61 Bresnahan BW, Rundell SD,

Dagadakis MC, Sullivan SD, Jarvik JG,

Nguyen H et al. A systematic review to

assess comparative effectiveness studies

in epidural steroid injections for lumbar

spinal stenosis and to estimate

reimbursement amounts. PM&R 2013;

5: 705–14.

62 Radcliff K, Kepler C, Hilibrand A,

Rihn J, Zhao W, Lurie J et al. Epidural

steroid injections are associated with

less improvement in patients with

lumbar spinal stenosis: a subgroup

analysis of the spine patient

outcomes research trial. Spine 2013;

38: 279–91.

63 Mandel S, Schilling J, Peterson E,

Rao DS, Sanders W. A retrospective

analysis of vertebral body fractures

following epidural steroid injections.

J Bone Joint Surg Am 2013; 95: 961–4.

64 Slipman CW, Bhat AL, Gilchrist RV,

Issac Z, Chou L, Lenrow DAA. Critical

review of the evidence for the use of

zygapophysial injections and

radiofrequency denervation in the

treatment of low back pain. Spine J

2003; 3: 310–16.

65 Carette S, Marcoux S, Truchon R,

Grondin C, Gagnon J, Allard Y et al. A

controlled trial of corticosteroid

injections into facet joints for chronic

low back pain. N Engl J Med 1991; 325:

1002–7.



Appendix S1. Recommendations included in the survey considered to be low-value care, but did not make the ‘topfive’ recommendations



143

doi:10.1111/imj.13667

Underappreciation of non-alcoholic fatty liver diseaseby primary care clinicians: limited awareness of surrogatemarkers of fibrosisPreya J. Patel ,1,2 Xuan Banh,2 Leigh U. Horsfall,1,2 Kelly L. Hayward ,2,3 Fabrina Hossain,4 Tracey Johnson,4

Katherine A. Stuart,1 Nigel N. Brown,5 Nivene Saad,6 Andrew Clouston,2 Katharine M. Irvine,2,7

Anthony W. Russell,3,8 Patricia C. Valery,2,9 Suzanne Williams4* and Elizabeth E. Powell 1,2*

Departments of 1Gastroenterology and Hepatology, 6Radiology, and 8Diabetes and Endocrinology, Princess Alexandra Hospital, 2Centre for LiverDisease Research, Translational Research Institute, Faculty of Medicine and Biomedical Science, and 3School of Clinical Medicine, The University ofQueensland, 4Inala Primary Care, 5Pathology Queensland, 7Mater Research, Translational Research Institute, University of Queensland, and 9QIMRBerghofer Medical Research Institute, Brisbane, Queensland, Australia

Key wordsnon-alcoholic steatohepatitis, non-alcoholic fattyliver disease, general practice, transientelastography, Enhanced Liver Fibrosis test.

CorrespondenceElizabeth E. Powell, Department ofGastroenterology and Hepatology, PrincessAlexandra Hospital, Brisbane, Ipswich Road,Woolloongabba, Qld 4102, Australia.Email: [email protected]

Received 3 July 2017; accepted23 October 2017.

Abstract

Background: Non-alcoholic fatty liver disease (NAFLD) is a common cause of inciden-

tal liver test abnormalities. General practitioners (GP) have a key role in identifying

people with NAFLD at risk of significant liver disease. Recent specialist guidelines

emphasise the use of fibrosis algorithms or serum biomarkers rather than routine liver

tests, to assess advanced fibrosis.

Aim: To evaluate primary care clinicians’ current approach to diagnosis, management

and referral of NAFLD.

Methods: A cross-sectional survey of primary care clinicians was undertaken through

a structured questionnaire about NAFLD. A convenience sample of general practice

clinics and general practice conferences in Metropolitan Brisbane and regional south

east Queensland was selected.

Results: A total of 108 primary care clinicians completed the survey (participation rate

100%). Fifty-one percent of respondents considered the prevalence of NAFLD in the gen-

eral population to be ≤10%. Twenty-four percent of respondents felt that liver enzymes

were sufficiently sensitive to detect underlying NAFLD. Most respondents were unsure

whether the Fibrosis 4 score (62.7% unsure) or Enhanced Liver Fibrosis score (63.7%

unsure) could help to identify advanced fibrosis or cirrhosis. Although 47% of respondents

said they would refer a patient to a Gastroenterologist/Hepatologist if they suspect the

patient has NAFLD, 44.1% do not make any referrals. Of concern, 70.6% of clinicians said

they were unlikely to refer a patient to Hepatology unless liver function tests are abnormal.

Conclusion: Our findings demonstrate that many primary care clinicians underesti-

mate the prevalence of NAFLD and under-recognise the clinical spectrum of NAFLD

and how this is assessed.

Introduction

Non-alcoholic fatty liver disease (NAFLD), encompassingboth simple steatosis and non-alcoholic steatohepatitis(NASH), is a condition characterised by excessive fataccumulation in the liver in the presence of metabolic

risk factors and the absence of significant alcohol intake,

or other causes of hepatic steatosis. The diagnosis of

NASH relies upon a liver biopsy to demonstrate liver cell

injury and inflammation. The prevalence of NAFLD is

increasing in association with the widespread presence

of obesity and type 2 diabetes1,2 and is now the most

common cause of incidental liver test abnormalities in

primary care.3

A study of over 1000 adult patients from eight primary

care practices in the UK reported that abnormal liver

function tests were commonly found incidentally during

routine review of other chronic diseases, such as diabe-

tes, cardiovascular disease and hypertension, and that

*These authors contributed equally as senior authors.Funding: This study was funded by the Pathology Queensland –

Study, Education and Research Trust Fund. P. C. Valery wassupported by the Australian National Health and MedicalResearch Council (Career Development Fellowship #1083090).Conflict of interest: None.


144

NAFLD accounted for over 25% of cases.3 Of concern, astudy from a single Veterans Administration primarycare centre showed that most patients at high risk ofNAFLD were not being acknowledged or evaluated forthis condition.4 Among 251 patients with NAFLD identi-fied by the study investigators, only 21.5% had a diagno-sis of NAFLD in primary care, 14.7% were counselledabout diet and exercise and 10.4% were referred to aspecialist.4 The study found that the only factors associ-ated with receiving NAFLD-related care were the magni-tude (with at least one alanine aminotransferase (ALT)value >80 IU/mL) and proportion (where ≥50% of apatient’s ALT values were >40 IU/mL) of ALT elevation.4

Among people with NAFLD, recognition of advancedliver fibrosis is important because these patients mayrequire specialist care and surveillance for liver cancerand liver decompensation. Most people with NAFLDhave traditional normal-range liver blood tests5 and liverenzyme levels do not reflect the presence or severity ofnecroinflammation or fibrosis.6,7 The recently releasedUK National Institute of Clinical Excellence guidelines(NAFLD: assessment and management) advocate thatroutine liver blood tests alone should not be used to rule

out NAFLD nor identify advanced fibrosis in people withNAFLD.8 Clinical practice guidelines for the diagnosisand management of NAFLD produced by internationalsubspecialist committees9 recommend that ‘… Surrogatemarkers of fibrosis (NAFLD fibrosis score (NFS), Fibrosis4 test (FIB-4), Enhanced Liver Fibrosis (ELF) test orFibroTest) should be calculated for every NAFLD patient,in order to rule out significant fibrosis (≥F2). If significantfibrosis cannot be ruled out, patients should be referredto a Liver Clinic for transient elastography …’.9 Both theNFS and FIB-4 have a high negative predictive value,10

are low to no cost as they are calculated from routineblood tests (platelets, albumin, ALT and aspartate amino-transferase (AST)), age and body mass index (BMI) andcan be run repeatedly without additional costs. The com-mercially available ELF test is calculated from combiningthree measured direct markers of liver matrix metabo-lism in serum. The ELF panel has good diagnostic accu-racy for advanced fibrosis11,12 and has been shown topredict disease progression13 and the development ofliver-related clinical events in patients with chronic liverdisease.13,14 In Australia, the FibroTest is currently onlyavailable in a research capacity. Transient elastography(TE) provides a non-invasive assessment of liver fibrosis;however, this technology is not ideal for stratifying liverdisease severity in the community, as it requires special-ist expertise to obtain and interpret data.Primary care clinicians have a key role in identifying

patients with NAFLD who are at risk of significant liverdisease and who may require specialist referral for

further evaluation, or closer management of metaboliccomorbidities and lifestyle interventions. However, it iscurrently unclear whether GP are aware of the recom-mendations regarding NAFLD risk stratification andmanagement. We have previously shown that fatty liverand abnormal liver tests accounted for around 10% ofreferrals for hepatology outpatient consultation at amajor tertiary referral hospital.15 Clinical informationprovided by referring clinicians was often incompleteand only a minority of referrals provided informationabout BMI or alcohol intake.15,16 There are no currentAustralian guidelines for NAFLD management in pri-mary care, although the RACGP website (http://www.racgp.org.au/afp/2013/july/fatty-liver-disease/) containsa practical guide and algorithm for management of sus-pected NAFLD.17

The purpose of this study was to assess primary careclinicians’ current understanding of NAFLD, along withtheir approach to diagnosis, management and criteria forreferral. In order to address this, a cross-sectional surveyof primary care clinicians from the greater Brisbane areawas undertaken through a structured questionnaireabout NAFLD.

Methods

A cross-sectional study was conducted in south-eastQueensland between August 2016 and March 2017. Aconvenience sample of general practice clinics and gen-eral practice conferences in metropolitan Brisbane andregional south east Queensland were selected. All pri-mary care clinicians who attended general practice unitmeetings or the conferences were invited to participatein the study.The self-administered structured survey (available on

request) developed by a research group (comprised of ahepatologist, epidemiologist and GP) for this study con-sisted of a series of closed-ended questions and oneopen-ended question. The survey was divided into fourparts: (i) a brief introductory paragraph defining NAFLDand the spectrum of disease, including simple steatosisand NASH; (ii) a short demographic section comprisingvariables, such as Practice postcode, level of appoint-ment, an estimate of the number of different patientsseen per week, and an estimate of the proportion of sub-jects with NAFLD within their patient population;(iii) questions assessing the respondents’ knowledgeabout the prevalence of NAFLD, risk factors and percep-tions about morbidity, diagnosis, risk stratification andtreatment and (iv) respondents were asked about thereferrals they made to hepatology services for investiga-tion or management of NAFLD/NASH.

Underappreciation of NAFLD in primary care


145

http://www.racgp.org.au/afp/2013/july/fatty-liver-disease/

http://www.racgp.org.au/afp/2013/july/fatty-liver-disease/

Face and content validity of the survey was assessedby delivering the questionnaire to three GP to determinewhether they understood the questions and thought thatthe content of questions was relevant. Retest reliabilitywas not assessed as it was not feasible to deliver thequestionnaire to the same primary care clinicians twicedue to their busy schedules.

This study was approved by the Princess AlexandraHospital and The University of Queensland HumanResearch Ethics Committees (HREC/15/QPAH/301;UQ:2015001047).

Statistical methods

Basic descriptive details (total numbers and percentages)are presented. Continuous variables were summarised asmean (standard error of the mean) and median (range).The Chi-squared test was used to compare categoricalvariables and Student’s t-test to compare continuous var-iables. SPSS version 24.0 (SPSS, Chicago, IL, USA) was usedfor all analyses and a value of P < 0.05 was consideredstatistically significant. Microsoft Excel (Microsoft OfficeProfessional Plus 2010, Redmond, WA, USA) was usedto conduct inductive thematic analysis, classifying quali-tative data obtained through an open-ended questioninto meaningful answers. The proportion of respondentsreporting the most common responses was described.

Results

Demographic characteristics of the surveyrespondents

The survey was completed by 108 primary care clini-cians. Of these, 87 were GP, 7 were GP registrars, 9 werepractice nurses and 5 did not provide their level ofappointment. The respondents’ practices were located inmetropolitan Brisbane (n = 68), regional/remoteQueensland (n = 30), locum positions or outsideQueensland (n = 5) and five respondents did not providetheir Practice post-code. A surveyor was present duringadministration of the survey and the participation ratewas 100%. Six respondents completed less than 50% ofthe survey and were excluded from all further analyses.The remainder of the respondents completed >65% ofthe survey. The median number of different patientsseen by a respondent each week was 80.

Characteristics of the patient populations

Overall, respondents estimated that 15.5 � 13.5% oftheir patient population had type 2 diabetes, 19.1 �13.1% had dyslipidaemia or hypertriglyceridaemia, 21.7� 14.4% had hypertension, 28.6 � 17.2% were

overweight or obese and 16.6 � 12.2% consumed excessalcohol. A substantial proportion of the respondentsindicated that NAFLD was not common among theirpatients, with 38.2% estimating the prevalence to be≤10% (Fig. 1). There was no statistical difference in theestimated prevalence of NAFLD/NASH or metabolic co-morbidities between rural and metropolitan regions(data not presented, P > 0.05).

Awareness of prevalence of and factorsassociated with NAFLD/NASH

Overall, 51% of respondents considered the prevalenceof NAFLD in the general population to be ≤10%(Fig. 2A) and 54% of primary care clinicians consideredthe prevalence of NAFLD in the obese population to≤50% (Fig. 2B). There was no significant differencebetween metropolitan and rural practitioners’ responses(data not presented, P > 0.05).

More than 90% of respondents recognised that over-weight/obesity, type 2 diabetes, metabolic syndrome andhypertriglyceridaemia were strongly associated withNAFLD. Half of the respondents (51.0%) considered thathypertension was associated with NAFLD, and most(73.5%) considered alcohol consumption to be strongly

Figure 1 Proportion of primary care clinicians estimating different rates

of non-alcoholic fatty liver disease (NAFLD) in their patient population.

( ), <10%; ( ), 10–30%; ( ), >30%.

Patel et al.


146

associated with NAFLD. Most respondents were awareof current recommendations for ‘safe’ alcohol consump-tion. Approximately, 87.3% of respondents consideredthat the lifetime risk of harm from alcohol-related dis-ease is reduced by drinking ≤2 standard drinks per day.

Perception of morbidity/mortality associatedwith NAFLD

Respondents were aware that simple steatosis is associ-ated with an increased incidence of cardiovascular disease(65.7%), and future development of type 2 diabetes(73.5%). In contrast, the relative absence of liver-relatedoutcomes associated with simple steatosis was not as wellappreciated. 61.8% of respondents considered that simplesteatosis is associated with increased liver-related mortal-ity and 47.1% considered these subjects at significantlyhigher risk of cirrhosis. Most respondents were aware ofthe increased risk of cirrhosis and liver-related mortalityin subjects with NASH (71.6 and 75.5% respectively).

Perception of tests for diagnosis and riskstratification of NASH

Many primary care clinicians incorrectly felt that a diag-nosis of NASH can be made using serum liver enzymes(52.9%), liver imaging (70.6%) or FibroScan (62.7%).Of concern, 24.5% of respondents felt that liver enzymes(ALT and AST) are sufficiently sensitive to detect under-lying NAFLD–NASH and a further 25.5% were unsure.The majority of respondents agreed that liver enzymes

(ALT, AST) (79.4%), platelet count (67.6%), serumalbumin (80.4%), prothrombin time (69.6%), NAFLDFibrosis score (81.4%), abdominal ultrasound (77.5%)and FibroScan (71.6%) can help to identify NAFLDpatients with advanced fibrosis/cirrhosis. However, mostrespondents were unsure whether the FIB-4 score(62.7% unsure) or ELF score (63.7% unsure) could helpto identify advanced fibrosis or cirrhosis.Although 87.3% of respondents felt that 6 monthly

liver enzyme tests can help to monitor disease progression

in patients with NAFLD, there was less clarity about theuse of other tests. Many clinicians were unsure whetherplatelet count (26.5% unsure), FibroScan (36.3%),NAFLD Fibrosis score/Fib-4 test (43.1%) or the ELF test(56.9%) can help to monitor disease progression.

Assessment of clinical and referral practice

Respondents were asked whether they utilise certaintools in their clinical practice to assess their patients withNAFLD (Fig. 3). A total of 92 and 87.3% use liverenzymes and abdominal ultrasound, respectively, forNAFLD assessment. However, the majority of cliniciansdo not use FibroScan (80.4%), NAFLD Fibrosis score(76.5%), Fib-4 score (85.3%), AST to Platelet RatioIndex (APRI) score (78.4%) or ELF test (74.5%).With respect to clinical management, the majority of

clinicians would provide information on optimising dietand exercise (94.1%), provide a GP management planand team care arrangements (74.5%) and refer to a die-tician (90.2%). Forty-seven percent of respondentsstated that they would refer a patient to a Gastroenterol-ogist/Hepatologist if they suspect the patient has NAFLD.Despite this, when asked how many referrals they maketo Hepatology each month for an opinion regarding sus-pected NAFLD/NASH, 44.1% do not make any referralsand a further 44.1% make less than 1–2 referrals eachmonth (Fig. 4). Common reasons provided for not refer-ring patients to Hepatology included: ‘The patients donot want referral’ (18.8%), ‘There is no specific pharma-cotherapy available’ (6.3%), ‘I manage them myself byoptimising lifestyle’ (38.5%), ‘The waiting list is too long’(12.5%), ‘I don’t see many patients with NAFLD/NASH’(31.3%), ‘I do not think it is necessary’ (4.2%). Of con-cern, 70.6% of clinicians said they were unlikely to refera patient to Hepatology unless liver function tests areabnormal. A total of 22 respondents volunteered addi-tional written comments, of whom 31.8% highlightedtheir own lack of knowledge regarding NAFLD, forexample ‘After doing this survey I realise I don’t knowvery much about this important topic…’.

(A) (B)

Figure 2 Proportion of primary care clinicians estimat-

ing different rates of non-alcoholic fatty liver disease in

(A) the general population and (B) the obese population.

(A) ( ), ≤10%; ( ), 11–30%; ( ), 50%; ( ), unsure. (B) ( ),

≤25%; ( ), 26–50%; ( ), 70%; ( ), unsure.



147

Discussion

Primary care clinicians have a key role in the initial identi-fication and management of NAFLD and in recognisingpatients at risk of significant liver disease, who mayrequire specialist referral for further evaluation. Recentclinical guidelines have emphasised the use of fibrosisalgorithms or serum biomarkers rather than routine liverblood tests, to assess advanced fibrosis in people who havebeen diagnosed with NAFLD. The present study indicatedthat the majority (>85%) of primary care clinicians do notuse TE, fibrosis biomarkers or algorithms in their clinicalpractice, to assess their patients with NAFLD. In addition,the majority of clinicians (70.6%) said they were unlikelyto refer a NAFLD patient for a Hepatology opinion unlessliver function tests are abnormal. These findings are con-cerning because liver enzyme levels do not correlate withhistological findings7 and reliance on abnormal liver testsmay fail to identify patients with significant liver disease.10

Similar to our findings, a recent survey among 64 GP inThe Netherlands found that serum markers/scores werenever (73%) or rarely (22%) used to assess NAFLD/NASHand referral to specialist care was highly dependent onelevated liver tests.18 Not surprisingly, the authors foundthat GP were not familiar with NAFLD clinical practiceguidelines prepared by the hepatology medical societies.18

Most of the liver-related morbidity and mortality associ-ated with NAFLD occurs in patients with advanced fibro-sis, who are at risk of developing complications of end-

Figure 3 Proportion of primary care clinicians utilising certain tools in their clinical practice, to assess their patients with non-alcoholic fatty liver dis-

ease. Region of practice: ( ), Metropolitan; ( ), rural.

Figure 4 Approximate number of referrals made to hepatology each

month for an opinion regarding suspected non-alcoholic fatty liver dis-

ease/non-alcoholic steatohepatitis. ( ), None; ( ), 1–2 per month; ( ), 3–5

per month.

Patel et al.


148

stage liver disease and hepatocellular carcinoma.19 Assess-ment for liver fibrosis within primary care is important todetermine prognosis, monitor disease progression and todecide if and when to refer to specialist care. At present, apractical approach to NAFLD diagnosis and staging in thecommunity is recommended, using locally availabletests.20 These include simple non-invasive fibrosis algo-rithms (NAFLD fibrosis and FIB-4 scores), commercialnon-invasive fibrosis tests (the Enhanced Liver Fibrosistest), liver ultrasound and TE (if available). In patientswith NAFLD, liver stiffness measurements (LSM) have ahigh negative predictive value and a modest positive pre-dictive value for detecting advanced fibrosis.21 Table 1summarises the components of locally available risk strati-fication tools and their interpretation.Our findings demonstrate that many primary care clini-

cians underestimate the overall prevalence of NAFLD intheir community. Approximately half of respondents con-sidered the prevalence of NAFLD in the general popula-tion to be ≤10 and 38.2% considered the prevalence ofNAFLD in their patient population to be ≤10%. These per-ceptions conflict with recent reports estimating that GPare likely to encounter more than 300 cases of NAFLD forevery 1000 patients that are seen.22 In a recent large,population-based study from The Netherlands (n = 3041individuals ≥45 years), 32.8% of participants had evidenceof NAFLD on ultrasonography, in the absence of second-ary causes of steatosis.23 Clinically relevant fibrosis(defined by LSM ≥8 kPa) was present in 8.4% of theNAFLD subgroup and in 17.2% of participants with bothdiabetes and steatosis.23 Underappreciation of the

prevalence of NAFLD may contribute to many affectedindividuals remaining undiagnosed, in part because thecondition is usually asymptomatic and associated with rel-atively normal or only mildly elevated liver enzyme levels.Our findings also demonstrate lack of recognition of

the clinical spectrum of NAFLD and how this is assessed.NAFLD represents a spectrum of disease from bland stea-tosis to the necro-inflammatory form, NASH, which ischaracterised by inflammation and cellular damage(hepatocyte ballooning). Although fibrosis may developin both steatosis and NASH, fibrosis progression occurs ata more rapid rate in patients with NASH.24 A meta-analysis of NAFLD studies assessing paired liver biopsiesfound that on average, fibrosis progressed by 1 stage over7.1 years for patients with NASH and by 1 stage over14.3 years for patients with bland steatosis.24 In this sur-vey, >50% of respondents incorrectly considered that adiagnosis of NASH could be made using serum liverenzymes, liver imaging and/or FibroScan. The diagnosisof NASH requires a liver biopsy, because clinical features,biochemical or imaging tests cannot distinguish steatohe-patitis from bland steatosis, and non-invasive tests are notcurrently validated for this purpose.9 Recent data fromlongitudinal studies suggest that fibrosis stage, rather thanthe presence of NASH, is the most important histologicalfeature associated with liver-related outcome.19,25,26

Conclusion

Despite recognition that these patients are at risk of pro-gressive liver disease, approximately 45% of primary

Table 1 Locally available non-invasive tests that can be used to assess advanced fibrosis in people who have been diagnosed with non-alcoholic fattyliver disease (NAFLD)

Test/Score Specific components Interpretation/Comments

Platelet count — • Low platelet count suggestive of advanced fibrosis• Late finding

Serum albuminlevel

— • Low serum albumin level may be seen in advanced liver disease• Late finding; not specific for chronic liver disease

AST/ALT ratio10 AST/ALT ratio • >0.8 – higher risk of advanced fibrosis• ALT falls and AST remains stable or rises with " fibrosis

NAFLD fibrosisscore10

Age, impaired fasting glucose or diabetes, BMI,platelet count, albumin, AST/ALT ratio

• <−1.455 – low risk for advanced fibrosis• >0.675 – suggestive of advanced fibrosis• Age dependent – interpret with caution <35 years or >65 years.• High rate of intermediate scores

FIB-4 score10 Age, AST, ALT, platelet count • < 1.3 – low risk for advanced fibrosis• >2.67 – suggestive of advanced fibrosis• Age dependent – interpret with caution <35 years or >65 years.

FibroScan 21 Uses pulse-echo ultrasoundTrained operator required

• Provides a liver stiffness measurement (LSM) as a surrogate marker offibrosis

• LSM < 8.2 kPa – advanced fibrosis unlikely in NAFLDEnhanced LiverFibrosis test11

Commercial panel comprising tissue inhibitor ofmatrix metalloproteinase 1, hyaluronic acid,

procollagen III amino terminal peptide

• Manufacturer recommends a cut-off ≥9.8 for severe fibrosis• Recommended by UK NICE guidelines8 as first-line test for advanced

fibrosis in NAFLD

ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; LSM, liver stiffness measurements.



149

care clinicians do not make any referrals to hepatologyservices for an opinion regarding suspected NAFLD/NASH. This may result in many affected subjectsremaining undiagnosed, consequently leading topatients presenting late with decompensated cirrhosis orhepatocellular cancer. Indeed, we have previouslyshown up to 46% of new referrals for NAFLD can pre-sent with advanced disease.27 Recognition of subjectswith early or compensated cirrhosis is important, asthese patients require surveillance for the developmentof gastroesophageal varices and hepatocellular cancer.Several respondents volunteered written comments onthe questionnaire form referring to their perceived lack

of knowledge regarding NAFLD. In surveys from TheNetherlands18 and an urban western US population,28

84% of GP and 83% of largely primary care providers,respectively, endorsed the need for increased awarenessand knowledge on NAFLD. Our Australian data in con-junction with others, support the global need for educa-tional strategies to improve the recognition andassessment of NAFLD by primary care clinicians. Practi-cal approaches to address this include the provisionof educational workshops to increase recognition.22

Another approach would be to utilise existing models ofcare to improve patient management by upskilling pri-mary care practices.29

References

1 Brunt EM, Wong VW, Nobili V, Day CP,

Sookoian S, Maher JJ et al.

Nonalcoholic fatty liver disease. Nat Rev

Dis Primers 2015; 1: 15080.

2 Bjornsson E, Angulo P. Non-alcoholic

fatty liver disease. Scand J Gastroenterol

2007; 42: 1023–30.

3 Armstrong MJ, Houlihan DD,

Bentham L, Shaw JC, Cramb R, Olliff S

et al. Presence and severity of non-

alcoholic fatty liver disease in a large

prospective primary care cohort. J

Hepatol 2012; 56: 234–40.

4 Blais P, Husain N, Kramer JR,

Kowalkowski M, El-Serag H, Kanwal F.

Nonalcoholic fatty liver disease is

underrecognized in the primary care

setting. Am J Gastroenterol 2015; 110:

10–14.

5 Browning JD, Szczepaniak LS,

Dobbins R, Nuremberg P, Horton JD,

Cohen JC et al. Prevalence of hepatic

steatosis in an urban population in the

United States: impact of ethnicity.

Hepatology 2004; 40: 1387–95.

6 Wong VW, Wong GL, Tsang SW,

Hui AY, Chan AW, Choi PC et al.

Metabolic and histological features of

non-alcoholic fatty liver disease patients

with different serum alanine

aminotransferase levels. Aliment

Pharmacol Ther 2009; 29: 387–96.

7 Mofrad P, Contos MJ, Haque M,

Sargeant C, Fisher RA, Luketic VA et al.

Clinical and histologic spectrum of

nonalcoholic fatty liver disease

associated with normal ALT values.

Hepatology 2003; 37: 1286–92.

8 Glen J, Floros L, Day C, Pryke R. Non-

alcoholic fatty liver disease (NAFLD):

summary of NICE guidance. BMJ 2016;

354: i4428.

9 European Association for the Study of

the Liver (EASL), European Association

for the Study of Diabetes (EASD),

European Association for the Study of

Obesity (EASO). EASL-EASD-EASO

clinical practice guidelines for the

management of non-alcoholic fatty liver

disease. Diabetologia 2016; 59: 1121–40.

10 McPherson S, Stewart SF, Henderson E,

Burt AD, Day CP. Simple non-invasive

fibrosis scoring systems can reliably

exclude advanced fibrosis in patients

with non-alcoholic fatty liver disease.

Gut 2010; 59: 1265–9.

11 Rosenberg WM, Voelker M, Thiel R,

Becka M, Burt A, Schuppan D et al.

Serum markers detect the presence of

liver fibrosis: a cohort study.

Gastroenterology 2004; 127: 1704–13.

12 Guha IN, Parkes J, Roderick P,

Chattopadhyay D, Cross R, Harris S

et al. Noninvasive markers of fibrosis in

nonalcoholic fatty liver disease:

validating the European liver fibrosis

panel and exploring simple markers.

Hepatology 2008; 47: 455–60.

13 Irvine KM, Wockner LF, Shanker M,

Fagan KJ, Horsfall LU, Fletcher LM et al.

The enhanced liver fibrosis score is

associated with clinical outcomes and

disease progression in patients with

chronic liver disease. Liver Int 2016; 36:

370–7.

14 Parkes J, Roderick P, Harris S, Day C,

Mutimer D, Collier J et al. Enhanced

liver fibrosis test can predict clinical

outcomes in patients with chronic liver

disease. Gut 2010; 59: 1245–51.

15 Horsfall L, Macdonald G, Scott I,

Skoien R, Khatun M, Moss C et al. Use

of standardised assessment forms in

referrals to hepatology outpatient

services: implications for accurate

triaging of patients with chronic

hepatitis C. Aust Health Rev 2013; 37:

218–22.

16 Fagan KJ, Irvine KM, Kumar S, Bates A,

Horsfall LU, Feeney GF et al. Assessment

of alcohol histories obtained from

patients with liver disease: opportunities

to improve early intervention. Intern

Med J 2013; 43: 1096–102.

17 Iser D, Ryan M. Fatty liver disease – a

practical guide for GPs. Aust Fam

Physician 2013; 42: 444–7.

18 van Asten M, Verhaegh P, Koek G,

Verbeek J. The increasing burden of

NAFLD fibrosis in the general

population: time to bridge the gap

between hepatologists and primary

care. Hepatology 2017; 65: 1078.

19 Ekstedt M, Hagstrom H, Nasr P,

Fredrikson M, Stal P, Kechagias S et al.

Fibrosis stage is the strongest predictor

for disease-specific mortality in NAFLD

after up to 33 years of follow-up.

Hepatology 2015; 61: 1547–54.

20 Dyson JK, Anstee QM, McPherson S.

Non-alcoholic fatty liver disease: a

practical approach to diagnosis and

staging. Frontline Gastroenterol 2014; 5:

211–18.

21 Cassinotto C, Boursier J, de

Ledinghen V, Lebigot J, Lapuyade B,

Cales P et al. Liver stiffness in

nonalcoholic fatty liver disease: a

comparison of supersonic shear

imaging, FibroScan, and ARFI with liver

biopsy. Hepatology 2016; 63: 1817–27.

22 Grattagliano I, D’Ambrosio G,

Palmieri VO, Moschetta A,

Palasciano G, Portincasa P et al.

Improving nonalcoholic fatty liver

disease management by general

practitioners: a critical evaluation and

impact of an educational training

program. J Gastrointestin Liver Dis 2008;

17: 389–94.

Patel et al.


150

23 Koehler EM, Plompen EP, Schouten JN,

Hansen BE, Darwish Murad S, Taimr P

et al. Presence of diabetes mellitus and

steatosis is associated with liver stiffness

in a general population: the Rotterdam

study. Hepatology 2016; 63: 138–47.

24 Singh S, Allen AM, Wang Z, Prokop LJ,

Murad MH, Loomba R. Fibrosis

progression in nonalcoholic fatty liver

vs nonalcoholic steatohepatitis: a

systematic review and meta-analysis of

paired-biopsy studies. Clin Gastroenterol

Hepatol 2015; 13: 643-54.e1-9 quiz

e39-40.

25 Angulo P, Kleiner DE, Dam-Larsen S,

Adams LA, Bjornsson ES,

Charatcharoenwitthaya P et al. Liver

fibrosis, but no other histologic features,

is associated with long-term outcomes

of patients with nonalcoholic fatty liver

disease. Gastroenterology 2015; 149: 389–

97.e10.

26 Dulai PS, Singh S, Patel J, Soni M,

Prokop LJ, Younossi Z et al. Increased

risk of mortality by fibrosis stage in

nonalcoholic fatty liver disease:


Hepatology 2017; 65: 1557–65.

27 El-Atem NA, Wojcik K, Horsfall L,

Irvine KM, Johnson T, McPhail SM et al.

Patterns of service utilisation within

Australian hepatology clinics: high

prevalence of advanced liver disease.

Intern Med J 2016; 46: 420–6.

28 Wieland AC, Quallick M, Truesdale A,

Mettler P, Bambha KM. Identifying

practice gaps to optimize medical care

for patients with nonalcoholic fatty liver

disease. Dig Dis Sci 2013; 58: 2809–16.

29 Burridge LH, Foster MM, Donald M,

Zhang J, Russell AW, Jackson CL. A

qualitative follow-up study of diabetes

patients’ appraisal of an integrated

diabetes service in primary care. Health

Soc Care Community 2017; 25: 1031–40.

doi:10.1111/imj.13612

Acute oxygen therapy: an audit of prescribing and deliverypractices in a tertiary hospital in Perth, Western AustraliaAnam Kamran ,1 Elisa Chia1 and Claire Tobin1,2

1Acute Medical Unit, and 2Department of Respiratory Medicine, Royal Perth Hospital, Perth, Western Australia, Australia

Key wordsoxygen inhalation therapy, prescription practice,inappropriate prescribing, oxygen, WesternAustralia.

CorrespondenceAnam Kamran, Acute Medical Unit, Royal PerthHospital, Wellington Street, Perth, WA 6000,Australia.Email: [email protected]

Received 3 April 2017; accepted17 August 2017.

Abstract

Background: Oxygen is a widely used drug in the hospital setting. However, interna-

tional audits suggest that oxygen administration practices are often not compliant with

prescribed standards. This can place patients at risk and cause serious adverse events.

Aim: To analyse data related to recent practices of oxygen prescription and administra-

tion at Royal Perth Hospital (RPH), Western Australia. The results of this audit aim to

guide further research on possible interventional studies implementing key solutions.

Methods: All patients who received care in the Acute Medical Unit at RPH between

1 September and 14 September 2015 were included in this audit. Patients who were

given supplemental oxygen during their admission were selected for further review of

records. Appropriate medically indicated target oxygen saturations for each patient

were judged under consultation with a respiratory specialist.

Results: A total of 65 patients received oxygen supplementation within the study

period; 36 of these patients (55.4%) had target oxygen saturations prescribed by doc-

tors, and 25% of the prescribed targets were judged to be inappropriate. In total,

49 patients (75.4%) were exposed to a potential risk from oxygen therapy due to pre-

scription error and/or delivery error. A real risk was identified in 19 patients (29.2%)

as they received oxygen at levels outside their appropriate medically indicated target

range.

Conclusion: The current practices of oxygen prescription and administration within

RPH are suboptimal. Patients are placed at risk of oxygen toxicity due to deviation from

oxygen prescription guidelines.



151

Introduction

Oxygen supplementation is an essential component ofthe management of hypoxia. It aims to ensure that suffi-cient oxygen is provided at a tissue level to meet meta-bolic demand. However, if provided inappropriately,oxygen administration can lead to significant adverseeffects.1 Oxygen is generally not indicated in breathlesspatients who do not have hypoxia, even in the setting ofa cerebral vascular accident or acute coronary syn-drome.2 Over-oxygenation can result in airway paren-chymal injury, local vasoconstriction, ventilation–perfusion mismatch, adverse cardiovascular events andaccentuation of pre-existing hypercapnia.3–6 The lattercan lead to significant end-organ damage and carbondioxide narcosis.3 Over-oxygenation can artificiallyincrease saturation levels, masking subtle clinical deteri-oration and diverting focus from underlying causes ofhypoxia.5

A number of guidelines has been published interna-

tionally to standardise safe and effective oxygen ther-

apy.2,7 Royal Perth Hospital (RPH), a tertiary hospital in

Perth, Western Australia, is governed by the Western

Australian (WA) oxygen therapy guidelines,2 which are

similar to the British Thoracic Society guidelines.7 The

guidelines highlight the indications for oxygen, including

guidance on correct target saturation levels for different

patient groups. The aim to actively wean stable patients

off oxygen is emphasised through regular monitoring

and titration of oxygen therapy.The accurate completion of a dedicated oxygen pre-

scription chart is instituted in the WA guidelines, withdaily medical review of the prescription chart recom-mended. The RPH Oxygen Prescription chart was devel-oped in 2012 (Fig. S1, Supporting information)following the WA guidelines. Doctors at RPH arerequired to use this chart to prescribe clear parametersfor inpatient oxygen therapy.

Several international audits on oxygen therapy have

been conducted, showing relatively widespread substan-

dard practices,8–10 with the potential to result in signifi-

cant adverse patient outcomes.1,11 In 2010, the UK

National Patient Safety Agency reported 281 serious inci-

dents related to oxygen use over 5 years, including

9 deaths and 35 contributions to death.11 Despite the

importance of this topic, a review of literature demon-

strates limited research on recent oxygen therapy prac-

tices in Australian tertiary hospitals. The aim of this audit

is to create a snapshot of current practices within our

institution to help guide future interventions and audits.The Acute Medical Unit (AMU) at RPH admits all

medical patients from the Emergency Department.Patient care continues under AMU if the length of stay is

anticipated to be less than 72 h; otherwise, care is trans-ferred to other hospital specialties. At the time of thisstudy, RPH AMU comprised 43 beds, with an averagelength of stay for AMU patients of 1.6 days and a dailypatient turnover of 72%. Due to the acuity and variednature of patients in the AMU, together with the highpatient turnover and regular involvement of differentspecialties, the unit is a good base for conductingresearch to gauge hospital-wide practices.

Methods

All patients cared for in AMU between 1 September and14 September 2015 were selected for this audit. For eachpatient, notes and observation charts were reviewed.Patients who were provided supplemental oxygen at anypoint of their hospital stay were selected for further anal-ysis. Those who were not administered oxygen were notincluded for further analysis in this audit. A question-naire was developed to standardise data collection whilereviewing all available records for these patients. Areasof focus for data collection included clinical informationrelated to oxygen delivery and relevant patient back-ground, particularly any history of chronic respiratorydisease or smoking.

Patients’ notes were checked for the presence of filledoxygen prescription charts (Fig. S1, Supporting informa-tion) to determine the target saturation level prescribed(P). Filled charts were also examined for any errors oromissions in the eight required fields, as indicated inFigure S1 (Supporting information). If an oxygen pre-scription chart was not utilised, any target saturationlevel described by medical staff in patient notes was con-sidered the prescribed oxygen target (P) for nurses to fol-low. Observation charts (Fig. 1) were assessed todetermine each patient’s average oxygen saturation(S) while on oxygen supplementation. As per the obser-vation chart, oxygen saturations were classified intothree groups: > 94%, 88–92% and <88%. Non-parametric techniques were then used to determine themean oxygen saturation obtained.

Royal Perth Hospital healthcare software (iSoft) wasused to search for previous arterial or venous blood gasesfor each patient. A history of CO2 retention in conjunc-tion with a background of chronic lung disease was usedto determine a predisposition to type 2 respiratory fail-ure. These data were reviewed by a Respiratory Consul-tant who led the consensus on confirming thispredisposition and determining the appropriate medi-cally indicated target oxygen saturation that should havebeen prescribed for each patient (IP). As per WAguidelines,2 patients with a predisposition to type 2 respi-ratory failure were determined to have an ideal

Kamran et al.


152

medically indicated target oxygen saturation of 88–92%,whilst those without this predisposition had an idealmedically indicated target oxygen saturation of 94–98%.The prescribed targets (P), medically indicated targets

(IP) and average oxygen saturations achieved (S) wereevaluated and compared to determine any potential orreal risk to each patient. A potential risk was consideredto be a ‘prescription error’ if prescriptions were missing(no P) or incorrect (P not equal to IP) or a ‘deliveryerror’ if there was inappropriate provision of oxygen(S not equal to P). Real risk to the patient was consideredto have occurred when oxygen delivery (S) differed fromthe medically indicated target saturation for that patient(IP), resulting in either insufficient oxygenation (S < IP)or over-oxygenation (S > IP).

Results

During the study period, 327 patients were admitted tothe RPH AMU; 19.9% of these patients (65) receivedoxygen therapy during their admission. Oxygen wasoften initiated in the Emergency Department prior toAMU admission and continued in AMU on transfer(44.6% of patients). The mean age of patients given oxy-gen was 69 years (SD = 19 yrs). Table 1 outlines otherkey characteristics, including risk of Type 2 respiratoryfailure, for the patients who received oxygen.Of the 65 patients receiving oxygen therapy, 53.9%

had a dedicated oxygen prescription chart. Two patientshad target saturations documented in their medical noteswithout the use of a prescription chart. In total, 44.6%of patients received oxygen supplementation without

any prescription. Oxygen saturations were not docu-mented prior to oxygen supplementation in 37%.Of the completed prescription charts, 94.3% were not

filled appropriately (Fig. 2). On average, two errors/omissions were noted per prescription chart, most com-monly in the field related to weaning instructions. Docu-mentation with regards to weaning oxygen therapyoccurred in only 20%.Of those patients who had target saturations pre-

scribed, half were in the range of ‘88–92%’, and the

Figure 1 Excerpt from RPH Inpatient observation chart showing oxy-

gen saturation and oxygen delivery; S, Patient’s oxygen saturation while

on supplementation.

Table 1 Characteristics of patients receiving oxygen therapy(total n = 65)

Patient characteristicn (% out of patients

receiving oxygen therapy)

Male gender 40 (62.0)Background of• Chronic obstructive pulmonarydisease (COPD)

20 (30.8)

• Tuberculosis/sarcoidosis 4 (6.2)• Lung cancer 2 (3.1)• Asthma 12 (18.5)• Interstitial lung disease 2 (3.1)Background/risk of type 2 respiratoryfailure

21 (32.3)

On home oxygen prior to presentation 4 (6.2)Smoking status:• Not recorded 9 (13.9)• Non-smokers 22 (39.3)• Current smokers 18 (32.1)• Ex-smokers 16 (28.6)Respiratory condition as primarydiagnosis in current admission

29 (44.6)

Purely AMU admission (no otherspecialty as treating team)

26 (40.0)

AMU, Acute Medical Unit.

Figure 2 Completion of oxygen prescription chart in patients receiving

oxygen therapy (see Figure S1 (Supporting information) for fields con-

sidered mandatory). ( ), No chart used; ( ), chart filled, but with errors/

omissions; ( ), chart filled without errors.

Oxygen therapy practices in WA


153

other half were ‘>94%’. Analysis of clinical informationto determine the appropriate medically indicated target(IP) for each patient identified a risk of developing Type2 Respiratory Failure in 32.3%. Comparing prescribedtarget saturations (P) with medically indicated targets(IP) showed that 25% of prescriptions were inappropri-ate. Figure 3 illustrates the total number of prescriptionerrors causing potential risk to patients.

Comparing patient saturations achieved during oxy-gen therapy (S) with prescribed targets (P) showed that38.9% of patients received oxygen above the prescribedlevel (S > P). No patient had an oxygen saturation belowthe prescribed level (S < P). Figure 4 illustrates deliveryerrors causing potential risk to patients.

Of the patients, 49 (75.4%) were put at potential riskof harm from oxygen supplementation, identified asmissing or inappropriate oxygen prescriptions or incor-rect oxygen titration, as detailed above. In total,19 patients (29.2%) were exposed to a real risk of com-plications as a result of inappropriate oxygen therapy,

where oxygen saturations (S) deviated from the medi-cally indicated target (IP); 18 patients were over-oxy-genated, and 1 patient was under-oxygenated (Fig. 5).

The mean length of stay for all patients, includingpatients not requiring oxygen supplementation, admit-ted to AMU during the study period was 1.6 days. Theaverage admission duration of patients requiring oxygensupplementation (including patients transferred to otherspecialties) was 6.7 days; 40% of patients on oxygenwere discharged home from AMU without the transferof care to any other specialty. The average length of stayfor these patients on oxygen who were discharged homedirectly from AMU was 2.5 days. So, patients requiringoxygen supplementation appear to have a longerhospital stay.

Two patients were discharged home with a new pre-scription for home oxygen, and one was transferred toanother hospital still requiring oxygen therapy. Theremaining 62 patients returned to their baseline (pre-admission) oxygen status.

Figure 3 Prescription errors causing potential patient risk – Oxygen prescription (P) compared to appropriate medically indicated target (IP). ( ),

Potential risk.

Figure 4 Delivery errors causing potential patient risk – Patient’s oxygen saturation while on supplementation (S) compared to Prescribed target (P).

( ), Potential risk.

Kamran et al.


154

Discussion

Results show that the prescription, administration andmonitoring of acute oxygen therapy is suboptimal at ourinstitution, with a significant proportion of patientsplaced at potential risk of oxygen-related adverse events.The concept of an oxygen prescription chart has been

instituted in guidelines to emphasise the role of oxygen as

a formally prescribed drug and to create awareness of the

potential for oxygen toxicity. It is clear from the data pre-

sented that the use of oxygen prescription charts within

RPH needs improvement. In total, 55% of the patients

receiving oxygen therapy had prescribed target satura-

tions, and only 3.1% of them had fully completed oxygen

prescription charts. A 2005 audit showed that, at North

Shore Hospital, New Zealand, 2% of patients receiving

oxygen had it adequately prescribed ‘with respect to

guidelines’,12 while a 2013 audit in Waikato Hospital,

New Zealand showed a 51.7% correct prescription rate,7

which was defined as ‘containing flow rate, device and

appropriate oxygen saturation aims, considering the

patients’ medical history and risk of hypercapnia’. The

BTS has conducted audits of oxygen-prescribing practices

within the National Health Service (NHS) since the imple-

mentation of the 2008 guidelines. The most recently

available data from a 2013 audit demonstrated that 55%

of patients using oxygen had some form of written order,

an improvement from only 32% in 2008.13 Therefore,

international data suggest that the standards of practice at

RPH are generally similar to those in other countries.

There is a pressing need to audit regularly and improve

oxygen administration practices in hospitals worldwide.The low portion of correctly filled prescription charts

found in this study may be due to a variety of factors,including the number of fields required to be completed,time constraints and inadequate education relating to

correct prescribing practices. It was found that manydoctors completed the flow rate, delivery system, appli-cation and target saturation fields on the prescriptionchart but did not complete the weaning instructions. Theprescription chart appears to be unnecessarily compli-cated; thus, it is our recommendation that the prescrip-tion chart be simplified for ease of use. The field forweaning instructions may be omitted from revised pre-scription charts as administration should be titratedbased on oxygen requirements.The guidelines for appropriate oxygen administration

advise targeting oxygen saturations of 94–98% forpatients who are hypoxaemic, with the exception ofpatients who are at risk of type 2 respiratory failure andcertain types of poisoning.2 Patients at risk include thosewith prior hypercapnic respiratory failure, known severechronic obstructive pulmonary disease, morbid obesity,chest deformities or neuromuscular disorders. In theacutely unwell patient in the non-emergency setting,blood gases should be performed to assess the risk ofhypercapnic respiratory failure to aid oxygen titrationaccordingly. This is particularly true for patients withCOPD given that a proportion of these will not be at riskof type 2 respiratory failure. This is relevant to our studypopulation, where we found that doctors tended to bebiased towards under-oxygenation when prescribing.Two-thirds of the prescription errors were due to targetsaturations prescribed below appropriate medically indi-cated targets. This is likely due to insufficient researchinto patients’ backgrounds before prescribing, resultingin an inaccurate estimation of oxygen requirements. Inpractical terms, this could be due to the time constraintsfaced by prescribers working in a busy department com-bined with insufficient stress on the importance of accu-rate oxygen prescription. AMU’s nature as a unitdependent on shift work and rapid turnover makes it

Figure 5 Patients at real risk of oxygen-related adverse events – Patient’s oxygen saturation while on supplementation (S) compared to medically indi-

cated target (IP). ( ), Real risk.

Oxygen therapy practices in WA


155

hard for doctors to maintain continuity of care andobtain a deep knowledge of each patient’s background.

Although prescriptions often leaned towards under-oxygenation, the administration of oxygen leanedtowards oversupply. Oxygen is often not down-titratedor weaned to meet prescribed targets, resulting in over-oxygenation. This may be due to insufficient focus onthe prescription chart target while filling the observationchart. As there is no marker on the observation chart toalert staff to a filled prescription chart, it is likely that thisis easily overlooked. The counterbalancing nature ofthese errors explains why not all the patients found to beat potential risk from oxygen therapy in our study(75.4%) were at real risk.

The real risk to patients in this study was still signifi-cant; 18 patients (27.7%) were at risk of oxygen toxicity,and 1 patient was under-oxygenated and at risk of hyp-oxia. This risk profile illustrates a significant non-compliance with oxygen administration standards.

This study highlights several areas requiring improve-ment, mostly surrounding inadequate prescription andinaccurate titration of oxygen. As oxygen was frequentlycontinued without reassessment after being initiatedprior to admission, any quality improvement pro-grammes should also include the Emergency Depart-ment. Previous studies have shown that interventionsfocusing on visual reminders of oxygen guidelines, rede-sign of inpatient observation charts and nurse-initiatedoxygen prescription education have resulted in signifi-cant improvements in oxygen therapy practices.1,14,15

These encouraging results demonstrate the positivepotential impact of instituting oxygen-related interven-tions, which can be quantified in future re-audits.

There are clearly areas of limitation within this study.Data collection was retrospective in nature and relied on

accurate documentation by staff in patient notes andobservation charts. Lapses in documentation resulted inloss of data. Logistical problems related to file retrievalforced the exclusion of four study patients who mayhave received oxygen. Patients who were not adminis-tered oxygen were not further assessed to determinetheir oxygen requirements. The magnitude of variationsbetween prescribed, delivered and ideal oxygen satura-tions was not analysed in this study.

Data collection was mainly based in one medical wardand restricted to a short study period. However, as almostall medical patients are admitted through the AMU, wefeel the data provide valid information on acute oxygentherapy practices at the hospital and are likely to be repro-ducible in other Australian hospitals that use a similarprocess of admission. It is hoped that these data provide auseful basis for further research and intervention relatedto oxygen delivery practices in Australian hospitals.

Conclusion

The current practices of oxygen prescription and admin-istration within RPH Acute Medical Unit are suboptimal.Patients are placed at risk of oxygen toxicity due to devi-ation from oxygen prescription guidelines. Based on thedata from this study, recommendations have been madeto the executive team at Royal Perth Hospital to improveoxygen therapy practices, including simplification of theprescription chart, in order to reduce patients’ risk ofoxygen-related adverse events.

Acknowledgment

The authors thank Dr Roesia Miranda (Box Hill Hospital)for her contributions during the initial phase of the project.

References

1 Austin M, Wills K, Blizzard L, Walters E,

Wood-Baker R. Effect of high flow

oxygen on mortality in chronic

obstructive pulmonary disease patients

in prehospital setting: randomised

controlled trial. Br Med J 2010; 341:

c5462.

2 Snowball K. Guidelines for Acute Oxygen

Therapy for Western Australian Hospitals.

2012. WA Health Operational Directive

OD 0397/12.

3 Jackson RM. Pulmonary oxygen

toxicity. Chest 1985; 88: 900–905.

4 Thomson L, Paton J. Oxygen toxicity.

Paediatr Respir Rev 2014; 15: 120–123.

5 Beasley R, Aldington S, Robinson G. Is

it time to change the approach to

oxygen therapy in the breathless

patient? Thorax 2007; 62:

840–1.

6 Sjobery F, Singer M. The medical use of

oxygen: a time for critical reappraisal. J

Intern Med 2013; 274: 505–28.

7 O’Driscoll BR, Howard LS, Davison AG.

Guideline for emergency oxygen use in

adult patients. Thorax 2008; 63:

vi1–vi68.

8 Pease P. Oxygen administration: is

practice based on evidence? Paediat Nurs

2006; 18: 14–18.

9 Veduri A et al. Poor adherence to

guidelines for long-term oxygen therapy

(LTOT) in two Italian university hospitals.

Intern Emerg Med 2014; 9: 319–24.

10 Al-Mobeireek A, Abba A. An audit of

oxygen therapy on the medical ward in

2 different hospitals in Central Saudi

Arabia. Saudi Med J 2002; 23: 716–720.

11 Lamont T, Dagmar L, Scarpello J,

O’Driscoll R, Connew S. Improving the

safety of oxygen therapy in hospitals:

summary of a safety report from the

National Patient Safety Agency. Br Med

J 2010; 340: c187.

12 Boyle M, Wong J. Prescribing oxygen

therapy. An audit of oxygen prescribing

practices on medical wards at North

Shore Hospital, Auckland, New Zealand.

NZ Med J 2006; 119: U2080.

13 O’Driscoll BR. British Thoracic Society

Emergency Oxygen Audit 2013 (National

Audit Period 15 August–1 November 2013.)

2014 [cited 2016 Apr 1]. Available from

URL: https://www.brit-thoracic.org.uk/

document-library/audit-and-quality-

Kamran et al.


156

https://www.brit-thoracic.org.uk/document-library/audit-and-quality-improvement/audit-reports/bts-emergency-oxygen-audit-report-2013/


improvement/audit-reports/bts-

emergency-oxygen-audit-report-2013/

14 Medford A, Bowen J, Harvey J.

Improved oxygen prescription using a

nurse-facilitated reminder. Br J Nurs

2009; 18: 730–4.

15 Rudge J, Odedra S, Harrison D. A new

oxygen prescription produces real

improvements in therapeutic oxygen

use. BMJ Qual Improv Rep 2014; 3:

u204031.w1815.



Figure S1. RPH oxygen prescription chart; P, Prescribed target saturation level; 1–8, Fields required for a fully completechart.

doi:10.1111/imj.13679

Incidence of in-hospital and post-discharge diagnosedhospital-associated venous thromboembolism using linkedadministrative dataJoanne M. Stubbs ,1 Hassan Assareh,1 Jennifer Curnow,2 Kerry Hitos3,4 and Helen M. Achat1

1Epidemiology and Health Analytics, Western Sydney Local Health District, 2Department of Haematology, and 3Westmead Research Centre forEvaluation of Surgical Outcomes, Westmead Hospital, Westmead, and 4Discipline of Surgery, The University of Sydney, Sydney, New South Wales,Australia

Key wordsvenous thromboembolism, incidence, diagnosis,risk factors, complications.

CorrespondenceJoanne Stubbs, Epidemiology and HealthAnalytics, Western Sydney Local Health District,Locked Bag 7118, Parramatta BC, NSW 2124,Australia.Email: [email protected]

Received 13 March 2017; accepted9 November 2017.

Abstract

Background: Hospital-associated venous thromboembolism (HA-VTE) is a serious

adverse event, preventable with appropriate care during and post-admission. Accurate

measurement of in-hospital and post-discharge incidences is essential for implementa-

tion and evaluation of prevention strategies and monitoring.

Aims: To estimate in-hospital and post-discharge diagnosed VTE, trends and risk factors.

Methods: This was a population-based study in New South Wales, Australia, using

linked hospital admission and emergency department data for 2010–2013 of adult

patients with a minimum stay of 48 h. HA-VTE were diagnosed in-hospital or post-

discharge (within 90 days). Multi-level modelling schemes produced adjusted rates and

ratios for patient, admission and hospital-related characteristics.

Results: From 1 865 059 admissions, the HA-VTE incidence rate was 9.7 per 1000

admissions; 71% were diagnosed post-discharge, and 4.3% died with a greater risk for

VTE diagnosed in hospital compared to post-discharge (8.4% vs 2.6%, P < 0.001). Com-

pared with surgical patients, medical patients developed fewer HA-VTE (IRR = 0.60, 95%

CI: 0.58–0.63) but were more likely to be diagnosed post-discharge (OR = 2.19; 95% CI:

2.00–2.40). HA-VTE increased 6.5% over the period, driven by the 44% increase in in-

hospital diagnoses and not by the 9% decrease in post-discharge diagnoses.

Conclusions: HA-VTE is a continuing burden, and diagnosis after recent hospital dis-

charge is notably high. Incidence varies across patients and facilities, highlighting the

need for individual VTE risk assessment. Inclusive measures and routine monitoring of

HA-VTE incidence and mortality are essential for implementing best practice and asses-

sing effectiveness of prevention strategies.


Internal Medicine Journal 48 (2018) 157–165© 2017 Commonwealth of Australia. Internal Medicine Journal © 2017 Royal Australasian College of Physicians

157



Introduction

Venous thromboembolism (VTE) is a major, potentiallypreventable cause of morbidity and mortality.1 On a cost-per-case basis, VTE is the second ranked health conditionin Australia, but including the cost of premature mortal-ity elevates it to first place.2 Hospital admission is a majorrisk factor for VTE and accounts for almost 50% of theattributable risk.3 Many predisposing factors, includingprolonged immobilisation, malignancy, major surgery,multiple trauma and chronic heart failure,4 commonlyoccur among hospital patients. Not all VTE associatedwith hospitalisation either occur or are diagnosed duringthe hospital stay,5–9 and consequently, hospital-associ-ated VTE (HA-VTE) underreporting is common.10 VTErisk persists after surgery and beyond discharge of medi-cal patients, with most events developing within90 days.11 The term HA-VTE encompasses VTE acquiredand diagnosed during the index admission or followingdischarge,12 reflecting this association between current orrecent hospitalisation and the development of VTE. Dataon readmission rates are also lacking.10 Thus, HA-VTErates calculated only on VTE cases diagnosed during ahospital stay may underestimate the true prevalence,and burden, of HA-VTE.10,13–15 Enhanced monitoringand reporting of VTE incidents, together with awarenessof its preventable nature, are key to improving theuptake of VTE risk assessment and prophylaxis.10,15

Compared to overseas studies,7,9,14,16,17 few Australianstudies have examined the incidence of HA-VTE, andmost have relied on inpatient data.2,13,15,18 This studysupplemented inpatient data with emergency depart-ment (ED) presentations across New South Wales (NSW)facilities to estimate more accurately the incidence ofHA-VTE in NSW, including VTE diagnosed during theindex admission and after discharge. Factors associatedwith the increased likelihood of a post-discharge diagno-sis and the time of VTE diagnosis were also assessed.

Methods

Data source

Two data sources capture all admissions to NSW hospitalsand presentations to contributing public ED in NSW forpersons 18 years or over who were discharged between1 July 2010 and 30 June 2014 (2010–2013 financialyears). Linked records from the Admitted Patients DataCollection (APDC; 11 354 936 records) and EmergencyDepartment Data Collection (EDDC; 9 273 285 records)were obtained from the NSW Admitted Patient, Emer-gency Department Attendance and Deaths Register,established under the public health and disease registers

provisions of the NSW Public Health Act 2010 and main-tained by the NSW Ministry of Health. Record linkagewas carried out by The Centre for Health Record Linkage(CHeReL). Use of the data for this study was approved bythe data provider. The APDC collects information for alladmitted patient services in NSW hospitals. Principal diag-nosis and comorbidities are coded using the ICD-10,Australian Modification (ICD-10-AM).19 The accompany-ing ‘Condition Onset Flag’ variable indicates whethereach condition arose during the episode of care or waspresent on admission.20 The EDDC collects informationabout presentations to contributing NSW public ED. EDdiagnosis may be coded using either ICD-10-AM, ICD-9-CM or Systematised Nomenclature of Medicine-ClinicalTerms (SNOMED-CT).21

VTE identification

VTE was defined by the ICD-10-AM codes for pulmo-nary embolism (PE) – I26.0, I26.8, I26.9, O88.2 – anddeep vein thrombosis (DVT) – I80.1, I80.2, I80.3, I80.8,I80.9, I82.2, I82.8, I82.9, O22.3 and O87.1 – based onrecognised quality and safety indicators.22,23 Corre-sponding ICD-9-CM and SNOMED-CT codes weredetermined using mapping tables provided by theAustralian Consortium for Classification Develop-ment24 and the National eHealth Transition Authority(personal communication).

We identified HA-VTE and examined its associationwith an index admission, that is, an admission to an NSWacute public hospital (n = 82), with a minimum length ofstay (LOS) of 48 h (Fig. 1). HA-VTE with condition onsetand diagnosis during the index admission were classifiedas in-hospital diagnosed VTE (IH-VTE). HA-VTE presenton admission or ED presentation to any NSW public orprivate hospital within 90 days3,11,15 of an index admis-sion were classified as post-discharge diagnosed VTE (PD-VTE). Index admission details of these VTE were ana-lysed. Any VTE diagnosed within 90 days of a previouslydiagnosed VTE (capturing readmissions, transfers, caretype changes) were excluded. Isolated HA-VTE was thusidentified and differentiated from community-acquiredVTE (VTE present on admission, no hospital admissionwithin the previous 90 days15 with minimum 48 h LOS).We reported on public hospital incidence only due tolower data availability and quality in private hospitals.25

For VTE unclassified due to missing condition onsetdata, logistic models to predict onset status were devel-oped based on patient and admission characteristics fromcomplete data. Public and private hospitals were mod-elled separately. Models with minimum Akaike Informa-tion Criterion26 values and maximum area under areceiver operating characteristic curve27 (0.89 and 0.98

Stubbs et al.


158

with 11 and 7 predictors for public and private hospitals,respectively) predicted the onset status of 3824 VTE(7.6% of all isolated VTE) and subsequently identified411 IH-VTE (10.7%), 1440 PD-VTE (37.7%) and 1973community-acquired VTE (51.6%).We examined data for 3 months before and after our

study period to resolve time-censoring effects. Dischargestatus (alive/dead) at the last VTE-related hospitalisationof each isolated VTE determined case fatality. Cause ofdeath information was not available.

Description of covariates

Patient characteristics included gender, age and coun-try of birth. Admission characteristics included admis-sion urgency, admission type based on The AustralianClassification of Health Interventions28 procedurecodes, surgery type, the Charlson Comorbidity Index(derived from ICD-10-AM codes for the principal diag-nosis and comorbidities as defined by Quan et al.),29,30

major principal diagnosis disease groups based on theCharlson Comorbidity Index categories among VTEcases, LOS and year of separation (grouped by finan-cial year, reported as year at 1 July). Hospital charac-teristics consisted of health district location and peergroup. Peer groups comprised hospitals of a similartype and size.31

Statistical analysis

Analyses were based on the index admission. Mixedmodelling schemes were used to derive adjusted rates foroutcomes and ratios for all patient, admission and hospitalcovariates while controlling for hospital clustering effects.Negative binomial mixed models were employed for

HA-VTE and associated case fatality rates. A logisticmixed model assessed the likelihood of in-hospital ver-sus post-discharge diagnosed HA-VTE. Yearly adjustedincidence and fatality rates were derived by multiplyingyearly adjusted rate ratios and the crude rates observedin the reference year (2010). IH-VTE and PD-VTEannual adjusted trends were obtained using an odds-to-rate ratio conversion method.32

Data preparation was conducted using SAS EnterpriseGuide V.6.133 through SAPHaRI,34 and analyses wereperformed using R package V.3.1.2.35

Results

Hospital-associated VTE

Of the 1 865 059 (13.3% of all APDC records) eligibleadmissions between 1 July 2010 and 30 June 2014, with

a median LOS of 5 days (1st and 3rd quartiles: 3 and9 days, respectively), 18 171 isolated HA-VTE were iden-tified, yielding an incidence rate of 9.7 per 1000 admis-sions (median LOS = 8 days, quartiles: 5 and 16 days).Patients were diagnosed with one (n = 16 407) or multi-ple isolated HA-VTE; 834 patients contributed 1764 cases(9.7% of all HA-VTE). DVT is more common (11 378;62.6%) than PE. There were 7693 HA-VTE identifiedand excluded as repeat presentations of an HA-VTE diag-nosed within the previous 90 days. An additional 32 058diagnosed VTE were classified as communityacquired (Fig. 1).The majority of HA-VTE were diagnosed post-

discharge (12 837; 70.6%). In-hospital diagnosed HA-VTE had a longer median LOS (18 days, quartiles:11 and 33) than PD-VTE (6 days, quartiles: 4 and 11).Excluding PD-VTE involving a hospital transfer or caretype change (n = 2907), median time between indexadmission discharge and subsequent diagnosis admissionwas 17 days (quartiles: 6 and 40 days).The risk of HA-VTE is shown in Table S1 (Supporting

information). Males had 9% (IRR = 0.91) lower risk ofHA-VTE than females; risk increased with age. Asian-born patients had 36% less risk of developing an HA-VTE than patients born in Australia or New Zealand.Non-emergency admissions had 17% less risk of HA-

VTE than emergency admissions. Medical and obstetricadmissions were associated with markedly lower rates ofHA-VTE compared to surgical admissions (40% and77% lower risk respectively). Risk varied according tosurgery type: compared to cholecystectomy, patientsundergoing orthopaedic procedures had 1.79–3.6 timesthe risk of HA-VTE. Patients with cancer were 1.5 timesas likely to develop HA-VTE as patients with vasculardisease. Conversely, patients with cardiac or chronic pul-monary disease had 17% and 14% lower risk respec-tively. Higher Charlson Comorbidity Index scores andlonger hospital LOS were associated with increased riskof HA-VTE. Patients with a Charlson score of 5 or morewere 1.6 times as likely to develop HA-VTE as patientsscoring 0 or 1. A hospital stay of 10 or more daysincreased the risk of HA-VTE by 136% compared topatients with maximum LOS of 4 days. Large principalreferral hospitals and hospitals in metropolitan areas hada higher likelihood of HA-VTE than smaller hospitalsand those located in rural areas.

Case fatality

There were 775 in-hospital deaths of patients with HA-VTE (case fatality 4.3%) during the study period;413 (53.3%) were PE, and 447 (57.7%) occurred withinthe index admission. Case fatality was higher for PE than

HA-VTE incidence from linked data


159

DVT (6.1% vs 3.2%, P < 0.001) and for in-hospital com-pared to PD-VTE (8.4% vs 2.6%, P < 0.001).

Time trends

The adjusted incidence rate of HA-VTE increased by11% from 9.2 in 2010 to 10.2 in 2012, with a non-

significant decline to 9.8 in 2013 (6.5% increase from2010) (Fig. 2). The adjusted rate for IH-VTE alsoincreased (44%) throughout the study period. Most ofthe increase (20%) occurred between 2012 and 2013,mirrored by a marked drop (1 per 1000 patients; 14%)in the adjusted rate of PD-VTE, resulting in a 9% overalldecline (Fig. 2). PE as a proportion of VTE decreased

Admission at risk of VTE

Admission with VTE

All NSW acute hospital adult admissions and

Emergency presentations2010 – 2013

Emergency presentation

with VTE

Admission with LOS 48hrs+

Admission excluded

Condition present on admission

In-hospital diagnosedHA-VTE

Index admission had VTE

Within 90

VTE identified as community acquired; excluded

days of an (index)

admission

Presentation excluded

Post-discharge diagnosed HA-VTE;

index admission identified & included

Index admission had LOS 48hrs+

NoYes

No

Yes

Yes

No

Yes

No No Yes

No

Admission with LOS 48hrs+

Yes

No

VTE a re-presentation of

index case; excluded

No

Yes

Yes

Figure 1 Hospital-associated VTE case identification.

Stubbs et al.


160

from 39.3% to 35.7% across the study period (Table S1,Supporting information). Adjusted hospital case fatalityremained stable at 4.0%.

In-hospital versus PD-VTE

Patient, case mix and hospital factors affected the odds ofa post-discharge diagnosis (Table S1, Supporting infor-mation). The likelihood of PD-VTE increased with age,being a medical patient (OR = 2.19), coronary arterybypass graft (OR = 2.34) or hip replacement surgery(OR = 1.61), cancer (OR = 1.79) and chronic pulmonarydisease (OR = 2.07) and admission to smaller districthospitals (OR = 2.32) or hospitals in rural and regionalareas (OR = 1.60). Planned admissions (OR = 0.86), kneereplacement surgery (OR = 0.56), mid-range Charlsonscores (OR = 0.86), increased hospital LOS (OR = 0.05)and more recent year of admission (OR = 0.69) wereassociated with reduced odds of PD-VTE.

Discussion

This analysis of almost two million admissions to NSWacute public hospitals over 4 years identified 18 171 iso-lated HA-VTE events, an incidence rate of 9.7 per 1000admissions. HA-VTE incidence amongst surgical admis-sions was almost twice that of medical admissions and

varied across surgical types and diagnoses groups. Lessthan one-third of HA-VTE was diagnosed during theindex admission. The overall HA-VTE case fatality ratewas 4.3% and was more likely for patients with in-hospital compared to PD-VTE (8.4% vs 2.6%).Our findings of a persistent risk of VTE following dis-

charge and HA-VTE most commonly diagnosed post-discharge (59% of surgical, 83% of medical, 71% of allpatients) are consistent with previous research fromAustralia15 and overseas.7,9,14,17 Our higher proportionsof HA-VTE diagnosed within 3 months of a previousadmission (compared to 54%14 of medical patients and53%15 of all patients) may reflect patient and methodo-logical differences. Differences in measuring time todiagnosis and case mix may have contributed to ourshorter delay to PD-VTE identification (17 days) com-pared to some (51 days),7 but not all, (16 days)8 reports.Capturing VTE diagnosed post-discharge expanded

our study population to include patients with anincreased likelihood of later diagnosis, particularly medi-cal, compared to surgical, admissions and patientsattending smaller hospitals, compared to principal refer-ral hospitals. Consequently, our estimates of HA-VTEwithin the Australian context are more accurate thanprevious reports,13,18 which – based only on in-hospitaldiagnosed VTE – yielded an incidence rate of 11.5 per1000 admissions, and update other results, which are

Figure 2 Hospital-associated VTE

incidence rate and associated

case fatality in NSW public hospi-

tals. ( ), Adjusted HA-VTE inci-

dence rate; ( ), Crude HA-VTE

incidence rate; ( ), Adjusted

in-hospital diagnosed HA-VTE inci-

dence rate; ( ), Crude in-hos-

pital diagnosed HA-VTE incidence

rate; ( ), Adjusted post--

discharge diagnosed HA-VTE

incidence rate; ( ), Crude

post-discharge diagnosed HA-VTE

incidence rate; ( ), Adjusted

HA-VTE case fatality; ( ),

Crude HA-VTE case fatality.



161

more than 10 years out of date (2–3 per 1000 admis-sions).15 The marked difference in the incidence rates formedical patients (8.1 vs 4.5), compared to the smallerdifference for surgical patients (15.5 vs 12.9), betweenour and previous studies13 highlights the impact ofincluding post-discharge VTE. The high incidence of VTEdiagnosed after hospital discharge is cause for concern.While not all HA-VTE will become clinically apparentduring an admission, especially with the trend towardsshortened LOS, our results support the urgent call formore individualised VTE risk assessment, appropriatethromboprophylaxis and an extension of thrombopro-phylaxis into the post-discharge period.36–38

There was no evidence of a decrease in HA-VTE duringthe study period despite national- and state-wide guide-lines and policy directives to increase awareness and pre-vention of HA-VTE through routine risk assessment andappropriate prophylaxis.39,40 More recently, NSW Health’sClinical Excellence Commission established a state-wideVTE Prevention Program,11 including updated strategies,41

resources, support and advice, to assist clinicians and facil-ities to implement VTE risk assessment and prophylaxisinto the clinical workflow and ultimately reduce HA-VTEin NSW hospitals. This study provides a baseline for pro-gramme evaluation against which anticipated reductionsin incidence can be measured.

Insufficient dosage or duration of prophylaxis mayincrease rates of both IH-VTE and PD-VTE. Quantifyingthe risk of post-discharge diagnosis reveals those patient,admission and hospital factors where greater vigilanceand attention to assessing patient risk and prescribingappropriate thromboprophylaxis may be warranted. Thetraditional view that VTE is primarily a complication ofsurgery has been discounted by research documentingthe high incidence of VTE amongst medicalpatients3,5,9,13,14 and the underutilisation of appropriateprophylactic modalities in both the hospital and post-discharge setting amongst this group.3,5,9 The elevatedodds of PD-VTE amongst medical, compared to surgical,patients in our study suggests that risk stratification forPD-VTE and extended prophylaxis among medicalpatients is advantageous. However, medications are notrecorded in the administrative data analysed in ourstudy, precluding examination of prophylaxis and itsinfluence on HA-VTE. Investigation into the appropriateduration of prophylaxis is warranted.

Encouragingly, the adjusted incidence rate for post-discharge VTE decreased in the final year of the study,with a corresponding increase in VTE identified duringthe index admission. Increased awareness of VTE as apreventable adverse event may have had an initial effectin increasing the number of VTE identified while thepatient was still in hospital.2 The crude trend towards a

decreasing proportion of PE observed in our study mayimply earlier diagnosis. Awareness is integral to behav-iour change and adoption of guidelines42 but needs to betranslated into action. The NSW Clinical ExcellenceCommission’s initiatives11 are aimed at facilitatingbehaviour change. Reassessment of our measures andprocess parameters following full implementation ofthese initiatives will reveal whether the desired out-comes have been achieved.

Our higher HA-VTE incidence in large principalreferral hospitals may relate to the severity and com-plexity of patient caseload, which was not fully cap-tured by our case mix.13 In contrast, an inability todetect asymptomatic VTE in smaller district hospitals(mostly located in rural areas) due to limited healthresources, in combination with the characteristic opti-mism and stoicism of rural Australians, which contrib-utes to delayed help-seeking behaviour,43 may havecontributed to the elevated risk of a PD-VTE diagnosisand resulted in a larger proportion of PE. PE, as indi-cated by higher case fatalities in smaller hospitals,13 isassociated with poorer outcomes. Clinical practiceguidelines promote evidence-based practice aimed atovercoming unwarranted variation and improvingpatient outcomes,39,44 but their success is dependenton implementation and uptake.

Post-discharge VTE can be difficult to measure becausepatients may not re-present to the original hospital; maybe treated solely as an outpatient, including by their localdoctor or may not seek medical attention.12 Many stud-ies investigating PD-VTE have restricted their sample to aspecific group of patients.5–7,12,14,17,37 Our study includedall acute public hospital patients and identified allpatients readmitted to a NSW hospital or presenting toED following index admission. Diagnoses onset statusonly applies to the current admission and is prone toincompleteness or biases15,45,46; it is not completely accu-rate when used alone to estimate HA-VTE.13,15,18,45,46 Weused classification methods to enhance onset data com-pleteness, and longitudinal follow up of patients enabledidentification of HA-VTE versus community-acquiredVTE or consecutive re-presentations of the same VTE.Identifying isolated VTE events avoided double countingrepeated VTE presentations. Accordingly, our rates ofboth hospital-associated and post-discharge VTE havegreater validity and more accurately depict the extent ofthis problem, filling the gap in prevalence estimatesbased on administrative datasets.10 Although supportedby literature,9,11,14,15 our decision to include all VTE diag-nosed within 90 days of discharge as PD-VTE may havemissed some late onset HA-VTE but also potentiallyoverestimated PD-VTE by incorrectly including somecommunity-acquired VTE.

Stubbs et al.


162

Reporting on HA-VTE in private settings providesadditional insight, although lower quality in theiradministrative data requires more caution.25 Our post-discharge diagnosis rate, based on hospital data, islikely a conservative estimate as we were unable toidentify patients diagnosed and treated solely at outpa-tient clinics or by their local doctor or presenting tonon-contributing ED. Current methodologies used torecord this information in NSW are disparate and arenot amenable to linkage with hospital data,10 makingit difficult to accrue these data. In the United Kingdom,linked primary care and hospital administrative dataidentified an additional 38% of post-surgical VTE.16

Differences in healthcare systems and study methodol-ogies prevent applying this rate directly to our resultsbut suggest a noteworthy underestimation of HA-VTE.The inability to capture the small proportion ofpatients who may have re-presented to hospitals instates bordering NSW also potentially underestimatesthe post-discharge rate. Case mix variables, such assmoking, obesity, history of VTE, risk assessment andprophylaxis, were not included in our modelling.

Future analyses incorporating such details might pro-vide insight into influences on the onset and diagnosisof HA-VTE.

Conclusion

HA-VTE incidence increased over time and varied acrosspatients and facilities. Recognition of variation in riskadvances efforts to improve targeted risk assessment anddiagnosis. The significantly elevated incidence of post-discharge HA-VTE suggests the need for improvedprevention strategies. Adopting inclusive measures androutine monitoring and reporting of HA-VTE incidenceand associated mortality are essential to the implementa-tion of best practice and assessment of effectiveness.

Acknowledgement

Record linkage was carried out by the Centre for HealthRecord Linkage (www.CHeReL.org.au).

References

1 Geerts WH, Bergqvist D, Pineo GF,

Heit JA, Samama CM, Lassen MR et al.

Prevention of venous thromboembolism:

American College of Chest Physicians

Evidence-Based Clinical Practice

Guidelines. Chest 2008; 133: 381S–453S.

2 Access Economics. The burden of

venous thromboembolism in Australia.

Report for the Australia and

New Zealand working party on the

management and prevention of venous

thromboembolism. 2008 [cited 2015

Oct 15]. Available from URL: www.

accesseconomics.com.au

3 Heit JA, O’Fallon WM, Petterson TM,

Lohse CM, Silverstein MD, Mohr DN

et al. Relative impact of risk factors for

deep vein thrombosis and pulmonary

embolism: a population-based study.

Arch Intern Med 2002; 162: 1245–8.

4 Anderson FA Jr, Spencer FA. Risk

factors for venous thromboembolism.

Circulation 2003; 107: I-9–I-16.

5 Amin AN, Lin J, Thompson S,

Wiederkehr D. Real-world rates of in-

hospital and postdischarge deep-vein

thrombosis and pulmonary embolism in

at-risk medical patients in the United

States. Clin Appl Thromb Hemost 2011;

17: 611–19.

6 Amin AN, Lin J, Thompson S,

Wiederkehr D. Inpatient and outpatient

occurrence of deep vein thrombosis

and pulmonary embolism and

thromboprophylaxis following selected

at-risk surgeries. Ann Pharmacother

2011; 45: 1045–52.

7 Spyropoulos AC, Hussein M, Lin J,

Battleman D. Rates of symptomatic

venous thromboembolism in US

surgical patients: a retrospective

administrative database study. J Thromb

Thrombolysis 2009; 28: 458–64.

8 Spyropoulos AC, Anderson JFA,

FitzGerald G, Decousus H, Pini M,

Chong BH et al. Predictive and

associative models to identify

hospitalized medical patients at risk for

VTE. Chest 2011; 140: 706–14.

9 Spencer FA, Lessard D, Emery C, Reed G,

Goldberg RJ. Venous thromboembolism

in the outpatient setting. Arch Intern Med

2007; 167: 1471–5.

10 Clinical Excellence Commission. Clinical

focus report: hospital-associated venous

thromboembolism. Sydney: Clinical

Excellence Commission; 2015; [cited

22 Feb 2016]. Available from: http://

www.cec.health.nsw.gov.au/__data/

assets/pdf_file/0008/278612/Clinical-

Focus-Report-Hospital-Associated-

Venous-Thromboembolism.pdf.

11 Clinical ExcellenceCommission.

Medication Safety andQuality: Venous

ThromboembolismPrevention [webpage].

Sydney: Clinical ExcellenceCommission;

[cited 2016 Feb 17]. Available fromURL:

http://www.cec.health.nsw.gov.au/

programs/vte-prevention

12 Khanna R, Maynard G, Sadeghi B,

Hensley L, Medvedev S, Danielsen B

et al. Incidence of hospital-acquired

venous thromboembolic codes in

medical patients hospitalized in

academic medical centers. J Hosp Med

2014; 9: 221–5.

13 Assareh H, Chen J, Ou L, Hillman K,

Flabouris A. Incidences and variations

of hospital acquired venous

thromboembolism in Australian

hospitals: a population-based study.

BMC Health Serv Res 2016; 16: 511.

14 Spyropoulos AC, HusseinM, Lin J,

Battleman D. Rates of venous

thromboembolism occurrence in medical

patients among the insured population.

Thromb Haemost 2009; 102: 951–7.

15 National Institute of Clinical Studies.

The incidence and risk factors for

venous thromboembolism in hospitals

in Western Australia 1999–2001.

Prepared by The School of Population

Health, University of Western Australia.

Melbourne: NICS; 2005 [cited 2015 Oct

7]. Available from URL: http://www.

nhmrc.gov.au/_files_nhmrc/file/nics/

material_resources/The%20incidence%

20of%20venous%20thromboembolism

%20in%20Western%20Australian%

20hospitals.pdf



163

http://www.cherel.org.au

http://www.accesseconomics.com.au

http://www.accesseconomics.com.au

http://www.cec.health.nsw.gov.au/__data/assets/pdf_file/0008/278612/Clinical-Focus-Report-Hospital-Associated-Venous-Thromboembolism.pdf





http://www.cec.health.nsw.gov.au/programs/vte-prevention

http://www.cec.health.nsw.gov.au/programs/vte-prevention

http://www.nhmrc.gov.au/_files_nhmrc/file/nics/material_resources/The%20incidence%20of%20venous%20thromboembolism%20in%20Western%20Australian%20hospitals.pdf






16 Bouras G, Burns EM, Howell A-M,

Bottle A, Athanasiou T, Darzi A. Risk of

post-discharge venous

thromboembolism and associated

mortality in general surgery: a

population-based cohort study using

linked hospital and primary care data in

England. PLoS One 2016; 10: e0145759.

17 Imberti D, Bianchi C, Zambon A, Parodi A,

Merlino L, Gallerani M et al. Venous

thromboembolism after major orthopaedic

surgery: a population-based cohort study.

Intern Emerg Med 2012; 7: 243–9.

18 Assareh H, Chen J, Ou L, Hollis SJ,

Hillman K, Flabouris A. Rate of venous

thromboembolism among surgical

patients in Australian hospitals: a

multicentre retrospective cohort study.

BMJ Open 2014; 4: e005502.

19 National Centre for Classification in

Health (NCCH). The International

Statistical Classification of Diseases and

Related Health Problems, Tenth Revision,

Australian Modification (ICD-10-AM).

Sydney: NCCH, Faculty of

Health Sciences, The University of

Sydney; 2004.


Welfare (AIHW). METeOR, Metadata

Online Registry [webpage]. AIHW,

Australian Government; [cited 2015

Nov 15]. Available from URL: http://

meteor.aihw.gov.au/content/index.

phtml/itemId/181162

21 The International Health Terminology

Standards Development Organisation

(IHTSDO). SNOMED CT [webpage].

IHTSDO; [cited 2015 Nov 15]. Available

from URL: http://www.ihtsdo.org/

22 The Organisation for Economic Co-

operation and Development (OECD).

Definitions for Health Care Quality

Indicators: 2012-2013. HCQI Data

Collection; 2013.

23 Victorian Government Health

Information. Patient Safety Indicators-

Translated Technical Specifications.

Melbourne: Victorian State

Government DoH; 2006.

24 Australian Consortium for Classification

Development. Mapping Tables (ICD-9-

CM/ICD-10-AM/ACHI) Backward and

Forward Maps [cited 2015 Sep 14].

Available from URL: https://www.accd.

net.au/Downloads.aspx?type=Archive

25 Assareh H, Achat HM, Stubbs JM,

Guevarra VM, Hill K. Incidence and

variation of discrepancies in recording

chronic conditions in Australian

hospital administrative data. PLoS One

2016; 11: e0147087.

26 Akaike H. A new look at the statistical

model identification. IEEE Trans Autom

Control 1974; 19: 716–23.

27 Hanley JA, McNeil BJ. The meaning

and use of the area under a receiver

operating characteristic (ROC) curve.

Radiology 1982; 143: 29–36.

28 Australian Consortium for Classification

Development (ACCD). Australian

Classification of Health Interventions

(ACHI), Tabular List of Interventions.

Sydney: Independent Hospital Pricing

Authority; 2015.

29 Charlson ME, Pompei P, Ales KL,

MacKenzie CR. A new method of

classifying prognostic comorbidity in

longitudinal studies: development and

validation. J Chronic Dis 1987; 40:

373–83.

30 Quan H, Sundararajan V, Halfon P,

Fong A, Burnand B, Luthi JC et al.

Coding algorithms for defining

comorbidities in ICD-9-CM and ICD-10

administrative data. Med Care Res Rev

2005; 43: 1130–9.

31 Demand and Performance Evaluation.

NSW Health Services Comparison Data Book

2008/2009. NSW Health: Sydney; 2010.

32 Grant RL. Converting an odds ratio to a

range of plausible relative risks for

better communication of research

findings. BMJ 2014; 348: f7450.

33 SAS Institute. SAS Enterprise Guide.

Cary: SAS Institute; 2013.

34 Centre for Epidemiology and

Evidence. Secure Analytics for

Population Health Research and

Intelligence (SAPHaRI). Sydney: NSW

Ministry of Health; 2014 [cited 2014

Nov 20]. Available from URL: http://

www.health.nsw.gov.au/

epidemiology/Pages/saphari.aspx

35 R Core Team. R: A Language and

Environment for Statistical Computing.

Vienna: The R Foundation for Statistical

Computing; 2013.

36 Friedman RJ. Optimal duration of

prophylaxis for venous

thromboembolism following total hip

arthroplasty and total knee

arthroplasty. J Am Acad Orthop Surg

2007; 15: 148–55.

37 Fleming FJ, Kim MJ, Salloum RM,

Young KC, Monson JR. How much do

we need to worry about venous

thromboembolism after hospital

discharge? A study of colorectal surgery

patients using the National Surgical

Quality Improvement Program database.

Dis Colon Rectum 2010; 53: 1355–60.

38 Goldhaber SZ. Eradication of hospital-

acquired venous thromboembolism.

Thromb Haemost 2010; 104: 1089–92.

39 National Health andMedical Research

Council. Clinical Practice Guideline for the

Prevention of Venous Thromboembolism

(Deep Vein Thrombosis and Pulmonary

Embolism) in Patients Admitted to Australian

Hospitals. Melbourne: National Health

andMedical Research Council; 2009.

40 NSW Ministry of Health. Policy

Directive. Prevention of Venous

Thromboembolism (PD2010_077).

Sydney: NSW Ministry of Health; 2010

[cited 2016 Feb 22]. Available from

URL: http://www0.health.nsw.gov.au/

policies/pd/2010/pdf/PD2010_077.pdf

41 NSW Ministry of Health. Policy

Directive. Prevention of Venous

Thromboembolism (PD2014_032).

Sydney: NSW Ministry of Health; 2014

[cited 2016 Feb 22]. Available from

URL: http://www0.health.nsw.gov.au/

policies/pd/2014/pdf/PD2014_032.pdf

42 Cabana MD, Rand CS, Powe NR,

Wu AW, Wilson MH, Abboud P-AC

et al. Why don’t physicians follow

clinical practice guidelines? A

framework for improvement. JAMA

1999; 282: 1458–65.

43 Emery JD, Walter FM, Gray V,

Sinclair C, Howting D, Bulsara M et al.

Diagnosing cancer in the bush: a

mixed-methods study of symptom

appraisal and help-seeking behaviour

in people with cancer from rural

Western Australia. Fam Pract 2013; 30:

294–301.

44 National Institute for Health and Care

Excellence (NICE). Venous

thromboembolism: reducing the risk for

patients in hospital. Clinical guideline

[CG92]. NICE; 2015 [cited 28 February

2017]. Available from URL: https://

www.nice.org.uk/guidance/cg92

45 Jackson TJ, Michel JL, Roberts R,

Shepheard J, Cheng D, Rust J et al.

Development of a validation algorithm

for ’present on admission’ flagging. BMC

Med Inform Decis Mak 2009; 9: 1–8.

46 Khanna R, Kim SB, Jenkins I, El-

Kareh R, Afsarmanesh N, Amin AN

et al. Predictive value of the present-on-

admission indicator for hospital-

acquired venous thromboembolism.

Med Care 2015; 53: e31–6.

Stubbs et al.


164




http://www.ihtsdo.org

https://www.accd.net.au/Downloads.aspx?type=Archive

https://www.accd.net.au/Downloads.aspx?type=Archive

http://www.health.nsw.gov.au/epidemiology/Pages/saphari.aspx



http://www0.health.nsw.gov.au/policies/pd/2010/pdf/PD2010_077.pdf




https://www.nice.org.uk/guidance/cg92

https://www.nice.org.uk/guidance/cg92



Table S1. Incidence and time of diagnosis of HA-VTE in 82 NSW public hospitals, 2010–2013

doi:10.1111/imj.13554

Off-label use of rituximab in autoimmune disease in the Top Endof the Northern Territory, 2008–2016Sarah Wongseelashote ,1 Vipin Tayal2 and Peter Francis Bourke2,3

1General and Acute Care Medicine, and 2Division of Medicine (Rheumatology), Royal Darwin Hospital, Tiwi, Northern Territory and 3Division ofIntegrated Medicine and Emergency Services (Clinical Immunology), Cairns Hospital, Cairns, Queensland, Australia

Key wordsrituximab, autoimmune diseases, systemic lupuserythematosus, off-label, Aboriginal Australian.

CorrespondencePeter Bourke, Division of Integrated Medicineand Emergency Services, Cairns Hospital,PO Box 902, Cairns, Qld 4870, Australia.Email: [email protected]

Received 7 April 2017; accepted 18 July 2017.

Abstract

Background: Rituximab, an anti-CD20 B-cell depleting monoclonal antibody, is

increasingly prescribed off-label for a range of autoimmune diseases. There has not pre-

viously been an audit of off-label rituximab use in the Northern Territory, where the

majority of patients are Aboriginal.

Aims: To evaluate retrospectively off-label rituximab use in autoimmune diseases in

the Top End of the Northern Territory.

Methods: We performed a retrospective audit of 8 years of off-label rituximab use at

the Royal Darwin Hospital, the sole tertiary referral centre for the Darwin, Katherine

and East Arnhem regions. Electronic and paper records were reviewed for demographic

information, diagnosis/indication for rituximab, doses, previous/concomitant immuno-

suppression, clinical outcomes and specific adverse events.

Results: Rituximab was prescribed off-label to 66 patients for 24 autoimmune diseases.

The majority of patients (62.1%) were Aboriginal and 60.6% female. The most com-

mon indications were refractory/relapsing disease despite standard therapies (68.7%)

or severe disease with rituximab incorporated into an induction immunosuppressive

regimen (19.4%). Systemic lupus erythematosus was the underlying diagnosis in

28.8% of cases. A clinically significant response was demonstrated in 74.2% of cases

overall. There were 18 clinically significant infections; however, 13 were in patients

receiving concurrent immunosuppressive therapy. There was a total of nine deaths

from any cause.

Conclusion: Rituximab has been used off-label for a range of autoimmune diseases in

this population with a high proportion of Aboriginal patients successfully and safely in

the majority of cases.

Introduction

Rituximab is a monoclonal antibody directed at theCD20 surface antigen expressed on precursor andmature B cells, but not plasma cells. Rituximab depletesB cells from the peripheral circulation through a varietyof mechanisms, including antibody-mediated and

complement-dependent cytotoxicity and B-cell apopto-sis.1 Rituximab was initially developed for the treatmentof non-Hodgkin lymphoma and was first made availableon the Australian Pharmaceutical Benefits Scheme(PBS) for this indication in 2003.2 It is also licensed foruse in CD20+ chronic lymphocytic leukaemia.Rituximab has increasingly been utilised in the treat-

ment of autoimmune diseases. A number of proposedmechanisms for its efficacy in this setting relate to the dis-ruption of B- and T-cell interactions thought important inFunding: None.

Conflict of interest: None.


165

generating autoimmunity.3,4 In 2007, rituximab gainedPBS listing for use in severe rheumatoid arthritis in com-bination with methotrexate 5 and in early 2016 for induc-tion of remission in granulomatosis with polyangiitis andmicroscopic polyangiitis.6 These are the only indicationsother than B-cell neoplasms for which rituximab islicensed by the Therapeutic Goods Administration.

Off-label prescribing of rituximab is expanding to awide range of autoimmune diseases, in most cases with-out high-quality evidence to support its use.7 A prospec-tive data collection in Australian public hospitals in 2012recorded 364 instances of off-label rituximab use pre-scribed for 63 different conditions over a 6-monthperiod.8 Forty percent of the indications for which it wasprescribed had level 4 evidence of benefit (case series,case–control or historically controlled studies).8 This 2012study did not include data from the Northern Territory.

The Northern Territory population has the youngestmedian age (32 years) and the highest proportion ofAboriginal Australians (30%) of all states and terri-tories.9 The majority of the population lives in the TopEnd region, which has only one tertiary referral hospitalwith a catchment area of over 500 000 km2.10 AboriginalAustralians are overrepresented in the Northern Terri-tory public health system accounting for 70% of hospitalpatients in 2013–2014.11 The burden of infectious dis-ease is high with a fourfold greater incidence of sepsisamongst Aboriginal people compared with the non-Aboriginal population living in the Top End.10 Over 80%of the Northern Territory’s Aboriginal population lives inan area classified as remote or very remote.12 Thesedemographics and the distances over which healthcaremust be delivered create unique challenges for theNorthern Territory health system.

This audit provides a descriptive retrospective analysisof 8 years of off-label use of rituximab for autoimmunedisease in the Top End of the Northern Territory, includ-ing evaluation of its efficacy and safety. It adds to theAustralian national data on off-label rituximab use andprovides the first substantial information on its use inAboriginal Australians.

Methods

The Human Research Ethics Committee (HREC) of theNorthern Territory Department of Health and MenziesSchool of Health Research granted ethical approval forthis audit, which met National Health and MedicalResearch Council criteria for a quality assurance activityand was deemed negligible/low risk. Patient informationwas deidentified in line with HREC requirements.

Cases of off-label rituximab use between April 2008and August 2016 were identified from a database kept

by the Royal Darwin Hospital pharmacy containing allcases of off-label rituximab prescribed for patients resid-ing in the Darwin, Katherine and East Arnhem healthregions. Demographic information, diagnosis/indicationfor rituximab, doses given, previous and concomitantimmunosuppressants, clinical outcomes and specificadverse events (see below) were recorded by searchingelectronic and paper medical records for each patient, upto 1 November 2016. Cases where rituximab was givenfor a haematological malignancy or rheumatoid arthritiswere excluded. Anti-neutrophil cytoplasmic antibody-associated vasculitis-associated vasculitis was included;however, all cases either predated PBS approval or didnot meet approval criteria.

Rituximab was prescribed for a heterogeneous array ofautoimmune diseases and validated disease-specific mea-sures of response not always documented, hence the fol-lowing definitions of response were devised:

1 Clinically significant response (CSR): Patient demon-strated clinical benefit from rituximab in terms ofimprovement in organ function or symptoms, or haltingof disease progression or prevention of further relapses,such that

a the patient did not require any further immunosup-pression or

b the initial clinical benefit of rituximab was main-tained with stable immunosuppressive therapy,including not more than 10 mg prednisolone/day or

c further doses of rituximab were given to treat dis-ease relapses or to maintain the initial clinicalbenefit.

2 No response (NR): Patient had no clear clinical benefitfrom rituximab in terms of symptomatic improvementor organ function as reflected by the applicable diseaseactivity tests, requiring the commencement of a differentimmunosuppressive agent in an attempt to gain diseasecontrol.

In addition, five subcategories of CSR were derivedbased on duration of response, the number of courses ofrituximab required, need for other immunosuppressiveagents and ultimate clinical outcome (Table 1).

We sought the following specific adverse events:(i) infusion reactions, (ii) anaphylaxis or hypersensitivityreactions, (iii) infections requiring inpatient treatmentoccurring up to 12 months since the last administereddose of rituximab, (iv) hepatitis B reactivation, (v) pro-gressive multifocal leukoencephalopathy, (vi) new diag-noses of malignancy, (vii) hypogammaglobulinaemia and(viii) all-cause mortality.

Wongseelashote et al.


166

Results

Sixty-six patients met inclusion criteria (Figure 1, Sup-porting Information Table S1). One of the 66 patientsresponded to rituximab for systemic lupus erythemato-sus (SLE)-associated pancytopenia initially, but not attime of relapse 2 years later, and therefore contributed aCSR followed by NR (discussed further below). The aver-age time from first dose of rituximab to the end of thestudy period was 3 years and 4 months. A total of ninepatients died and three patients were lost to follow-upbefore the end of the study period.

Patient demographics

The median age of the patients at time of first rituximabadministration was 38.5 (range 13–79) years and40 (60.6%) were female. Of the females, 31 (77.5%)were Aboriginal, with a median age of 26 (range 13–60)years compared with a median age of 53 (range 25–73)years in the non-Aboriginal female group. Of the26 males, 10 (38.5%) were Aboriginal, with a medianage of 47 (range 20–60) years compared with a medianof 55 (range 22–79) years in the non-Aboriginal malegroup. The majority of Aboriginal patients lived inremote communities (87.8%) compared with 1.5% (onepatient) from the non-Aboriginal group.

Indications and efficacy

In over two-thirds of episodes (68.7%), rituximab wasgiven for disease refractory to or relapsing on other

immunosuppressive therapies. In 19.4% of cases,rituximab was used in conjunction with other immu-nosuppressants (intravenous immunoglobulin andhigh-dose corticosteroids in the majority) as inductiontherapy in an acute illness deemed severe and life-threatening. In 11.9% of cases, rituximab was pre-scribed because of side effects from standard therapiesor variable adherence to prescribed immunosuppres-sive medication.Rituximab was given for 24 different diagnoses. Table 2

summarises the number of cases and responses for eachdiagnosis. The most common diagnoses were thromboticthrombocytopenic purpura (TTP) (CSR in seven of sevencases), inflammatory myopathy (CSR in six of sevencases), immune thrombocytopenic purpura (ITP) (CSR intwo of seven cases), anti-neutrophil cytoplasmic antibody-associated vasculitis-associated vasculitis (CSR in five offive cases), autoimmune encephalopathy (CSR in four offour cases), SLE – not otherwise subcategorised (CSR infour of eight cases) and lupus nephritis (CSR in three offour cases). A total of 19 patients had an underlying diag-nosis of SLE having met Systemic Lupus InternationalCollaborating Clinics classification criteria.13 CSR wasdemonstrated in 14 of the 19 SLE patients (73.7%).Table S1 provides further details about the clinical sce-nario and outcome for each patient. Overall, a clinicallysignificant response to rituximab was demonstrated in49 of 66 patients (74.2%).In three cases of ITP, assessment of response was con-

founded by splenectomy within a month of rituximabadministration. These cases have been included in theNR category.

Adverse events

Table 3 summarises adverse events. There were 18 casesof infection requiring inpatient treatment. There wereno cases of hepatitis B reactivation. Hepatitis B serologyhad been documented prior to rituximab in each caseand any patient with chronic hepatitis B infection com-menced on antiviral therapy. There was only one case ofhypogammaglobulinaemia identified in a patient whowas on concurrent chemotherapy for multiple myeloma.However, testing of serum immunoglobulins wasinconsistent.There were nine deaths during the study period. Fur-

ther details of each death are given in Table S1. Of note,a patient with ITP died of multi-organ failure attributedto disseminated cytomegalovirus infection 2 weeks fol-lowing rituximab, however, this patient had alsoreceived both high-dose dexamethasone and methyl-prednisolone in the weeks prior to her death.

Table 1 Categories of clinically significant response (CSR) to rituximab

Category DescriptionNo.cases

CSR 1 No ongoing immunosuppression requiredfollowing one or two courses of rituximab.Steroids weaned completely. ≥2 years sincelast rituximab

15

CSR 2 No ongoing immunosuppression requiredfollowing one or two courses of rituximab.Steroids weaned completely. <2 years sincelast rituximab

8

CSR 3 Relapsing disease responsive to furtherrituximab, or rituximab used as maintenanceagent to prevent relapses

14

CSR 4 Remission induced or significant improvementwith rituximab, remains stable on anotherimmunosuppressive agent (not more than10 mg/day prednisolone)

5

CSR 5 Clinically significant response to rituximab, butdied or lost to follow-up before end of studyperiod

7

Off-label rituximab use in the Top End


167

Also of note, a patient who had received rituximab2 years prior for SLE-related pancytopenia with subse-quent normalisation of cell counts and no ongoingrequirement for immunosuppression, represented withpancytopenia (including severe thrombocytopenia), butdid not respond to rituximab on this occasion. Threemonths later he presented with a spontaneous intracra-nial haemorrhage and received methylprednisolone,intravenous immunoglobulin and vincristine. He alsounderwent splenectomy, but died a short time later inthe community (cause of death not known).

Discussion

This retrospective audit adds to our current knowledgeof the frequency and outcome of off-label use of rituxi-mab for autoimmune diseases in Australia. It is the firstaudit of rituximab use presented from the Northern Ter-ritory and the largest documented cohort of AboriginalAustralians to have received rituximab.

The disproportionately high number of young Aborigi-nal females in this audit (46.3% of the total, median age26 years) reflects the increased prevalence and severityof autoimmune disease generally amongst this demo-graphic. The prevalence of SLE in Aboriginal people

living in the Darwin, Katherine and East Arnhemregions of the Northern Territory was estimated 20 yearsago to be twice that of the non-Aboriginal Australianpopulation.14 Several other studies focusing on remoteAustralian populations have also identified a higherprevalence of SLE in Aboriginal compared with non-Aboriginal people living in the same regions.15 One ofthe authors (PB) collected a database of SLE patientsfrom 2009 to 2015, and estimated prevalence of SLEamongst Top End Aboriginal Australians to be 140 per100 000 people.

Although the number of cases per diagnosis was small,overall there was evidence of a clinically significantresponse to rituximab in close to three quarters of cases.TTP had an excellent response rate (seven of sevencases). In each case, TTP disease was either refractory,relapsing or deemed severe. Evidence from observationaland uncontrolled studies supports the use of rituximabin TTP in patients with suboptimal response to standardtherapy.16–18

SLE was the underlying diagnosis in a significant pro-portion of patients in this audit (28.8%), with threeof four cases of lupus nephritis and 11 of 15 cases ofSLE – other manifestations, demonstrating CSR to rituxi-mab. Two young SLE patients who did not respond had

66 patients

[13-79]Median age 38.5

40 female (60.6%)

[13-73]Median age 29

31 Aboriginal

Median age 26 [13-60]

23 CSR 8 NR

ITP (5)

MN (1)

SLE (other) (2)

9 non-Aboriginal

[25-73]Median age 53

6 CSR

(1)

Autoimmune

TTP (2)

hemolytic anaemia Autoimmune

encephalopathy (1)

CAPLS (3)

CIDP (2)

Lupus nephritis (1)

SLE (other) (3)

MN (1)

Pemphigus vulgaris (1)

Inflammatory myopathy (5)

TTP (3)

ITP (2)

Transverse myelitis (1)

FSGS (2)

NMO (1)

3 NR

26 Male (39.4%)

[20-79]Median age 49.5

Lupus nephritis (1)

Recurrent parotitis (1)

Systemic sclerosis (1)

10 Aboriginal

[20-60]

Median age 47

7 CSR†

Autoimmune encephalopathy (2)

SLE (other) (1)

4 NR†

Inflammatory

MCD (1)

myopathy (1)TTP (2)

Acquired hemophilia A (1)

Lupus nephritis (1) (1)Myasthenia gravis

SLE (other) (1)

Median age 55 [22-

79]

16 non-Aboriginal

13 CSR

ANCA-AV (5)

Acquired hemophilia A (1)

Autoimmune encephalopathy (1)

MN (1)

Mixed cryoglob. (1)

MPGN (1)

NMO (1)

Inflammatory myopathy (1)

Lupus nephritis (1)

3 NR

Acquired vWD (1)

Chronic

SLE (other) (1)

seronegative polyarthritis(1)

Figure 1 Response to rituximab by demographic group and diagnosis. †One of the 66 patients, with SLE-related pancytopenia, was assessed as hav-

ing had a CSR in the first episode, but NR in the second. ANCA-AV, anti-neutrophil cytoplasmic antibody-associated vasculitis; CAPLS, catastrophic anti-

phospholipid antibody syndrome; CIDP, chronic inflammatory demyelinating polyneuropathy; CSR, clinically significant response; FSGS, focal

segmental glomerulosclerosis; ITP, immune thrombocytopenic purpura; MCD, minimal change disease; MN, membranous nephropathy; NMO, neuro-

myelitis optica; NR, no response; SLE, systemic lupus erythematosus – ‘other’ refers to manifestations not subcategorised elsewhere; TTP, thrombotic

thrombocytopenic purpura; vWD, von Willebrand disease.



168

advanced and likely irreversible end-organ damagebefore receiving rituximab as ‘last hope’ therapy. Other-wise, the apparent effectiveness of rituximab in SLE inthis demographic may support the hypothesis of a B-cell-dependent phenotype.19 The role of rituximab inlupus remains controversial. Two randomised clinical tri-als failed to demonstrate a benefit of adding rituximab tostandard therapies in patients with either renal20 or non-renal21 lupus, however, this contrasts with the results ofnumerous preceding uncontrolled studies.22 Several ofthese studies examined the use of rituximab in patientswith lupus nephritis who had failed to achieve remissionfrom standard therapies, and supported a role for thisagent in refractory disease.23 The majority of the casesdescribed in this audit reflect this practice of prescribingrituximab in SLE resistant to standard immunosuppres-sion. In other cases, rituximab was prescribed whenstandard immunosuppression presented too great a riskfor infection in patients remote from acute medical care,such as in the case of a patient with lupus nephritis whorequired aeromedical evacuation back to Darwin due to

respiratory sepsis following his first dose ofcyclophosphamide.Interestingly, two patients with pulmonary arterial

hypertension in the setting of SLE showed dramaticreversal of pulmonary artery pressures after rituximab.In the first case, a female patient received two courses ofrituximab 1 year apart for chronic inflammatory demye-linating polyradiculoneuropathy (CIDP) on a back-ground of SLE with class V lupus nephropathy. Herestimated pulmonary artery systolic pressure on trans-thoracic echocardiogram progressively improved from116 to 31 mmHg over a 32-month period. Given herclinical improvement, she was discharged from PalliativeCare services after earlier being referred for presumedterminal severe pulmonary hypertension. Her protein-uria was also noted to have completely resolved. In thesecond case, the patient’s estimated pulmonary arterysystolic pressure fell from >80 mmHg prior to rituximabto 43 mmHg on transthoracic echo 15 months later.

Table 2 Summary of diagnoses and responses to rituximab

DiagnosisNo.cases CSR NR

Acquired haemophilia A 2 2Acquired von Willebrand factor deficiency 1 1ANCA-associated vasculitis 5 5Autoimmune encephalopathy 4 4Autoimmune haemolytic anaemia 1 1Catastrophic antiphospholipid syndrome† 3 3Chronic seronegative polyarthritis 1 1CIDP† 2 2Focal segmental glomerulosclerosis 2 2Hep C assoc. mixed cryoglobulinaemia 1 1Inflammatory myopathy 7 6 1ITP† 7 2 5‡Lupus nephritis 4 3 1Membranous nephropathy 3 2 1Minimal change disease 1 1Membrano-proliferative glomerulonephritis 1 1Myasthenia gravis 1 1Neuromyelitis optica 2 2Pemphigus vulgaris 1 1Recurrent parotitis 1 1SLE – not otherwise subcategorised(Table 3)

8 4 4

Systemic sclerosis 1 1Transverse myelitis† 1 1TTP† 7 7

†A number of patients in these categories had known SLE.‡Assessment of response in three cases of ITP confounded by splenec-tomy; these have been included in NR category. ANCA, anti-neutrophilcytoplasmic antibody-associated vasculitis; CIDP, chronic inflammatorydemyelinating polyradiculoneuropathy; CMV, cytomegalovirus; ITP,immune thrombocytopenic purpura; SLE, systemic lupus erythematosus.

Table 3 Adverse events following rituximab

Adverse event Number

Infusion reaction 5Anaphylaxis 3Hypersensitivity reaction (rash) 2Infection 18Lower respiratory tract 11Skin/soft tissue 3Meningitis – Listeria monocytogenes 1Gastroenteritis – Shigella spp. 1Disseminated CMV infection 1Dental abscess (1) 1

Hepatitis B reactivation 0Progressive multifocal leukoencephalopathy 0New diagnosis of malignancy 5Acute myeloid leukaemia (secondary) 1Thymoma 1Breast 1Squamous cell carcinoma, skin 1Small cell carcinoma of lung 1

Hypogammaglobulinaemia 1Death from any cause 9Disseminated CMV infection 1Spontaneous intracranial haemorrhage

in the setting of ITP1

Haemorrhage secondary to assault inthe setting of SLE-related hepaticfailure with coagulopathy

1

Haemolytic-uraemic syndrome 1Gallstone pancreatitis 1Secondary acute myeloid leukaemia 1Small cell lung cancer 1Multi-organ failure presumed secondary

to sepsis on background of SLE1

Circumstances not known 1

CMV, cytomegalovirus; ITP, immune thrombocytopenic purpura; SLE,systemic lupus erythematosus.



169

This audit also included a number of cases of rituxi-mab use in autoimmune disease primarily affecting thecentral nervous system, with demonstrated response inall nine cases (neuromyelitis optica, transverse myelitis,CIDP, autoimmune encephalopathy). A recently pub-lished systematic review and meta-analysis of the use ofrituximab in neuromyelitis optica spectrum disordersdemonstrated a reduction in frequency of relapses andneurological disability, although there were no rando-mised trials available for inclusion.24 Success with rituxi-mab in CIDP has been documented in a handful of casereports and small case series.25–30 Level IV evidence alsoexists for rituximab mainly as a second-line treatment inrefractory autoimmune encephalopathy.31–33

Another group identified in this audit as havingresponded well to rituximab was inflammatory myopa-thies, comprising four cases of polymyositis, two caseswhere myositis was not able to be subclassified and onecase of statin-induced myopathy. Only the latter did notrespond (noting that anti-HMG-CoA reductase antibodieswere not tested), with all other cases achieving a clinicallysignificant response in terms of improvement in weaknesswith restoration of function, and return of creatininekinase to normal range in most cases. Evidence for thebenefit of rituximab in dermatomyositis and polymyositisrefractory to conventional therapy comes from small caseseries34,35 and one randomised trial involving 195 paediat-ric and adult patients, all with disease refractory to gluco-corticoids plus at least one other agent.36

Depletion of circulating B cells generally begins withindays of rituximab administration and B-cell recoverytakes 6–12 months in the majority of patients, thoughthere is significant individual variability and longerrepletion times of over 2 years have been documented.1

The relationship between B-cell recovery and diseaserelapse is variable, and in a subset of responders rituxi-mab produces a more enduring state of remission despiteB-cell recovery.3 In 15 of the 49 cases (30.6%) of CSR inthis audit, response was achieved with the use of eitherone or two courses of rituximab, such that no furtherimmunosuppression was required during the studyperiod (and for a period of at least 2 years) (Table 1).Several patients had notably long periods of clinicalremission still enduring at the end of the study period,including two cases of acquired haemophilia A (5 and7 years) and four cases of TTP (5 years in two cases,8 years in two cases). Data on CD19+ B-cell recovery,particularly in relation to disease relapse or ongoingremission, were mostly not available. Where CD19+ B-cell recovery was documented, it was difficult to assessdisease activity at the same point in time for that patient.

Within the limits of a retrospective audit, this reportalso demonstrates the apparent safety of rituximab.

Although there were 18 documented infections requiringinpatient treatment, in 13 of these cases the patient wason concurrent immunosuppression (>10 mg predniso-lone/day in four cases; mycophenolate or azathioprine �prednisolone 5 mg/day in five cases; prednisolone 5 mg/day in three cases; cytotoxic chemotherapy in one case).The majority of these infections were of the respiratorytract and treated successfully with standard empiric anti-biotic therapy. Notably, one patient who was not on otherconcurrent immunosuppression presented with Listeria

monocytogenes meningitis 1 month following rituximab.Another patient died of disseminated cytomegalovirusinfection (see Adverse Events section above).

The clinical benefit of rituximab documented in thisaudit should also be supported economically. Althoughrituximab is costly (approximately $2000 per 500 mg),this expense would be offset by a number of benefits,including its apparent safety as an immunosuppressantin a demographic with a high prevalence of infection;supplanting variable medication adherence and its long-lived clinical efficacy. It was not in the scope of this auditto compare the rituximab-treated cohort of patients witheither a contemporary or historical conventional immu-nosuppression control group. Nonetheless, each of thebenefits attributed to rituximab here would be expectedto contribute an economic advantage through reducedhospital presentations for both the primary indicationand infectious complications of immunosuppression.

The main limitations of our study are its retrospective,uncontrolled nature and the inconsistent documentationof objective disease activity measures from which todraw conclusions about response to treatment. Theimportance of clinical governance and accountability inhealth care is gaining recognition world-wide. Mostmajor hospitals in Australia have therapeutics commit-tees tasked with providing clinical governance overspecial-case expenditure. This was the finding of O’Con-nor and Liddle in their survey of off-label rituximab usein Australian public hospitals, with the majority ofrequests approved through a local therapeutics commit-tee (78%).8 Such committees formalise review of evi-dence supporting off-label use of specific therapeuticsand promote good clinical practice by requiring cliniciansto document patient consent and specify how outcomeswill be assessed and recorded. Clinical governance overspecial-case expenditure on rituximab in the Top End ofthe Northern Territory was not formalised during thestudy period.

Conclusion

The gold standard of double-blind randomised placebocontrolled trials to determine efficacy and safety are



170

unlikely to occur for medications used in rare diseases,and even less likely to include the demographic subjectof this audit. Off-label prescribing of rituximab in autoim-mune disease is increasing, particularly when standardimmunosuppression has failed or is poorly tolerated. Thisaudit of off-label rituximab use in the Top End of theNorthern Territory demonstrated that rituximab isincreasingly utilised in this population effectively andsafely in the majority of cases. Formal clinical governancestructures, such as a therapeutics committee would facili-tate future auditing, research and quality assurance.

Acknowledgements

We would like to acknowledge the following cliniciansfor clarification of certain patient case details, advice on

specific clinical measures of outcome following rituxi-mab therapy or contribution in the early stages of assem-bling this audit (in alphabetical order): Dr Jim Burrow(neurologist, Royal Darwin Hospital (RDH)), Dr Pasqua-lina Coffey (public health registrar, Northern TerritoryCentre for Disease Control), Ms Genevieve Francais(pharmacist, Katherine Hospital), Ms Bianca Heron(pharmacist, RDH), Dr Akash Kalro (haematologist,RDH), Dr Akshay Kulkarni (medical registrar, RDH), DrCheng Lu (medical registrar, RDH), Dr William Majoni(nephrologist, RDH), Dr Babu Philip (immunologist,RDH), Ms Tristen Pogue (pharmacist, RDH), Dr SimonQuilty (general physician, Katherine Hospital), Dr Mad-hivanan Sundaram (nephrologist, RDH), Dr FerencSzabo (haematologist, RDH) and Ms Joanna Wallace(pharmacist, RDH).

References

1 Kado R, Sanders G, McCune WJ.

Suppression of normal immune

responses after treatment with

rituximab. Curr Opin Rheumatol 2016;

28: 251–8.

2 Yu XQ, Chen WH, O’Connell DL.

Improved survival for non-Hodgkin

lymphoma patients in New South

Wales, Australia. BMC Cancer 2010;

10: 231.

3 Dorner T, Radbruch A, Burmester GR.

B-cell-directed therapies for

autoimmune disease. Nat Rev Rheumatol

2009; 5: 433–41.

4 Cooper N, Arnold DM. The effect of

rituximab on humoral and cell

mediated immunity and infection in the

treatment of autoimmune diseases. Br J

Haematol 2010; 149: 3–13.


Welfare. The Use of Disease-Modifying

Anti-Rheumatic Drugs for the

Management of Rheumatoid Arthritis.

Arthritis Series No. 16. Canberra: AIHW;

2011 [cited 2017 Feb 11]. Available from

URL: http://www.aihw.gov.au/

WorkArea/DownloadAsset.aspx?id=

10737418551&libID=10737418550

6 Randall KL. Rituximab in

autoimmune diseases. Aust Prescr

2016; 39: 131–4.

7 Ramos-Casals M, Garcia-Hernandez FJ,

de Ramon E, Callejas JL, Martinez-

Berriotxoa A, Pallares L et al. Off-label

use of rituximab in 196 patients with

severe, refractory systemic autoimmune

diseases. Clin Exp Rheumatol 2010; 28:

468–76.

8 O’Connor K, Liddle C. Prospective data

collection of off-label use of rituximab

in Australian public hospitals. Intern Med

J 2013; 43: 863–70.

9 Australian Bureau of Statistics. Regional

Statistics, Northern Territory, Mar 2011

Canberra [2017 Jan 31]. Available from

URL: http://www.abs.gov.au/ausstats/

[email protected]/Latestproducts/1362.7Feature%

20Article1Mar%202011?opendocument

10 Davis JS, Cheng AC, McMillan M,

Humphrey AB, Stephens DP,

Anstey NM. Sepsis in the tropical Top

End of Australia’s Northern Territory:

disease burden and impact on

Aboriginal Australians. Med J Aust 2011;

194: 519–24.

11 Northern Territory Government.

Department of Health Annual Report

2013–2014. Darwin: Department of

Health, Northern Territory

Government; 2014 [cited 2017 Feb 15].

Available from URL: http://

digitallibrary.health.nt.gov.au/

prodjspui/handle/10137/613

12 Australian Bureau of Statistics.

Population Distribution, Aboriginal and

Torres Strait Islander Australians.

Canberra: Australian Bureau of

Statistics; 2006 [cited 2017 Jan 17].

Available from URL: http://www.abs.

gov.au/ausstats/[email protected]/mf/4705.0

13 Petri M, Orbai AM, Alarcon GS,

Gordon C, Merrill JT, Fortin PR et al.

Derivation and validation of the

Systemic Lupus International

Collaborating Clinics classification

criteria for systemic lupus

erythematosus. Arthritis Rheum 2012;

64: 2677–86.

14 Anstey NM, Bastian I, Dunckley H,

Currie BJ. Systemic lupus

erythematosus in Australian Aborigines:

high prevalence, morbidity and

mortality. Aust N Z J Med 1993; 23:

646–51.

15 Nikpour M, Bridge JA, Richter S. A

systematic review of prevalence, disease

characteristics and management of

systemic lupus erythematosus in

Australia: identifying areas of unmet

need. Intern Med J 2014; 44: 1170–9.

16 Froissart A, Buffet M, Veyradier A,

Poullin P, Provot F, Malot S et al. Efficacy

and safety of first-line rituximab in

severe, acquired thrombotic

thrombocytopenic purpura with a

suboptimal response to plasma exchange.

Experience of the French Thrombotic

Microangiopathies Reference Center. Crit

Care Med 2012; 40: 104–11.

17 Scully M, Cohen H, Cavenagh J,

Benjamin S, Starke R, Killick S et al.

Remission in acute refractory and

relapsing thrombotic thrombocytopenic

purpura following rituximab is

associated with a reduction in IgG

antibodies to ADAMTS-13. Br J

Haematol 2007; 136: 451–61.

18 Benhamou Y, Paintaud G, Azoulay E,

Poullin P, Galicier L, Desvignes C et al.

Efficacy of a rituximab regimen based

on B cell depletion in thrombotic

thrombocytopenic purpura with

suboptimal response to standard

treatment: results of a phase II,

multicenter noncomparative study.

Am J Hematol 2016; 91: 1246–51.

19 Vincent FB, Bourke P, Morand EF,

Mackay F, Bossingham D. Focus on



171

http://www.aihw.gov.au/WorkArea/DownloadAsset.aspx?id=10737418551&libID=10737418550



http://www.abs.gov.au/ausstats/[email protected]/Latestproducts/1362.7Feature%20Article1Mar%202011?opendocument



http://digitallibrary.health.nt.gov.au/prodjspui/handle/10137/613



http://www.abs.gov.au/ausstats/[email protected]/mf/4705.0

http://www.abs.gov.au/ausstats/[email protected]/mf/4705.0

systemic lupus erythematosus in

Aboriginal Australians: towards a better

understanding of autoimmune diseases.

Intern Med J 2013; 43: 227–34.

20 Rovin BH, Furie R, Latinis K, Looney RJ,

Fervenza FC, Sanchez-Guerrero J et al.

Efficacy and safety of rituximab in

patients with active proliferative lupus

nephritis: the Lupus Nephritis

Assessment with Rituximab study.

Arthritis Rheum 2012; 64: 1215–26.

21 Merrill JT, Neuwelt CM, Wallace DJ,

Shanahan JC, Latinis KM, Oates JC

et al. Efficacy and safety of rituximab in

moderately-to-severely active systemic

lupus erythematosus: the randomized,

double-blind, phase II/III systemic lupus

erythematosus evaluation of rituximab

trial. Arthritis Rheum 2010; 62: 222–33.

22 Ramos-Casals M, Soto MJ,

Cuadrado MJ, Khamashta MA.

Rituximab in systemic lupus

erythematosus: a systematic review of

off-label use in 188 cases. Lupus 2009;

18: 767–76.

23 Weidenbusch M, Rommele C,

Schrottle A, Anders HJ. Beyond the

LUNAR trial. Efficacy of rituximab in

refractory lupus nephritis. Nephrol Dial

Transplant 2013; 28: 106–11.

24 Damato V, Evoli A, Iorio R. Efficacy and

safety of rituximab therapy in

neuromyelitis optica spectrum disorders:


JAMA Neurol 2016; 73: 1342–8.

25 Benedetti L, Briani C, Franciotta D,

Fazio R, Paolasso I, Comi C et al.

Rituximab in patients with chronic

inflammatory demyelinating

polyradiculoneuropathy: a report of

13 cases and review of the literature. J

Neurol Neurosurg Psychiatry 2011; 82:

306–8.

26 Benedetti L, Franciotta D, Beronio A,

Cadenotti L, Gobbi M, Mancardi GL

et al. Rituximab efficacy in CIDP

associated with idiopathic

thrombocytopenic purpura. Muscle Nerve

2008; 38: 1076–7.

27 Knecht H, Baumberger M, Tobon A,

Steck A. Sustained remission of CIDP

associated with Evans syndrome.

Neurology 2004; 63: 730–2.

28 Querol L, Rojas-Garcia R, Diaz-

Manera J, Barcena J, Pardo J, Ortega-

Moreno A et al. Rituximab in treatment-

resistant CIDP with antibodies against

paranodal proteins. Neurol Neuroimmunol

Neuroinflamm 2015; 2: e149.

29 Sanz PG, Garcia Mendez CV, Cueto AL,

Silva VB, Walther JC, Diez RA et al.

Chronic inflammatory demyelinating

polyradiculoneuropathy in a patient

with systemic lupus erythematosus and

good outcome with rituximab

treatment. Rheumatol Int 2012; 32:

4061–3.

30 Briani C, Zara G, Zambello R, Trentin L,

Rana M, Zaja F. Rituximab-responsive

CIDP. Eur J Neurol 2004; 11: 788.

31 Lee WJ, Lee ST, Byun JI, Sunwoo JS,

Kim TJ, Lim JA et al. Rituximab

treatment for autoimmune limbic

encephalitis in an institutional cohort.

Neurology 2016; 86: 1683–91.

32 Nosadini M, Mohammad SS,

Ramanathan S, Brilot F, Dale RC.

Immune therapy in autoimmune

encephalitis: a systematic review.

Expert Rev Neurother 2015; 15:

1391–419.

33 Titulaer MJ, McCracken L, Gabilondo I,

Armangue T, Glaser C, Iizuka T et al.

Treatment and prognostic factors for

long-term outcome in patients with

anti-NMDA receptor encephalitis: an

observational cohort study. Lancet

Neurol 2013; 12: 157–65.

34 Mahler EA, Blom M, Voermans NC,

van Engelen BG, van Riel PL, Vonk MC.

Rituximab treatment in patients with

refractory inflammatory myopathies.

Rheumatology 2011; 50: 2206–13.

35 Levine TD. Rituximab in the

treatment of dermatomyositis: an

open-label pilot study. Arthritis Rheum

2005; 52: 601–7.

36 Oddis CV, Reed AM, Aggarwal R,

Rider LG, Ascherman DP, Levesque MC

et al. Rituximab in the treatment of

refractory adult and juvenile

dermatomyositis and adult

polymyositis: a randomized, placebo-

phase trial. Arthritis Rheum 2013; 65:

314–24.



Table S1. Details of patients, clinical scenario and outcomes.



172

doi:10.1111/imj.13651

Characteristics of adrenal incidentalomas in a New ZealandcentreZiwei Goh ,1 Ian Phillips,2 Penny J. Hunt,1,3 Steven Soule1,3 and Tom J. Cawood1

1Department of Endocrinology, Christchurch Hospital, 2Endocrine and Steroid Laboratory, Canterbury Health Laboratories, and 3Department ofMedicine, University of Otago, Christchurch, New Zealand

Key wordsadrenal tumour, adrenal cancer, incidentaloma.

CorrespondenceZiwei Goh, Department of Endocrinology,Christchurch Hospital, Private Bag 4710,Christchurch, New Zealand.Email: [email protected]

Received 21 May 2017; accepted17 October 2017.

Abstract

Background: Management of adrenal incidentalomas (AI) is becoming more conserva-

tive, based on international data showing a low incidence of functional or malignant

lesions. The clinical characteristics of AI in New Zealand are unknown. Therefore,

whether the AI guidelines apply to the New Zealand population is also unknown.

Aims: To investigate the clinical characteristics of patients with AI presenting to a

tertiary-care centre in New Zealand.

Method: This study prospectively evaluated consecutive patients aged 18 or older with

AI, 1 cm or larger, diagnosed in Canterbury, New Zealand. A standardised nurse-led

dedicated AI clinic was used, including clinical assessment, hormonal evaluation and

imaging.

Results: From January 2010 to April 2016, 306 patients were referred to the AI clinic,

228 met the inclusion criteria. Most of those excluded were not true AI, due to imaging

performed for known or suspected non-adrenal malignancy.

The most common reason for imaging was abdominal pain (46%). Most cases were

benign (96.5%) and 88.6% of all cases were non-functional. Of the functioning

tumours (26 patients), 18 had subclinical Cushing syndrome, four had late-onset con-

genital adrenal hyperplasia, two had phaeochromocytoma and one had primary hyper-

aldosteronism. Three patients had primary adrenal cancer, one of whom was secreting

excess cortisol. One adrenal metastasis was diagnosed.

Conclusion: This study found a similar prevalence of functional and malignant AI as

international centres, although mild cortisol excess and primary aldosteronism may

be under-represented. Therefore, the conservative approach to management of AI

recommended in current guidelines is likely to be applicable to New Zealand

population.

Introduction

Adrenal incidentalomas (AI) are defined as adrenalmasses that are discovered unexpectedly on imagingperformed for an unrelated reason.1,2 Adrenal massesare common and the reported prevalence has beenincreasing due to the continued advances in imagingtechnology and widespread use of imaging in clinicalpractice.2

The main clinical concerns in AI are the risk of malig-nancy and presence of hormone overproduction. How-ever most AI are non-functional, benign corticaladenomas that require no treatment.1,3

Based on international studies, the prevalence ofAI is reported as 3–10% and 93–98% of AI arebenign.1,2,4 About 5–15% of the benign adenomas arefunctional tumours which produce excess hormones.4

Previous guidelines have involved significant investiga-tion and follow up of AI with associated costs.2,5 Inview of this more recent literature, guidelines haveevolved to a more conservative approach to manage-ment of AI, particularly regarding follow up if the AI isthought to be benign and non-functioning at presenta-tion. 4

It is unclear if the guidelines are applicable to localpatients as the clinical characteristics of AI in theNew Zealand population may be different from interna-tional centres.This paper aims to investigate the clinical characteris-

tics of AI patients who presented to a tertiary-care centrein New Zealand.



173

Methods

Patients

This study prospectively evaluated consecutive patientsaged 18 or older with AI, 1 cm or larger, discovered onimaging in the Canterbury region of New Zealandbetween January 2010 and April 2016. If the AI wasdiscovered by abdominal ultrasound, it was confirmedby computed tomography (CT). The patients wereassessed in a dedicated nurse-led AI clinic following astandardised protocol (in Supporting InformationAppendix S1).

Patients were excluded if imaging was initiated due tosigns or symptoms of adrenal disease (e.g. severe hyper-tension, hypokalaemia, Cushingoid features), knownmalignancy under surveillance in the last 5 years (otherthan cured skin basal cell carcinoma or squamous cellcarcinoma), metastatic malignancy found on referringCT scan, incomplete hormonal evaluation or hormonalresults that could not be interpreted due to patientsbeing on long-term, high-dose prednisone.

If there were abnormal findings on hormonal testingor imaging, the patients were seen by an endocrinologistand diagnosis was based on further investigations andclinical review. Adrenalectomy was offered to patientswith clinically significant hormonal hypersecretion orradiographic features suspicious of malignancy.

CT findings

Adrenal tumours characterised by attenuation value of≤10 Hounsfield units (HU) or rapid contrast mediumwashout (absolute contrast medium washout of morethan 60%, 10 min after administration of contrast) wereconsidered benign adrenocortical adenomas. If therewere multiple adrenal nodules, CT characteristics of eachnodule were assessed. Dedicated adrenal CT scan wasperformed in those whose initial scan had only contrast-enhanced images. The CT findings were reported byradiologists from one District Health Board who areexpert at abdominal imaging.

Data collection

Data were prospectively collected on patients assessed inthe AI clinic, including medical history, blood pressureand anthropometric data. All patients had a nurse-ledclinical examination for signs of Cushing syndrome,including proximal muscle weakness, pathologic striae,bruising and skin thickness.

Biochemical analysis

Investigations included plasma sodium and potassium,random plasma glucose, renal function, HbA1c, plasmametanephrine and normetanephrine and 24 h urine col-lection for cortisol, metanephrine and normetanephrine.Where possible, blood tests were performed seated andin the morning. Due to geographical issues (the popula-tion catchment area is over 45 000 km), there wereoccasional patients who had their tests performed in theafternoon. Urine testing was chosen in preference to alow dose overnight 1 mg-dexamethasone suppressiontest partly due to geographical reasons as this allowed alltests to be done with one clinic visit and the subsequenturine collection was then handed to a laboratory nearthe patient’s home address.

Plasma testosterone and DHEA-S were assayed infemale patients with hirsutism. If the CT showed bilat-eral adrenal masses, the patient was evaluated for late-onset congenital adrenal hyperplasia (CAH) (Synacthentest for 17-OH progesterone). Serum testosterone(female reference range 0.3–2.7 nmol/L), DHEA-sulfate(reference range 0.5–12.0 μmol/L), 17-OH progesterone(reference range post Synacthen test <30 nmol/L) weremeasured by in-house ELISA method (CanterburyHealth Laboratory).6–8

Autonomous cortisol secretion was consideredexcluded if the 24 h urine cortisol excretion was normal.If the 24 h urinary cortisol excretion (reference range:100–400 nmol) was elevated, a 1-mg dexamethasonesuppression test (reference range < 50 nmol/L) was per-formed. Serum cortisol and 24 h urine cortisol excretionwere measured by in-house ELISA method (CanterburyHealth Laboratory).9,10 Patients with both an elevated24 h urine cortisol and post 1 mg-dexamethasone corti-sol underwent clinical review by an Endocrinologist.Patients were diagnosed with subclinical Cushing syn-drome (SCS), as per the commonly used definition inthe literature at the time, that is, the absence of clinicalsigns of cortisol excess plus at least two biochemicalabnormalities of the hypothalamic–pituitary–adrenalaxis.11–13

Plasma aldosterone and renin were measured if thepatient was hypertensive (defined as the use of anti-hypertensive drugs, or blood pressure > 140/90 mmHg)or was hypokalaemic. Anti-hypertensive medicationswere not changed prior to the screening test. Prior toJuly 2014, plasma aldosterone was assessed by direct dis-placement radioimmunoassay (units ng/L) and plasmarenin measured in an antibody trapping assay (unitsnmol/L/h).14,15 After July 2014, plasma aldosterone andrenin were measured by chemiluminescence providedby Immunodiagnostics Systems, Specialty Immunoassay

Goh et al.


174

System (IDS-iSYS). Where possible, blood tests were per-formed seated and in the morning, however due to geo-graphical issues, there were occasional patients who hadtests performed in the afternoon. A raised aldosterone torenin ratio (ARR) (>800 prior to July 2014 or >30 withthe newer assay) was considered suggestive of primaryhyperaldosteronism, and those patients went on toreview by an Endocrinologist and, where appropriate,confirmatory testing with recumbent saline suppressiontest was performed. An aldosterone level of >210 pmol/Lpost saline infusion was considered as diagnostic of pri-mary hyperaldosteronism. Other centres may use alower cut-off level of aldosterone for saline suppressiontest. Recent primary aldosteronism guidelines suggest alevel of 190 pmol/L, which would have a higher sensitiv-ity, but a lower specificity.16

Plasma metanephrine (reference range <500 pmol/L)and normetanephrine (reference range <900 pmol/L)and 24 h urine metanephrine (referencerange <1.5 μmol/24 h) and normetanephrine (referencerange < 2.9 μmol/24 h) were measured by liquid chro-matography (LC)–mass spectrometry detection.17

Results

From January 2010 to April 2016, 306 patients with anadrenal mass were referred to the AI clinic, of whom228 patients met the inclusion criteria. Seventy-eightpatients were excluded – 14 due to incomplete biochem-ical testing (Table 1), 15 as their adrenal lesion was smal-ler than 1 cm; 46 as they were not true incidentalomas(36 due to CT scans being performed for known malig-nancy or metastatic malignancy found on referring CTscan, five due to hypertension under investigation, fivedid not have adrenal mass). Other reasons for exclusionincluded three patients on prednisone.Table 2 shows the clinical characteristics of patients

with AI. The most common reason for abdominal imag-ing was abdominal pain (46%) followed by detection ofan abnormality found on CT chest (11.8%). The major-ity were female (59.6%). The mean age was 60.6 years,range 20–86 years. Most of the AI were on the left

(55.3%), 30.7% on the right and 14% were bilateral.The mean diameter was 21 mm, with median diameter18 mm, range 10–82 mm.Table 3 and Figure 1 show the final diagnosis of the

AI. Most cases were benign as defined by CT imaging cri-teria (96.5%). Twenty-six patients had functioningtumours, of whom 18 had subclinical Cushing syn-drome, four had late onset CAH, two had phaeochromo-cytoma and one had primary hyperaldosteronism. Therewere no cases of Cushing syndrome.Primary adrenal cancer was diagnosed in three

patients, one of which was functional with cortisolexcess and the other two were non-secretory. One adre-nal metastasis was found in a patient having been trea-ted surgically and considered cured for Merkel cellcarcinoma 6 years earlier.Three out of 18 patients with SCS underwent surgery.

Fifteen patients with SCS were managed conservatively,three of whom were felt not to be surgical candidates.The other patients either had no associated medicalproblem or they were able to be managed medically.

Table 1 Reasons for exclusions of patients

Excluded patients n

Metastatic malignancy or known non-adrenal malignancy 36Lesion smaller than 1 cm 15Incomplete biochemical assessment 14Not adrenal lesion 5Hypertension under investigation 5On prednisone 3

Table 2 Clinical characteristics of patients with adrenal incidentalomas

Characteristic Number of patients

Sex, n (%)Male 92 (40.4)Female 136 (59.6)

Mean age (years) (SD) 60.6 (13.0)Age range (years) 20–86Location of lesion, n (%)Right 70 (30.7)Both 32 (14.0)Left 126 (55.3)

Size of adrenal incidentaloma (mm)Mean diameter (range) 21 (10–82)

Median diameter 18Reason for imaging, n (%)Abdominal pain 105 (46.1)Abnormality on CT chest 27 (11.8)Haematuria 11 (4.8)Abnormality on ultrasound abdomen 11 (4.8)Anaemia 10 (4.4)CT angiogram 10 (4.4)Abnormality on MRI abdomen 9 (3.9)Altered bowel habit 9 (3.9)Rectal bleed 7 (3.1)Liver abnormalities 6 (2.6)Trauma scan 5 (2.2)Weight loss 4 (1.8)Possible cancer† 3 (1.3)Others 11 (4.8)

†Patients undergoing CT scan for possible cancer were included as can-cer was not found on investigation. CT, computed tomography; MRI,magnetic response imaging; SD, standard deviation.

Clinical characteristics of AI patients


175

Plasma aldosterone and renin were assessed in152 patients (those with BP < 140/90 at assessment, orthose on anti-hypertensive medication), of whom89 patients were on interfering medications, such asangiotensin converting enzyme-inhibitors, angiotensin IIreceptor blockers diuretics and three patients were tak-ing spironolactone. One patient was diagnosed withprobable primary hyperaldosteronism based on clinicalgrounds and ARR. Her plasma renin activity was0.3 (nmol/L)/h (0.4–2.3), with aldosterone 245 pmol/L(100–800), which is highly likely of primary hyperaldos-teronism as she was on candesartan 32 mg daily. She didnot have confirmatory testing as she was aged 86 yearsand surgery was not thought to be appropriate. Sheresponded well to spironolactone. Four other patients

had raised ARR ratios, two of whom were unfit for fur-ther investigations, one was a false positive and the lastpatient declined further testing. There were no patientswith positive confirmatory saline suppression test.

Discussion

This study describes the baseline characteristics of228 patients assessed in a dedicated AI clinic in Canter-bury, New Zealand, over a 6-year period.

Of the study cohort, 190 patients (83.3%) were diag-nosed with benign non-functioning adrenal adenomas,18 patients (7.5%) with SCS, 4 patients (1.7%) with lateonset CAH, 2 patients (0.9%) with phaeochromocytomaand 1 patient (0.4%) with primary hyperaldosteronism.Four patients (1.7%) had a malignant condition, three ofwhich were primary adrenal carcinoma. The prevalencerates for these conditions are similar to a review of ninehigher quality studies, with the possible exception ofphaechromocytomas which we found in 0.9% ofpatients, compared to 0–5.3%, mean 3.1%, median3.0%, found in these other studies.4

The mean diameter of AI in this study was 21 mm,which is smaller than in a review of the nine higherquality studies (e.g. mean 32 mm).4 This may be due toincreased use of imaging technology leading to more AIbeing discovered at a smaller size, improved resolutionof newer imaging technology and/or greater awarenessof reporting radiologists for adrenal pathology.

The prevalence of patients diagnosed with primaryhyperaldosteronism in this study was 0.4%, which maybe falsely low due to our exclusion criteria and the lackof cessation of all anti-hypertensive medication. How-ever, the prevalence in this study was similar to thatfound in those studies that likely best reflect the trueprevalence of hyperaldosteronism in AI (e.g. range0–1.8%, mean 0.6%. median 0%).4

There was a very low detection rate of adrenal metas-tasis, only one patient, which likely relates to the prede-termined protocol, focusing on true AI and excludingpatients with a history of cancer in the last 5 years orthose diagnosed with metastatic cancer on the referringCT scan.

New European guidelines for the assessment of AIwere published after this study was commenced. TheEuropean Society of Endocrinology clinical practiceguideline no longer uses the term ‘subclinical Cushingsyndrome’ and instead classifies patients as having ‘pos-sible autonomous cortisol secretion’ or ‘autonomouscortisol secretion’ based on the results of the 1 mg-dexamethasone suppression test.1 As a result, our centrenow uses the 1 mg-dexamethasone suppression test asthe initial assessment of cortisol autonomy in AI patients,

Table 3 Diagnosis of adrenal incidentalomas

Diagnosis Number of patients (%)

Benign non-functioning adenoma 190 (83.3)SCS 18 (7.9)Late onset CAH 4 (1.8)Indeterminate† 4 (1.8)Adrenocortical carcinoma‡ 3 (1.3)Ganglioneuroma‡ 2 (0.9)Phaeochromocytoma‡ 2 (0.9)Myelolipoma‡ 2 (0.9)Primary hyperaldosteronism 1 (0.4)Lymphangioma‡ 1 (0.4)Merkel cell metastasis‡ 1 (0.4)

†Patients with indeterminate CT adrenal scan either had interval followup CT scan that showed no change in size of adrenal lesion, or haddeclined surgical intervention. ‡All patients with adrenocortical carci-noma, merkel cell carcinoma, ganglioneuroma, lymphangioma andphaeochromocytoma underwent surgical excision and diagnosis wasconfirmed on histology. One of two patients with myelolipoma under-went surgery which confirmed the diagnosis and the other patient wasdiagnosed based on scan characteristics. CAH, congenital adrenalhyperplasia; CT, computed tomography; SCS, subclinical Cushingsyndrome.

88.6%-

11.4%-functional

nonfunctional

Functionality of tumour

96.5%-Benign‡

1.8%-malignant 1.8%-indeterminate†

Benign vs malignant

Figure 1 Incidence of functional and malignant lesions. †Patients with

indeterminate initial CT adrenal scan either had interval follow up CT

scan which showed no change in size of adrenal lesion, or had declined

surgical intervention. ‡Phaeochromocytoma was included in the benign

group as there were no malignant features on baseline assessment;

however, follow up is required due to the potential for malignancy.

Goh et al.


176

despite the geographical challenges. It is possible that byusing urinary free cortisol instead of 1 mg-dexamethasone suppression test, we may have missedsome patients with possible autonomous cortisolsecretion.This study does not provide any data on whether ini-

tially non-functional benign adenomas gain functionalityover time or whether patients with AI should have fol-low up imaging. The European Society of Endocrinologyclinical practice guideline has suggested routine followup of patients with non-functional adenomas smallerthan 4 cm is not indicated.1

Of note, all patients in this study were diagnosed at asingle, tertiary-care medical centre in New Zealand, rais-ing the possibility that the data may not be representa-tive of the wider New Zealand population. According tothe 2013 census data, 8.1% of the Canterbury popula-tion identifies as Maori, compared to 14.9% of the over-all New Zealand population.17 In addition, 86.9% ofpeople in Canterbury belong to the European ethnicgroup, compared with 74% for New Zealand.18 The

study however offers the advantage of being the largestsource of AI data for New Zealand population.

Conclusion

The characteristics of AI in Canterbury, New Zealand,are similar to international centres, although mild corti-sol excess and primary aldosteronism may have beenunder-represented in this study. This study provides sup-port that adapting the recent changes to managementguidelines would likely provide similar outcomes tointernational centres, at least for the Canterbury popula-tion. A follow up study of these patients with AI to assessoutcomes further is ongoing.

Acknowledgements

We acknowledge the work of our Endocrine nurses,especially Ms Janet Gilbert (Department of Endocrinol-ogy, Christchurch Hospital), in the acquisition of data forthis study.

References

1 Fassnacht M, Arlt W, Bancos I, Dralle H,

Newell-Price J, Sahdev A et al.

Management of adrenal incidentalomas:

European Society of Endocrinology

Clinical Practice Guideline in

collaboration with the European

network for the study of adrenal

tumors. Eur J Endocrinol 2016; 175:

G1–34.

2 Mansmann G, Lau J, Balk E,

Rothberg M, Miyachi Y, Bornstein SR.

The clinically inapparent adrenal mass:

update in diagnosis and management.

Endocr Rev 2004; 25: 309–40.

3 Kastelan D, Kraljevic I, Dusek T,

Knezevic N, Solak M, Gardijan B et al.

The clinical course of patients with

adrenal incidentaloma: is it time to

reconsider the current

recommendations? Eur J Endocrinol

2015; 173: 275–82.

4 Cawood TJ, Hunt PJ, O’Shea D, Cole D,

Soule S. Recommended evaluation of

adrenal incidentalomas is costly, has

high false-positive rates and confers a

risk of fatal cancer that is similar to the

risk of the adrenal lesion becoming

malignant; time for a rethink? Eur J

Endocrinol 2009; 161: 513–27.

5 Zeiger MA, Siegelman SS,

Hamrahian AH. Medical and surgical

evaluation and treatment of adrenal

Incidentalomas. J Clin Endocrinol Metab

2011; 96: 2004–15.

6 Yeo KH, Elder PA, Lewis JG. An

automated solid-phase 17 alpha-

hydroxyprogesterone ELISA method

using a microtitre plate. Ann Clin

Biochem 1988; 25: 293–7.

7 Lewis JG, Bason LM, Elder PA.

Production and characterization of

monoclonal antibodies to

dehydroepiandrosterone sulfate:

application to direct enzyme-linked

immunosorbent assays of

dehydroepiandrosterone sulfate and

androsterone/epiandrosterone sulfates

in plasma. Steroids 1996; 61: 682–7.

8 Elder PA, Lewis JG. An enzyme-linked

immunosorbent assay (ELISA) for

plasma testosterone. J Steroid Biochem

1985; 22: 635–8.

9 Lewis JG, Manley L, Whitlow JC,

Elder PA. Production of a monoclonal

antibody to cortisol: application to a

direct enzyme-linked immunosorbent

assay of plasma. Steroids 1992; 57: 82–5.

10 Lewis JG, Manley L, Townsend JC,

Elder PA. An enzyme-linked

immunosorbent assay (ELISA) for

urinary free cortisol. Clin Chim Acta

1986; 159: 205–9.

11 Mantero F, Terzolo M, Arnaldi G,

Osella G, Masini AM, Alì A et al. A survey

on adrenal incidentaloma in Italy. J Clin

Endocrinol Metab 2000; 85: 637–44.

12 Sippel RS, Chen H. Subclinical

Cushing’s syndrome in adrenal

incidentalomas. Surg Clin North Am

2004; 84: 875–85.

13 Zografos GN, Perysinakis I, Vassilatou E.

Subclinical Cushing’s syndrome:

current concepts and trends. Hormones

2014; 13: 323–37.

14 PoulsenK, Jorgensen J.Aneasy radioimm

unologicalmicroassay of renin activity,

concentration and substrate inhumanand

animal plasmaand tissues based on

angiotensin I trapping by antibody. J Clin

EndocrinolMetab1974;39: 816–25.

15 Lun S, Espiner EA, Nicholls MG,

Yandle TGA. Direct radioimmunoassay

for aldosterone in plasma. Clin Chem

1983; 29: 268–71.

16 Funder JW, Carey RM, Mantero F,

Murad MH, Reincke M, Shibata H et al.

The Management of Primary

Aldosteronism: case detection, diagnosis,

and treatment: an Endocrine Society

clinical practice guideline. J Clin Endocrinol

Metab 2016 May; 101: 1889–916.

17 Whiting MJ. Simultaneous

measurement of urinary metanephrines

and catecholamines by liquid

chromatography with tandem mass

spectrometric detection. Ann Clin

Biochem 2009; 46: 129–36.

18 Statistics New Zealand. 2013 Census.

[cited 2017 Jan 4]. Available from URL:

stats.govt.nz

Clinical characteristics of AI patients


177

http://stats.govt.nz



Appendix S1. Nurse clinical assessment.

doi:10.1111/imj.13647

United we stand, divided we conquer: pilot study ofmultidisciplinary General Medicine Heart Failure Care ProgramOmar Wahbi-Izzettin ,1 Ingrid Hopper,1,2 Edward Ritchie,1 Vathy Nagalingam1 and Ar Kar Aung1,2

1Department of General Medicine, Alfred Hospital, and 2School of Public Health and Preventive Medicine, Monash University, Melbourne, Victoria,Australia

Key wordshealthcare quality, heart failure, patienteducation, care plan, nursing.

CorrespondenceAr Kar Aung, Department of General Medicine,Alfred Hospital, 55 Commercial Road,Melbourne, Vic. 3004, Australia.Email: [email protected]

Received 16 May 2017; accepted3 October 2017.

Abstract

Background: Heart failure care and education require a multifaceted approach to

ensure appropriate transition from inpatient to outpatient care.

Aims: To explore the feasibility of a multidisciplinary heart failure care model, General

Medicine Heart Failure Care Program (GM-HFCP), within a General Medical

Unit (GMU).

Methods: Prospective non-randomised before-and-after observational quality

improvement intervention over a 6-month period was conducted. All consecutive

patients admitted to GMU at Alfred Hospital, Melbourne with a diagnosis of acute

decompensated heart failure were included. Main outcome measures included changes

in rates of pharmacologic prescription, non-pharmacologic ward-based management,

patient education and action plan provision after intervention.

Results: In total, 108 patients were included (median age 84 (inter-quartile range

80–89) years, 47(44%) females). Significant improvements were noted in non-

pharmacologic management for patient education regarding fluid restriction (12–30%,

P = 0.04), weight monitoring (10–28%, P = 0.03), heart failure action plan on discharge

(4–28%, P = 0.002) and salt restriction (6–32%, P = 0.002). The rates of prescription of

heart failure medications remained similar between the pre- and post-implementation

periods, particularly in patients with reduced ejection fraction by ‘appropriateness of

prescription’ criteria. There were no differences in inpatient mortality or 30-day read-

mission rates in both groups.

Conclusions: This prospective observational study demonstrated that it is possible to

share the roles of a heart failure nurse amongst members of the multidisciplinary team,

with similar rates of delivery of pharmacologic and non-pharmacologic management

aspects. However, further innovative improvements are needed to address certain

aspects of heart failure care.

Introduction

Management of heart failure is complex. National HeartFoundation of Australia guidelines strongly recom-mend a multidisciplinary approach to both

pharmacologic and non-pharmacologic aspects of heartfailure management.1

In specialist heart failure care units, the role of a dedi-cated heart failure nurse improves patients’ knowledge,and is associated with improved self-care behavioursand reduction in hospital readmissions.2–5 However, inGeneral Medical wards, which also manage a significantnumber of heart failure patients with complex comor-bidities, there often is no dedicated heart failure nurse



178

to address all recommended components of heartfailure care.We hypothesise that heart failure management can be

effectively shared amongst members of multidisciplinaryteams within General Medicine (e.g. medical, nursingand pharmacy), each taking ownership in addressingspecific pharmacologic and non-pharmacologic aspects.The aim of this quality improvement project was toexplore the feasibility of such shared responsibilityapproach in improving education and management ofinpatients with heart failure, in lieu of a dedicated heartfailure nurse, using process measures as outcomes.

Methods

Setting

This was a single-centre prospective non-randomisedbefore-and-after observational quality improvementintervention, evaluating changes in process measuresafter implementation of a general medicine heart failurecare programme at The Alfred hospital, Melbourne,Australia. The hospital is a 350-bed university affiliatedquaternary hospital, with the General Medical Unit(GMU) consisting of five medical inpatient teams havingapproximate bed occupancy of 20–30% of total acutehospital beds.

Study population

All consecutive patients admitted to GMU with a clinicaldiagnosis of acute decompensated heart failure (ADHF)were included in this study. All eligible patients wereidentified from the daily admission lists by the studyinvestigators (OI and VN). The pre-implementation(baseline) and post-implementation (intervention) auditperiods were over 8 weeks each between 21 March 2016and 15 May 2016, and between 1 August 2016 and26 September 2016 respectively. Patients were excludedfrom the final comparative analyses if they received end-of-life care during the hospital admission. Approval toconduct this study was granted by the institutionalHuman Research Ethics Committee (approval number:Alfred 182/16).

Intervention

A multidisciplinary programme, named General Medi-cine Heart Failure Care Program (GM-HFCP), wasdesigned specifically for general medicine inpatientsadmitted with ADHF. GM-HFCP was designed by a steer-ing committee within the GMU as a collaborative qualityimprovement project between general medicine

physicians, a heart failure physician, heart failure nurse,general medical nurses and pharmacists. The individualcomponents of GM-HFCP and task assignments wereoutlined in Table 1. The programme was developed overa 6-week period from mid-May 2016 (at the end of base-line audit period) to early July 2016.Programme implementation period was over 2 weeks

from 18 July 2016 to 31 July 2016. This involved pro-viding education to members of the GMU team includ-ing general medicine doctors (interns and registrars),nursing and pharmacists on heart failure managementas per Australian National Heart Foundation Heart Fail-ure guidelines1 and task assignment for GM-HFCP.Implementation was conducted by the study investiga-tors, and a one-off reinforcement session was alsoprovided on 15 August 2016, the first 2 weeks into thepost-implementation period. A heart failure manage-ment pack was developed, which included patient infor-mation, daily weight monitoring chart and heart failureaction plan based on the guidelines,1 to be provided topatients during admission and upon discharge.

Data collection

Demographic and clinical details were extracted frommedical records. Data on non-pharmacologic and phar-macologic management, and provision of heart failureeducation and/or action plan were collected using a casereport form at two-time points: 1 day after admissionand at discharge. Non-pharmacologic ward-based man-agement data were collected including daily weightrecordings, daily fluid balance recordings, daily renalfunction and electrolyte (UEC) monitoring, orthostaticblood pressure (BP) monitoring, salt restriction statusand whether patient education and action plan wereprovided and documented. Pharmacologic managementdata that were collected included use of beta-blockers,angiotensin converting enzyme inhibitor/angiotensin IIreceptor blocker (ACEi or ARB), frusemide,

Table 1 Overview of General Medicine Heart Failure Care Programmodel

Medicalteam

Empower patients with self-management strategiesthrough:

• Symptom education• Lifestyle modification (including referral to cardiac

rehab programme)• Discharge action plan

Nursing Educate patients with regards to:• Importance of baseline dry weight• Daily weight monitoring• Fluid restriction status

Pharmacy Provide medication education on discharge

Heart Failure Care Program


179

spironolactone and digoxin. If these medications werenot prescribed, reasons were recorded.

Definitions

Left ventricular ejection fraction (LVEF) of ≥50% wasdefined as heart failure with preserved ejection fraction(HFPEF) and LVEF of <50% was defined as heart failurewith reduced ejection fraction (HFREF), in keeping withAmerican Heart Association (AHA) classification.6 Lastavailable transthoracic echocardiogram result on elec-tronic medical records was used to determine heart failureclassification. ‘Appropriate prescription/non-prescriptionof medications’ was defined as patients being on appropri-ate classes of medications (ACEi/ARB, beta-blockersand aldosterone antagonists) for correct indication(i.e. HFREF), and if not prescribed, with appropriate rea-sons or contraindication documented in notes/results(e.g. acute kidney injury, hyperkalaemia or side effects) orobservation charts (e.g. bradycardia, hypotension ororthostatic intolerance).

Outcome measures

Outcome measures included changes in rates of pharma-cologic prescription, adherence to non-pharmacologicward-based management components and patienteducation and action plan provision between pre-implementation and post-implementation periods.Appropriate use of pharmacologic management inHFREF, inpatient mortality rates and 30-day readmissionrates were also measured.


All categorical variables were expressed as counts andproportions. Numerical variables were assessed for nor-mality of distribution, and expressed as median andinterquartile ranges. Comparisons between pre- andpost-implementation process measures were performedusing bivariate models employing Mann–Whitney U-tests, Chi-squared and Fisher exact tests. A two-tailed P-value of <0.05 was taken as a statistically significantresult.

Results

A total of 108 patients with ADHF was included overthe entire study period (52 during pre-implementationand 56 during post-implementation periods). Themedian age of patients was 84 (80–89) years and47 (44%) were female (Table 2). Nine (8%) patientsreceived end-of-life care during acute admission(Table 2) and their results were censored from compar-ative analyses.

Comparison of clinical characteristics between thepatients in two study periods is summarised in Table 2.

Outcomes

After GM-HFCP implementation, there were statisticallysignificant improvements in non-pharmacologic man-agement especially patient education regarding fluidrestriction, salt restriction, weight monitoring and provi-sion of heart failure action plan on discharge (Table 3).Improvements were also noted in daily weight and fluid

Table 2 Clinical characteristics and outcomes of patients (overall, pre- and post-implementation periods)

Overall (n = 108) Pre-implementation (n = 52) Post-implementation (n = 56) P-value

Clinical characteristicsMedian age with IQR 84 (80–89) 83 (80–89) 85 (80–89) 0.84Female, n (%) 47 (44) 29 (56) 18 (30) 0.01HFREF, n (%) 41 (38) 15 (29) 26 (47) 0.07HFPEF, n (%) 54 (50) 31 (60) 23 (41) 0.08No TTE report, n (%) 13 (12) 6 (11) 7 (12) 1.00Admission from home, n (%) 98 (91) 45 (86) 53 (95) 0.34Dementia, n (%) 18 (17) 9 (17) 9 (16) 1.00

NYHA functional classI, n (%) 1 (1) 1 (2) 0 0.48II, n (%) 57 (53) 41 (79) 16 (28) 0.0001III, n (%) 48 (44) 10 (19) 38 (68) 0.0001IV, n (%) 2 (2) 0 2 (4) 0.49

OutcomesEnd-of-life care referral, n (%) 5 (5) 2 (3) 3 (5) 1.00Deaths, n (%) 4 (4) 1 (2) 3 (5) 0.61

HFREF, heart failure with reduced ejection fraction; HFPEF, heart failure with preserved ejection fraction; IQR, inter-quartile range; NYHA, New YorkHeart Association; TTE, transthoracic echocardiogram.

Wahbi Izzettin et al.


180

balance documentations and orthostatic blood pressuremeasurements, although not statistically significant.The overall rates of medication prescription on dis-

charge also improved modestly after GM-HFCP imple-mentation, particularly ‘appropriate use’ of beta blockersand ACEi/ARB in patients with HFREF (P > 0.05 for allvariables) (Table 3). There was no difference in inpatientmortality or 30-day readmission rates between bothgroups (n = 24; 12, 25% vs 12, 24% in each period,P = 1.0).

Discussion

This pilot project showed that it is possible to share the

responsibilities of a heart failure nurse amongst individ-

ual members of the multidisciplinary care teams within

general medicine to address specific management aspects

of heart failure care in a decentralised but efficient and

effective manner. Overall, we found improvements in

multiple aspects of patient care between pre- and post-

implementation periods, lending support to this model

of care.In patients who have been hospitalised with ADHF,

specialist heart failure nurse input has been shown to beassociated with better health outcomes.2–5 However,there is often no dedicated heart failure nurse in GMUsto address and reinforce all management aspects, poten-tially disadvantaging these patients. Other studies utilis-ing similar multidisciplinary strategies to ours haveshown comparable improvements, in terms of patienteducation, medication management, nutritional guid-ance and adherence to cardiac rehabilitation.7,8 Indeed,education interventions, regardless of delivery by a dedi-cated heart failure nurse or by multidisciplinary teams,have been shown to improve heart failure knowledge,quality of life, enhance self-care behaviours and reducereadmission rates.2–9 Taken together, these studies and

Table 3 Non-pharmacologic management, patient education and pharmacologic management results of pre- and post-implementation periods

Pre-implementation (n = 49) Post-implementation (n = 50) P-value

Non-pharmacologic managementDaily weight documentation, n (%) 42 (85) 45 (90) 0.55Daily fluid balance chart, n (%) 46 (93) 47 (94) 1.00Daily UEC checks, n (%) 42 (85) 46 (92) 0.35Orthostatic BP ordered, n (%) 3 (6) 8 (16) 0.19Orthostatic BP checked, n (%) 3 (6) 8 (16) 0.19Salt restriction ordered, n (%) 3 (6) 16 (32) 0.002

Patient educationFluid restriction, n (%) 6 (12) 15 (30) 0.04Weight monitoring, n (%) 5 (10) 14 (28) 0.03Action plan given on discharge, n (%) 2 (4) 14 (28) 0.002

Pharmacologic management – OverallBeta blockers, n (%) 25 (51) 33 (66) 0.15ACEi/ARB, n (%) 32 (65) 26 (52) 0.22Frusemide, n (%) 45 (91) 49 (98) 0.20Spironolactone, n (%) 10 (20) 17 (34) 0.17Digoxin, n (%) 13 (26) 13 (26) 1.00

Pharmacologic management – HFREF Pre-implementation (n = 13) Post-implementation (n = 24) P-value

Beta blockers, n (%) 7 (53) 17 (71) 0.47ACEi/ARB, n (%) 6 (46) 12 (50) 1.00Furusemide, n (%) 12 (92) 23 (95) 1.00Spironolactone, n (%) 4 (30) 9 (37) 0.73Digoxin, n (%) 5 (38) 6 (25) 0.46

Appropriate prescription/non-prescriptionof medications in HFREF†

Beta blockers, n (%) 11 (84) 21 (87) 1.00ACEi/ARB, n (%) 10 (76) 19 (79) 1.00Spironolactone, n (%) 6 (46) 16 (67) 0.30

†Total number of patients who had beta-blockers, ACEi/ARB and spironolactone prescribed for correct indication (i.e. HFREF) and if not prescribed, con-traindications for non-prescription (e.g. hyperkalaemia, orthostatic hypotension, etc.) were clearly documented. Total ‘n’ for appropriateness is largerthan the ‘n’ for pharmacologic management as patients with appropriate non-prescription have been taken into account. ACEi/ARB, angiotensin con-verting enzyme inhibitor/angiotensin II receptor blocker; BP, blood pressure; HFREF, heart failure with reduced ejection fraction; UEC, urea andelectrolytes.



181

ours support a model of care in which members of themultidisciplinary teams can be empowered to addressand educate effectively specific patient goals in heart fail-ure management. Further, such an approach may easilybe extended to include other members of Allied Health.However, in order for this approach to be successful, sus-tained strong multidisciplinary leadership is required anda mature organisational culture of collaboration forpatient safety and improvement in care delivery.

Our study demonstrated high compliance rates withdaily weight and fluid balance monitoring by ward nurses,even before implementation of the care programme. Thisis likely due to well-established nursing practices, whichare already ingrained into the routine care of patientswith ADHF. Weight and fluid intake monitoring isan important aspect of heart failure management, a pro-cess which starts as inpatient and continues post-discharge.10–12 We found that opportunities exist whereroutine nursing practices can be translated into patienteducation. As patients’ acute clinical condition improves,ward nurses can start encouraging patients to monitortheir own weight and fluid balance independently undersupervision whilst still in hospital, thereby further reinfor-cing knowledge and confidence in self-management.7,12

Alongside medical doctors, ward nurses can therefore playa vital role educating the ‘action plan’ to the patient priorto discharge.

With regards to medications, in-hospital initiation ofneurohormonal antagonists has been demonstrated toincrease their continued long-term use after hospital dis-charge.13,14 Although our pharmacologic prescriptionrates were lower than published national data,15 due toour unique patient cohort of predominant elderly withmultiple co-morbidities, they were comparable to pre-scription rates from other major centres serving similartype of patient population.16 Orthostatic blood pressuremeasurements may help guide medication titration inthese patients, who are also at risk of falls.

Despite improvements in these process measures, wefound that implementation of GM-HFCP did notimprove readmission and mortality rates in our study.However, readmission rates in our study were similar topreviously published data, which vary between 20 and49% within 1–6 months of discharge.17,18

Our study has several limitations. First, the sample sizewas relatively small, which may limit the ability to dem-onstrate differences after the intervention. Second, theimplementation period was relatively short. As with allquality improvement studies, delivery of well-resourcededucation programmes, repeated reinforcement of mes-sages delivered over a longer period and promoting

organisation culture in patient safety and care improve-ment are the key to success.19 Third, we conservativelyassumed that lack of documentation on medical notesmeant education was not provided whilst this may nothave been the case. Fourth, only objective clinical out-comes were captured. Other patient-centred outcomessuch as degree of understanding and satisfaction withthe education content, attitude toward medication bur-den and confidence levels in managing heart failure onown were not measured, neither was staff satisfactionnor staff time involved. These outcomes could poten-tially influence the success rates of programme imple-mentation. Finally, although this study was conducted ina well-resourced institution based on national guidelines,this may not necessarily reflect routine practice orresource availability at other institutions.

Despite these limitations, we showed that a multidisci-plinary care programme in a busy general medical wardimproved many aspects of heart failure care. Weemployed a before-and-after observational study designto avoid the Hawthorne effect by care providers, whichmay be associated with cross-over or step-wedge trialdesigns. Improvements to this model of care should befurther explored in well-structured pragmatic rando-mised controlled trials, and practice incorporated intoroutine heart failure care or care bundles, which are cur-rently being implemented at our institution.

Conclusion

This prospective observational study demonstrated that itis possible to share the roles of a heart failure nurseamongst members of the multidisciplinary team, withsimilar rates of delivery of pharmacologic and non-pharmacologic management aspects. This study alsofound that routine nursing practices can be translatedinto patient education. Future models can be furtherrefined based on our model, which takes the ‘divide andconquer’ approach, sharing responsibility and account-ability toward a common treatment goal together as aunit. Structured quality improvement research employ-ing a cluster randomised control design can further iden-tify strategies to encourage multidisciplinary teams todeliver innovative heart failure care.

Acknowledgement

We thank Mr Thomas Martin (Monash University) andMs Kellie Easton (Alfred Hospital) for their invaluablehelp in the project.

Wahbi Izzettin et al.


182

References

1 Heart Foundation. A Systematic Approach

to Chronic Heart Failure Care: A Consensus

Statement. Melbourne: The Foundation;

2013 [cited 2017 Mar 10]. Available

from URL: https://heartfoundation.org.

au/images/uploads/publications/HF_

CHF_consensus_web_FINAL_SP.pdf.

2 Kommuri NV, Johnson ML,

Koelling TM. Relationship between

improvements in the heart failure

patient disease specific knowledge and

clinical events as part of a randomized

controlled trial. Patient Educ Couns 2012;

86: 233–8.

3 Manning S. Bridging the gap between

hospital and home: a newmodel of care

for reducing readmissions rates in chronic

failure. J Cardiovasc Nurs 2011; 26: 368–76.

4 Goodlin SJ, Trupp R, Bernhardt P,

Grady KL, Dracup K. Development and

evaluation of the "advanced heart

failure clinical competence survey": a

tool to assess knowledge of heart failure

care and self-assessed competence.

Patient Educ Couns 2007; 67: 3–10.

5 Anderson JH. The impact of using

nursing presence in a community heart

failure program. J Cardiovasc Nurs 2007;

22: 89–94.

6 Yancy CW, Jessup M, Bozkurt B,

Butler J, Casey DE Jr, Drazner MH et al.

2013 ACCF/AHA guideline for the

management of heart failure: executive

summary: a report of the American

College of Cardiology Foundation/

American Heart Association Task Force

on practice guidelines. Circulation 2013;

128: 1810–52.

7 Kinugasa Y, Kato M, Sugihara S,

Yanagihara K, Yamada K, Hirai M et al.

Multidisciplinary intensive education in

the hospital improves outcomes for

hospitalized heart failure patients in a

Japanese rural setting. BMC Health Serv

Res 2014; 14: 351–9.

8 Williams G, Akroyd K, Burke L.

Evaluation of the transitional care

model in chronic heart failure. Br J Nurs

2010; 19: 1402–7.

9 Kent B, Cull E, Phillips NM. A

systematic review of the effectiveness of

current interventions to assist adults

with heart failure to comply with

therapy and enhance self-care

behaviours. JBI Libr Syst Rev 2011; 9:

2572–626.

10 Gibbs CR, Jackson G, Lip GY. ABC of

heart failure, non-drug management.

BMJ 2000; 320: 366–9.

11 Krum H, Driscoll A. Management of

heart failure. Med J Aust 2013; 199:

334–9.

12 Heart Foundation. Living Well with

Chronic Heart Failure. Melbourne: The

Foundation; 2017 [cited 2017 Mar 10].

Available from URL: https://www.

heartfoundation.org.au/images/

uploads/publications/Living_well_with_

heart_failure_2017.pdf.

13 Gattis WA, O’Connor CM, Gallup DS,

Hasselblad V, Gheorghiade M, IMPACT-

HF Investigators and Coordinators.

Predischarge initiation of carvedilol in

patients hospitalized for decompensated

heart failure: results of the Initiation

Management Predischarge: Process for

Assessment of Carvedilol Therapy in

Heart Failure (IMPACT-HF) trial. J Am

Coll Cardiol 2004; 43: 1534–41.

14 Lappe´ JM, Muhlestein JB, Lappe´ DL,

Badger RS, Bair TL, Brockman R et al.

Improvements in 1-year cardiovascular

clinical outcomes associated with a

hospital-based discharge medication

program. Ann Intern Med 2004; 141:

446–53.

15 Lefkovits J, Brennan A, Dinh D,

Brien R, Driscoll A, Reid C. The

Victorian Cardiac Outcomes Registry

annual report 2015. Monash University,

DEPM 2016. Report No 3, pages

74 [cited 2017 Mar 12]. Available from

URL: http://www.vcor.org.au/sites/

default/files/VCOR_files/2015_vcor_

annual_report.pdf

16 Parmar KR, Xiu PY, Chowdhury MR,

Patel E, Cohen M. In-hospital

treatment and outcomes of heart

failure in specialist and non-specialist

services: a retrospective cohort study

in the elderly. Open Heart 2015; 2:

e000095.

17 Stewart S, Carrington MJ,

Marwick TH, Davidson PM,

Macdonald P, Horowitz JD et al. Impact

of home versus clinic-based

management of chronic heart failure

the WHICH? (Which Heart Failure

Intervention Is Most Cost-Effective &

Consumer Friendly in Reducing

Hospital Care) multicenter,

randomized trial. J Am Coll Cardiol

2012; 60: 1239–48.

18 Krumholz HM, Merrill AR, Schone EM,

Schreiner GC, Chen J, Bradley EH et al.

Patterns of hospital performance in

acute myocardial infarction and heart

failure 30-day mortality and

readmission. Circ Cardiovasc Qual

Outcomes 2009; 2: 407–13.

19 Gilbert G, Sawyer R, McNeill EB. Health

Education: Creating Strategies for School

and Community Health. London: Jones

and Barlett Publishers; 2011.



183

https://heartfoundation.org.au/images/uploads/publications/HF_CHF_consensus_web_FINAL_SP.pdf



https://www.heartfoundation.org.au/images/uploads/publications/Living_well_with_heart_failure_2017.pdf




http://www.vcor.org.au/sites/default/files/VCOR_files/2015_vcor_annual_report.pdf



doi:10.1111/imj.13551

Delays in presentation and diagnosis of pulmonarytuberculosis: a retrospective study of a tertiary health servicein Western Melbourne, 2011–2014Eloise Williams ,1 Allen C. Cheng,2 Garry P. Lane1,3 and Stephen D. Guy1,3

1Department of Infectious Diseases, Western Health, 2Department of Epidemiology and Preventive Medicine, Infectious Diseases, Monash UniversityFaculty of Medicine Nursing and Health Sciences, and 3Department of Medicine, Melbourne Medical School, Western Precinct, The University ofMelbourne, Melbourne, Victoria, Australia

Key wordstuberculosis, pulmonary, pneumonia,tuberculosis, epidemiology, communicabledisease control.

CorrespondenceEloise Williams, Department of InfectiousDiseases, Alfred Health, 55 Commercial Road,Prahran, Melbourne, Vic. 3004, Australia.Email: [email protected]


Abstract

Background: Effective tuberculosis (TB) control relies on early diagnosis and prompt

treatment commencement.

Aim: To investigate delays in presentation and diagnosis of pulmonary TB (PTB) in a

low incidence setting in Western Melbourne.

Methods: A single-centred retrospective observational cohort study of symptomatic

patients ≥ 18 years newly diagnosed with PTB that were commenced on treatment

between 1 December 2011 and 1 December 2014 at a tertiary teaching hospital in

Western Melbourne. Main outcome measures included median duration of patient,

health system and total delays to diagnosis of PTB and clinical factors associated with

prolonged patient (>35 days) and health system (>21 days) delay.

Results: A total of 133 patients were included. The median (range) duration of patient,

health system and total delay to diagnosis were 28 (0–610), 18 (0–357) and 89 (1–730)

days respectively. Prolonged patient delay was associated with being from a country with

an annual TB incidence of <50/100 000 (odds ratio (OR) 5.98, 95% confidence interval

(CI) 1.19, 29.98) and diabetes mellitus (OR 3.02, 95% CI 1.04, 8.78) in multivariate analy-

sis. Being Australian-born or a resident of Australia ≥6 years (OR 0.03, 95% CI 0.12, 0.74;

OR 0.30, 95% CI 0.00, 0.033 respectively) was associated with reduced patient delay.

Conclusions: In this low-incidence, high-resource setting, patient delays contribute

most to total delay in PTB diagnosis. Strategies addressing this aspect of the TB diagnosis

pathway, such as health literacy and promotion programmes for new migrants and

raised primary healthcare awareness, could have the largest impact on reducing total

PTB diagnosis delays.

Introduction

Tuberculosis (TB) is a major global health problem; anestimated 10.4 million people developed TB in 2015 and1.4 million people died from the disease.1 Early diagnosisand prompt commencement of treatment are importantfor effective TB control. Delayed diagnosis may worsenthe severity of disease and increase the risk of death.2

Australia has maintained an annual incidence rate ofactive TB of 5.2–7.0 cases per 100 000 since the mid-1980s.3 Of the 1263 notified cases (nationally) in 2013,58% had pulmonary tuberculosis (PTB).3 Over the pastdecade, Australia has experienced a modest increase inTB incidence, from a rate of 5 per 100 000 in 2003 to 6.2

per 100 000 in 2011. This is likely due to an increasingmigrant population from high-incidence regions of theworld; 88% of patients diagnosed with TB in Australia in2013 were born overseas.3 In comparison to the risk ofactive TB (over the past 10 years) in the Australian-bornpopulation of approximately 1 in 100 000,3 the risk ofAustralian immigrants reflects the TB incidence in theindividual’s country of origin.4 The risk of active TB ishighest for individuals in the first 6 years after migration,and approximates 100–150 per 100 000 person-years forthose from Asia, Africa and the Pacific during this period.Risk then falls to approximately 50 per 100 000 person-years by 12 years.4

Several studies in various international settings haveinvestigated time delays in the diagnosis of PTB. Delaysmay occur before medical presentation (‘patient’ delays)or after medical presentation (‘health system’ delay).



184

However, there is no consistent definition of ‘delayed’(i.e. prolonged time until) diagnosis for PTB in the inter-national literature. Previous studies have arbitrarily nomi-nated a certain number of days from onset of symptomsto differentiate ‘delayed’ from ‘acceptable’ time periods;the number of days constituting ‘delayed diagnosis’, how-ever, has varied widely. Definitions of total delay havevaried from 28 to 98 days5–10; patient delays have beendefined variously from 30 to 60 days,7,8,10 and health sys-tem delays from 7 to 56 days.7,8,10,11

Since the last study of delayed diagnosis in Australia15 years ago,9 there have been major sociodemographicchanges in the Australian population and new rapid testsfor TB based on nucleic acid detection. We sought toexplore factors associated with diagnostic delays in alarge community hospital in Western Melbournebetween 2011 and 2014.

Methods

Study design and setting

A retrospective cohort study of patients diagnosed andtreated for PTB at a single health service in Victoria wasundertaken between 1 December 2011 and 1 December2014. Western Health is a multi-campus tertiary healthservice in the western suburbs of Melbourne (Victoria)that provides healthcare services to a population ofapproximately 800 000 people. Western Health treatsapproximately 25–30% of Victoria’s approximately350 cases of TB each year.

Case finding, inclusion and exclusion criteria

Patients were identified by two methods: a hospital data-base of discharge coding and notification data from theVictorian Tuberculosis Program. In Victoria, the notifica-tion of confirmed cases of TB is required under state legis-lation. The medical records of all additional casesidentified by Western Health Australian Discharge RelatedGroup (A-DRG) codes for pulmonary or miliary TB (ICD-10-AM codes A15, A16 and A19) were reviewed.Patients were included if they were symptomatic with

PTB, including disseminated TB with lung involvementor extra-pulmonary TB at a contiguous site combinedwith PTB, and commenced on treatment between1 December 2011 and 1 December 2014. Patients under18 years, those with extra-pulmonary TB without pul-monary involvement and asymptomatic patients wereexcluded.

Definitions of delays and prolonged delays

Initial healthcare contact was defined as the date thepatient reported attending a healthcare provider, whetherin primary care or hospital-based secondary or tertiaryhealth services. The date of ‘diagnosis’ was defined as thedate of commencing anti-TB therapy consistent with pre-vious studies. Patient delay was defined as the time ofonset of symptoms suggestive of PTB until initial health-care contact; health system delay as the time from first ini-tial healthcare contact until diagnosis and hospital delay astime from hospital admission to diagnosis. Total delay wasthe sum of patient delay and health system delay. Wedefined a prolonged patient delay and prolonged healthsystem delay as a duration greater than the median of thesample population, rounded up to the nearest 1 week.Countries of origin were defined as high risk if the

annual incidence of TB was ≥50/100 000 and low risk ifthe annual incidence of TB was <50/100 000 accordingto the WHO Global TB report, 2014.1 The nucleic acidtest (NAT) used for rapid molecular diagnosis of TB inthis population was the GeneXpert MTB/RIF (Cepheid,Sunnyvale, CA, USA), a fully automated DNA-PCR tech-nique for the detection of TB and rifampicin-resistancemutations.


Univariate and multivariate regression analyses wereperformed to determine factors associated with pro-longed delays. Univariate analyses were performed first,and factors with P < 0.2 were considered for furtheranalysis in the multivariate analysis. For final modelselection, a modified backwards stepwise selection pro-cess was used, checking the effect of dropping variableson other odds ratios to obtain a parsimonious model. Forall tests, two-sided P-values with an α ≤ 0.05 level of sig-nificance were used.

Ethical considerations

This study was approved by the Western Health LowRisk Human Research Ethics Panel prior to its com-mencement (project number LNR/15/WH/24).

Results

Patient inclusion

Of 331 patients identified, 133 patients met eligibility cri-teria and were included in the study (Fig. 1). Notably,37 patients were excluded due to asymptomatic PTB,which may be more common at this health service due

Delays in diagnosis of tuberculosis


185

to its role as the coordinating centre for visa screening ofchest X-ray (CXR) abnormalities in Victoria.

Patient demographics

Of these 133 patients, median age was 35 years, and 67%were male (Table 1). Most spoke a language other thanEnglish at home (90%) and were born overseas (92%),most commonly from the Western Pacific Region, partic-ularly from Vietnam (Fig. 2); 85% (85%) had migratedfrom a high-risk country, and 44% had resided inAustralia <6 years. Only two (2%) were HIV positive.

Radiology

Of the patients, 130 had a CXR available, with mediantime from CXR to diagnosis of 13 days (IQR 4–36 days);106 (80%) had CXR features suspicious of active PTB(Table 2). Of the patients, 82 (62%) had computed

tomography (CT) of the chest performed, with a mediantime from CT until diagnosis of 9 days (IQR 2–29). All82 patients who underwent CT chest had abnormalitiesfound.

Microbiology

Of the patients, 110 had expectorated sputa sent foracid-fast bacilli (AFB) smear and mycobacterial culture;39 (36%) patients had a positive AFB smear and86 (78%) a positive sputum mycobacterial culture(Table 2). Median time from sputum assay to diagnosiswas 4 days (IQR 2–29 days). Of the patients, 57 (43%)underwent bronchoscopy; median time from bronchos-copy until diagnosis was 4 days (IQR 1–13). Of thepatients, 57 patients had NAT performed on sputum;47 (83%) of these were positive. In those who had NATperformed on sputa, the yield of positive NAT in AFBsmear-positive sputa was 33 of 35 (94%), whereas the

Victorian Department of Health

Database

(n = 219)

A-DRG Database

(n = 112)

Total number of patients(n = 133)

Records excluded

(n = 110) • 1 outside timeframe • 13 extrapulmonary TB • 15 alternative final

diagnosis • 81 accounted for

from Victorian

Department of Health

Database (duplicate)

Total patients

included

(n = 131)

Total patients

included

(n = 2)

Patients excluded

(n = 88)

• 39 outside timeframe • 4 less than 18 years

old • 4 managed at

different site • 37 asymptomatic • 3 extra-pulmonary TB • 1 left Australia prior

to diagnosis

Figure 1 Patient inclusion flow diagram.

Williams et al.


186

Table 1 Patient demographic characteristics

Characteristic Patients (n = 133), n (%)

AgeMedian [IQR] 35 [26–57]

DemographicsMale 89 (67)Native Language English 13 (10)Employed 47 (35)

Time since migrationAustralian born 11 (8)Less than 6 years 59 (44)6 or more years 63 (47)

Migration riskAustralian-born 11 (8)Low-risk country 9 (7)High-risk country 113 (85)

SymptomsCough 124 (93)Productive cough 86 (65)Haemoptysis 25 (19)Dyspnoea 38 (29)Chest pain 38 (29)Weight loss 75 (56)Sweats 50 (38)Fevers 75 (56)Fatigue 31 (23)

ComorbiditiesCurrent smoking 48 (36)Excessive alcohol intake 18 (14)Diabetes mellitus 31 (23)Chronic liver disease 12 (9)Asthma/COPD 18 (14)Malignancy 7 (5)HIV positive 2 (2)Injecting drug use 5 (4)

COPD, chronic obstructive pulmonary disease; HIV, human immunodefi-ciency virus; IQR, interquartile range.

Table 2 Diagnostic and outcome data

Characteristic Patients (n = 133), n (%)

MicrobiologySputum 110/133AFB smear positive 39/110 (36)Mycobacterial culture positive 86/110 (78)NAT positive 47/57 (83)

Bronchoscopy specimen 57/133 (43)AFB smear positive 9/57 (16)Mycobacterial culture positive 42/53 (79)NAT positive 29/43 (67)

Overall microbiology 122/133 (92)Positive 11/133 (8)Negative

Radiology†CXR available 130/133 (98)CXR suggests active PTB 106/130 (82)CXR – cavitation 29/130 (22)CT chest performed 82/133 (62)CT chest abnormal 82/82 (100)

DiagnosisPulmonary TB 97 (73)Disseminated TB 16 (12Pulmonary TB with contiguousextra-pulmonary involvement

20 (15)

OutcomeCompleted treatment 118 (89)Continued treatment 1 (1)Treated at another site 7 (5)Lost to follow up 3 (2)Died 4 (3)

†Radiological features not mutually exclusive. AFB, acid fast bacilli; CXR,chest X-ray; NAT, nucleic acid test; PTB, pulmonary tuberculosis; TB,tuberculosis.

WHO Region

0

5

10

15

20

25

30

35

Ind

ia

Nep

al

My

anm

ar

Ch

ina

Ban

gla

des

h

Dem

ocr

atic

Peo

ple

's…

Eas

t T

imo

r

Pak

ista

n

Sri

Lan

ka

Th

aila

nd

Vie

tnam

Ph

illi

pin

es

Au

stra

lia

My

anm

ar

Cam

bo

dia

Ind

on

esia

Ch

ina

Mal

aysi

a

New

Zea

lan

d

Sam

oa

Tai

wan

Afg

han

ista

n

Pak

ista

n

Su

dan

Eri

trea

Eth

iop

ia

Ken

ya

So

mal

ia

Cro

atia

Ger

man

y

Bo

snia

Ro

man

ia

Sco

tlan

d

Th

e F

orm

er Y

ug

osl

av…

Uru

gu

ay

OHAPORUEORFAORMEORPWORAES

Num

ber

of p

eopl

e

Country of Birth

Figure 2 Patient country of birth, categorised by WHO Region. WHO Regions are SEARO: South-East Asia Regional Office; WPRO: Western Pacific Regional

Office; EMRO: Eastern Mediterranean Regional Office; AFRO: Africa Regional Office; EURO: Europe Regional Office; PAHO: Americas Regional Office.



187

yield of positive NAT on smear-negative sputa was 14 of27 (52%). Of the patients who had a bronchoscopy,46 (81%) had NAT performed on the bronchoscopyspecimen. The yield of positive NAT on smear-positivebronchoscopy specimens was 6 of 6 (100%), whereasthe yield of positive NAT on smear-negative bronchos-copy specimens was 23 of 40 (58%). Of the 26 patientswith a prospectively collected NAT, median time fromNAT to diagnosis was 1 day (IQR 0–1 days).

Delays

Patient delay was the predominant delay, with a medianof 28 days from symptom onset until first health servicecontact (Table 3). Median health system delay was18 days. We therefore defined a prolonged patient delayas more than 35 days and a prolonged health systemdelay as more than 21 days. According to this definition,55 of 133 (41.3%) patients in this study had a prolongedpatient delay, and 59 of 133 (44.4%) had a prolongedhealth system delay.

Prolonged patient delay

In univariate analysis, prolonged patient delay was asso-ciated with weight loss (odds ratio (OR) 2.16, 95% confi-dence interval (CI) 1.06, 4.43) and inversely associatedwith presence of fatigue (OR 0.41, 95% CI 0.16, 0.99)(Table 4). In multivariate analysis, migration from a low-risk annual TB incidence country (OR 5.98, 95% CI1.19, 29.98) was the strongest predictor of patient delay(Fig. 3). Weight loss (OR 2.23, 95% CI 1.00, 4.96) anddiabetes mellitus (OR 3.02, 95% CI 1.04, 8.78) were alsoassociated with prolonged patient delay. Conversely,migration to Australia ≥ 6 years prior or beingAustralian born was associated with reduced patientdelay (OR 0.30, 95% CI 0.12, 0.74; OR 0.03, 95% CI0.00, 0.33 respectively).

Prolonged health system delay

In univariate analysis, patient age of 65 years or older(OR 3.07, 95% CI 1.21, 7.79) (Fig. 4) and patient reviewin outpatient clinic (OR 7.67, 95% CI 3.51, 16.77) wereassociated with prolonged health system delay (Table 5).Factors associated with reduced risk of prolonged health

system delay were presence of cough (OR 0.09, 95% CI0.01, 0.72), review in the emergency department(OR 0.33, 95% CI 0.16, 0.69) or admission to hospital(OR 0.17, 95% CI 0.06, 0.46). Additionally, for patientswho were able to expectorate sputum, AFB smear posi-tivity (OR 0.23, 95% CI 0.09, 0.56) and NAT positivity(OR 0.14, 95% CI 0.03, 0.62) were associated withreduced risk of prolonged health system delay. Patientswho had positive sputum NAT had a median health sys-tem delay of 9 days, compared with a median health sys-tem delay of 41 days in those who had a negativesputum NAT and 21 days in those who did not have asputum sent for NAT. Having cavitating lesions on CXR(OR 0.25, 95% CI 0.09, 0.67) was also associated withreduced health system delay.

In multivariate analysis, patient age of 65 or older wasassociated with prolonged health system delay (OR 4.29,95% CI 1.55, 11.89), whereas presence of cough(OR 0.09, 95% CI 0.01, 0.79) and being admitted to hos-pital were associated with reduced health system delays(OR 0.28, 95% CI 0.09, 0.92).

Discussion

This single-centre retrospective review of delays in diag-nosis of PTB in Western Melbourne illustrated thatdelays between symptom onset and treatment com-mencement are predominantly due to patient delay inpresentation for care. The median duration of delaysfound in this study align with those found in an interna-tional systematic review of the topic in 2009, whichincluded 52 studies and found an average patient delayof 28.7 days and average health system delay of25 days.12 Findings are also comparable to the lateststudies conducted in Australia on this topic.9,13

A study published in 20019 included 782 symptomatic,culture-positive cases of PTB based on QueenslandTuberculosis Control Centre TB case notification formdata between 1985 and 1998. Median patient delay inthis study was 29 days and median health system delaywas 22 days. Despite advances in rapid diagnosis, delaysare only marginally shorter than those described 15 yearsago. A recently published study by Dale et al.13 involved5106 Victorian patients identified through notificationform data between 2002 and 2015 with PTB and extra-pulmonary TB and compared the management ofpatients receiving private versus public healthcare. Themedian patient delay was 18 days and median health-care delay was 22 days in those receiving public health-care in this cohort. The duration of health care delayfound by Dale et al. correlates well with our study. Thediscrepancy in patient delay between our study and thatby Dale et al. may be a factor of information bias, with

Table 3 TB diagnosis delays

Primary outcome (days) Median (IQR) Range

Patient delay 28 (13–90) 0–610Health system delay 18 (7–53) 0–357Hospital delay 4 (2–9) 0–159Total delay 89 (33–151) 2–730

IQR, interquartile range; TB, tuberculosis.

Williams et al.


188

Table 4 Risk factors for prolonged patient delay

VariableUnivariate analysis Multivariate analysis

n (%) OR (95% CI) P OR (95% CI) P

Age≥65 years 12/24 (50) 1.53 (0.63, 73) 0.34<65 years 43/109 (39)

GenderFemale 17/44 (39) 0.85 (0.40, 1.77) 0.66Male 38/89 (43)

English spoken at homeYes 3/13 (23) 0.39 (0.10, 1.50) 0.16 NINo 52/120 (43)

EmploymentWorking 18/47 (38) 0.82 (0.40, 1.70) 0.60Not working 37/86 (43)

Time since migration<6 years 29/59 (49) 1.00 NA 1.00≥6 years 24/63 (38) 0.63 (0.29, 1.39) 0.30 (0.12, 0.74) 0.01No migration 2/11 (18) 0.22 (0.02, 1.27) 0.03 (0.00, 0.33) <0.01

TB risk in country of migrationHigh risk 47/113 (42) 1.00 NA 1.00 0.03Low risk 6/9 (67) 2.81 (0.56, 18.1) 5.98 (1.19, 29.98)No migration 2/11 (18) 0.31 (0.03, 1.62) —

CoughYes 52/124 (42) 1.44 (0.35, 6.04) 0.61No 3/9 (33)

HaemoptysisYes 12/25 (48) 1.40 (0.58, 3.34) 0.45No 43/108 (40)

DyspnoeaYes 12/38 (32) 0.56 (0.25, 1.24) 0.15 NINo 43/95 (45)

Chest painYes 11/38 (29) 0.47 (0.21, 1.06) 0.07 NINo 44/95 (46)

Weight lossYes 37/75 (49) 2.16 (1.06, 4.43) 0.03 2.23 (1.00, 4.96) 0.05No 18/58 (31)

SweatsYes 18/50 (36) 0.70 (0.34, 1.44) 0.33No 37/83 (45)

FeversYes 29/75 (39) 0.78 (0.39, 1.56) 0.47No 26/58 (45)

FatigueYes 8/31 (26) 0.41 (0.16, 0.99) 0.05 NINo 47/102 (46)

Diabetes mellitusYes 12/21 (57) 2.14 (0.83, 5.50) 0.11 3.02 (1.04, 8.78) 0.04No 43/112 (38)

Baseline eGFR < 60 mL/minYes 3/5 (60) 2.19 (0.35, 13.59) 0.39No 52/128 (41)

Chronic liver diseaseYes 6/12 (50) 1.47 (0.45, 4.82) 0.52No 49/121 (40)



189

the duration of symptoms prior to healthcare contactperhaps being less accurately recorded on a notificationform compared to a more detailed file review, or it maybe due to socioeconomic factors specific to the commu-nity served by this single-centre health service.

Of significant interest in our study population was thewide range of patient and health system delays thatoccurred. Whilst this will not impact the median delaysfor each locus of delay, when total delay is considered, asubstantial increase in delay occurs. A quarter of patientshad patient delays between 90 and 610 days and healthsystem delays between 53 and 357 days. Whilst in oursubsequent analysis, various delays have been analysedregarding their potential to reduce time to diagnosis, per-haps the biggest impact overall could be achieved byexpediting diagnosis of the quartile of individuals whoexperience very long delays.

In this study, migration from a country with annualTB incidence of <50/100 000 population was the stron-gest predictor of patient delay, although only 7% ofpatients fell into this group. Whilst several putativeexplanations for this exist, it is possible that people fromthese ‘low-risk’ countries, especially those residing inAustralia <6 years, may be less aware of TB as a diseasestate whilst still in the higher risk time period to developactive disease post-migration. Recent migrants fromhigh-risk countries also had a longer delay thanAustralian-born patients, whereas Australian-bornpatients and those residing in Australia >6 years hadreduced risk of delay. This may be due to greater healthliteracy and English language skills in longer-term resi-dents and Australian-born individuals, facilitating accessto healthcare. We speculate that poor access to health-care, poor health literacy, fear that TB diagnosis could

Table 4 Continued


n (%) OR (95% CI) P OR (95% CI) P

Asthma/COPDYes 4/13 (31) 0.60 (0.18, 2.06) 0.42No 51/120 (43)

Current smokingYes 24/48 (50) 1.74 (0.85, 3.57) 0.13 2.27 (0.99, 5.18) 0.05No 31/85 (64) 1.00

Hazardous alcohol intakeYes 11/18 (61) 2.53 (0.92, 7.03) 0.07 NINo 44/115 (38)

Bold values are statistically significant. CI, confidence interval; COPD, chronic obstructive pulmonary disease; eGFR, estimated glomerular filtrationrate; NA, not appropriate; NI, not included in final model; OR, odds ratio; TB, tuberculosis.

Figure 3 Kaplan–Meier curve of delay in patients by TB risk in country

of birth. ( ), Australian born; ( ), low risk; ( ), high risk.

p=0.18

0.00

0.25

0.50

0.75

1.00

pro

po

rtio

n d

iag

no

se

d

0 30 60 90 120

Health system delay (days)

<65 years

>=65 years

Figure 4 Kaplan–Meier curve of health system-related delays by age.

( ), <65 years; ( ), ≥65 years.

Williams et al.


190

Table 5 Risk factors for prolonged health system delay


n (%) OR (95% CI) P OR (95% CI) P

Age≥65 years 16/24 (67) 3.07 (1.21, 7.79) 0.02 4.29 (1.55, 11.89) 0.01<65 years 43/109 (39)

GenderFemale 20/44 (46) 1.07 (0.52, 2.21) 0.86Male 39/89 (43)

English spoken at homeYes 5/13 (39) 0.76 (0.24, 2.47) 0.65No 54/120 (45)

Australian bornYes 3/11 (8) 0.44 (0.11, 1.75) 0.23No 56/122 (46)

EmploymentWorking 16/47 (34) 0.52 (0.25, 1.08) 0.08 NINot working 43/86 (50)

CoughYes 51/124 (41) 0.09 (0.01, 0.72) 0.01 0.09 (0.01, 0.79) 0.03No 8/9 (89)

HaemoptysisYes 7/25 (28) 0.42 (0.16, 1.08) 0.07 NINo 52/108 (48)

DyspnoeaYes 17/38 (45) 1.02 (0.48, 2.18) 0.96No 42/95 (44)

Chest painYes 12/38 (32) 0.47 (0.21, 1.04) 0.06 NINo 47/95 (50)

Weight lossYes 33/75 (44) 0.97 (0.49, 1.93) 0.92No 26/58 (45)

SweatsYes 22/50 (44) 0.98 (0.48, 1.98) 0.95No 37/83 (62)

FeversYes 30/75 (40) 0.67 (0.33, 1.33) 0.25No 29/58 (50)

FatigueYes 9/31 (29) 0.43 (0.18, 1.01) 0.05 NINo 50/102 (47)

Local medical officer reviewYes 44/94 (47) 1.41 (0.66, 30.2) 0.38No 15/39 (25)

Emergency department reviewYes 27/80 (34) 0.33 (0.16, 0.69) <0.01 0.43 (0.17, 1.07) 0.07No 32/53 (60)

Outpatients reviewYes 39/54 (72) 7.67 (3.51, 16.77) <0.001 NINo 20/79 (25)

Hospital admissionYes 39/107 (36) 0.17 (0.06, 0.46) <0.001 0.28 (0.09, 0.92) 0.04No 20/26 (77)

Current smokingYes 16/48 (33) 0.49 (0.23, 1.01) 0.05 NINo 43/85 (51)



191

impact visa and residency status within Australia, fear ofassociated costs of healthcare (even though TB servicesare provided free to all patients within Victoria) orstigma and misunderstanding about TB diagnosis andtreatment may disproportionately affect recent migrants,and that structural barriers to TB care may exist.

Prolonged health system delay was more common inolder patients, a finding common to a number of similarstudies, including the previous Australian study.9 Pre-dictably, there was reduced risk of prolonged healthcaredelay in those with classical symptoms of cough, cavitat-ing features on CXR and sputum AFB smear positivity. Apositive sputum NAT was also associated with a shorterhealth system delay in this study. While delays weremuch shorter in patients admitted to hospital for investi-gation, this is likely to be confounded by the severity ofillness and reverse causation (where patients are admit-ted because of a higher likelihood of TB and/or positivetests). Recommendations regarding where patientsshould be assessed for TB workup is therefore not possi-ble from this dataset.

NAT and other molecular diagnostics appear to offersome utility in reducing diagnostic delays and may bethe first of several new diagnostic tools available overthe coming years. Vital to any strategy to reduce healthsystem delay will be clinician education and awareness,consideration of PTB as a disease process and appropri-ately resourced facilities for diagnosis and managementof PTB.

The limitations of this study include its retrospectivenature, making it vulnerable to recall bias, with data

concerning primary outcomes based on hospital recordsof symptom onset and nature recorded by medical offi-cers rather than directly from patients. It was also limitedto a single centre, which raises the possibility that find-ings are specific for this site; however, it is reassuringthat a similar duration of health system delay was foundin the recently conducted study by Dale et al.13

Conclusion

This study demonstrates that patient delay remains thebiggest delay in diagnosis of PTB, consistent with pub-lished literature. For greatest overall impact on reduc-ing delays until TB diagnosis, strategies should continueto target migrants, especially those from high-riskcountries, at several time points within their first yearsof residence within Australia. These may include healthliteracy programmes that assist new migrants with nav-igating the Australian healthcare system and informa-tion regarding TB disease, treatment and costs. Thesemessages need to be framed in a culturally and linguis-tically contextualised manner to reach the populationsmost at risk. Engagement of community leaders withthis issue to help disseminate these messages would behighly valuable. A letter to new migrants in the appro-priate language explaining the importance of TB, thesymptoms of active disease and that diagnosis andtreatment will not incur costs or affect visa status couldbe another useful tool to address prolonged patientdelays. The Australian Government Department ofImmigration and Border Protection has developed an

Table 5 Continued


n (%) OR (95% CI) P OR (95% CI) P

Excessive alcohol intakeYes 6/18 (33) 0.59 (0.21, 1.67) 0.31No 53/115 (46)

CXR shows cavityYes 6/29 (21) 0.25 (0.09, 0.67) <0.01 NINo 53/104 (51)

Sputum AFB smear positive (n = 110)Yes 8/39 (21) 0.23 (0.09, 0.56) <0.01 NINo 38/71 (54)

Sputum mycobacterial culture positive (n = 110)Yes 33/86 (38) 0.53 (0.21, 1.31) 0.17 NINo 13/24 (54)

Sputum NAT positive (n = 58)Yes 13/47 (28) 0.14 (0.03, 0.62) 0.01 NINo 8/11 (73)

Bold values are statistically significant. AFB, acid fast bacilli; CI, confidence interval; CXR, chest X-ray; NAT, nucleic acid test; NI, not included in finalmodel; OR, odds ratio.

Williams et al.


192

excellent information web page14 in a number of lan-guages for international students that could be morewidely disseminated to all new migrants to Australia.Given that only 19.5% of patients in this cohort identi-fied as students, disseminating this information morewidely could have a powerful effect. Additionally,healthcare providers to this group of patients requireknowledge on TB disease, presentation and diagnostics

as likelihood differs significantly from Australian-bornindividuals.

Acknowledgements

We acknowledge the Victorian Tuberculosis Program fortheir assistance with case finding through the provisionof tuberculosis notification data.

References

1 World Health Organization. Global

Tuberculosis Report 2016. Geneva:

World Health Organization; 2016 [cited

2016 Dec 12]. Available from URL:

http://www.who.int/tb/publications/

global_report/en/

2 TB CARE I. International Standards for

Tuberculosis Care, 3rd edn. The Hague:

TB CARE I; 2014 [cited 2015 Aug 15].

Available from URL: http://www.who.

int/tb/publications/ISTC_3rdEd.pdf

3 Toms C, Stapledon R, Waring J,

Douglas P, the National Tuberculosis

Advisory Committee for the

Communicable Diseases Network

Australia, the Australian

Mycobacterium Reference Laboratory

Network. Tuberculosis notifications in

Australia, 2012 and 2013. Commun Dis

Intell 2015; 39: E217–35.

4 McBryde ES, Denholm JT. Risk of active

tuberculosis in immigrants: effects of

age, region of origin and time since

arrival in a low-exposure setting. Med J

Aust 2012; 197: 458–61.

5 Phoa LL, Teleman MD, Wang YT,

Chee CBE. Characteristics of patients

with delayed diagnosis of infectious

pulmonary tuberculosis. Respirology

2005; 10: 196–200.

6 Jurcev-Savicevic A, Mulic R, Kozul K,

Bozica B, Valic J, Bacun-Ivcek L et al.

Health system delay in pulmonary

tuberculosis treatment in a country with

an intermediate burden of tuberculosis:

a cross-sectional study. BMC Public

Health 2013; 13: 250–63.

7 Lin CY, Lin WR, Chen TC, Lu PL,

Huang PM, Tsai ZR et al. Why is in-

hospital diagnosis of pulmonary

tuberculosis delayed in Southern

Taiwan? J Formos Med Assoc 2010; 109:

269–77.

8 Ohmori M, Ozasa K, Mori T,

Wada M, Yoshiyama T, Aoki M et al.

Trends of delays in tuberculosis case

finding in Japan and associated

factors. Int J Tuberc Lung Dis 2005; 9:

999–1005.

9 Ward J, Siskind V, Kostantinos A.

Patient and health care system delays in

Queensland tuberculosis patients,

1985–1998. Int J Tuberc Lung Dis 2001;

5: 1021–7.

10 Sendagire I, Van der Loeff MS,

Mubiru M, Konde-Lule J, Cobelens F.

Long delays and missed opportunities in

diagnosing smear-positive pulmonary

tuberculosis in Kampala, Uganda: a

cross-sectional study. PLoS One 2010; 5:

e14459.

11 Craig SE, Bettinson SH, Sabin CA,

Gillespie SH, Lipman MCI. Think TB!

Is the diagnosis of pulmonary

tuberculosis delayed by the use of

antibiotics? Int J Tuberc Lung Dis 2009;

13: 209–13.

12 Sreeramareddy CT, Panduru KV,

Menten J, Van den Ende J. Time delays

in diagnosis of pulmonary tuberculosis:

a systematic review of the literature.

BMC Infect Dis 2009; 9: 91–101.

13 Dale KD, Tay EL, Trauer JM,

Trevan PG, Denholm JT. Comparing

tuberculosis management under public

and private healthcare providers:

Victoria, Australia, 2002–2015. BMC

Infect Dis 2017; 17: 324–36.

14 Australian Government, Department of

Immigration and Border Protection.

Tuberculosis Warning for International

Students. 2017 [cited 2017 Jun 1].


border.gov.au/Trav/Stud/More/

tuberculosis



193

http://www.who.int/tb/publications/global_report/en/

http://www.who.int/tb/publications/global_report/en/

http://www.who.int/tb/publications/ISTC_3rdEd.pdf

http://www.who.int/tb/publications/ISTC_3rdEd.pdf

https://www.border.gov.au/Trav/Stud/More/tuberculosis



doi:10.1111/imj.13693

BRIEF COMMUNICATIONS

Metformin-induced encephalopathy: the role of thiamineCaoimhe McGarvey ,1 Catherine Franconi,1 David Prentice2 and Michael Bynevelt3

Departments of 1Neurology, 2General Medicine, and 3Neurological Intervention and Imaging Service of WA, Royal Perth Hospital, Perth, WesternAustralia, Australia

Key wordsmetformin, encephalopathy, thiamine deficiency,haemodialysis, end-stage renal failure.

CorrespondenceCaoimhe McGarvey, Department of Neurology,Royal Perth Hospital, 197 Wellington Street,Perth, WA 6000, Australia.Email: [email protected]

Received 19 February 2017; accepted19 March 2017.

Abstract

A case of metformin encephalopathy is presented in a patient on haemodialysis for

end-stage diabetic renal failure. The patient presented with frequent falls and clinical

signs of Parkinsonism, on a background of recent anorexia and significant weight loss.

Magnetic resonance imaging showed bilateral, symmetrical abnormalities centred on

the lentiform nuclei. Metformin was withheld and signs and symptoms quickly

resolved. We hypothesise that metformin may cause thiamine deficiency in patients

with end-stage renal failure resulting in a specific metabolic encephalopathy.

Metformin is a rare cause of metabolic encephalopathy.Most reported cases occur in patients with end-stagerenal failure (ESRF). Haemodialysis patients can presentwith various neurological disorders, including uraemicencephalopathy, disequilibrium syndrome, dialysis-related dementia, stroke and infections.1 Thiamine defi-ciency as a cause of metformin encephalopathy has notbeen previously reported. We describe a case ofmetformin-induced encephalopathy and discuss thia-mine deficiency as a potential aetiological mechanism.

A 44-year-old woman was transferred to our hospitalfrom another health facility for investigation of frequentfalls. The patient had ESRF secondary to diabeticnephropathy and had commenced haemodialysis3 months previously. The patient’s past medical historyincluded type 2 diabetes mellitus with associated reti-nopathy, hirsutism, hypothyroidism, hypertension andobesity.

The patient had experienced frequent falls and gaitinstability for 4 weeks prior to presentation. The fallswere not associated with dizziness, chest pain, change invision or altered sensation. Her family reported reducedspeech volume. The patient had recent poor appetite andhad lost 27.2 kg in weight over the preceding 3 months,her body mass index on admission was 36.1 kg/mm2.She denied smoking and any alcohol consumption. Thepatient had been recently commenced on metformin 1 g

once daily but had been taking 1 g twice daily for7 weeks. Other medications on admission were rosuvas-tatin, amlodipine, aspirin, moxonidine, prazosin, telmi-sartan and thyroxine. The patient had no known drugallergies.

On examination, respiratory rate was 18 breaths/min,oxygen saturation was >94% on 3 L O2 through nasalprongs, heart rate was 75 beats/min, blood pressure was185/80 mmHg with no postural drop and the tempera-ture was 37.7�C. The patient was alert and orientatedwith a Glasgow Coma Scale of 15/15. Cardiovascular,respiratory and abdominal examinations did not revealany abnormal findings. On neurological examination, allcranial nerves were intact with normal eye movementsand no facial asymmetry or droop. However, markedhypophonia, facial hypomimia and dysarthria werenoted. There was moderate cogwheel rigidity at bothwrists and power was 5/5 throughout all four limbs.Reflexes were diminished but present in the upper limbsbut ankle and knee jerks were absent bilaterally. Pin-prick sensation was intact throughout and propriocep-tion was normal. She was initially unable to stand, laterdemonstrating retropulsion and a shuffling gait. It wasconcluded her falls were related to her clinical extrapyra-midal signs.

Haematology and biochemistry screen on admission:haemoglobin: 108 g/dL, white blood cell count:6.7 × 103/μL, platelet count: 216 × 109/L, C-reactiveprotein: 43 mg/L, creatinine kinase: 251 U/L, creati-nine: 345 mmol/L, urea: 9.1 mmol/L, potassium:3.7 mmol/L, sodium: 140 mmol/L and bicarbonate:



194

25 mmol/L. Lactate on admission was 2.1 mmol/L.Liver function tests were deranged: bilirubin: 38 μmol/L, ALT: 92 U/L, ALP: 127 U/L, gamma gt: 58 U/L,albumin: 42 g/dL and total protein: 75 g/dL. Glucosewas 9.7 mmol/L and a HbA1c 4 weeks prior to admis-sion was 42 mmol/mol. The patient tested negative forhuman immunodeficiency virus and hepatitisB and C. Autoantibody and vasculitic screen was nega-tive. Serum thiamine level was taken after replace-ment on Day 11 of admission and was 269 nmol/L(66–200 nmol/L).A non-contrast enhanced computed tomography

(CT) scan of the brain (Fig. 1) on the day of admissiondemonstrated symmetric low attenuation and swellingof the lentiform nuclei with surrounding oedema involv-ing the internal and external capsule regions, head ofthe caudate nuclei and white matter of the temporalstems. There was no evidence of haemorrhage or hydro-cephalus. A magnetic resonance imaging (MRI) of thebrain performed 4 days later (Fig. 2) confirmed the basalganglia and adjacent capsular abnormalities which wereT2 hyperintense, also seen to involve the sub-insularregions, hypothalamus, optic tracts and chiasm. Therewas also a diffusion restriction along the posterior puta-mina bilaterally and magnetic resonance spectroscopy(two-dimensional chemical shift imaging at the level ofthe basal ganglia, echo time: 135 ms) demonstrated apersistent inverted doublet at 1.3 ppm consistent withthe presence of lactate.Suspecting the patient’s encephalopathy was related

to metformin, the drug was withheld on admission andhaemodialysis was arranged the day after hospital admis-sion to facilitate the elimination of administered metfor-min. A detailed dietary diary revealed borderline dietaryconsumption of thiamine as a result of significantlyreduced appetite since commencing dialysis. The patientrecovered over the next few days without neurologicalsequelae and continues to have haemodialysis threetimes per week.

Discussion

Metformin-induced encephalopathy is a rare conditiontypically described in patients with ESRF.2 Reportedfindings in metformin encephalopathy include Parkin-sonism and vasogenic oedema with T2 hyperintensity inthe basal ganglia. Symptoms and signs improve on with-drawal of metformin.2–5

Metformin is considered the first-line agent for thetreatment of type 2 diabetes mellitus and has beenshown to have numerous benefits,2,6 including a signifi-cant reduction in cardiovascular morbidity and mortalityand is associated with a lower body weight and incidence

of clinically significant hypoglycaemia compared to othertreatments, such as sulfonylureas and insulin. The avoid-ance of metformin in conditions, such as chronic renalfailure, cardiovascular disease and chronic hepatic andpulmonary conditions, has been recommended as it maypotentially increase the risk of tissue anoxia and causelactic acidosis.6

However, a Cochrane review published in 2010, foundno evidence that metformin increases the risk of thedevelopment of lactic acidosis.6 A more recent Cochraneregistered controlled trial, published last year studiedmetformin use in chronic kidney disease (CKD) andfound metformin was well tolerated in diabetes mellitusand CKD (stage III–V).7 No correlation was foundbetween serum lactate and metformin dose or CKDstage. The finding of this report may lead to increaseduse of metformin in CKD with the potential to increasethe incidence of metformin-induced encephalopathy.Metformin exists primarily as a cation at physiological

pHs, diffusing poorly across membranes. Metformin istherefore reliant on several cell membrane transportersfor absorption, distribution and elimination. Of note,metformin is a substrate and competitive inhibitor of thehuman intestinal thiamine transporter, THTR-2(SLC19A3), the major absorptive transporter of thiaminein the intestine.8 THTR-2 is similarly competitively inhib-ited by fedratinib, a Janus kinase 2 inhibitor, trialled as atreatment in patients with myelofibrosis.9 Fedratinib waswithdrawn in the final stages of clinical trials after sev-eral patients developed Wernicke’s encephalopathy. Theputative cause of thiamine deficiency was from the inhi-bition of THTR-2-mediated thiamine uptake.10 So far,thiamine levels have not been extensively measuredwith metformin use, but have however been showed tobe reduced in CKD and diabetes mellitus.11

After a review of the literature, we hypothesise thatthiamine deficiency is a possible mechanism inmetformin-induced encephalopathy as THTR-2 plays amajor role in mediating the first step in thiamine absorp-tion from the small bowel lumen to the enterocyte.6 It isconceivable that metformin may inhibit intracerebralthiamine transport, THTR-2 is also found in the brain soit is possible that metformin can locally interfere withthiamine transport.The classic triad of Wernicke encephalopathy includes

encephalopathy, gait ataxia and oculomotor dysfunc-tion.12 Other clinical findings include acute confusion,delirium, ataxia, ophthalmoplegia, memory disturbance,hypothermia with hypotension, delirium tremens andamnesia. Typical MRI findings with Wernicke encepha-lopathy include T2 and fluid-attenuated inversion recov-ery hyper-intensity in the mammillary bodies,periventricular thalamus and periaqueductal grey



195

matter.12 CT imaging may show symmetrical, low-densityabnormalities in periventricular areas, the diencephalonand the midbrain. These symmetrical lesions are uncom-mon in other cerebral encephalopathies.13 Atypical clini-cal and radiological features generally tend to occur innon-alcoholic Wernicke encephalopathy and have beendescribed in the setting of chemotherapy, hyperemesisgravidarum, malnutrition and renal failure.14,15

Our patient’s imaging findings are typical of those seenin metformin and other metabolic encephalopathies par-ticularly in the setting of metabolic acidosis.15 Imagingfindings as seen in our patient have been described inconfirmed thiamine deficiency in haemodialysis patients.Hung et al. reported thiamine deficiency causing choreain two patients on haemodialysis. Particular note is madeof one patient having lost several kilograms of bodyweight in the previous 6 months due to poor appetite,while the other was severely malnourished on examina-tion and thiamine deficiency was confirmed in theserum.16 Both patients presented with dysarthria andunsteady gait initially before progressing to chorea.Imaging in these patients was similar to that seen in ourpatient with CT Brain showing low-density changes inthe basal ganglia and with MRI (in one patient) demon-strating T2 signal hyperintensity in the basal ganglia.

Dialysis patients are at particular risk of thiaminedeficiency due to poor intake and accelerated loss ofwater-soluble vitamins during dialysis.17 Our patienthad borderline dietary thiamine intake as per detaileddietary diary on a background of anorexia over a3-month period. A case of metformin encephalopathywith similar clinical and radiological features to ourcase described by Kang et al. in 2013 also reportedanorexia for 1 month prior to presentation in a patienton haemodialysis.2

Hung et al. went on to investigate further haemodia-lysis patients presenting with altered mental statusover a 1-year period and found that 30% of patientshad confirmed thiamine deficiency that responded tointravenous thiamine. Their presentations includedconfusion, chorea, acute visual loss, rapidly progressivedementia, myoclonus, convulsions and coma.16

Although all patients in the study had elevated plasmalactate levels, severe lactic acidosis was only observed in2 out of 10 patients with proven thiamine deficiency,one of whom had concurrent liver cirrhosis but theother had been administered metformin for glycaemiccontrol.16

There is a genetic model of impaired intracerebral thi-amine transport recognised in the paediatric population.

Figure 1 Unenhanced axial computed tomography scanning of brain.

Symmetric low attenuation and swelling of the lentiform nuclei and

head of the caudate nuclei with surrounding oedema involving the

internal and external capsule regions.

Figure 2 Axial T2-weighted magnetic resonance imaging demonstrates

hyperintensity in the lentiform nuclei and head of the caudate nuclei

with surrounding oedema in the internal, external and extreme capsule

regions.



196

As reported by Ortigoza-Escobar et al. in 2014, thiaminetransporter 2 deficiency due to a recessively inheritedmutation in the SLC19A3 gene, presents with acute dys-tonia and features of Leigh encephalopathy.18 These clin-ical symptoms and signs improve significantly with earlyadministration of intravenous thiamine. This diagnosiswas suggested by a characteristic MRI pattern of T2hyperintensities with striatal and medial thalamicinvolvement in association with lactic acid accumulationand high organic acid excretion in affected infants.18

We conclude that clinicians should have ahigh index of suspicion for metformin-inducedencephalopathy in patients with ESRF on haemodia-lysis presenting with neurological abnormalities. Werecommend large-scale studies on the effects of met-formin in ESRF to characterise the true risk of lacticacidosis in regard to metformin use in states of tissueanoxia. Further studies are also needed to determinethe prevalence of thiamine deficiency in all patientsreceiving metformin.

References

1 Denker BM, Chertow GM, Owen WF

Jr. Haemodialysis. In: Brenner BM,

ed. The Kidney. Philadelphia, PA:

Saunders; 2000; 2373–454.

2 Kang YJ, Bae EJ, Seo JW, Jeon DH,

Cho HS, Kim HJ et al. Two additional

cases of metformin-associated

encephalopathy in patients with end

stage renal disease undergoing

haemodialysis. Hemodial Int 2013; 17:

111–15.

3 Vander T, Hallevy H, Ifergane G,

Herishanu YO. Metformin-induced

encephalopathy without lactic acidosis.

Diabet Med 2003; 21: 194–5.

4 Jung EY, Cho HS, Seo JW, Kim DW,

Kim HJ, Chang SH et al. Metformin-

induced encephalopathy without lactic

acidosis in a patient with a

contraindication to metformin. Hemodial

Int 2009; 13: 172–5.

5 Fernandes GC, Koltermann T,

Campos L, Vedolin L, Rieder CR.

Teaching neuroimages: the lentiform

fork sign: an MRI pattern of metformin

associated encephalopathy. Neurology

2015; 84: e15.

6 Salpeter SR, Greyber E, Pasternak GA,

Salpeter Posthumous EE. Risk of fatal

and nonfatal lactic acidosis with

metformin use in type 2 diabetes

mellitus. Cochrane Database Syst Rev

2010: CD002967.

7 Mareddy AS, Prabhu AR, Nagaraju SP,

Rangaswamy D, Parthasarathy R,

Kosuru S et al. Metformin use in

diabetes mellitus with chronic kidney

disease – is lactic acidosis a real

concern? Nephrol Dial Transplant 2015;

30: iii534.

8 Liang X, Chien HC, Yee SW,

Giacomini MM, Chen EC, Piao M et al.

Metformin is a substrate and inhibitor

of the human thiamine transporter,

THTR-2 (SLC19A3). Mol Pharm 2015;

12: 4301–10.

9 Pardanani A, Harrison C, Cortes JE,

Cervantes F, Mesa RA, Milligan D et al.

Safety and efficacy of fedratinib in

patients with primary or secondary

myelofibrosis. A randomised clinical

trial. JAMA Oncol 2015; 1: 643–51.

10 Zhang Q, Zhang Y, Diamond S,

Boer J, Harris JJ, Li Y et al. Janus

kinase 2 inhibitor fedratinib inhibits

thiamine uptake: a putative

mechanism of the onset of Wernicke’s

encephalopathy. Drug Metab Dispos

2014; 42: 1656–62.

11 Thornalley PJ, Babaei-Jadidi R, Al AH,

Rabbani N, Antonysunil A, Larkin J

et al. High prevalence of low plasma

thiamine concentration in diabetes

linked to a marker of vascular disease.

Diabetologia 2007; 50: 2164–70.

12 Sechi GP, Serra A. Wernicke’s

encephalopathy: new clinical settings

and recent advances in diagnosis and

management. Lancet Neurol 2007; 6:

442–55.

13 Antunez E, Estruch R, Cardenal C,

Nicholas JM, Fernandez Sola J, Urbano-

Marquez A. Usefulness of CT and MRI

imaging in the diagnosis of acute

Wernicke’s encephalopathy. AJR

Am J Roentgenol 1998; 171: 1131–7.

14 Galvin R, Bråthen G, Ivashynka A,

Hillbom M, Tanasescu R, Leone MA.

EFNS guidelines for diagnosis, therapy

and prevention of Wernicke’s

encephalopathy. Eur J Neurol 2010; 17:

1408–18.

15 Kumar G. Lentiform fork sign: a unique

MRI picture. Is metabolic acidosis

responsible? Clin Neurol Neurosurg 2010;

112: 805–12.

16 Hung SC, Hung SH, Tarng DC,

Yang WC, Huang TP. Chorea induced

by thiamine deficiency in haemodialysis

patients. Am J Kidney Dis 2001; 39:

427–30.

17 Hung SC, Hung SH, Tarng DC,

Yang WC, Chen TW, Huang TP.

Thiamine deficiency and unexplained

encephalopathy in haemodialysis and

peritoneal dialysis patients. Am J Kidney

Dis 2001; 38: 941–7.

18 Ortigoza-Escobar JD, Serrano M,

Molero M, Oyarzabal A, Rebollo M,

Muchart J et al. Thiamine transporter-2

deficiency: outcome and treatment

monitoring. Orphanet J Rare Dis 2014;

9: 92.



197

doi:10.1111/imj.13692

Assessment of potential opioid toxicity and responseto naloxone by rapid response teams at an urban MelbournehospitalBharathy Gunasekaran ,1 Jennifer Weil,2 Tom Whelan,2 John Santamaria3 and Mark Boughey2

1Department of Medicine, 2Centre for Palliative Care, and 3Intensive Care Unit, St Vincent’s Hospital, Melbourne, Victoria, Australia

Key wordsnaloxone, opioids, opioid toxicity, rapidresponse system.

CorrespondenceBharathy Gunasekaran, 6 Gertrude Sreet,Fitzroy, Vic. 3065, Australia.Email: [email protected]

Received 22 March 2017; accepted14 May 2017.

Abstract

Opioid prescriptions have significantly increased in recent years and are used for a wide

variety of indications. Electronic medical records of 45 patients who received naloxone

by a rapid response team over an 18-month period were retrospectively reviewed. This

study found inconsistencies in the management of possible opioid toxicity with varia-

tion in the total naloxone dose and number of doses administered. This highlights the

importance of a standardised protocol for recognition and management of opioid

overdose.

Although pharmaceutical opioids have been recom-mended in the treatment of acute pain and cancer pain,the use of opioids has expanded in the last decade toinclude the treatment of acute and chronic non-cancerpain despite the limited evidence of the long-term bene-fits of opioid use for any indication.1–3 In Australia, therewas a 15-fold increase in the number of PharmaceuticalBenefit Scheme listed opioid dispensing episodes(500 568–7 495 648) between 1992 and 2012.1

One of the main concerns about increased opioid pre-scriptions, coupled with prolonged use in patients withchronic non-cancer pain, is the potential for opioid-related harm. Between 2002 and 2011, the number ofaccidental deaths due to pharmaceutical opioids andillicit drugs in Australia increased from 151 to 266, repre-senting a 1.7-fold rise.1 People at higher risk of opioidoverdose include those with opioid dependence, thosewho inject opioids, patients who take more than 100 mgmorphine (or equivalent) daily or who use opioids withother sedating substances, the elderly and those withcomorbid health conditions.3,4

Opioid overdose can be identified by a combination ofsigns, including decreased conscious state, respiratorydepression and miosis.3,5 Naloxone, an antidote to opioidpoisoning, completely reverses the effects of opioid over-dose if administered in time.3

Although opioids are widely prescribed in the acutehospital, specific methods to evaluate opioid toxicity in

this setting are lacking. The aim of our study was todescribe the assessment of potential opioid toxicity andits management. As naloxone is usually stored in crashcarts, its use is not easily audited. Consequently, weassessed naloxone use by rapid response teams as a sur-rogate for cases of suspected opioid toxicity.

The study population included patients in an urbanhospital in Melbourne, Australia who received naloxoneby the rapid response team during a Code Blue or Medi-cal Emergency Team (MET) call between January 2012and June 2013. Electronic medical records of thesepatients were retrospectively reviewed. Data analysiswas carried out using STATA version 11.2 (StataCorp,College Station, TX, USA). Ethics approval was obtainedfrom the institutional ethics committee.

In total, there were 49 codes where naloxone wasused involving 45 patients over this 18-month period.Forty (81.6%) of these codes were MET calls, five(11.1%) were Code Blues and four (8.9%) were METcalls subsequently escalated to Code Blues. The charac-teristics of these patients are described in Table 1.

Reduced conscious state was the main reason or oneof the reasons for the code in 44 episodes (89.8%).Other documented reasons included hypoxia, hypoten-sion and reduced respiratory rate. The respiratory rateprior to naloxone administration was clearly documen-ted in 34 episodes. Of these, only 4 (11.8%) episodeshad patients with a consistently documented respiratoryrate of less than 12, 18 patients (52.9%) had a consis-tently normal respiratory rate (12–20) and 9 patients(26.5%) had a consistently increased respiratory rategreater than 20. Pupil size was commented on in

Funding : None.Conflict of interest: None.


198

37 episodes and of these, 17 (45.9%) episodes made anote of pupils being constricted.Opioids taken in the 24 h prior to the emergency code

were also reviewed. Opioid consumption was not clearlydocumented in three episodes and in one episode, thepatient had not received any opioids in the preceding24 h. Of the remaining episodes, we were able to convertthe opioids consumed to an oral morphine equivalent in33 episodes. Of these, patients in 16 episodes (48.5%)had more than 100 mg of oral morphine equivalent.Overall, patients in 24 of the episodes (49.0%) had alsoreceived another sedating substance, such as an antipsy-chotic, antidepressant or benzodiazepine.There was a wide range of naloxone doses adminis-

tered. Nine episodes were excluded, as the naloxonedose administered was unclear either because it was notcharted in the medication chart or there was a discrep-ancy between the medication chart and the patient’snotes. Of the remaining 40 episodes, the total naloxonedose given during the code itself ranged from 40 to 2000micrograms with a mean dose of 367 μg. The number ofdoses given during the code was clearly documented in23 episodes and ranged from one to six doses. One doseof naloxone was given in 16 episodes (69.6%). Two,three and four doses were given in two episodes eachand six doses of naloxone were given in one episode.The route of naloxone administration was clearly

documented in 34 episodes. In 30 episodes (88.2%), nal-oxone was given intravenously only. Naloxone was

given through both intramuscular and intravenousroutes in three episodes (8.8%) and subcutaneously inone episode (2.9%). Only one episode included a statdose followed by an infusion of naloxone.The effectiveness of naloxone was documented in

46 episodes. Table 2 shows a comparison between theperceived effectiveness of naloxone for example, if therapid response team thought there was an improvementin the patient’s conscious state, and whether the teambelieved opioids were implicated as a cause for thepatient’s deterioration.There was no significant correlation between the per-

ceived effectiveness of naloxone and gender, presence ofcomorbidities, opioid naïve status prior to admission,concurrent use of another sedative, respiratory depres-sion, or the dose and route of naloxone administration.The adverse effects of naloxone were difficult to ascer-

tain. Agitation was reported in seven episodes, onepatient had severe rebound pain and one patient experi-enced vomiting post naloxone. Pain scores post-codewere not documented in 29 episodes.With regards to the outcomes of these 45 admissions,

21 (46.7%) patients were discharged home including toa supported residential service, 14 (31.1%) patients weretransferred to another hospital or facility and 10 (22.2%)patients died with three of these patients transferred tothe palliative care unit prior to their deaths. The durationfrom the code to time of death in these 10 patients var-ied between 0 and 25 days with a mean of 7.5 days.There were no discharge summaries available for four

patients. One patient was not formally admitted andtherefore did not have a discharge summary. In the32 patients where opioids were implicated as one of thecauses for the deterioration, opioid toxicity was docu-mented as a complication in 10 patients (31.2%), alteredconscious state was noted in 5 patients (15.6%) and nal-oxone was mentioned in 2 patients (6.3%).

Discussion

The assessment of potential opioid toxicity in hospital ischallenging as there are no systems in place that accu-rately track opioid prescribing or administration errors.

Table 1 Characteristics of patients

Characteristic n (%)

GenderMale 29 (64.4)Female 16 (35.6)

Age (years)23–50 9 (20)51–70 20 (44.4)71–93 16 (35.6)

Admission unitSurgical 27 (60)Medical 17 (37.8)Emergency short stay 1 (2.2)

Comorbid conditions (renal, hepatic or centralnervous system impairment; mentalhealth issues)

No 10 (22.2)Yes 35 (77.8)

Known documented malignancyNo 36 (80)Yes 9 (20)

Documented opioid use prior to admissionNo (opioid naïve) 22 (48.9)Yes 23 (51.1)

Table 2 Comparison between documented effectiveness of naloxoneand documented implication of opioids as a cause for the code by therapid response team

Effectivenessof naloxone

Opioids notimplicated

Opioidsimplicated

Notdocumented

No 4 1 2Yes 4 33 2Not documented 1 0 2



199

Our study demonstrates a need for improved educationand practice guidelines around the recognition and man-agement of opioid overdose. The opioid overdose symp-toms of reduced consciousness, miosis and respiratorydepression may not always be present. Respiratorydepression is the sine qua non of opioid intoxication with arespiratory rate of 12 breaths/min or less strongly sugges-tive of acute opioid overdose.5 Miosis alone is insufficientto infer the diagnosis of opioid overdose particularly as itmay be present in chronic opioid use, and polysubstanceuse may produce normally reactive or mydriatic pupils.5

Interestingly, a consistently reduced respiratory rate wasdocumented in only four patients in this study. However,naloxone was noted to be effective in 39 episodes andopioids were implicated in 34 episodes. This raises con-cern regarding the diagnostic assessment of opioid toxicityand the appropriateness of naloxone administered. Thedeterioration of these patients could potentially be attrib-uted to other factors such as disease progression or sepsis.It is important to note that patients on opioids willbecome more alert post-naloxone but this does not neces-sarily imply opioid toxicity.

The inconsistencies in the management of possibleopioid overdose found in this study with wide varia-tions in the total naloxone dose and number of nalox-one doses administered are likely because of the lackof consensus regarding the definition of opioid over-dose and its management. This study highlights theimportance of having a standardised medical protocolin place to identify the clinical criteria for naloxoneadministration, as well as the recommended route andadministration dose.

Choosing the effective dose of naloxone to be admin-istered can be challenging as it depends on multiple fac-tors including the amount of opioids received, thepatient’s weight, and degree of penetrance of the opioidanalgesic into the central nervous system.5 Boyer et al.recommends an initial naloxone dose of 0.04 mg foradults.5 If there is no response, the dose should be

slowly increased every 2 min to a maximum of 15 mg.5

Naloxone has a shorter half-life than that of many opi-oids and some patients may require repeated doses toachieve satisfactory clinical outcomes.5 If multiple dosesof naloxone are required, a naloxone infusion should beconsidered. Opioid toxicity is an unlikely cause if respi-ratory depression continues.

Naloxone may precipitate a short period of acute with-drawal symptoms which include hypertension, tachycar-dia, tremor, convulsion, confusion, headache andvomiting.4 Consequently, patients who have receivednaloxone should be monitored closely for side effectsand pain scores post administration of naloxone shouldbe appropriately documented.

Although MET calls and Code Blues are significantevents during a patient’s hospital admission, there wasinadequate documentation during and after these eventsparticularly with regards to the adverse effects of nalox-one. The poor documentation of these events and poten-tial opioid toxicity as a complication in dischargesummaries were also concerning. It is vital that compli-cations during the admission are communicated to thelocal doctor as it may reduce irrational or inappropriateopioid prescribing in the community. The relatively smalldata set and the lack of adequate documentation in someepisodes were potential limitations; the latter was animportant finding highlighting the need for educationabout the importance of adequate, accurate and struc-tured documentation during rapid response systems aswell as in discharge summaries.

In conclusion, true and significant opioid toxicity inacute hospitals is relatively infrequent; however, educa-tion and guidelines about diagnosis, management anddocumentation of opioid overdose need improvement.This study presents a useful approach to assessing theuse of naloxone in the acute hospital setting during rapidresponse systems as a surrogate for suspected opioid tox-icity, and will allow comparisons and change to beassessed over time.

References

1 BlanchB,PearsonS-A,HaberPS.An

overviewofthepatternsofprescriptionopioid

use,costsandrelatedharmsinAustralia.BrJ

ClinPharmacol2014;78:1159–66.

2 Roxburgh A, Bruno R, Larance B,

Burns L. Prescription of opioid

analgesics and related harms in

Australia. Med J Aust 2011; 195:

280–4.

3 World Health Organization. Information

Sheet on Opioid Overdose. 2014 [cited

2017 Mar]. Available from URL: http://

www.who.int/substance_abuse/

information-sheet/en/

4 Wermeling DP. Review of naloxone

safety for opioid overdose: practical

considerations for new technology and

expanded public access. Ther Adv Drug Saf

2015; 6: 20–31.

5 Boyer EW. Management of opioid

analgesic overdose. N Engl J Med 2012;

367: 146–55.



200

http://www.who.int/substance_abuse/information-sheet/en/



doi:10.1111/imj.13691

Lipid-lowering therapy use and achievement of cholesteroltargets in an Australian diabetes clinicKaterina V. Kiburg ,1,2 Glenn M. Ward,1,3 David N. O’Neal1,3 and Richard J. MacIsaac1,3

1Department of Endocrinology and Diabetes, St Vincent’s Hospital Melbourne, 2Department of Endocrinology and Diabetes, St Vincent’s Institute ofMedical Research, and 3Department of Medicine, University of Melbourne, Melbourne, Victoria, Australia

Key wordsdiabetes, lipid-lowering therapy, endocrinology,cholesterol, cardiovascular disease.

CorrespondenceKaterina Kiburg, Department of Endocrinologyand Diabetes, Level 4, Daly Wing, St Vincent’sHospital Melbourne, 35 Victoria Parade,Melbourne, Vic. 3065, Australia.Email: [email protected]

Received 13 April 2017; accepted4 October 2017.

Abstract

We documented temporal changes in the use of lipid-lowering medications and

achievement of cholesterol targets in an Australian diabetes clinic. The number of

patients using lipid-lowering therapy for primary or secondary cardiovascular preven-

tion increased from 6 to 69% between 1993–1995 and 2014–2016, which corre-

sponded to a decrease in low-density lipoprotein cholesterol levels from 3.7 to

2.4 mmol/L (P < 0.01).

Patients with diabetes are at high risk for the develop-ment of cardiovascular (CV) disease. Statins have beenshown to reduce CV events in direct relationship to theirlow-density lipoprotein cholesterol (LDL-C)-loweringeffect.1 However, there have been no long-term studiesdocumenting the achievement of LDL-C targets and theuptake of lipid-lowering medications in patients attend-ing diabetes clinics in Australia. Our objective was todocument temporal changes in LDL-C levels and the useof lipid-lowering medications in patients attending dia-betes clinics at an Australian university teachinghospital.We studied patients attending diabetes clinics at St

Vincent’s Hospital Melbourne, Australia with ethicalapproval for this quality assurance project being obtainedfrom St Vincent’s Research and Governance DepartmentMelbourne. Information was extracted from a clinicdatabase from 1993 to 2016. The patient population hada mean age of 62.8 years (62.8 � 19.11), of which,75.9% had type 2 diabetes and 18.2% had a history ofCV disease (CVD). We determined the percentage ofpatients on lipid-lowering therapy and LDL-C levels andgrouped the data by number of patients results in orderto ensure that a minimum overall sample size wasgreater than 30 patients per 3-year observation periodand then represented the results as a line of best fit using

lowess smoothing.2 All analysis was done using STATA14.1 (StataCorp, College Station, TX, USA).Figure 1 shows that the use of anti-lipid therapy

increased from 6 to 69% of patients between the timeperiods 1993–1995 and 2014–2016 (P < 0.01). This cor-responded to a decrease in LDL-C levels from 3.7 to2.4 mmol/L (P < 0.01). There was a strong negative cor-relation between anti-lipid therapy use and LDL-C levels(r = −0.88, P < 0.01). The publication dates of threemajor ‘statin-trials’, arbitrarily selected for illustrativepurposes only, in subjects with and without diabetes arealso shown on the diagram, that is, The ScandinavianSimvastatin Survival Study (4S),3 Primary prevention forCVD with atorvastatin in type 2 diabetes (CARDS)4 andthe CV benefits and diabetes risks of statin therapy inprimary prevention (JUPITER).5 In patients with andwithout a history of CVD, the use of anti-lipid therapyand LDL-C levels achieved at eight different observationperiods during the study period is shown in Table 1.Patients with or without a history of CVD had a decreasein mean LDL-C from 1993–1995 to 2014–2016 of 3.55to 2.29 (P < 0.01) and 3.67 to 2.52 mmol/L (P < 0.01),respectively. There was a similar decrease in non-high-density lipoprotein cholesterol levels in patients from4.66 to 3.22 mmol/L (P < 0.01) between 1993–1995 and2014–2016, respectively. In general, the percentage ofpatients who had achieved lipid-treatment targets thatwere current at the time the data were collected werefound to have increased. For 1992–1995, 35.8% weremeeting the 1992 American Diabetes Association (ADA)



201

LDL-C target of <3.36 mmol/L, in 2004–2006 this rose to46.3%, reaching the target of <2.60 mmol/L set in 2003by the ADA. In 2013, targets were introduced by theADA with specific reference to CVD history with 66.5%of our patients without a history of CVD meeting a LDL-Ctarget of 2.60 mmol/L during 2013–2015. However, only29.4% with a history of CVD met the target of<1.80 mmol/L.6–8 Similar results have been reported forthe community-based Australian, Fremantle DiabetesStudy 1 and 2 which ran during 1993–1996 and2009–2011 and saw a reduction in LDL-C from 3.3 to2.3 mmol/L. These studies were also able to show a simi-lar uptake of lipid-lowering therapy from 11 to 68%.9 Inthe years 2010–2016, the breakdown of lipid-loweringmedication was 79% for statins, 14% for fibrates, 5% for

ezetemibe and 2% for other agents. During the years2010–2016, the majority of patients (44%) taking a statinwere on a dosage greater than or equal to 40 mg/day,23% were taking a dose between 20 and 40 mg/day and33% took a dosage less than 20 mg/day. Informationregarding the type of lipid-lowering therapy prescribedfor patients attending our clinic was not available priorto 2010.

The results of our audit suggest that there has been asubstantial increase in the use of lipid-lowering therapiesby patients with diabetes over the last 20 years, which hasresulted in a significant improvement in LDL-cholesterollevels. It should be noted that our clinic populationincludes patients with type 1 and type 2 diabetes. Thereare known limitations in the use of the Friedewald equa-tion for the calculation of LDL-C in patients with elevatedtriglyceride levels.10 For this reason, patients with a triglyc-eride level >4.5 mmol/L were not included in our audit.Although statin therapy is recommended for the vastmajority of patients with diabetes, the use of statins in veryyoung patients with type 1 diabetes may not be warranted.The above trends appear to be influenced by the publica-tion of the results from major statin trials. A preliminaryanalysis of the HbA1c and blood pressure data over timefrom our clinic found a lack of improvement compared tothe significant reduction in LDL-C. In the STENO-2 study,patients with type 2 diabetes in the conventional arm onlyexperienced minor reductions in HbA1c (0.2� 0.3%), sys-tolic blood pressure (−3 � 3 mmHg) and diastolic bloodpressure (−8� 2 mmHg) over 8 years. The STENO-2 studyshowed a reduction of 0.72 � 0.33 mmol/L in LDL-C inthe conventional therapy arm.11

The risks and benefits associated with statin use havebeen a recent topic for discussion both in the medicaland broader community.12 A recent review has sug-gested that for every 10 000 patients treated with aneffective statin regimen for 5 years, there are five casesof myopathy, 50–100 of new onset diabetes and 5–10 of

2.5

33

.5

02

0

LD

L-C

(m

mol/L)

40

60

80

1996

1999

2002

2005

2008

2011

2014

1993

4S S

tudy

CAR

DS S

tudy

Year

JUPIT

ER S

tudy

Figure 1 Uptake of lipid-lowering therapy and mean LDL-cholesterol

results over time. The publication dates of major trials in subjects with

and without diabetes are also shown. 4S Study – The Scandinavian Sim-

vastatin Survival Study (4S).3 CARDS Study – Primary prevention of car-

diovascular disease with atorvastatin in type 2 diabetes.4 JUPITER Study

– Cardiovascular benefits and diabetes risks of statin therapy in primary

prevention.5

Table 1 LDL-C measurements and cardiovascular disease history

Year History of CV disease† No history of CV disease Total

Mean LDL-C n Mean LDL-C n

1993–1995 3.55 (1.7, 6.3) 37 3.67 (1.40, 8.90) 118 1551996–1998 2.85 (2.46, 3.24) 64 3.18 (1.17, 4.44) 40 1041999–2001 2.30 (2.36, 3.64) 212 3.19 (2.20, 4.12) 105 3172002–2004 2.50 (1.17, 3.71) 378 2.38 (1.32, 4.23) 174 5522005–2007 2.48 (0.97, 3.38) 301 2.64 (1.32, 3.82) 100 4012008–2010 2.14 (0.98, 4.4) 478 2.43 (0.99, 4.13) 237 7152011–2013 2.09 (0.27, 5.55) 427 2.27 (0.23, 5.54) 477 9042014–2016 2.29 (0.03, 6.72) 831 2.52 (0.14, 6.09) 192 1023

Only the first LDL-C level available for an individual patient per time period was included in this analysis. †Patients were classified as having a history ofCV disease if they had one of the following: acute myocardial infarction, stroke, coronary artery bypass graft, angioplasty or carotid artery disease.CV, cardiovascular; LDL-C, low-density lipoprotein cholesterol.



202

haemorrhagic strokes.13 It is estimated that for everyyear an individual takes a statin, the risk of a major vas-cular event or the need for coronary revascularisationdecreases by 25% for each mmol/L reduction in LDL-C,with absolute benefits being dependent on an individ-ual’s background CV risk profile.13 Although these ratesmay be lower than seen in clinical practice, our clinicdoes not routinely systematically collect information onstatin myopathy incidence.There is a range of varied recommendations for the

commencement for lipid-lowering therapy and targetsfor LDL-C levels. These range from assessments based onan individual’s risk factors to group-based risk-levelassessment as advocated by The Canadian Cardiovascu-lar Society.14 These Canadian Cardiovascular Societyrecommendations advocate the use of statins in anypatients with diabetes over the age of 40, diabetespatients with microvascular disease or type 1 patientsover the age of 30 who have had had diabetes for morethan 15 years.14 These guidelines also recommend tar-geting a LDL-C level of below 2.0 mmol/L in the primaryprevention setting. The US Preventative Services TaskForce released an updated version of their 2008 recom-mendations regarding statin use in November 2016.They recommend the use of low-to-moderate dose sta-tins in adults aged 40–75 years without a history of CVDbut with one or more risk factors, including diabetes,and a 10% or greater risk of experiencing a CV eventduring the next 10 years. The guidelines suggest usingthe Pooled Cohort Equations to calculate 10-year risk ofCVD as developed by the American College of

Cardiology and American Heart Association.15 TheAustralian College of General Practitioners bases theirrecommendation for the use of statins in patients withtype 2 diabetes on an assessment of each individual’sabsolute CV risk and also recommend a LDL-C target of<2 mmol/L.16 The National Heart Foundation ofAustralia and Cardiac Society of Australia andNew Zealand recommend a general primary preventiontarget of <2.0 mmol/L and a secondary prevention targetof <1.8 mmol/L.17,18Our results suggest that there hasbeen a substantial decrease in LDL-C in patients withdiabetes who do not have a clinical history of CVD. Thismost likely reflects the appreciation of the increase CVDrisk that a diagnosis of diabetes conveys and the provenbenefits of statins for the primary prevention of CVD indiabetes.19

The release of generic statins, and regulatory interven-tions aimed at reducing the cost of many medications,should improve access to statin therapy for manypatients with diabetes, who currently do not fulfil thePharmaceutical Benefit Scheme indications for subsi-dised prescriptions. The impact of these changes on theactual uptake of statins and expected corresponding dropin LDL-C levels await to be documented by future audits.The use of newer lipid-lowering therapies, such as thePCSK9 inhibitors, although not currently routinely usedin patients with diabetes shows promise, with some earlystudies showing a greater mean reduction in LDL-Cwhen treated with a PCSK9 inhibitor compared to stan-dard treatment with statins in patients with type2 diabetes.20

References

1 Morris PB, Ballantyne CM, Birtcher KK,

Dunn SP, Urbina EM. Review of clinical

practice guidelines for the management

of LDL-related risk. J Am Coll Cardiol

2014; 64: 196–206.

2 Cleveland SW, Devlin JS. Locally

weighted regression: an approach to

regression analysis by local fitting. J Am

Stat Assoc 1988; 83: 596–610.

3 Pederson TR, Kjekshus J, Berg K,

Haghfelt T, Faergeman O, Faergeman G

et al. Randomised trial of cholesterol

lowering in 4444 patients with coronary

heart disease: the Scandinavian

Simvastatin Survival Study (4S). Lancet

1994; 344: 1383–9.

4 Colhoun H, Betteridge D, Durrington P,

Hitman G, Neil H, Linvingstone S et al.

Primary prevention of cardiovascular

disease with atorvastatin in type

2 diabetes in the Collaborative

Atorvastatin Diabetes Study (CARDS):

multicentre randomised placebo-

controlled trial. Lancet 2004; 364: 685–96.

5 Ridker P, Pradhan A, MacFadyen J,

Libby P, Glynn R. Cardiovascular

benefits and diabetes risks of statin

therapy in primary prevention: an

analysis from the JUPITER trial. Lancet

2012; 380: 565–71.

6 Garber A, Vinik A, Crespin S. Detection

and management of lipid disorders in

diabetic patients. Diabetes Care 1992; 15:

1068–74.

7 American Diabetes Association.

Management of dyslipidaemia in adults

with diabetes. Diabetes Care 2003; 26

(Suppl 1): S83–6.

8 American Diabetes Association.

Standards of Medical Care in Diabetes –

2013. Diabetes Care 2013; 36(Suppl 1):

S11–66.

9 Baba M, Davis WA, Norman PE,

Davis TME. Temporal changes in the

prevalence and associates of diabetes-

related lower extremity amputations in

patients with type 2 diabetes: the

Fremantle Diabetes Study. Cardiovasc

Diabetol 2015; 14: 152.

10 Hirsch G, Vaid N, Blumenthal RS. The

significance of measuring non-HDL-

cholesterol. Prev Cardiol 2002; 5: 156–9.

11 Gaede P, Vedel P, Larsen N, Jensen G,

Parving HH, Pedersen O. Multifactorial

intervention and cardiovascular disease

in patients with type 2 diabetes. N Engl

J Med 2003; 348: 383–93.

12 Wendling P. Statin safety claims in

Lancet reignite acrimony, scientific

divide. Medscape. 2016 [cited 2017 Feb

8]. Available from URL: http://www.

medscape.com/viewarticle/869125

13 Collins R, Reith C, Emberson J,

Armitage J, Baigent C, Blackwell L et al.

Interpretation of the evidence for the

efficacy and safety of statin therapy.

Lancet 2016; 388: 2532–61.

14 Anderson TJ, Gregoire J, Pearson GJ,

Barry AR, Couture P, Dawes M et al.

2016 Canadian cardiovascular society



203

http://www.medscape.com/viewarticle/869125

http://www.medscape.com/viewarticle/869125

guidelines for the management of

dyslipidemia for the prevention of

cardiovascular disease in the adult. Can

J Cardiol 2016; 32: 1263–82.

15 US Preventive Services Task Force,

Bibbins-Domingo K, Grossman DC,

Curry SJ, Davidson KW, Epling JW Jr

et al. Statin use for the primary

prevention of cardiovascular disease in

adults: US preventive services task force

recommendation statement. JAMA

2016; 316: 1997–2007.

16 Royal Australian College of General

Practitioners. General Practice

Management of Type 2 Diabetes.

Melbourne: RACGP; 2016.

17 National Vascular Disease Prevention

Alliance. Absolute Cardiovascular Disease

Management. Quick Reference Guide

for Health Professionals. Melbourne:

National Stroke Foundation; 2012.

18 National Heart Foundation of Australia

and the Cardiac Society of Australia and

New Zealand. Reducing Risk in Heart

Disease: An Expert Guide to Clinical Practice

for the Secondary Prevention of Coronary

Heart Disease. Melbourne: National

Heart Foundation of Australia; 2012.

19 Taylor F, Huffman MD, Macedo AF,

Moore THM, Burke M, Davey Smith G

et al. Statins for the primary prevention

of cardiovascular disease. Cochrane

Database Sys Rev 2013; 1: 64–5.

20 Sattar N, Preiss D, Robinson JG,

Djedjos CS, Elliott M, Somaratne R et al.

Lipid-lowering efficacy of the PCSK9

inhibitor evolocumab (AMG 145) in

patients with type 2 diabetes: a meta-

analysis of individual patient data. Lancet

Diabetes Endocrinol 2016; 4: 403–10.

doi:10.1111/imj.13703

Management of neurosyphilis: time for a new approach?Olivia C. Smibert ,1,2 Adam W. J. Jenney1,2 and Denis W. Spelman1,2

1Department of Infectious Diseases, Microbiology Unit, The Alfred Hospital, and 2Monash University, Melbourne, Victoria, Australia

Key wordssyphilis, neurosyphilis, syphilis serology, rapidplasma reagin, cerebrospinal fluid.

CorrespondenceOlivia C. Smibert, Department of InfectiousDiseases, Microbiology Unit, The Alfred Hospitaland Monash University, 55 Commercial Road,Melbourne, Vic. 3004, Australia.Email: [email protected]

Received 18 April 2017; accepted4 October 2017.

Abstract

Given the long term sequelae of untreated neurosyphilis and insensitive tests to detect

treponemes in the cerebrospinal fluid, questions regarding the utility of a lumbar punc-

ture and cerebrospinal fluid analysis either to confirm or exclude neurosyphilis are

raised.

Syphilis has been afflicting humans for over 500 yearsand no manifestation of the infection has been as trou-blesome as that of neurosyphilis. Existing definitions ofneurosyphilis require evidence of Treponema pallidum

invasion of the central nervous system (CNS) and yetthere is no single available cerebrospinal fluid (CSF) testthat is both sensitive and specific enough for this pur-pose.1 We are currently in the midst of a local Australiansyphilis epidemic with notification rates increasing from5.1 cases per 100 000 men per year in 2005 to 15.9 casesper 100 000 men in 2014, particularly in the populationof men who have sex with men.2 Similarly, notificationsare increasing across Europe and the United States.3,4

Whilst syphilis is notifiable, in most countries, a diagno-sis of neurosyphilis is not, so there has been no publisheddata documenting a rise in neurosyphilis diagnoses per

se; however, anecdotally this appears to be the case. Cli-nicians everywhere are more and more likely to seepatients presenting with a possible diagnosis of neurosy-philis. Now is a useful time to re-examine the limitationsand pitfalls of available diagnostics and possibly considera new approach of treating such patients with consistenteye, ear or other neurological symptoms with 15 days ofintravenous penicillin without the requirement for CSFexamination.

The neurotropic nature of T. pallidum has been appre-ciated for well over a century with early attempts toabort dissemination and CNS disease, even by excisionof primary chancres, proving futile! Experimentally,T. pallidum is detectable within the bloodstream within



204

48 h and in the CSF within 2 weeks following cutaneousinoculation.5,6 During the early 20th century, largecohorts of patients underwent lumbar puncture and CSFanalysis in an attempt to address the question regardingthe prevalence of neuroinvasion of T. pallidum. It wasfound that 15–30% of all stages of syphilis were associ-ated with CSF abnormalities (predominantly elevationsin protein and white cell counts) – the overwhelmingmajority without localising symptoms to the CNS.7,8

These findings have been replicated in contemporarystudies in the human immunodeficiency virus (HIV) eraleading to the concept of ‘asymptomatic neurosyphilis’, asyndrome believed to be the precursor of symptomaticneurosyphilis.9,10 However, the variables promoting pro-gression from asymptomatic neurosyphilis to symptom-atic neurosyphilis are unknown, although advancedimmunosuppression with low CD4 count in the HIVpopulation has long been thought to be a risk factor.11–13

The value of screening HIV patients with low CD4 countand positive serum syphilis serology for evidence ofasymptomatic neurosyphilis and subsequent treatmentof patients with a ‘CSF paretic formula’ (characterised byelevated protein and white cell count but, negative serol-ogy) as for neurosyphilis is still a cause for ongoingdebate.14,15 An open-label, prospective randomised clini-cal trial at the University of Washington (Clinical TrialsRegistry number NCT02031146) hopes to answer thisquestion by demonstrating that a strategy of immediatelumbar puncture followed by therapy based on CSF find-ings results in better serological and functional outcomesin patients at 6 and 12 months.While several diagnostic algorithms exist for neurosy-

philis possibly the most widely accepted is described bySena et al. in the Manual of Clinical Microbiology.16 Thisalgorithm mandates that a patient must have a positiveserum treponemal test and a clinical syndrome compatiblewith neurosyphilis with one of three CSF tests positive;a reactive CSF Venereal Disease Research Laboratory(VDRL) test, a positive T. pallidum polymerase chain reac-tion (PCR) or identification of treponemes in nervous tis-sue by histological methods.16 Therefore, based on thesecriteria, a diagnosis of neurosyphilis is possible at any clin-ical stage of syphilis infection. This is in keeping with thescientific literature, including local experience at our insti-tution, where CNS invasion with treponemes has beendemonstrated at all clinical stages (i.e. primary, secondary,tertiary) of infection.1,7–9 Unfortunately, despite being thebest available, there are still significant limitations withSena et al.’s definition. We have not widely used the CSFVDRL in Australia for more than 20 years and the rapidplasma reagin that has largely replaced the VDRL is notvalidated for use in CSF nor is it as sensitive as the VDRL(67 vs 58%).16 Whilst hope was held that a T. pallidum

PCR would have sufficient sensitivity to increase the diag-nostic accuracy of CSF analysis, this has not proven to bethe case. A study published in 2016 analysed 40 CSF sam-ples from patients with documented neurosyphilis using anested PCR targeting the tpp47 gene. Disappointingly thePCR gave a sensitivity of only 42.5% with a specificity of97% when compared with a diagnosis based on clinicalassessment and existing CSF diagnostic tests.17 T. pallidumPCR has found its place in the diagnosis of primary syphi-lis with superficial swabs of lesions of primary syphiliscontaining sufficient concentrations of treponemes fordetection. Finally, examination of nervous tissue to dem-onstrate invasion of treponemes is neither justifiable norpractical in the majority of cases and is likely to be limitedin the future to animal and autopsy studies.The search for more sensitive and specific markers of

CNS invasion of T. pallidum continues. Recentapproaches include the use of quantitative CSF : serumratios of treponemal antibodies (i.e. FTA-ABS) withhigher ratios presumed to be indicative of intrathecalantibody production.18 These approaches attempt toimprove sensitivity also by estimating permeability of theblood brain barrier with an albumin quotient. Whilethese techniques, and others, appear to improve the sen-sitivity of existing CSF serology they are considerablymore labour intensive than existing techniques and areyet to be validated or widely applied. CSF cytokine andchemokine profiling of patients with neurosyphilis hasled to an expanding library of novel markers that corre-late with CNS inflammatory responses and invasion ofthe CNS.18–20 A prospective study looking at the utility ofCSF levels of chemokine CXCL 13 and T. pallidum DNAPCR to improve diagnostic accuracy in patients with sus-pected neurosyphilis has recently begun recruiting inShanghai (Clinical Trials Registry number ChiCTR-DDD-16009591). Validation of these indicators will be difficultand further work is required before these tests can beconfidently incorporated into any existing neurosyphilisdiagnostic algorithm.So, neurosyphilis continues to present both trouble-

some clinical syndromes for patients and a diagnosticdilemma for physicians. We are limited by a lack ofunderstanding of disease pathogenesis, the unavailabilityof CSF tests with both an adequate sensitivity and speci-ficity to confirm or exclude the diagnosis and considerablebarriers to further clinical research. Furthermore, it is aheterogeneous syndrome, that may occur at any stage ofthe otherwise traditional temporal sequence of syphilisinfection (i.e. primary, secondary, latent and tertiary),with manifestations that can include meningitis, menin-govasculitis with stroke, uveitis and visual disturbanceand otosyphilis that may present with vertigo, tinnitus orhearing loss. Late complications of untreated



205

neurosyphilis include cognitive impairment, dementia,psychosis, general paresis and tabes dorsalis. Add to thisthe insensitive and non-specific CSF tests available, and itis no wonder that clinicians struggle to confirm confi-dently or exclude a diagnosis of neurosyphilis. Further-more, later stages of syphilis are associated with low titrenon-treponemal tests and an even lower chance of CSFabnormalities. Therefore, CSF examination in thesepatients is likely to be even less useful. Armed with anunderstanding of the limitations of CSF serology is nowthe time to have the discussion and collectively consideran alternative approach to patients who present with aclinically compatible syndrome with positive serum syphi-lis serology? We know of instances at our institution

where patients with a high pre-test probability of neuro-syphilis but negative CSF serology have been treated pre-sumptively with 15 days of intravenous benzylpenicillin.Given the late consequences of untreated neurosyphilisand the current limitations with CSF analysis and tests,perhaps patients, at any stages of syphilis infection, withsymptoms consistent with neurosyphilis and positiveserum serology should be offered treatment with 15 daysof benzylpenicillin? This then raises the question regard-ing the value of lumbar puncture and CSF analysis as partof a diagnostic evaluation in this cohort. In light of ourcurrent local syphilis epidemic perhaps now is the time toreview collectively and reconsider our approach to thistroublesome clinical syndrome.

References

1 Chang CC, Leslie DE, Spelman D,

Chua K, Fairley CK, Street A et al.

Symptomatic and asymptomatic early

neurosyphilis in HIV-infected men who

have sex with men: a retrospective case

series from 2000 to 2007. Sex Health

2011; 8: 207–13.

2 The Kirby Institute. HIV, Viral Hepatitis

and Sexually Transmissible Infections in

Australia Annual Surveillance Report

2014. Sydney: The Kirby Institute, UNSW

[cited 2016 Nov 5]. Available from URL:

http://kirby.unsw.edu.au/sites/default/

files/hiv/resources/ASR2014.pdf

3 European Centre for Disease Prevention

and Control. Annual Epidemiological

Report 2016 – Syphilis. Stockholm:

ECDC [cited 2016 Dec 12]. Available

from URL: http://ecdc.europa.eu/en/

healthtopics/Syphilis/Pages/Annual-

epidemiological-report.aspx. See more at:

http://ecdc.europa.eu/en/healthtopics/

syphilis/Pages/Annual-Epidemiological-

Report.aspx#sthash.ArNRIOxp.dpuf

4 Centers for Disease Control and

Prevention. Sexually Transmitted

Disease Surveillance 2015. Atlanta:

U.S. Department of Health and Human

Services [cited 2016 Dec 12]. Available

from URL: https://www.cdc.gov/std/

stats15/std-surveillance-2015-print.pdf

5 Marra C, Baker-Zander SA, Hook EW,

Lukehart SA. An experimental model of

early central nervous system syphilis. J

Infect Dis 1991; 163: 825–9.

6 Marra CM, Castro CD, Kuller L,

Dukes AC, Centurion-Lara A,

Morton WR et al. Mechanisms of

clearance of Treponema pallidum from

the CSF in a nonhuman primate model.

Neurology 1998; 51: 957–61.

7 Mills CH. Routine examination of the

Cerebro-spinal fluid in syphilis: its value

in regard to more accurate knowledge,

prognosis, and treatment. Br Med J

1927; 2: 527–32.

8 Fildes P, Parnell RJG, Maitland HB. The

occurrence of unsuspected involvement

of the central nervous system in

unselected cases of syphilis. Brain 1918;

41: 255–301.

9 Lukehart SA, Hook EW 3rd, Baker-

Zander SA, Collier AC, Critchlow CW,

Handsfield HH. Invasion of the central

nervous system by Treponema pallidum:

implications for diagnosis and treatment.

Ann Intern Med 1988; 109: 855–62.

10 Rolfs RT, Joesoef MR, Hendershot EF,

Rompalo AM, Augenbraun MH,

Chiu M et al. A randomized trial of

enhanced therapy for early syphilis in

patients with and without human

immunodeficiency virus infection. The

syphilis and HIV study group. N Engl J

Med 1997; 337: 307–14.

11 Marra CM, Maxwell CL, Smith SL,

Lukehart SA, Rompalo AM, Eaton M

et al. Cerebrospinal fluid abnormalities

in patients with syphilis: association

with clinical and laboratory features. J

Infect Dis 2004; 189: 369–76.

12 Ghanem KG, Moore RD, Rompalo AM,

Erbelding EJ, Zenilman JM, Gebo KA.

Neurosyphilis in a clinical cohort of

HIV-1-infected patient with syphilis and

no neurologic symptoms. Clin Infect Dis

2009; 48: 816–21.

13 Firlag-Burkacka E, Swiecki P,

Cielniak I, Siwak E, Gizinska J,

Bakowska E et al. High frequency of

neurosyphilis in HIV positive patients

diagnosed with early syphilis. HIV Med

2016; 17: 323–6.

14 Ho EL, Spudich SS. Neurosyphilis and

the impact of HIV infection. Sex Health

2015; 12: 148–54.

15 Ghanem KG. REVIEW: Neurosyphilis:

a historical perspective and review. CNS

Neurosci Ther 2010; 16: 157–68.

16 Sena AC, Pillay A, Cox DL, Radolf JD.

Treponema and brachyspira, human

host-associated spirochetes. In:

Jorgensen JH, Pfaller MA, Carroll KC,

et al., eds. Manual of Clinical Microbiology,

11th edn. Washington, DC: ASM press;

2015.

17 Vanhaecke C, Grange P, Benhaddou N,

Blanche P, Salmon D, Parize P et al.

Clinical and biological characteristics of

40 patients with Neurosyphilis and

evaluation of Treponema pallidum nested

polymerase chain reaction in

cerebrospinal fluid samples. Clin Infect

Dis 2016; 63: 1180–6.

18 Salamano R, Ballesté R, Perna A,

Rodriguez N, Lombardo D, García N

et al. Cerebrospinal fluid examination

may be useful in diagnosing

neurosyphilis in asymptomatic HIV+

patients with syphilis. Arq Neuropsiquiatr

2016; 74: 128–32.

19 Wang C, Wu K, Yu Q, Zhang S, Gao Z,

Liu Y et al. CXCL13, CXCL10 and

CXCL8 as potential biomarkers for the

diagnosis of Neurosyphilis patients. Sci

Rep 2016; 6: 33569.

20 RX H, Lu C, Lu S, Hu Y, Ma H, Lai W

et al. Value of CXCL13 in diagnosing

asymptomatic Neurosyphilis in HIV-

infected patients. Int J STD AIDS 2015;

27: 141–6.



206

http://kirby.unsw.edu.au/sites/default/files/hiv/resources/ASR2014.pdf

http://kirby.unsw.edu.au/sites/default/files/hiv/resources/ASR2014.pdf

http://ecdc.europa.eu/en/healthtopics/Syphilis/Pages/Annual-epidemiological-report.aspx



http://ecdc.europa.eu/en/healthtopics/syphilis/Pages/Annual-Epidemiological-Report.aspx#sthash.ArNRIOxp.dpuf



https://www.cdc.gov/std/stats15/std-surveillance-2015-print.pdf

https://www.cdc.gov/std/stats15/std-surveillance-2015-print.pdf

doi:10.1111/imj.13698

High urinary interleukin-8 levels is associated with poorprognosis in idiopathic membranous nephropathyJilin Chen , Xiaoye Fu, Yanling Sun, Shengkun Zhang, Hua Xie and Hongli Lin

Department of Nephrology, The First Hospital of Dalian Medical University, Dalian, China

Key wordsidiopathic membranous nephropathy, cytokines,interleukin-8, prognosis.

CorrespondenceHua Xie and Hongli Lin, Department ofNephrology, The First Hospital of Dalian MedicalUniversity, 222 Zhongshan Road, Dalian 116011,China.Email: [email protected]; [email protected]

Received 23 March 2017; accepted22 June 2017.

Abstract

Biomarkers required to assess accurately the prognosis of idiopathic membranous

nephropathy (IMN) are still unavailable. A retrospective study on 156 IMN patients

showed only urinary IL-8 was associated with the achievement of initial complete

remission (CR) in IMN patients. A urinary IL-8 level of less than 61.25 pg/mL was more

sensitive for prediction of CR in IMN patients. Therefore, urinary IL-8 may be a poten-

tial biomarker for evaluating short-term prognosis of IMN patients.

Idiopathic membranous nephropathy (IMN) is a com-mon immune-mediated glomerular disease and remainsthe leading cause of nephrotic syndrome in adults.1 Theaetiology and optimal biomarkers required for evalua-tion of IMN prognosis remain unclear. Patients whoachieved long-term complete remission (CR) have alower risk for receiving renal replacement treatment.2

Many studies have failed to provide early, sensitive andspecific biological markers for assessing the prognosisof IMN.3

More recently, it has been demonstrated that cyto-kines play a critical role as mediators of inflammationand as progressive factors in idiopathic nephrotic syn-drome (INS).4 We recently examined serum levels ofrepresentative pro-inflammatory cytokines, includingIL-8,5 IL-6,6 and IL-177 as well as the anti-inflammatorycytokine IL-108 was also examined as it is implicated insteroid resistance. Our studies indicated that onlyserum IL-8 levels were significantly increased inpatients with steroid-resistance when compared tosteroid-sensitive INS. Further studies indicated thatintermittent high-volume haemofiltration (IHVHF) pro-moted remission in steroid-resistant INS patients withvolume overload and accompanied with IL-8 serumlevels decreased,9 therapeutic efficacy of IHVHF may

partly be due to serum IL-8 clearance. However,whether IL-8 levels can be used to evaluate the progno-sis of IMN remains unknown.A total of 156 eligible patients with biopsy-proven

IMN was included in this retrospective cohort study fromJanuary 2009 to December 2011, while another46 healthy volunteers served as controls. The inclusioncriteria included patients aged 18–75 years with biopsy-proven IMN and nephrotic syndrome, estimated glomer-ular filtration rate (eGFR) ≥ 60 mL/min/1.73 m2, admin-istration of glucocorticoids, angiotensin converting-enzyme inhibitors (ACEI) or angiotensin receptorblockers (ARB). The exclusion criteria included second-ary membranous nephropathy, infections, inflammatorydiseases, cancers, acute kidney injury, acute liver injuryand severe heart failure.Initial therapy included the administration of gluco-

corticoids, which were administered to all patients fol-lowed by methylprednisolone pulse therapy with500 mg/day for 3 days and oral prednisone 40 mg oncedaily in patients younger than 60 years of age. Alterna-tively, they were given methylprednisolone pulse ther-apy (250 mg/day) for 3 days, followed by oralprednisone administration (30 mg once) for those aged≥60 years. After 8 weeks of prednisone treatment,patients were given cyclophosphamide (CTX) (0.6 g ×2, pulse therapy per 15 days, cumulative dose to150 mg/kg), Tacrolimus (FK506) (0.05 mg/kg daily), ormycophenolate mofetil (1500 mg/day) for 6–18 months.The primary outcome measures were CR (defined as uri-nary protein excretion <0.3 g/day with serum albumin

Funding: This study was supported by grants from the NationalNatural Science Foundation of China (grant number:81200522) and Science and Technology Plan Projects ofLiaoning province (grant number: 2014225018).Conflict of interest: None.


207

>3.5 g/dL and normal serum creatinine) and partialremission (PR) (defined as urinary protein excretion of0.3–3.5 g/day and stable serum creatinine). Patients whoachieved CR were examined every 3 months, while allothers were evaluated on a monthly basis.

We enrolled 156 patients with IMN, of which110 were male patients (70.5%). The median age was54 (interquartile range: 43–61) years of age. Patientswere divided into three groups according to immuno-suppressive therapy type. The groups did not signifi-cantly differ in sex, systolic and diastolic blood pressure,serum creatinine, eGFR, serum uric acid, fibrinogen ortime from diagnosis to biopsy. The median observationperiod for the entire cohort was 23 months. A total of57 patients (36.5%) achieved CR, 82 patients (52.6%)achieved PR, 17 patients (10.9%) were classified as noresponse. The mean time to CR was 6.0 months. Thecumulative probability of CR was not significantly dif-ferent between treatment groups (log Rank X2 = 3.696,P = 0.158). And the cumulative probability of PR wasnot significantly different between treatment groups(log Rank X2 = 0.256, P = 0.88). Urinary IL-8 and IL-17levels in all three groups of patients were significantlyhigher than in healthy control subjects (P < 0.01). Uni-variate Cox proportional hazards models and multivari-ate Cox proportional hazards model analysis indicatedonly urinary IL-8 levels were associated with initial CR(HR, 0.957; 95% CI, 0.937–0.977; P = 0.000) in patientswith IMN.

The receiver operating characteristic curve analysisindicated the area under the receiver operating

characteristic curve (AUROC) in urinary IL-8 levels gen-erated an optimal cut-off of 61.65 pg/mL (Youdenpoint). The AUC was 0.677 (95% CI 0.586–0.768; P =0.000), with the sensitivity and specificity 0.83 and 0.51respectively (Fig. 1A). Patients with IMN were dividedinto two groups: urinary IL-8 levels >61.65 pg/mL (n =110), defined as the high-IL-8 group, and urinary IL-8levels <61.65 pg/mL (n = 46), defined as the low-IL-8group. The cumulative probability of CR in the low-IL-8group patients was significantly higher than that of thehigh-IL-8 group patients (log Rank X2 = 17.79, P =0.000)(Fig. 1B). With 61.65 pg/mL used as a threshold,63% of patients with urinary IL-8 levels less than 61.65pg/mL achieved CR. This contrasted to only 25.5% ofpatients with urinary IL-8 levels greater than 61.65pg/mL achieving CR (P < 0.001).

Discussion

This study is the first to evaluate the association of cyto-kines and their effect on short-term outcomes of patientswith IMN. Previous studies identified several clinical pre-dictors of poor renal outcome in IMN patients includingmale sex, age > 50 years, hypertension, severity of pro-teinuria, elevated serum creatinine level at the time ofdiagnosis and chronic renal tubular interstitialchanges.10–12 Our study indicated that age, male sex, sys-tolic and diastolic blood pressure, serum creatinine, uri-nary protein excretion, serum albumin, eGFR, serumuric acid, serum cholesterol, serum triglyceride, haemo-globin, fibrinogen and ACEI (or ARB therapy) were not

Figure 1 Predictive value of baseline urinary IL-8 levels in complete remission (CR) and cumulative probability of CR in 156 idiopathic membranous

nephropathy (IMN) patients stratified by IL-8 level. (A) Urinary IL-8 concentration of 61.65 pg/mL was shown to be the optimal predictive cut-off

value for CR in patients with IMN. Area under the curve (AUC) was 0.677 (P < 0.001), with sensitivity and specificity were 83% and 51% respectively. (B).

The cumulative probability of CR in the low-IL-8 group (IL-8 < 61.65 pg/mL) patients was significantly higher than that of the high-IL-8 group (IL-8 >

61.65 pg/mL) patients (P < 0.001). ( ), urinary IL-8 Level < 61.65 pg/mL; ( ), urinary IL-8 Level > 61.65 pg/mL.



208

associated with initial CR. Therefore, the routine clinical,laboratory and pathology indicators do not evaluate suf-ficiently the short-term outcomes of patients with IMN.Some studies suggesting that IMN is primarily an

immune disease resulting from immunoregulatoryimbalance between T helper subtype 1 (Th1) and Thelper subtype 2 (Th2).13,14 The mechanisms throughwhich T-cells affect the course of IMN remain unclear.However, there may be circulating factors released fromactivated T-cells which affect the pathogenesis of IMN.Our study showed that only urinary IL-8 was associ-

ated with the achievement of initial CR. The patientswith CR had significantly lower IL-8 levels than thosepatients who had not achieved initial CR. IL-8 is aproinflammatory cytokine produced by endothelial cellsand macrophages that attracts neutrophils and lympho-cytes to the site of inflammation.5 Early studies revealedthat the IL-8 gene is predominantly expressed inperipheral blood mononuclear cells and that serum IL-8concentration is increased in patients with INS.15 Sev-eral studies have reported elevated serum IL-8 levels inthe nephrotic phase when compared with the remissionphase. Urinary IL-8 was significantly higher in relapsed

SR than in SS patients in remission. This increased uri-nary IL-8 levels was associated with local changes inglomerular permeability.16,17 Therefore, IL-8 may beinvolved in the pathogenesis of IMN, higher urinary IL-8 levels may be associated with severe inflammationreaction in renal tissues and predict a bad clinicaloutcomes.The study showed that urinary IL-8 concentration of

61.65 pg/mL was chosen as the optimal cut-off value forCR in patients with IMN. The AUROC was 0.677 withthe sensitivity and specificity 0.83 and 0.51 respectively.The cumulative probability of CR in the low-IL-8 group(IL-8 levels >61.65 pg/mL) was significantly higher thanthat in the high-IL-8 group patients (IL-8 levels <61.65pg/mL), therefore, a urinary IL-8 concentration greaterthan 61.65 pg/mL was considered higher risk to developpoor prognosis in IMN.This study is unique in that it identifies new predictors

for clinical outcomes among IMN patients. In conclusion,the study indicated that urinary IL-8 may be a significantshort-term prognostic biomarker for IMN. Patients withlower urinary IL-8 levels were more likely toachieve CR.

References

1 Glassock RJ. The pathogenesis of

idiopathic membranous nephropathy: a

50-year odyssey. Am J Kidney Dis 2010;

56: 157–67.

2 McQuarrie EP, Stirling CM, Geddes CC.

Idiopathic membranous nephropathy

and nephrotic syndrome: outcome in

the era of evidence-based therapy.

Nephrol Dial Transplants 2012; 27:

235–42.

3 Stahl RA, Harendza S, Helmchen U,

Hoxha E. Membranous nephropathy –

crucial developments for diagnostic and

treatment. Dtsch Med Wochenschr 2014;

139: 1433–6.

4 Araya CE, Wasserfall CH, Brusko TM,

Mu W, Segal MS, Johnson RJ et al. A

case of unfulfilled expectations.

Cytokines in idiopathic minimal lesion

nephrotic syndrome. Pediatr Nephrol

2006; 21: 603–10.

5 Baggiolini M, Loetscher P, Moser B.

Interleukin-8 and the chemokine family.

Int J Immunopharmacol 1995; 17: 103–8.

6 Wang LM, Chi YJ, Wang LN, Nie L,

Zou YH, Zhao TN et al. Expression of

interleukin-6 in rat model of

doxorubicin-induced nephropathy.

Chinese Journal of Contemporary Pediatrics

(Zhongguo Dang Dai Er Ke Za Zhi) 2010;

12: 912–4.

7 Ramesh R, Kozhaya L, McKevitt K,

Djuretic IM, Carlson TJ, Quintero MA

et al. Pro-inflammatory human Th17

cells selectively express P-glycoprotein

and are refractory to glucocorticoids. J

Exp Med 2014; 211: 89–104.

8 Hawrylowicz CM, O’Garra A. Potential

role of interleukin-10-secreting

regulatory T cells in allergy and asthma.

Nat Rev Immunol 2005; 5: 271–83.

9 Xie H, Fang M, Lin H, Li P, Chen J,

Sun Y et al. Intermittent high-

volume hemofiltration promotes

remission in steroid-resistant

idiopathic nephrotic syndrome. Ren

Fail 2015; 37: 966–73.

10 Yamaguchi M, Ando M, Yamamoto R,

Akiyama S, Kato S, Katsuno T et al.

Patient age and the prognosis of

idiopathic membranous nephropathy.

PLoS One 2014; 9: e110376.

11 Segal PE, Choi MJ. Recent advances

and prognosis in idiopathic

membranous nephropathy. Adv Chronic

Kidney Dis 2012; 19: 114–9.

12 Shiiki H, Saito T, Nishitani Y, Mitarai T,

Yorioka N, Yoshimura A et al. Prognosis

and risk factors for idiopathic

membranous nephropathy with

nephrotic syndrome in Japan. Kidney Int

2004; 65: 1400–7.

13 Lama G, Luongo I, Tirino G, Borriello A,

Carangio C, Salsano ME. T-lymphocyte

populations and cytokines in childhood

nephrotic syndrome. Am J Kidney Dis

2002; 39: 958–65.

14 Ronco P, Debiec H. Pathogenesis of

membranous nephropathy: recent

advances and future challenges. Nature

reviews. Nephrol Ther 2012; 8: 203–13.

15 Garin EH, Blanchard DK,

Matsushima K, Djeu JY. IL-8

production by peripheral blood

mononuclear cells in nephrotic patients.

Kidney Int 1994; 45: 1311–7.

16 Souto MFO, Teixeira AL, Russo RC,

Penido MGMG, Silveira KD,

Teixeira MM et al. Immune mediators in

idiopathic nephrotic syndrome:

evidence for a relation between

interleukin 8 and proteinuria. Pediatr

Res 2008; 64: 637–42.

17 Kanai T, Yamagata T, Momoi MY.

Macrophage inflammatory protein-

1beta and interleukin-8 associated with

idiopathic steroid-sensitive nephrotic

syndrome. Pediatr Int 2009; 51: 443–7.



209

doi:10.1111/imj.13707

POSIT ION PAPER

Considerations for pre-transfusion immunohaematology testingin patients receiving the anti-CD38 monoclonal antibodydaratumumab for the treatment of multiple myelomaHang Quach,1,2 Simon Benson,3 Helen Haysom,3,4 Anne-Marie Wilkes,5 Nicole Zacher,3,6

Merrole Cole-Sinclair,2,5 Henry Miles Prince,1,7 Peter Mollee,8 Andrew Spencer,9,10 Phoebe Joy Ho,11

Simon J. Harrison,1,12 Cindy Lee,13 Bradley Augustson14 and James Daly3,15

1Department of Medicine, The University of Melbourne, 2Department of Haematology, and 5Blood Transfusion Laboratory, St Vincent’s Hospital,4Department of Epidemiology and Preventive Medicine, and 9Australian Centre for Blood Diseases, Monash University, 7Department of Haematology,Epworth Health, 10Department of Haematology, Alfred Hospital, and 12Department of Haematology, Peter MacCallum Cancer Centre, Melbourne,Victoria, 3Australian and New Zealand Society of Blood Transfusion, and 11Department of Haematology, Royal Prince Alfred Hospital, Sydney, NewSouth Wales, 6Clinical Diagnostic Department, Capital Pathology, Canberra, Australian Capital Territory, 8Department of Haematology, PrincessAlexandra Hospital, and 15QML Pathology, Brisbane, Queensland, 13Department of Haematology, Royal Adelaide Hospital, Adelaide, South Austalia,and 14Department of Haematology, Sir Charles Gairdner Hospital, Perth, Western Australia, Australia

Key wordstransfusion, immunohaematology,daratumumab, multiple myeloma.

CorrespondenceHang Quach, Department of Haematology,St Vincent’s Hospital, 41 Victoria Parade, Fitzroy,Vic. 3065, Australia.Email: [email protected]

Received 30 November 2017; accepted30 November 2017.

Abstract

In recent years, the anti-CD38 monoclonal antibody daratumumab (Darzalex; Janssen-

Cilag Pty Ltd) has been shown to be highly efficacious in relapsed and refractory multiple

myeloma, with the final results of treatment in newly diagnosed patients awaited. Despite

awareness of the potential interference of daratumumab in pre-transfusion immunohae-

matology testing during phase I and II clinical studies, there was a degree of unprepared-

ness in the community upon the introduction of this drug into the clinics, particularly the

impact that it has on the operational processes in hospital transfusion laboratories and

timely issue of red blood cells (RBCs). Anti-CD38 interference in pre-transfusion immuno-

haematology tests is a particular problem in patients being treated with daratumumab for

multiple myeloma as many will require RBC transfusions during their disease treatment.

Panagglutination caused by anti-CD38 monoclonal antibody during the indirect antiglobu-

lin test may mask the presence of a clinically significant RBC alloantibody in the patient’s

plasma during the antibody screen and identification process, which may be overlooked,

particularly in urgent situations, subsequently resulting in a delayed or acute haemolytic

transfusion reaction. Here, we summarise daratumumab’s effects on pre-transfusion

immunohaematology testing and its impact on clinical practice and make practical recom-

mendations based on a consensus from medical and scientific transfusion experts and

myeloma specialists on behalf of the Australian and New Zealand Society of Blood Trans-

fusion and Myeloma Scientific Advisory Group to Myeloma Australia, respectively.

Introduction

In recent years, the anti-CD38 monoclonal antibody(mAb) daratumumab (Darzalex; Janssen-Cilag Pty Ltd)has been shown to be highly efficacious in relapsed andrefractory multiple myeloma (MM). In 2015, daratumu-mab was granted accelerated approval by the Food andDrug Administration in the United States for the treat-ment of relapsed/refractory MM, with Australia’s Thera-peutic Goods Administration (TGA) following suit in2017. These decisions were based on results only fromearly phase I/II clinical studies, in which heavily pre-treated patients with MM were shown to have an overall

Funding: Janssen-Cilag provided funding for the authorshipgroup to meet and discuss the guidelines. Janssen-Cilagreviewed the guidelines for scientific accuracy, but the finaldecision to publish was solely the authors.Conflict of interest: H. Quach has sat on scientific advisoryboards for Celgene, Janssen-Cilag and Amgen and has receivedresearch funding from Celgene and Amgen. H. M. Prince, A.Spencer, P. J. Ho, S. J. Harrison and B. Augustson have sat onscientific advisory boards for Janssen-Cilag Pty Ltd. H. Haysom,S. Benson and M. Cole-Sinclair have no conflicts of interest.


210

survival improvement of approximately 11 months fromsingle-agent daratumumab.1 As a result of this earlymove into the clinics, there was an underappreciation ofthe impact of daratumumab’s interference with pre-transfusion immunohaematology testing and, therefore,on hospital/pathology transfusion laboratory operationalprocesses, timely issuing of blood, potential blood trans-fusion reactions and, ultimately, patient safety.CD38 is an integral transmembrane glycoprotein that is

expressed on many cell types and highly expressed onplasma cells. It has diverse functions, including enzymeactivity, intracellular calcium regulation and receptor-mediated adhesion.2 It is also variably expressed on thesurface of red blood cells (RBCs). Anti-myeloma activityfrom daratumumab occurs though anti-CD38-mediatedimmune mechanisms, including complement-dependentcellular cytotoxicity (CDCC), antibody-dependent cellularcytotoxicity (ADCC), antibody-dependent cellular phago-cytosis (ADCP) and immunoregulatory depletion ofimmune suppressive regulatory T cells.3–5 In addition,direct tumouricidal activity occurs through pro-apoptoticsignalling pathways upon cross-linking of surface CD38. Asan off-target side-effect, when bound to CD38 on RBC,daratumumab interferes with the indirect antiglobulin tests(IAT), a technique routinely used in pre-transfusion test-ing. Anti-CD38 interference in immunohaematology testsis a particular problem in patients being treated for MM asmany will require blood transfusions as part of their sup-portive care during ongoing disease treatment. Panaggluti-nation caused by anti-CD38 may mask the presence of aclinically significant RBC antibody (Ab) in the patient’splasma, which may be overlooked, particularly in urgentsituations, and subsequently result in an acute or delayedhaemolytic transfusion reaction.Here, we summarise daratumumab’s impact on pre-

transfusion immunohaematology testing, its impact onclinical practice and provide practical recommendationsbased on a consensus from medical and scientific trans-fusion experts and myeloma specialists on behalf of theAustralian and New Zealand Society of Blood Transfu-sion (ANZSBT) and Myeloma Scientific Advisory Groupto Myeloma Australia (MSAG), respectively.

The nature of daratumumab’sinterference with pre-transfusion tests

The binding of daratumumab to CD38 on human RBC isdetected using the IAT (or indirect Coomb’s test) carriedout at 37�C, which is the primary antibody screeningmethod used to detect the presence of clinically significantalloantibodies. Secondary testing methods that may beused in antibody investigations, such as room temperaturetesting or immediate spin tests to check for ABO

compatibility, do not detect the effects of daratumumab.There is some variability of expression of CD38 on RBC,and the presence of daratumumab in the patient’s plasmatypically causes weak panagglutination in IAT used forpre-transfusion immunohaematology testing. In contrast,daratumumab does not interfere with ABO or RhDtyping.6,7

In the antibody screen and antibody identificationpanel, the plasma of patients treated with daratumumabexhibits weak (1+ or 2+, using 0–4 scoring) panaggluti-nation. This panagglutination occurs in all IAT tests, forexample, saline, low ionic-strength saline (LISS) andpolyethylene glycol (PEG), and all IAT methods, includ-ing column agglutination technology (CAT) and tubeand solid phase.6 Positive IAT may persist for up to6 months after discontinuation of daratumumabtherapy.7–9 The presence of panagglutination must beinvestigated at each testing episode as the reactivity maymask the presence of a clinically significant alloantibodyor the presence of autoimmune haemolytic anaemia.Interestingly, while daratumumab in the patient’s

plasma will cause agglutination in IAT with all reagentRBC and donor RBCs, reactivity with the patient’s ownRBC is not consistent, and the auto-control in the anti-body identification panel is frequently negative, as is thedirect antiglobulin test (DAT). This suggests that thepatient’s RBCs with high levels of CD38 may be clearedfrom the circulation and/or be subject to anti-CD38-mediated antigen downregulation,10 which may explainwhy, to date, clinical manifestations of daratumumab-related, immune-mediated haemolysis have not beenreported in daratumumab-treated patients. That observa-tion notwithstanding, interference by daratumumab has aserious impact on the ability of transfusion laboratories toperform timely pre-transfusion testing.11 The resolution ofthe interference requires time-consuming specialist inves-tigations that inevitably lead to delays in the provision ofblood for transfusion, especially if it is not know that thepatient is being or has been treated with daratumumab. Inaddition, clinically significant RBC alloantibodies may bemasked and overlooked, potentially resulting in an acuteor delayed haemolytic transfusion reaction. For urgent oremergency transfusions, however, it should be possible todetermine the patient’s ABO and RhD blood group andprovide ABO-compatible blood, but provision of this with-out further investigation is not without risks.12,13

Overcoming the interference of anti-CD38 therapy

Several methods have been proposed to overcome anti-CD38 interference in immunohaematology testing and tofacilitate alloantibody screening, thus reducing the risk of

Daratumumab and the transfusion laboratory


211

incompatible transfusions and the possibility of transfu-sion reactions. These include testing the patient’s plasmaagainst a panel of reagent RBC treated with dithiothreitol(DTT) or trypsin. In addition, extended RBC phenotypingor genotyping of the patient prior to the first dose of dara-tumumab enables transfusion laboratories to provide RBCwith a phenotype that matches the patient’s RBC pheno-type, with the aim of preventing or at least minimisingthe risk of incompatibility, particularly when daratumu-mab interference cannot be immediately resolved and/orthe RBC transfusion is urgent.6,14 Transfusion ofphenotype- or genotype-matched RBC will also reducethe risk of sensitisation and future alloantibody formation.

DTT is a thiol-reducing agent that denatures RBCsurface CD38 by disrupting the disulfide bonds in the mol-ecule’s extracellular domain, therefore preventing anti-CD38 from binding to the RBC.6 The use of DTT treat-ment is a recognised immunohaematological method. Thetest is robust and reproducible6 but not automated, and itis primarily used by specialist or reference laboratories.

Trypsin is a proteolytic enzyme not routinely used inAustralian laboratories and is less efficient than DTT treat-ment at cleaving cell-surface CD38.7 Other more com-monly used proteolytic enzymes, such as papain,bromelin or ficin, are used in immunohaematology test-ing as part of antibody identification protocols, to enhanceweak antibody activity or aid in the resolution of multipleantibody specificities. These enzymes may also be used aspart of the immunohaematology laboratory tool kit fordaratumumab interference investigations, but no valida-tion studies of the use of these enzymes in the resolutionof daratumumab interference have been published.

It must be noted that DTT and trypsin (along withother proteolytic enzymes) also denature or weaken thereactivity of some RBC antigens (see Box 1), and thisshould be taken into consideration when assessingresults from tests where these agents are used. In partic-ular, DTT is known to denature the Kell system antigens,and therefore, when used to resolve daratumumab inter-ference, patients should be transfused with K-negativeRBC unless they have been shown to be K-positive onprevious testing.6 At present, reagent RBC pre-treatedwith DTT or trypsin are not available from reagent man-ufacturers. Australian laboratories may not have accessto sufficient quantities of reagent RBC to prepare andmaintain DTT- or trypsin-treated antibody screening oridentification panels cells for regular routine use.

An alternative and the optimal approach to managingthe interference of the anti-CD38 antibody would be toneutralise the anti-CD38 antibody in the patient’splasma using soluble CD38 antigen or anti-CD38 idio-type antibody. However, both are expensive and notcurrently routinely available.

Cord blood cells do not bind anti-CD38 mAb. A sug-gestion has been made that these cells could be used, butmanufacturers of reagent RBC are constrained by limitedsupply. In a routine transfusion laboratory, other sourcesof suitable cord blood samples would not typically beavailable and would require registration as an in-housein vitro diagnostic (IVD). In addition, cord cells havealtered expression of some antigens, and this method isunlikely to be routinely offered by hospital or pathologylaboratories.14,17

Obtaining an extended RBC phenotype for thepatient prior to commencement of daratumumab ther-apy is important in the provision of phenotype-matchedRBC for future transfusions. Knowledge of the pheno-type means that donor RBCs negative for the commonclinically significant RBC antigens that the patient lackscan be selected for transfusion, thereby reducing thepossibility of RBC antibody formation.14 Patient RBCphenotyping should be performed by the transfusionlaboratory prior to the patient commencing daratumu-mab and at least 3 months after any recent blood trans-fusion (which otherwise may lead to misleadingresults). The patient sample could be sent for genotyp-ing where samples are unsuitable for phenotyping atany point pre- or post-commencement on daratumu-mab, but typing prior to treatment is recommended.The results are not received immediately, and this, inaddition to antibody investigation confounded by thepresence of daratumumab, might add to the delay inprovision of safe blood for transfusion. Ideally, thisinformation should be sought prior to commencement

Box 1 Antigens denatured orweakened by treatment with DTTorproteolytic enzymes15,16

DTT Trypsin Papain/Bromelin

Kell (K, k, Kpa, Kpb, Jsa,Jsb, Ku)

Cartwright (Yta)IndianJMHSciannaLWLutheranMER2Ge3DombrockSome DiegoCromer

Cartwright(Yta)

IndianJMHGe2,

Ge3, Ge4DombrockBpa

Ch/RgXga

MNEnaTSLutheranMer2Knops

DuffyMNSs, ‘N’IndianJMHBpa

Ch/RgXga

EnaFS EnaTSGe2, Ge4Fya, Fyb, Fy6Yta

Quach et al.


212

of treatment. As a minimum, the patient should betyped for Rh antigens, K, Jka, Jkb, Fya, Fyb and Ss.18 Tomanage workload and preserve reagents used, pheno-typing may be performed in regular, for example,weekly batches. Genotyping is currently only offered inAustralia by the Australian Red Cross Blood Service inBrisbane. Rapid genotyping testing may be available,but routinely, a 1-week turnaround time should betaken into consideration.A practical approach for immunohaematology testing

of RBCs in myeloma patients receiving treatment withthe anti-CD38 mAb, daratumumab, is detailed in the fol-lowing section. The real-world constraints are discussed,recognising that investigations to resolve anti-CD38interference are time consuming and labour intensiveand may not be available to all laboratories, especiallyregional or rural laboratories.

Pre-transfusion testing requirements

A: Prior to anti-CD38 therapy

Clear and timely communication between the treatingclinician, patient and transfusion laboratory is absolutelyvital when anti-CD38 therapy is planned. Patients andhealthcare providers must be made aware of the poten-tial interference of anti-CD38 in pre-transfusion testingand of the potential sequelae if appropriate immunohae-matological testing is not performed.The transfusion laboratory can be provided with a

request for phenotype if there has been no recent trans-fusion or RBC genotyping if the patient has beenrecently transfused or has a positive DAT, noting thatthe patient will receive anti-CD38 therapy. The clinicianshould provide the transfusion laboratory with a full andaccurate transfusion, obstetric and drug history for thepatient, and this may also require review of both hospitaland laboratory records.Routine pre-transfusion testing includes a blood group

(ABO/RhD) and antibody screen and will establish pre-treatment baseline results. An RBC phenotype(or genotype) is most valuable and, as a minimum,should include: Rh (C, c, E, e), K, Jka, Jkb, Fya, Fyb andSs antigens. Genotyping will be informative when phe-notyping is not possible due to recent transfusion (i.e. inthe last 3 months) or if the patient has a positive DAT orif suitable phenotyping reagents are not available. TheRBC phenotype and genotype can assist the laboratorynot only by suggesting what RBC alloantibodies thepatient may potentially form but also by enabling trans-fusion of phenotype- or genotype-matched RBC, whichwill minimise the risk of RBC incompatibility in situa-tions where underlying unexpected alloantibodies

cannot be excluded in the presence of daratumumab.Furthermore, phenotype- or genotype-matched RBCtransfusions will minimise the potential for sensitisationand future alloantibody formation. A clinical decisionmay be required on whether to limit or prioritise chosenphenotypes based on the urgency of the request and thedifficulty of providing matched units for transfusion.13

Information relating to the immunohaematology test-ing should be maintained in the patient’s clinical and lab-oratory files, and the patient should be provided with a‘patient alert card’, which can inform healthcare providersthat they are receiving anti-CD38 therapy. It is importantto consider that patients may attend several hospitals andbe tested at several transfusion laboratories, and it is alsoimportant to remember that in the absence of a jurisdic-tional or national alloantibody register, information aboutthe patient’s treatment with daratumumab and RBC phe-notype and prior RBC alloantibody history may not beaccessible by the transfusion laboratory or hospital atwhich the patient currently presents.

Prior to treatment with daratumumab:1 Communications from treating professional and transfu-sion laboratory to document that the patient is to start anti-CD38 mAb.2 Provide a full transfusion, obstetric and drug history.3 Perform a blood group (ABO, RhD).4 Perform an antibody screen and DAT.5 Perform an extended RBC phenotype (or genotype, whereindicated).6 Provide patient with an alert card (see Fig. 3).

B: Following commencement of anti-CD38therapy

It is extremely important for the transfusion laboratory toknow that a potential transfusion recipient is receivinganti-CD38 therapy. The treating clinician needs to under-stand the impact on pre-transfusion testing and to con-sider the timeframes for testing and provision of blood.Specimens from patients on anti-CD38 may need to bereferred to a reference laboratory for the more complexinvestigations necessary in these cases. The resourceimpacts on specialised reference services would be miti-gated if the neutralising antibody was listed on theAustralian Register of Therapeutic Goods (ARTG) and avail-able. This would also simplify and expedite pre-transfusiontesting and improve the relative safety of transfusion.The ABO/RhD typing is unaffected by the presence of

anti-CD38 and can be reported normally. The anti-CD38panagglutination typically results in a universally weak(1+ or 2+; using 0–4 scoring) positive antibody



213

screen.18,19 If one or more of the screening cells arestrongly reactive (3+ or 4+), this suggests the potentialpresence of an antibody, possibly an alloantibody, otherthan anti-CD38 (Fig. 1).

To overcome anti-CD38 interference, the antibodyscreen can be repeated using DTT- or trypsin-treatedreagent screening RBC. If this is negative, it may beassumed that no clinically significant RBC alloantibodiesare present, with the caveat that specificities directedagainst antigens denatured by the chosen enzyme cannotbe excluded. In the case where DTT-treated cells are used,the laboratory can select donor RBC that are ABO, RhDand K compatible, and these might be issued using thestandard institutional cross-match (XM) protocol for anegative antibody screen, for example, electronic (com-puter) or immediate spin (IS) XM. In the absence of anidentified RBC alloantibody using DTT-treated screeningcells, the decision to provide more extended phenotype-or genotype-matched RBC beyond RhD and K (includingRh Cc, Ee, Jka, Jkb, Fya, Fyb and Ss) will be influenced bythe availability of suitable units, clinical urgency of trans-fusion, anticipated current and future transfusion require-ments and local policy. If the patient is revealed to havean unexpected genotype with potential antibody forma-tion, this could be considered in planning.

Note that apart from DTT and trypsin, no validationstudies have been published for other enzymes ormethods for the purpose of resolving daratumumabinterference. Thus, if other enzymes or methods areused, our consensus is that blood matched to thepatient’s phenotype/genotype should be given, particu-larly if long-term transfusion support is anticipated.

A positive antibody screen using DTT- or trypsin-treated reagent RBC suggests that the patient has anadditional RBC alloantibody. The antibody specificitywill need to be determined using a DTT- or trypsin-treated RBC. Antigen-negative blood may then beselected for XM. RBC that match the patient’s extendedRBC phenotype/genotype should be selected for transfu-sion, with the degree of matching determined by clinicalurgency and the practicable availability of the desiredphenotyped donor blood. A full IAT XM is required, butthis will be incompatible unless DTT- or trypsin-treateddonor cells are used for the XM.

The flowchart (Fig. 2) represents the expert group’srecommendation for pre-transfusion testing in the pres-ence of anti-CD38. It is recognised that not all transfusionlaboratories in Australia and New Zealand will either rou-tinely use or have access to DTT- or trypsin-treatedreagent cells. The scope of testing will depend on institu-tional policy, clinical urgency and availability of appropri-ately phenotyped (or genotyped) donor RBC. Antibodiesdeveloped by patients to antigens such as Dombrock,

which are destroyed by DTT and without routine typingsera for donors or patients, will be missed. Clinicians needto pay careful attention for signs of acute or delayed hae-molytic transfusion reactions in patients on daratumumabafter any transfusion; the genotype might provide a cluewhere the phenotype is not available.

Pre-transfusion testing following commencement of daratumumab:1 Provide laboratory with a full transfusion, obstetric anddrug history.2 Order a blood group (ABO/RhD) and DAT.3 Perform antibody screen panel.4 If panagglutination is indicative of interference with anti-CD38 mAb on the antibody screen (see Fig. 1), perform anantibody screen using DTT- or trypsin-treated screening cells.(Other enzymes, e.g. papain, bromelain, ficin, may be used asan adjunct to help identify or exclude particular alloanti-bodies to RBC (Note: Methods other than DTT or trypsin have beenused but might not be validated for the purpose of resolving daratu-mumab interference. We suggest that if enzyme methods other thanDTT or trypsin are used, then extended phenotype-/genotype-matcheddonor RBC should be given (Rh Cc, Ee, Jka, Jkb, Fya, Fyb and Ss).))5 Perform an extended RBC phenotype (or genotype, whereindicated).6 Issue donor RBC.

C: Life-threatening bleeding and emergencytransfusions

For patients experiencing life-threatening bleeding or inemergency situations where transfusion is requiredwithin 2 h, there may not be time for the recommendedroutine pre-transfusion testing. Previous antibody his-tory, phenotype and genotype results are invaluable inthis circumstance.

There is a need to balance the clinical risks of transfu-sion versus those of not transfusing the patient, butunder no circumstances should transfusion be delayed inthe setting of a bleeding emergency.

The greatest risk to the patient is transfusion of ABO-incompatible blood. In emergency situations, the risk isnormally mitigated by transfusion of group O RhD-negativeblood; however, it should be noted that RhD-negativeblood is not necessarily the most appropriate in all cases,especially in patients that are Rh c-negative and or Rh e-negative.

ABO and RhD typing are not affected by the presence of anti-CD38 antibody in the patient’s plasma.

Transfusions should be in accordance with institutionalcritical bleeding or emergency transfusion policies. Fur-ther information on transfusion in emergency situations

Quach et al.


214

can be found in the ANZSBT’s ‘Guidelines for Transfusionand Immunohaematology Laboratory Practice’.20

Clinical considerations

Daratumumab is the first anti-CD38 mAb that receivedclinic approval by the FDA in 2015 and subsequent TGAapproval in Australia in 2017. Its use in combinationwith current therapeutics, such as lenalidomide or borte-zomib, increases the frequency of minimal residual dis-ease negative remissions in MM, which may translate toimprovement in survival outcome.21,22 Healthcare pro-viders have not been adequately prepared for the critical

interference of this drug in laboratory tests, particularlypre-transfusion testing. The problem will increase if dar-atumumab’s use expands to early-phase diseasetreatment.A crucial aspect in risk mitigation is education to

increase awareness and a robust procedure to enabletimely and routine communication with the blood trans-fusion laboratory. The patient and family members needto be aware of daratumumab’s interference in pre-transfusion tests and the potential impact this may haveon any blood transfusions. A patient alert card (see Fig. 3)is also useful for this purpose. All levels of medical care –

from nursing staff to doctors and transfusion laboratory

Figure 1 Typical red blood cell reactivity due to anti-CD38 in a patient’s plasma. (A) Indirect antiglobulin test (IAT) panel where all cells display reac-

tions consistent with anti-CD38 therapy; (B) IAT panel with some reactions consistent with anti-CD38 therapy; however, the pattern suggests that an

alloantibody is present; (C) Saline panel where there is no interference by anti-CD38 therapy in cells 1–3, while cells 4–11 are positive for alloantibody.



215

Blood transfusion required

Life threatening bleeding / Emergency transfusion

Non-urgent (required > 2 h)

Perform Phenotyping or Genotyping (if time allows)

• Phenotype (if not transfused in preceding 3 months)

• Genotype (if recently transfused or DAT positive)

Perform IAT antibody screen

using DTT or trypsin treated red cellsc

NEGATIVE

• Assume no clinically significant red cell alloantibody/ies

• Cannot exclude antibodies to antigens denatured by chosen treatment method (see Box 1)

• Transfuse ABO /RhD compatible blood and blood compatible for any significant antigens destroyed by the method used e.g. Kell compatible for DTT methods (see Box 1)

• Consider selecting blood matched to patient’s extended phenotype /genotype, particularly if long-term transfusion support anticipated

• Abbreviated crossmatch (eXM or IS) and issue blood by usual protocol

• If IAT crossmatch used – will be positive unless donor cells are DTT or trypsin treated

POSITIVE

• Suggests presence of red cell alloantibody/ies.

• Identify antibody/s using DTT or trypsin treated ID antibody panel – may require investigation by a Reference Laboratory

• Cannot exclude alloantibodies against antigens denatured by chosen treatment method (see Box 1)

• Select blood that is compatible for antibody/s and antigens denatured by chosen treatment method, e.g. Kell compatible for DTT methods (see Box 1)

• If alloantibody cannot be identified for any reason, consider selecting blood matched to patient’s extended phenotype/genotype, particularly if long-term transfusion support anticipatedd

• Full IAT crossmatch – will be positive unless donor cells are DTT or trypsin treated

Issue blood using institutional protocols for emergency

transfusion or Massive Transfusion Policy (MTP).a

• Emergency O RhD-negative red cellsb

• Switch to ABO/RhD and Kell compatible red cells where appropriate.

Take samples before transfusion for retrospective testing and cross matching

YES NO

POSITIVE

Is transfusion

urgent?

Routine IAT antibodyscreen

Provide phenotype compatible blood, if available or refer samples to reference laboratory for further evaluationd

Use of other routine laboratory tests (papain

or bromeline) to exclude the development of a new of existing alloantibody may assist in the selection process for phenotype compatible units

POSITIVEPAN-AGGLUTINATION TYPICAL OF ANTI-CD38 (DARATUMUMAB) INTERFERENCE, E.G.

1+ TO 2+ REACTIONS WITH ALL CELLS; OWN CELLS (‘AUTO CONTROL’) MAY BE NEGATIVE

NEGATIVE

Proceed as per usual institutional protocol

Does laboratory have capability for testing using DTT or trypsin treated red

cells?c

NO

YES

Pretransfusion sample

Blood group (ABO/RhD), antibody screen

NEGATIVE

Figure 2 Pre-transfusion testing recommendations. aRefer to ANZSBT Guidelines for Transfusion and Immunohaematology Laboratory Practice; bO-

negative blood is not without risk and may not be suitable in all circumstances, e.g. patient has anti-c or anti-e antibodies; cTests using DTT or trypsin

treated red cells are published methods for resolving anti-CD38 (daratumumab) interference; however, testing may not be available in all laboratories

and/or subject to regulatory restrictions; dExtended phenotype/genotype including as a minimum: Rh (C, c, D, E, e), K, Jka, Jkb, Fya, Fyb and Ss; ePapain

and bromeliad are not IAT methods for crossmatching purposes. DAT, direct antiglobulin test; DTT, dithiothreitol; IAT, indirect antiglobulin test.

Quach et al.


216

scientists – need to be educated to ensure effective com-munication and adequate documentation in the patientrecord and the transfusion laboratory information sys-tem (LIS). Every public and private haematology/oncol-ogy facility should have a procedure to automaticallynotify the relevant transfusion laboratory when a patientis about to commence daratumumab and provide theappropriate specimens for testing. This will allow for

baseline extended RBC phenotype (regardless of theimmediate need for blood transfusion). The transfusionslaboratory requires ongoing notification of daratumu-mab treatment when RBC transfusion is requested forup to 6 months post-treatment cessation. Updating bloodtransfusion requisition forms to include questions aboutantiCD38 mAb might be considered, as well as suitablealert notifications in electronic alert/chemotherapy

Figure 2 (Continued)



217

prescribing systems and alerts in the transfusion LIS tostate that the patient is receiving daratumumab.

In the transfusion laboratory, while both DTT and tryp-sin are widely recommended, these methods are notalways practical when laboratories rely heavily on auto-mation. These methods are manual and laborious andincur additional costs. Although robust and reproducible,6

in Australia, both DTT and trypsin used in these methodshave not been approved for use as in vitro diagnostics bythe TGA. There are no commercially available DTT or tryp-sin reagents listed on the ARTG, nor are DTT- or trypsin-treated reagent RBC screening or extended panels avail-able. Thus, both methods would be considered ‘in-house’methods and may not meet Australian IVD device regula-tions, despite being fully validated by laboratories beforeintroduction. There is no current prospect of commercialavailability of ATRG-listed soluble CD38 or anti-idiotypeantibodies to neutralise the effect of daratumumab.

In the face of these constraints, the default contin-gency for many laboratories will be to issue extended

phenotype- or genotype-matched blood where available.The ensuing impact of increased demands on the ARCBSand the increasing need for relevant immunohaematol-ogy expertise outside of large metropolitan laboratorieswill need to be considered. The establishment of anational RBC alloantibody register has been under con-sideration and might reasonably include relevant docu-mentation for these circumstances.

With respect to the impact on patients, the risk per-tains not only to possibility of missing a significant allo-antibody that may cause acute or delayed haematolytictransfusion reactions but also to the delay in issuing ofblood products. The potential for delay is present bothwhen transfusion laboratories are unaware that patientsare receiving daratumumab and when, if aware, arerequired to undertake increased testing. Haemolytictransfusion reactions because of daratumumab interfer-ence with pre-transfusion testing were not reported inthe two pivotal phase III CASTOR23,24 and POLLUX25,26

studies. The patients in these trials were in the relatively

Name: _____________________________________

I am taking the following medication:

• <<insert anti-CD38 antibody>> product for the treatment of multiple myeloma

Dear Healthcare Provider,

Indirect antiglobulin test [IAT; Indirect Coomb’s test] may show positive results in patients taking daratumumab, even in the absence of other clinically significant RBC antibodies in the patient’s plasma. The determination of a patient’s ABO and RhD blood type are not affected.

If an emergency transfusion is required, uncrossmatched, ABO/RhD compatible RBC’s can be given as per local institutional policies. As dithiothreitol (DTT) treatment also denatures Kell antigens, K-negative units must be provided unless the patient is known to be K-positive.

For more information, please contact <<insert company name, telephone number and email address>>

Additional information on interference with blood compatibility testing can be found in the <<insert anti-CD38 antibody>> product information leaflet at <<insert website>>.

Before starting <<insert anti-CD38 antibody>>, my blood test results collected on (date) ________________________________were:Blood type: A B AB O Rh+ Rh-

Indirect Antiglobulin Test IAT (Coomb’s) antibody screen was: Negative Positive for the following antibodies:

Other: _______________________________________

Contact details of institution where the blood tests were performed:

_____________________________________________________ Figure 3 Example of patient

alert card.

Quach et al.


218

early course of their disease (with a median of 1–2 priorlines of treatment) and were not commonly transfusion-dependent. Conversely, in the clinic, daratumumab iscurrently also FDA- and TGA-approved as monotherapyfor heavily pretreated patients who have had at leastthree prior lines of therapy. It is, therefore, expected thathigher transfusion requirements will be seen in theseend-stage patients, and we cannot be certain of thenotion that no haemolytic transfusion reactions havebeen observed in daratumumab-treated patients before.Clinicians and laboratories should be aware of the poten-tial for acute and delayed haemolytic transfusion reac-tions and should investigate, document and report anysuch reactions, or adverse events through their localhaemovigilance programme.

Future directions

As the use of mAb is becoming increasingly prevalent fortherapy of cancers and other medical conditions, theconcept of potential interference in critical laboratorytests needs to be recognised and appropriate antibodyneutralising solutions developed, preferably prior to thewidespread introduction of these agents into the com-munity. The introduction of daratumumab into clinicaluse in MM has indeed created a predicament in thetransfusion laboratory that is without precedent, butshould serve as a case in point to gain experience andprepare for similar scenarios in the future. Any mAb thattargets common antigens present on RBC have thepotential to interfere with pre-transfusion testing. Cur-rently, these include the other anti CD38 mAb, such asisatuximab and MOR202,27,28 both of which are under-going clinical studies for the treatment of MM. While thenature of interference of these monoclonal antibodies isanticipated to be similar to that of daratumumab, thismay not become clear until the drugs are more widelyused. It is unclear whether there is concurrent develop-ment of an antidote to neutralise any of their interfer-ence in critical tests within the core laboratory. Fordaratumumab, neutralisation methods (soluble anti-CD38 mAb or anti-CD38 idiotype antibody) have beenused successfully and are a fast and uniform way to dealwith the interference.11 Such kits could attain IVDapproval and reduce the need for labour-intensive test-ing within the transfusion laboratory. Cost has been abarrier, and currently, the only commercial kit available(DIRA; Sebia, Evry Cedex, France) is in use to resolve

daratumumab’s interference in serum protein electro-phoresis and immunofixation assays, which are methodsto quantitate and type monoclonal immunoglobulins(M-proteins), respectively, in the serum or urine. In theabsence of such a kit for pre-transfusion testing, otherways to resolve the problem, to minimise workflow dis-ruption to transfusion laboratories and mitigate risks topatients must be considered.If a transfusion laboratory is not aware that a patient is

receiving daratumumab, protracted investigation anddelays are likely to occur when unexpected panaggluti-nation is found in the routine antibody screen. Anational database (or register) of patients treated withdaratumumab or any other mAb that interferes withpre-transfusion tests could provide an easily accessiblesource of information for patients who may demonstrateinterference in immunohaematology testing. Such adatabase, if incorporated in an antibody register or data-base, could also potentially alert the local laboratory ser-vice when a patient is known to have RBC allo- orautoantibodies. This might reduce delays in immunohae-matology testing and time to appropriate transfusion.Such databases have been recommended in otherjurisdictions.14

At the hospital level, routine and automatic notifica-tion to the transfusion laboratory about a patient’s treat-ment status could be mandated. Automated alerts,through electronic medical record systems to the transfu-sion laboratory, for every patient on treatment that mayinterfere with immunohaematology tests or requireselection of specialised blood products could be imple-mented. Investment in the development of this infra-structure needs to happen now to prepare adequatelyfor the surge of mAb in clinical use in the near future.For future targeted therapies, we emphasise the need toexplore fully any potential interference with critical labo-ratory assays that may impact the other areas of clinicalpractice prior to their introduction into the clinics.

Acknowledgements

The authors thank Belinda Butcher, BSc (Hons), MBio-stat, PhD, CMPP, AStat (WriteSource Medical Pty Ltd.,Sydney, Australia) for providing medical writing support,which was funded by Janssen-Cilag Pty Ltd, Sydney,Australia, in accordance with Good Publication Practice(GPP3) guidelines (http://www.ismpp.org/gpp3).

References

1 Usmani SZ, Weiss BM, Plesner T,

Bahlis NJ, Belch A, Lonial S et al. Clinical

efficacy of daratumumab monotherapy

in patients with heavily pretreated

relapsed or refractory multiple myeloma.

Blood 2016; 128: 37–44.

2 Mehta K, Shahid U, Malavasi F. Human

CD38, a cell-surface protein with

multiple functions. FASEB J 1996; 10:

1408–17.



219

http://www.ismpp.org/gpp3)

3 Krejcik J, Casneuf T, Nijhof IS, Verbist B,

Bald J, Plesner T et al. Daratumumab

depletes CD38+ immune regulatory cells,

promotes T-cell expansion, and skews

T-cell repertoire in multiple myeloma.

Blood 2016; 128: 384–94.

4 de Weers M, Tai YT, van der Veer MS,

Bakker JM, Vink T, Jacobs DC et al.

Daratumumab, a novel therapeutic

human CD38 monoclonal antibody,

induces killing of multiple myeloma and

other hematological tumors. J Immunol

2011; 186: 1840–8.

5 Overdijk MB, Jansen JH, Nederend M,

Lammerts van Bueren JJ, Groen RW,

Parren PW et al. The therapeutic CD38

monoclonal antibody daratumumab

induces programmed cell death via

Fcgamma receptor-mediated cross-

linking. J Immunol 2016; 197: 807–13.

6 Chapuy CI, Aguad MD, Nicholson RT,

AuBuchon JP, Cohn CS, Delaney M

et al. International validation of a

dithiothreitol (DTT)-based method to

resolve the daratumumab interference

with blood compatibility testing.

Transfusion 2016; 56: 2964–72.

7 Chapuy CI, Nicholson RT, Aguad MD,

Chapuy B, Laubach JP, Richardson PG

et al. Resolving the daratumumab

interference with blood compatibility

testing. Transfusion 2015; 55: 1545–54.

8 Oostendorp M, Lammerts van

Bueren JJ, Doshi P, Khan I, Ahmadi T,

Parren PW et al. When blood

transfusion medicine becomes

complicated due to interference by

monoclonal antibody therapy.

Transfusion 2015; 55: 1555–62.

9 Hannon JL, Clarke G. Transfusion

management of patients receiving

daratumumab therapy for advanced

plasma cell myeloma. Transfusion 2015;

55: 2770.

10 Sullivan HC, Gerner-Smidt C,

Nooka AK, Arthur CM, Thompson L,

Mener A et al. Daratumumab (anti-

CD38) induces loss of CD38 on red

blood cells. Blood 2017; 129: 3033–7.

11 De Vooght KM, Oostendorp M, van

Solinge WW. New mAb therapies in

multiple myeloma: interference with

blood transfusion compatibility

testing. Curr Opin Hematol 2016; 23:

557–62.

12 Patient Blood Management Guidelines:

Module 1 Critical Bleeding/Massive

Transfusion. Canberra: National Blood

Authority, Australian Government,

National Health and Medical Research

Council; 2011.

13 Chou ST, Westhoff CM. Application of

genomics for transfusion therapy in

sickle cell anemia. Blood Cells Mol Dis

2017; 67: 148–54.

14 De Vooght KM, Oostendorp M, van

Solinge WW. Dealing with anti-CD38

(daratumumab) interference in blood

compatibility testing. Transfusion 2016;

56: 778–9.

15 Branch DR, Muensch HA, Sy Siok

Hian AL, Petz LD. Disulfide bonds are a

requirement for Kell and Cartwright

(Yta) blood group antigen integrity. Br J

Haematol 1983; 54: 573–8.

16 Reid ME, Lomas-Francis C, Olsson ML.

The Blood Group Antigen Factsbook.

Cambridge, MA: Academic Press,

Elsevier; 2012.

17 Schmidt AE, Kirkley S, Patel N,

Masel D, Bowen R, Blumberg N et al.

An alternative method to dithiothreitol

treatment for antibody screening in

patients receiving daratumumab.

Transfusion 2015; 55: 2292–3.

18 Australian and New Zealand Society of

Blood Transfusion. Guidelines for

Transfusion and Immunohaematology

Laboratory Practice. Sydney: The

Transfusion; 2016 [cited 2017 Jun 21].


anzsbt.org.au/pages/anzsbt-

guidelines.htm

19 AABB. Mitigating the Anti-CD38

Intereference with Serologic Testing.

AABB Association Bulletin No. 16-02;

2016 [cited 2017 Jun 21]. Available

from URL: http://www.aabb.org/

programs/publications/bulletins/

Documents/ab16-02.pdf

20 Australian and New Zealand Society of

Blood Transfusion. Guidelines for

Transfusion and Immunohaematology

Laboratory Practice, 1st edn. Sydney:

Australian and New Zealand Society of

Blood Transfusion; 2016.

21 Dimopoulos MA, Oriol A, Nahi H, San-

Miguel J, Bahlis NJ, Usmani SZ et al.

Daratumumab, lenalidomide, and

dexamethasone for multiple myeloma.

N Engl J Med 2016; 375: 1319–31.

22 Palumbo A, Chanan-Khan A, Weisel K,

Nooka AK, Masszi T, Beksac M et al.

Daratumumab, bortezomib, and

dexamethasone for multiple myeloma.

N Engl J Med 2016; 375: 754–66.

23 Chanan-Khan A, Lentzch S, Quach H,

Horvath N, Capra M, Ovilla R et al.

Daratumumab, Bortezomib, and

Dexamethasone Versus Bortezomib and

Dexamethasone Alone for Relapsed or

Refractory Multiple Myeloma Based on

Prior Treatment Exposure: Updated Efficacy

Analysis of CASTOR. San Diego:

American Society of Hematology;

2016.

24 Mateos M-V, Estell J, Barreto W,

Corradini P, Min C-K, Medvedova E

et al. Efficacy of Daratumumab, Bortezomib,

and Dexamethasone Versus Bortezomib and

Dexamethasone in Relapsed or Refractory

Multiple Myeloma Based on Prior Lines of

Therapy: Updated Analysis of CASTOR.

San Diego: American Society of

Hematology; 2016.

25 Moreau P, Kaufman JL, Sutherland H,

Lalancette M, Magen H, Iida S et al.

Efficacy of daratumumab, lenalidomide,

and dexamethasone versus

lenalidomide and dexamethasone alone

for relapsed or refractory multiple

myeloma among patients with 1 to

3 prior lines of therapy based on

previous treatment exposure: updated

analysis of POLLUX. Blood 2016;

128: 489.

26 Usmani SZ, Dimopoulos MA, Belch A,

White D, Benboubker L, Cook G et al.

1151 Efficacy of Daratumumab,

Lenalidomide, and Dexamethasone Versus

Lenalidomide and Dexamethasone in

Relapsed or Refractory Multiple Myeloma

Patients with 1 to 3 Prior Lines of

Therapy: Updated Analysis of POLLUX.

San Diego: American Society

Haematology; 2016.

27 Martin T, Baz R, Benson DM,

Lendvai N, Wolf J, Munster P et al. A

phase 1b study of isatuximab plus

lenalidomide and dexamethasone for

relapsed/refractory multiple myeloma.

Blood 2017; 129: 3294–303.

28 Raab MS, Chatterjee M, Goldschmidt H,

Agis H, Blau I, Einsele H et al. 1152 A

phase I/IIa study of the CD38 antibody

MOR202 alone and in combination

with pomalidomide or lenalidomide in

patients with relapsed or refractory

multiple myeloma. American Society of

Hematology 58th Meeting and Exposition.

San Diego, CA; 2016.

Quach et al.


220

https://www.anzsbt.org.au/pages/anzsbt-guidelines.htm



http://www.aabb.org/programs/publications/bulletins/Documents/ab16-02.pdf



doi:10.1111/imj.13696

LETTERS TO THE EDITOR


Benefit of routine testicularexamination: hypogonadismin a person with 47XYY

This case study is to remind readers of the benefit of car-rying out routine testicular examination in people whopresent for unrelated reasons. A 70-year-old Caucasianman presented with a 20-year history of type 2 diabetes,and treated with insulin for 15 years, after having hisdriver’s licence suspended due to poor compliance. Thelevel of HbA1c was 102 mmol/mol and he was adminis-tering 60 units ‘NovoMix’ insulin in the morning and50 units in the evening.He had a history of a speech disorder, which he said

was the cause of poor performance at school. Hereceived speech therapy in his early 20s, whichimproved his speech greatly. The patient described him-self as always being different to his classmates, and assomewhat of a loner. He had been an apprentice carpen-ter for 4 years, but had spent most of his working life asa public servant in an administrative capacity and wasnow retired. He was not able to tolerate alcohol, was anon-smoker, lived alone, had never had a partner orchildren and was homosexual. There was no history ofmacrovascular disease.Libido had been low especially after a diagnosis of

Burkitt lymphoma had been made 2½ years previously,with the patient receiving six cycles of DA-R- EPOCH(etoposide, prednisone, vincristine, cyclophosphamideand rituximab).On examination, the patient was tall with a height of

186 cm, a weight of 146 kg and a body mass index of42 kg/m2. He had asymptomatic peripheral neuropathy,mild non-proliferative diabetic retinopathy, bilateral cat-aracts, microalbuminuria (25 mg/mmol creatinine) butnormal renal function (estimated glomerular filtrationrate 86 mL/min). Testes were small in size measuring6–8 mL, pubic hair was adult and there was nogynaecomastia.

Investigations confirmed primary hypogonadism witha low testosterone of 2.6 nmol/L and raised luteinisinghormone of 22 IU/L and follicle stimulating hormonemeasuring 25 IU/L. Chromosomal analysis did not con-firm Klinefelter syndrome, which seemed clinically likely,but rather 47XYY, a condition not usually associated withhypogonadism1 or small testes.2 The alkylating agentcyclophosphamide is the most likely chemotherapeuticagent responsible for adversely affecting the testes, withshrinkage of the gonads, and primary hypogonadism.3,4

The patient was given exogenous testosterone in theform of ‘Testogel’, with resultant increase in libido akinto how he had felt prior to receiving chemotherapy.The patient had many of the features of XYY syn-

drome, a disorder, affecting perhaps 1 out of every 1000males.1 Most people with this disorder are diagnosed ser-endipitously, as in this case, since apart from tall stature,there are no phenotypic features.5 The exact cause oferror in cell division is not known. Intelligence is normal,but there is an increased incidence of learning disabilities,behavioural problems and speech disabilities, which ben-efit from speech therapy and tutoring.6 Aggression is nota feature, and fathering of children is the norm.In summary, this unusual case demonstrates three

points. First, routine testicular examination may revealotherwise unsuspected conditions, which may also betreatable. Second, the diagnosis of XYY, as is so often thecase, was unexpected, and third, some chemotherapeuticagents are toxic to the testis. A PubMed search for ‘valueof routine testicular or scrotal examination’ produced atotal of 45 citations, most relating to infertility and nonesimilar to our case.


Eman Negm1 and Bernard E. Tuch 1,2

1Specialist Medical Centre, and 2School of Medical Sciences,

Discipline of Physiology, University of Sydney,

Sydney, New South Wales, Australia

References

1 Gravholt CH. Sex chromosome

abnormalities. In: Pyeritz RE, Rimoin DL

and Korf BR, eds. Emery and Rimoin’s

Principles and Practice of Medical

Genetics, 6th edn. San Diego, CA:

Elsevier Academic Press; 2013;

1180–211.

2 Boisen E. Testicular size and shape of

47,XYY and 47,XXY men in a double-

blind, double-matched population

survey. Am J Hum Genet 1979; 31:

697–703.

3 Schilsky RL, Lewis BJ, Sherins RJ,

Young RC. Gonadal dysfunction in

patients receiving chemotherapy for

cancer. Ann Intern Med 1980; 93: 109–14.


221

4 Chapman RM. Effect of cytotoxic therapy

sexuality and gonadal function. Semin

Oncol 1982; 9: 84–94.

5 Ross J, Zeger M, Kushner H, Zinn AR,

Roeltgen DP. An extra X or Y

chromosome: contrasting the cognitive

and motor phenotypes in childhood in

boys with 47,XYY syndrome or 47,XXY

Klinefelter syndrome. Dev Disabil Res Rev

2009; 15: 309–17.

6 Leggett V, Jacobs P, Nation K, Scerif G,

Bishop DV. Neurocognitive outcomes of

individuals with a sex chromosome

trisomy: XXX, XYY, or XXY: a systematic

review. Dev Med Child Neurol 2010; 52:

119–29.

doi:10.1111/imj.13697

Boerhaave syndrome: a commonmanifestation of a rare disease

A 65-year-old woman with multiple cardiovascular riskfactors presented with sharp central chest pain, whichradiated to the back and was exacerbated by inspirationand lying supine, and relieved by leaning forward. Thepatient also reported six episodes of diarrhoea on the daybefore presentation and one episode of vomiting. ChestX-ray revealed congested lungs and small bilateral pleuraleffusions suggesting cardiac failure or fluid overload.

She was admitted and investigated for acute coronarysyndrome complicated by cardiac failure. However, serialtroponin assays were negative, while white cell countwas elevated and her pain persisted, prompting furtherinvestigations for an underlying cause. A computedtomography of the abdomen revealed lower mediastinalfree gas in a pattern suggestive of a ruptured oesophagus(Fig. 1). The patient was subsequently airlifted to a ter-tiary centre, where she underwent a thoracotomy,debridement and drainage of mediastinum and a lapa-rotomy for insertion of feeding jejunostomy. At the timeof surgery, an adjacent oesophageal diverticulum wasidentified. Post-operative gastrografin swallow con-firmed the ongoing presence of an oesophageal leak.

The leak was managed conservatively. The patientunderwent an extensive inpatient stay followed by a periodof rehabilitation. She has since been able to be recom-menced on a normal diet and has made a full recovery.

More than half of oesophageal perforations are iatro-genic, such as following endoscopic instrumentation,while approximately a third are spontaneous in nature.1

Spontaneous oeseophageal rupture, also known as Boer-haave syndrome, is attributed to barogenic trauma due toan increase in intraoesophageal pressure, typically duringvomiting.2 It is a rare yet potentially life-threatening con-dition with a mortality rate as high as 50% if delays indiagnosis and treatment are incurred.3 Therefore, earlydiagnosis of oesophageal rupture is critical.

Spontaneous oesophageal rupture classically presentswith a triad of symptoms including vomiting, chest painand subcutaneous emphysema, known as Mackler triad.However, the combination of symptoms is not alwayspresent. The typical presentation includes chest pain,while shortness of breath, vomiting and other non-specific features are common.

Cases of oesophageal perforation are most frequentlyassociated with raised intraluminal pressure, such as invomiting or retching. However, compromised oesophagealwall integrity, such as in chronic reflux,4 eosinophilicoesophagitis5 and dermatomyositis6 have also been identi-fied as potentially important predisposing factors. Ourpatient was eventually found to have an oesophagealdiverticulum. As far as we are aware, spontaneous rupturewith underlying diverticulum has not previously beendescribed. It is unclear whether oesophageal diverticulapredispose to spontaneous rupture, but it is feasible thatboth may represent a weakness in the oesophageal wall.

This case illustrates the importance for clinicians tohave spontaneous oesophageal rupture as part of theirdifferentials when managing patients who present withchest pain.

Received 10 April 2017; accepted 6 June 2017.

Adel Ekladious1 and Mark K. Tiong 2

1Mackay Base Hospital, Mackay, Queensland, and 2Department of

General Medicine, Mildura Base Hospital,

Mildura, Victoria, Australia

Figure 1 Computed tomography of the abdomen, soft axial. Arrow

indicates lower mediastinal free gas.



222

References

1 Bhatia P, Fortin D, Inculet RI,

Malthaner RA. Current concepts in the

management of esophageal perforations:

a twenty-seven year Canadian

experience. Ann Thorac Surg 2011; 92:

209–15.

2 Korn O, Onate JC, Lopez R. Anatomy of

the Boerhaave syndrome. Surgery 2007;

141: 222–8.

3 Brinster CJ, Singhal S, Lee L,

Marshall MB, Kaiser LR, Kucharczuk JC.

Evolving options in the management of

esophageal perforation. Ann Thorac Surg

2004; 77: 1475–83.

4 Salo JA, Seppälä KM, Pitkäranta PP,

Kivilaakso EO. Spontaneous rupture and

functional state of the esophagus. Surgery

1992; 112: 897–900.

5 Straumann A, Bussmann C, Zuber M,

Vannini S, Simon HU, Schoepfer A.

Eosinophilic esophagitis: analysis of food

impaction and perforation in

251 adolescent and adult patients.

Clin Gastroenterol Hepatol 2008; 6:

598–600.

6 Dougenis D, Papathanasopoulos P,

Paschalis C, Papapetropoulos T.

Spontaneous esophageal rupture in

adult dermatomyositis. Eur J

Cardiothorac Surg 1996; 10:

1021–3.

doi:10.1111/imj.13701

Role of radiotherapy in managementof gingival infiltration of chronicmyelomonocytic leukaemia

Chronic myelomonocytic leukaemia (CMML) is a malig-nant haemopoietic stem cell disorder with clinical andpathological features of both a myeloproliferative neo-plasm and myelodysplastic syndrome.1

Gingival infiltration is a presenting symptom in 5%of patients. It is commonly seen in acute monocyticleukaemia.2–4

A 67-year-old woman presented with 18 months his-tory of CMML and 15 months of leukaemic gingivitis.She had been treated with chemotherapy (oral thio-

guanine and azacitidine) due to moderate tumourburden and splenomegaly. White cell count and spleno-megaly improved.However, gingival infiltration progressed, as did asso-

ciated symptoms: oral discomfort and bleeding, anorexia,nausea and 10 kg weight loss. Bleeding did not respondto tranexamic acid (Fig. 1).

She was treated with low-dose palliative radiotherapy

4 Gy in two fractions5,6 to entire oral cavity without

acute or late toxicity. There was gradual response such

that by 12 weeks post radiation there was resolution of

oral discomfort and anorexia, reduced bleeding, and

75% reduction in gingival hypertrophy (Fig. 2).Response was maintained for further 7 months. She

was treated with further 4 Gy in two fractions to the oral

cavity but died 6 weeks after from disease progression.Gingival hyperplasia is characterised by progressive

enlargement of the interdental papilla as well as the

marginal and attached gingiva. Mucosal haemorrhages,

ulcerative gingivitis, infectious gingivitis and odontalgia

may be observed.Pallor, spontaneous haemorrhage, petechiae and

ulceration occur more frequently in acute than chronic

leukaemia.Dental caries and poor oral hygiene have not been

described as risk factors; however, they can predispose to

superinfection, necrosis, pain and bleeding.

Figure 1 Pre radiotherapy. Figure 2 Twelve weeks post radiotherapy.



223

Generally, gingival hyperplasia resolves completely orat least partly with effective chemotherapy.

Low-dose radiation has been shown to be an effectivetreatment with minimal morbidity in the setting of indo-lent non-Hodgkin lymphoma.5,6

In the event of failure of response of leukaemic gingi-vitis to chemotherapy, palliative low-dose radiation canbe used for local symptom control. It is likely to be effec-tive and well tolerated with minimal toxicity.

Low-dose palliative radiation should be considered forpatients with symptomatic chemotherapy resistant leu-kaemic deposits, such as leukaemic gingivitis.

Received 27 April 2017; accepted 20 September 2017.

Bahare Moradi , Stephen Thompson andTimothy Brighton

Nelune Comprehensive Cancer Centre, Prince of Wales Hospital,

Faculty of Medicine, UNSW, Sydney, New South Wales, Australia

References

1 Onida F, Barosi G, Leone G, Malcovati L,

Morra E, Santini V et al. Management

recommendations for chronic

myelomonocytic leukaemia: consensus

statements from the SIE, SIES, GITMO

groups. Haematologica 2013;

98: 1344–52.

2 Anil S, Smaranayake LP, Nair RG,

Beena VT. Gingival enlargement as a

diagnostic indicator in leukaemia. Aust

Dent J 1996; 41: 235–7.

3 Felix DE, Lukens J. Oral symptoms as a

chief sign of acute monoblastic

leukaemia. J Am Dent Assoc 1986; 113:

899–900.

4 McKenna SJ. Leukaemia. Oral Surg Oral

Med Oral Pathol Oral Radiol Endol 2000;

89: 137–9.

5 Hoskin P, Kirkwood AA, Popova B,

Smith P, Robinson M, Gallop-Evans E

et al. 4 Gy vs 24 Gy radiotherapy

for patients with indolent lymphoma:

randomized phase 3 non-inferiority trial.

Lancet Oncol 2014; 15: 457–63.

6 Illidge T, Specht L, Yahalom J, Aleman B,

Berthelsen AK, Constine L et al. Modern

radiation therapy for nodal non-Hodgkin

lymphoma-target definition and dose

guidelines from the International

Lymphoma Radiation Oncology Group.

Int J Radiation Oncol 2014; 89: 49–58.

doi:10.1111/imj.13704

Encephaloclastic cyst: a rarecomplication of a malfunctioningmethotrexate Ommaya reservoir

The Ommaya reservoir (OR) is an intraventricular cathe-ter system used to treat leptomeningeal malignancy. Itallows for the instillation of chemotherapy into the cere-brospinal fluid (CSF) providing consistent drug levels.1,2

Methotrexate (MTX) can be administered intrathecallyto treat leptomeningeal metastases. Known neurotoxic-ities include aseptic meningitis, seizures and polyradicu-lopathy.3 Encephaloclastic cysts are non-infectiouscomplications rarely seen in the OR resulting from thecystic dilatation of the brain parenchyma around a cath-eter in response to a chemotherapeutic agent.4 Thesecysts have been seldom reported and are believed toresult from extravasation or back-flow of CSF along thecatheter.1,4,5

A 67-year-old woman with breast cancer carcinoma-tous meningitis developed confusion, drowsiness andleft-sided hemiplegia 3 days after instillation of 12 mgMTX through her OR. The device was inserted 1 monthpreviously, through the right frontal cortex and had fourprior instillations (45 mg MTX total) without anyreported complications. There was no trauma to thedevice and review of chemotherapy records showed nountoward events. Her magnetic resonance image (MRI)

showed a cystic fluid collection, 24 mm × 16 mm ×27 mm in the right sub-cortical matter below the supe-rior frontal gyrus with extensive white matter increasedsignal change with sulcal and ventricular effacement, asshown in Figure 1, with no enhancing focus.

Treatment included high-dose intravenous dexametha-sone, with subsequent removal of the OR and intraven-tricular catheter. During surgical removal, CSF samplingshowed a pleocytosis consistent with aseptic meningitis.White cell count: 44, protein: 237 mg/L (0–400 mg/L),glucose: 4.0 mmol/L (2–4 mmol/L), culture negative. Abiopsy of the cystic lesion showed inflammatory changewith reactive gliosis and some haemorrhage; no evidenceof malignancy was found. There was no eosinophilic infil-tration or features of demyelination to suggest animmune reaction to the catheter. Testing of the removedOR ex vivo demonstrated a small leak from the undersideof the reservoir, suggesting needle penetration throughthe base of the reservoir. One week later, her left-sidedweakness and confusion began improving allowing forcommencement of rehabilitation. A 1-week progress MRIshowed regression of oedema and reduction in the cavitysize to 18 mm × 10 mm. Her dexamethasone was weanedand she returned home independently after 20 days.

MTX CNS toxicity results from the disruption of theblood–brain barrier leading to focal necrotising changesand subsequent development of local chemical



224

encephalitis.2,4 Removal of the OR is recommended,thereby eliminating CSF’s path of least resistance alongthe catheter.2 Corticosteroid treatment remains contro-versial and is supported if major necrosis is noted orlarge volumes of MTX are used. Improvement is gener-ally observed in most patients after reservoir removal.2

We report an unusual presentation of local toxic enceph-alitis due to a malfunctioning OR. Clinical features ofconfusion and hemiplegia prompted concern relating todevice malfunction, with expedited reservoir removaland corticosteroids resulting in neurological recovery.Clinicians should consider MTX toxicity and device

Figure 1 (A) Coronal T2 magnetic resonance imaging (MRI), (B) Axial T2 MRI: increased signal intensity with a cystic fluid collection, 24 mm × 16 mm ×

27 mm in the right sub-cortical matter below the superior frontal gyrus with extensive white matter increased signal change with sulcal and ventricular

effacement. No enhancing focus. (C) Haematoxylin and eosin stain of subcortical white matter demonstrating mild perivascular chronic inflammation and

haemorrhage (×2000 magnification). (D) Glial fibrillary acidic protein immunohistostaining of astrocytes, showing reactive gliosis (×1000 magnification).



225

malfunction in patients who receive intrathecal treat-ment. Awareness to the possibility of device malfunctionincluding early clinical review and imaging post instilla-tion of chemotherapy through the OR may be beneficialto detect early complications.

Received 6 May 2017; accepted 11 May 2017.

Michal Lubomski ,1,2 Malcolm Pell,3 Alistair Lochhead4

and Martin Jude5

Departments of 1Neurology, 3Neurosurgery, and 4Anatomical

Pathology, St Vincent’s Hospital, and 2School of Medicine,

The University of Notre Dame Australia, Sydney, and 5University of

New South Wales, Rural Clinical School, Wagga Wagga,

New South Wales, Australia

References

1 Chowdhary S, Chalmers LM,

Chamberlain PAM. Methotrexate-

induced encephaloclastic cyst: a

complication of intraventricular

chemotherapy. Neurology 2006; 67: 319.

2 Lemann W, Wiley RG, Posner JB.

Leukoencephalopathy complicating

intraventricular catheters: clinical,

radiographic and pathologic study of

10 cases. J Neurooncol 1988; 6:

67–74.

3 Geiser CF, Bishop Y, Jaffe N, Furman L,

Traggis D, Frei E 3rd. Adverse effects of

intrathecal methotrexate in children with

acute leukemia in remission. Blood 1975;

45: 189–95.

4 Mella DB, Kamiya-Matsuoka C, Liao B,

Tummala S, de Groot J. Recurrent

encephaloclastic cyst induced by

intraventricular topotecan. J Neurol Sci

2015; 349: 52–3.

5 Packer RJ, Zimmerman RA,

Rosenstock J, Rorke LB, Norris DG,

Berman PH. Focal encephalopathy

following methotrexate therapy.

Administration via a misplaced

intraventricular catheter. Arch Neurol

1981; 38: 450–2.

doi:10.1111/imj.13689


Time of administration of treatmentfor hypertension in renal patients

Wang et al.1 make a valuable contribution to an aspectof blood pressure management often ignored. In 1961,in blood pressure studies over 24 h, we observed bloodpressure did not fall at night in patients with malignant(accelerated) hypertension who had renal failure orwere at risk of this.2 Wang et al.1 now provide a meta-analysis of subsequent studies and support our sugges-tion that ‘non-dipping’ hypertension is important in thegeneration of morbidity, and that evening administra-tion of medication may be optimal. Their conclusionfrom the analysis is that evening dosing, when con-trasted with morning administration, improves two sur-rogate markers of long-term benefit: restoring ‘dipping’in ‘non-dipping’ hypertensives and decreasing urinaryalbumin. Information on renal function was providedin only two of six studies, totalling 138 patients. Therewas no significant change in kidney function whenmeasured, but there was no information on the rate ofprogression of renal dysfunction. The studies did notfollow patients for long enough to evaluate ‘hard’ end-points, death and dialysis, and longer follow-up mighthave demonstrated renal function benefit from eveningdosing. Another surrogate marker of prognosis inchronic kidney disease, urinary albumin excretion(UAE), was reported in only one study included in this

meta-analysis. A reduction in UAE with evening dosingwas highly correlated with the decrease in nocturnalblood pressure and with the increase in diurnal/noctur-nal ratio of blood pressure but no rate of change in UAEwas calculated.

Renal function is usefully considered as the rate of

deterioration. Plasma creatinine, expressed as the recipro-

cal of plasma creatinine, or the estimated glomerular fil-

tration rate, often demonstrates linear progression,

allowing calculation of rates of deterioration in individual

patients. Future studies could usefully record more fre-

quent measurements of plasma creatinine and UAE in

each individual and identify those responding to changes

in time of administration. Time series analysis, utilising

Bayesian statistics with Kalman Filtering, were developed

for my group to monitor progression of renal function

and to detect and evaluate the probability of change.3 If

this method is utilised in any prospective trial of change

in administration time (or retrospectively if enough data

points are available) many fewer subjects will be needed

to confirm or refute the conclusions of this meta-analysis.My personal clinical experience provides examples of

a change in the rate of progression of kidney failure in

some patients when there were therapeutic changes to

modify nocturnal blood pressure, I have proposed3,4 that

in the clinic records of renal function and of urinary

albumin/creatinine ratio should be considered



226

graphically and also analysed for changes in rate of dete-rioration when changes, including change in time ofadministration of medication, have been introduced.

Received 23 October 2017; accepted 20 November 2017.

Martin S. KnappRetired Physician and Nephrologist

Melbourne, Victoria, Australia

References

1 Wang C, Ye Y, Liu C, Zhou Y, Lv L,

Cheng C et al. Evening versus morning

dosing regimen drug therapy for chronic

kidney disease patients with

hypertension in blood pressure patterns:


Intern Med J 2017; 47: 900–6.

2 Shaw DB, Knapp MS, Davies DH. Changes

in the blood pressure of hypertensives

during sleep. Lancet 1963; 281: 797–9.

3 Knapp MS, Smith AFM, Trimble IM,

Pownall R, Gordon K. Mathematical

and statistical aids to evaluate data from

renal patients. Kidney Int 1983; 24:

474–86.

4 Knapp MS. A transverse and longitudinal

audit of pre-dialysis renal failure in a

non-metropolitan clinic. Nephrology 2003;

8(Suppl): A90.

doi:10.1111/imj.13695

Starting beta-blockers duringexacerbations of chronic obstructivepulmonary disease

We read the report of Neef et al. on the safety of startingbeta-blockers during exacerbations of chronic obstructivepulmonary disease (COPD) with interest.1 We have alsobeen concerned about the high risk of adverse cardiacevents during COPD exacerbations and whether beta-blockers can offer cardioprotection for such patients. Werecently undertook a prospective study to test the feasi-bility of starting a cardioselective beta-blocker in patientsadmitted with exacerbations.2 We stopped recruitmentafter screening more than 570 patients but only manag-ing to enrol 23 (the reasons for exclusion are listed else-where2). Of these 23, only 16 completed 3 monthstreatment and only 12 reached the target dose of 95 mgmetoprolol daily. Symptomatic hypotension was a fre-quent problem leading to two withdrawals and two fail-ures to attain the target dose. Half of those recruitedexperienced at least one serious adverse event, includingeight who were readmitted with further COPDexacerbations.The finding that beta-blockers were better tolerated in

Neef et al.’s retrospective study suggests that there wassome selection bias and their reassuring observationsmay not be generalisable to a wider COPD population. Itwould be interesting to know if their patients continuedto tolerate the beta-blocker after discharge. Patients withsevere COPD are a particularly vulnerable population.Hypotension (observed in both Neef et al.’s and our

study) and an increased risk of falls could have disastrousconsequences. However, we do agree with regard to therespiratory safety of cardioselective beta-blockers: noneof the patients in our study had a significant deteriora-tion in lung function after starting the beta-blocker. Car-dioselective beta-blockers do not appear to have acuteeffects on airway calibre in COPD.3

Although the retrospective evidence suggests thatcardioselective beta-blockers are probably safe, there isstill no prospective evidence on the balance of risk andbenefits in COPD patients. An American study is cur-rently recruiting patients to see whether beta-blockersreduce the risk of COPD exacerbations4 and we areplanning a study in Australia and New Zealand to assessthe cardioprotective effects of beta-blockers in COPDpatients. Until the evidence from these trials becomesavailable, we suggest that beta-blocker therapy isreserved for those with clear cardiovascular indicationsand that these patients are monitored carefully foradverse effects.

Received 9 October 2017; accepted 24 October 2017.

Robert J. Hancox ,1 Catherina L. Chang,2

Richard Beasley 3 and Christine Jenkins41Department of Preventive and Social Medicine, University of

Otago, Dunedin, 2Respiratory Research Unit, Department of

Respiratory Medicine, Waikato Hospital, Hamilton, and 3Medical

Research Institute of New Zealand, Wellington, New Zealand, and4The George Institute for Global Health, Sydney,

New South Wales, Australia

References

1 Neef PA, Burrell LM, McDonald CF,

Irving LB, Johnson DF, Steinfort DP.

Commencement of cardioselective

beta-blockers during hospitalisation

for acute exacerbations of

chronic obstructive pulmonary

disease. Intern Med J 2017; 47:

1043–50.

2 Chang CL, Wong C, Beckert L,

Shafuddin E, Beasley R, Young R et al.



227

β-blockers in exacerbations of COPD:

feasibility of a randomised controlled

trial. ERJ Open Res 2017; 3: 00090–2016.

3 Salpeter S, Ormiston T, Salpeter E.

Cardioselective beta-blockers for chronic

obstructive pulmonary disease. Cochrane

Database Syst Rev 2005; 4: CD003566.

4 Bhatt SP, Connett JE, Voelker H,

Lindberg SM, Westfall E, Wells JM

et al. β-blockers for the prevention of

acute exacerbations of chronic

obstructive pulmonary disease (βLOCK

COPD): a randomised controlled

study protocol. BMJ Open 2016; 6:

e012292.

doi:10.1111/imj.13700

Author reply

We thank Hancox and colleagues1 for their interest inour recently published paper.2 Our study of beta-blockersstarted during admission with acute exacerbation ofchronic obstructive pulmonary disease (AECOPD) dem-onstrated that these drugs appeared to be well-tolerated.2

In contrast, the feasibility study reported by Chang, Han-cox and colleagues highlighted these authors’ difficultiesin recruiting for a prospective study of beta-blockerscommenced in a similar context.3

Althoughwe acknowledge the limitations of our retro-spective study, our detailed reviewof 36patients inwhombeta-blockers were initiated during AECOPD (see table1, online supplement2) found they were generally safe.Our study included patients with severe airflow obstruc-tion, as well as those with significant bronchodilatorresponse. It is reassuring that respiratory adverse effectswere not prevalent in either our own study or in the pro-spective study of Chang et al.3 Unfortunately, given ourstudy was a retrospective audit, we were unable to assessthe long-term tolerability of beta-blockers in this cohort.Collection of such information in prospective trials will becritical to determining the utility of commencing beta-blockers in patients with chronic obstructive pulmonarydisease (COPD).

We concur with the opinion of Hancox and colleaguesthat cardioselective beta-blockers should not be withheldin patients with COPD and clear cardiac indications fortheir use, and our data suggest continuation of beta-blockers in patients admitted with respiratory illness isappropriate. We recently highlighted the prevailing per-ception among clinicians that beta-blockers are contrain-dicated in COPD through our finding that only 44.8% ofpatients with AECOPD and clear cardiac indications forbeta-blocker therapy were found to be receiving beta-

blockers.4 Unpublished data from this same cohort of1071 patients with AECOPD demonstrated significant

rates of in-hospital cardiovascular events, including atrial

fibrillation with rapid ventricular response (100 patients,

9.3%) and acute coronary syndrome (43 patients, 4.0%)

occurring in the total cohort. Furthermore, our data sug-

gest a statistically significant increase in rates of atrial

fibrillation with rapid ventricular response following the

cessation of beta-blockers during admission compared to

those in whom they were continued throughout the

hospital admission (Table 1).Toaddress these issues further,weareconductingapro-

spective study of the safety and tolerability of initiating

beta-blockers during AECOPD in patients with clear car-

diac indications.5 While beta-blocker use is associated

with reducedratesofAECOPDinnumerousobservational

studies, it is unclear if this is due to a direct effect onCOPD

or due to improved medical management of concomitant

cardiac disease.6,7 Our study will include only patients

with indications for beta-blocker use for which mortality

benefit isunequivocal. It remains tobe seenwhether feasi-

bility and tolerability mirror or differ from those reported

by Chang et al., and we look forward to presenting these

results.


Pieter A. Neef ,1 Louise M. Burrell,1,2

Christine F. McDonald,3,4 Louis B. Irving,4,5

Douglas F. Johnson1,4 and Daniel P. Steinfort4,5

Departments of 1General Medicine, and 3Respiratory and Sleep

Medicine, Austin Health, Departments of 2Medicine and Cardiology,

and 4Medicine, The University of Melbourne, and 5Department of

Respiratory and Sleep Medicine, Melbourne Health,

Melbourne, Victoria, Australia

Table 1 Univariate analysis of risk of cardiovascular events following cessation of beta-blocker therapy compared to those in whom it was continuedthroughout admission

Beta-blockersceased

Beta-blockerscontinued

Odds ratio (95% confidenceinterval)

P-value

Atrial fibrillation with rapid ventricularresponse

10/23 13/253 14.2 (5.24–38.4) <0.001

Acute coronary syndrome 0/23 9/253 0 0.62



228

References

1 Hancox RJ, Chang CL, Beasley R,

Jenkins C. Starting beta-blockers during

exacerbations of chronic obstructive

pulmonary disease. Intern Med J 2018;

48: 227–8.

2 Neef P, Burrell L, McDonald C, Irving L,

Johnson D, Steinfort D. Commencement

of cardioselective beta-blockers during

hospitalisation for acute exacerbations of

chronic obstructive pulmonary disease.

Intern Med J 2017; 47: 1043–50.

3 Chang C, Wong C, Beckert L,

Shafuddin E, Beasley R, Young R et al.

β- blockers in exacerbations of COPD:

feasibility of a randomised controlled

trial. ERJ Open Res 2017; 3:

00090–2016.

4 Neef P, McDonald C, Burrell L, Irving L,

Johnson D, Steinfort D. Beta-blockers are

under-prescribed in patients with chronic

obstructive pulmonary disease and co-

morbid cardiac disease. Intern Med J

2016; 48: 1336–40.

5 Australian New Zealand Clinical Trials

Registry. Safety of beta blocker

commencement during acute

exacerbations of chronic obstructive

pulmonary disease (COPD) in patients

with co-existing cardiovascular disease.

Identifier ACTRN12617001635358.

Sydney: NHMRC Clinical Trials Centre,

University of Sydney; 2017 [cited 2017

Dec 19]. Available from URL: https://

www.anzctr.org.au/Trial/Registration/

TrialReview.aspx?id=374045

6 Bhatt S, Wells J, Kinney G, Washko

Jr G, Budoff M, Kim Y et al. β-Blockersare associated with a reduction in

COPD exacerbations. Thorax 2016;

71: 8–14.

7 Du Q, Sun Y, Ding N, Lu L, Chen Y.

Beta-blockers reduced the risk of

mortality and exacerbation in patients

with COPD: a meta-analysis of

observational studies. PLoS One 2014; 9:

e113048.

doi:10.1111/imj.13690

Iron polymaltose infusion therapyduring pregnancy

We read, with interest, the article by Grzeskowiak et al. onoptimising intravenous (i.v.) iron dosing in pregnancy.1

The authors recommend dosing according to adjusted pre-pregnancy body weight and caution against the unknownharms of excessive i.v. iron resulting in adverse pregnancyoutcome. The practical problem in implementing therecommended dosage is the non-availability of the pre-pregnancy bodyweight to clinicians at treatment centresadministering iron infusion. We describe an alternativeapproach using a simpler dosing model based on thepatient’s haemoglobin level, but capped at 1000 mg.In our outpatient clinic, 105 pregnant women received

iron infusions between 2014 and 2017; 33 received theinfusion in the second trimester and 72 in the third.Their haemoglobin (Hb) level ranged from 84 to 130 g/Land the ferritin level ranged <8–30 μg/L.

Dosage schedule: 150 mg of iron polymaltose to raiseHb by 10 g/L, aiming to bring the patient’s Hb level to130 g/L, plus 200–500 mg to replenish the iron stores.The minimal dose was 500 mg and the maximum 1000mg. The infusion was given over a 2-h period.2

The treatment was very well tolerated with minimalside-effects in a small number of patients: (i) brief vaso-vagal episode (4), (ii) minor gastrointestinal symptoms(10) and (iii) mild arthralgia/myalgia (4).Our experience suggests that capping the dose at

1000 mg can achieve an effective and adequate responsein most of the patients and precludes overdosing and thepossible ‘unknown harms’ discussed by Grzeskowiaket al.1

Received 8 September 2017; accepted 5 November 2017.

Raj Ramakrishna 1,2 and Arumugam Manoharan1,2

1Southern Sydney Haematology, and 2University of Wollongong,

Wollongong, New South Wales, Australia

References

1 Grzeskowiak L, Qassima A, Jeffries B,

Grivell R. Approach for optimising

intravenous iron dosing in pregnancy:

a retrospective cohort study. Intern Med J

2017; 47: 747–53.

2 Manoharan A, Alexander W,

Ramakrishna R, Legge J, Uebel J.

Comparative rates of adverse events

with 2-hour versus 4-hour infusion

of total dose intravenous iron

polymaltose. Int J Clin Med 2014; 5:

145–8.



229

https://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=374045



doi:10.1111/imj.13688

Clinical need for standardisedmultidisciplinary meeting assessmentprocesses

We read with interest the article by Johnson et al. regard-ing the development of a peer-review framework for can-cer multidisciplinary meetings (MDM).1 MDM are a criticalcomponent of multidisciplinary oncological care, associatedwith improved patient outcomes and satisfaction, trialsrecruitment and interdisciplinary communication.2–5 Whilenational and international guidelines exist relating to theconduct of MDM,3,6,7 processes still vary within and acrossinstitutions, differing in MDM frequency, the patients pre-sented, team membership, leadership and communica-tion.2,4 There is currently no standardised process withinAustralia to monitor the quality of MDM across differentinstitutions and regional settings. However, variationsbetween MDMmay affect quality of care.

We recently reviewed neoadjuvant chemotherapy

delivery for muscle-invasive bladder carcinoma (MIBC)

in a metropolitan and regional setting and found MDM

discussion to be a determinant of patients receiving

neoadjuvant chemotherapy. Neoadjuvant chemotherapy

is a standard of care for patients with MIBC.8 Our retro-

spective audit analysed 19 metropolitan patients and

23 regional patients who underwent radical cystectomy

for MIBC. Regional patients had significantly lower rates

of both referral to a preoperative MDM (84 vs 42%, P =

0.023) and neoadjuvant chemotherapy (42 vs 13%, P =

0.043). Nine of the 11 patients (82%) who received

neoadjuvant chemotherapy had been discussed at a

MDM, demonstrating the association between MDM

referral and neoadjuvant treatment. There were no sig-

nificant differences identified in the patient populations

or type of chemotherapy. Optimising the MDM processes

could therefore improve delivery of standard of care

treatment for these patients.While it is intuitive to believe that an MDM that meets

regularly, with well-considered terms of reference,

strong leadership and good team communication will

make better decisions, there is minimal literature com-

paring variations or interventions in MDM processes and

their effect on patient outcomes. Lamb et al. demon-

strated that use of a checklist during MDM discussion to

ensure adequate information was presented and all team

members’ views discussed, improved efficiency and

treatment decisions. There was no measure as to

whether this translated into improved patient out-

comes.9 The MDM peer-review framework described by

Johnson et al. is another tool, designed to reduce varia-

tion in practice and provide opportunity for quality

improvement. It may be limited by the labour-intensive

process and the perceived relevance of its

recommendations.10

We strongly support ongoing research into improving

the quality and efficiency of the MDM process. We argue

that for any MDM intervention, improved patient out-

comes should be the primary measure of success. While

a peer-review process enables assessment of communi-

cation and leadership within an MDM team, we wonder

whether less costly strategies should be explored first.

Our research would suggest that process to ensure that

all patients are discussed at an MDM soon after their

diagnosis, which are attended (either in person or via

teleconference) by all appropriate specialists, would be a

welcome first step. It would be interesting to investigate

whether a simple checklist used during the MDM to

prompt discussion could further improve patient

outcomes.


Wing H. Yau ,1,2 Bianca Devitt,1,3 Soe Y. Aung,1

Phillip Parente,1,3 Sharad Sharma2 and Carmel Pezaro1,3

1Department of Medical Oncology, Eastern Health, and 3Eastern

Health Clinical School, Monash University, Melbourne, Victoria,

and 2Department of Oncology, Launceston General Hospital,

Launceston, Tasmania, Australia

References

1 Johnson CE, Slavova-Azmanova N,

Saunders C. Development of a peer-

review framework for cancer

multidisciplinary meetings. Intern Med J

2017; 47: 529–35.

2 Fleissig A, Jenkins V, Catt S,

Fallowfield L. Multidisciplinary teams in

cancer care: are they effective in the

UK? Lancet Oncol 2006; 7: 935–43.

3 Wright FC, de Vito C, Langer B,

Hunter A, Expert Panel on

Multidisciplinary Cancer Conference

Standards. Multidisciplinary cancer

conferences: a systematic review

and development of practice

standards. Eur J Cancer 2007; 43:

1002–10.

4 Taylor C, Munro AJ, Glynne-Jones R,

Griffith C, Trevatt P, Richards M et al.

Multidisciplinary team working in

cancer: what is the evidence? BMJ

2010; 340: c951.

5 Kesson EM, Allardice GM, George WD,

Burns HJG, Morrison DS. Effects of

multidisciplinary team working on

breast cancer survival: retrospective,

comparative, interventional cohort

study of 13 722 women. BMJ 2012; 344:

e2718.

6 National Cancer Action Team. The

Characteristics of an Effective



230

Multidisciplinary Team. London: UK

National Cancer Intelligence Network;

2010.

7 National Breast Cancer Centre.

Multidisciplinary Meetings for Cancer Care:

A Guide for Health Service Providers.

Sydney: National Breast Cancer Centre;

2005.

8 Liew MS, Azad A, Tafreshi A,

Eapen R, Bolton D, Davis ID et al.

USANZ: time-trends in use and impact

on outcomes of perioperative

chemotherapy in patients treated with

radical cystectomy for urothelial

bladder cancer. BJU Int 2013; 112

(Suppl 2): 74–82.

9 Lamb BW, Green JSA, Benn J,

Brown KF, Vincent CA, Sevdalis N.

Improving decision making in

multidisciplinary tumor boards:

prospective longitudinal evaluation of a

multicomponent intervention for 1,421

patients. J Am Coll Surg 2013; 217:

412–20.

10 Slavova-Azmanova NS, Johnson CE,

Platell C, Bydder S, Saunders CM. Peer

review of cancer multidisciplinary

teams: is it acceptable in Australia? Med

J Aust 2015; 202: 144–7.

doi:10.1111/imj.13705

Author reply

The response from Yau et al.1 to our article describing a peerreview framework for multidisciplinary team meetings(MDM)2 is consistent with that provided during our study.3

In interviews with five peer reviewers and 17 multidisci-plinary team (MDT) members in our study, it wasacknowledged that peer review may be an effective vehiclefor ensuring that MDM are conducted effectively but wasresource intensive. Participants concluded that it might bemore beneficial to identify how consistently MDM deci-sions reflected the clinical practice guidelines (CPG) andthat patient outcomes may be a better measure of the effec-tiveness of MDM. Given the high cost of running MDM,4 itis imperative that they are efficient and effective.In view of this feedback, we examined the adherence of

breast MDM recommendations to CPG and of treatment toguidelines and MDM recommendations. In 1028 womenwith breast cancer (carcinoma in situ and invasive breastcancer), MDM recommended guideline-adherent treat-ment in 98.6% of cases. MDM recommendations wereimplemented for 89.7% of patients and 90.4% receivedguideline-adherent treatment. Patient preferences (n = 81),physician decision outside MDM (n = 8) and comorbidities(n = 7) were the main reasons for deviation from CPG.5

Yau et al. observed that variations in membership andleadership of MDM can affect the quality of their work.Indeed, variability in MDM dynamics resulting from var-iation in team composition was noticed by the peerreviewers observing the video recordings of MDM in ourstudy. One MDT requested that an additional videorecording be reviewed because several MDT memberswere absent in the initial video.3

We agree with Yau et al. that improved patient out-comes should be the primary measure of success of anyhealthcare. While improved outcomes for people dis-cussed at MDM have been demonstrated,6–8 this researchis subject to multiple biases.9,10 Given the difficultiesassociated with demonstrating a causal relationship ofMDM on patient outcomes, trying to establish the effectof particular aspects of the MDM on outcomes is likely tobe problematic.Research suggests that one of the most effective

methods of improving the quality of care is to monitorprocesses and patient outcomes (including patientreported outcomes) and to provide regular feedbackagainst standardised benchmarks.11,12 The process ofmonitoring and providing feedback is growing in favourwith health administrators and policymakers as clinicaldata collection and methods of manipulating databecome more efficient. A movement towards ‘value-based’ healthcare13 is the impetus for many healthservices to monitor outcomes to promote care that isconsistent with best practice guidelines and to reducevariability. Such an approach may be cost-effective andprovide the requisite motivation for MDT to monitortheir own performance and to address concerns aboutthe quality of care.

Received 26 November 2017; accepted 1 December 2017.

Claire E. Johnson ,1,2,3 Neli Slavova-Azmanova 1 andChristobel Saunders1

1Department of Surgery, Medical School, The University of Western

Australia, Perth, Western Australia, and 2Monash Nursing and

Midwifery, Monash University, and 3Supportive and Palliative Care,

Eastern Health, Melbourne, Victoria, Australia

References

1 Yau WH, Devitt B, Aung SY, Parente P,

Sharma S, Pezaro C. Clinical need forstandardised multidisciplinary meeting

assessment processes. Intern Med J 2018;48: 230–1.

2 Johnson CE, Slavova-Azmanova N,

Saunders C. Development of a peer-

review framework for cancer

multidisciplinary meetings. Intern Med J

2017; 47: 529–35.3 Slavova-Azmanova N, Johnson CE,

Platell C, Bydder S, Saunders CM. Peer

review of cancer multidisciplinary



231

teams: is it acceptable in Australia? Med

J Aust 2015; 202: 144–7.4 Fosker C, Dodwell D. The cost of

the MDT [response]. BMJ 2010; 340:

c951.

5 Slavova-Azmanova N, Oo K, Lim R,

Nguyen K, Sodhi-Berry N, Lopez D et al.

Guideline-adherent treatment for

women with breast cancer: do they

receive what the multidisciplinary team

recommend and does this affect survival?

Asia Pac J Clin Oncol 2016; 12: 145.

6 Gillis CR, Hole DJ. Survival outcome of

care by specialist surgeons in breast

cancer: a study of 3786 patients in the

west of Scotland. BMJ 1996; 312: 145–8.

7 Stephens MR, Lewis WG, Brewster AE,

Lord I, Blackshaw GR, Hodzovic I et al.

Multidisciplinary team management is

associated with improved outcomes

after surgery for esophageal cancer. Dis

Esophagus 2006; 19: 164–71.

8 Ruhstaller T, Roe H, Thurlimann B,

Nicoll JJ. The multidisciplinary meeting:

an indispensable aid to communication

between different specialities. Eur J

Cancer 2006; 42: 2459–62.

9 Look Hong NJ, Wright FC,

Gagliardi AR, Paszat LF. Examining the

potential relationship between

multidisciplinary cancer care and

patient survival: an international

literature review. J Surg Oncol 2010;

102: 125–34.

10 Coory M, Gkolia P, Yang IA,

Bowman RV, Fong KM. Systematic

review of multidisciplinary teams in the

management of lung cancer. Lung

Cancer 2008; 60: 14–21.

11 Lilford R, Mohammed MA,

Spiegelhalter D, Thomson R. Use and

misuse of process and outcome data in

managing performance of acute medical

care: avoiding institutional stigma.

Lancet 2004; 363: 1147–54.

12 Eagar K, Sansoni J, Loggie C,

Elsworthy A, McNamee J. A Literature

Review on Integrating Quality and Safety

into Hospital Pricing Systems. Wollongong:

Centre for Health Service Development,

University of Wollongong; 2013.

13 Porter ME. A strategy for health care

reform – toward a value-based system.

N Engl J Med 2009; 361: 109–12.



232

doi:10.1111/imj.13732

Corrigendum

The authors would like to draw the readers’ attention to an error in the following article:

Penny Allen, Lucy Gately, Patricia Banks, Adrian A. Y. S. Lee, Garry Hamilton, Lavinia Tan and Sheryl Sim. Directaccess colonoscopy: impact of intervention on time to colorectal cancer diagnosis and treatment in North West Tasma-nia. Intern Med J 2017; 47: 1129–1135. https://doi.org/10.1111/imj.13514

The name of the author Adrian A. Y. S. Lee should be Adrian Y. S. Lee.

The authors of the article should thus be:

Penny Allen, Lucy Gately, Patricia Banks, Adrian Y. S. Lee, Garry Hamilton, Lavinia Tan and Sheryl Sim.

The authors apologise for the error.


233

https://doi.org/10.1111/imj.13514

40 SALEM STREET, LYNNFIELD MA 01940 • 781-295-0072

crossrefCROSSREF.ORG | THE CITATION LINKING BACKBONE

reliable reference linking,powered by CrossRef

Linking to and from this journal is powered by CrossRef.CrossRef links millions of articles and chapters, from

thousands of publications, from hundreds of publishers.So your reference linking works. Permanently. Take a

tour at www.crossref.org. And make sure all yourpublished work is CrossRef-enabled.

hell is the place where nothing connects — T.S. Eliot

P720.indd 1P720.indd 1 20/06/17 10:20 AM20/06/17 10:20 AM

Volu

me 48, Issu

e 2, Febru

ary 2018, Pages 107–234IN

TE

RN

AL

ME

DIC

INE

JOU

RN

AL

Climate change allergens and allergic diseases

Documenting patient body weight

EVOLVE: low value practices in rheumatology

Non-alcoholic fatty liver disease

Hospital-associated venous thromboembolism

Adrenal incidentalomas

Management of neurosyphilis

INTERNALMEDICINEJOURNAL

Volume 48

Issue 2

February 2018

ISSN 1444-0903

Total-pages: Spine-width: Paper-type:previous Total-pages:128 Spine-width: 4.2mm Paper-type: UPM Classic 64Current

Editorial 113 What we know and do not know about women and kidney

diseases: questions unanswered and answers unquestioned: reflection on World Kidney Day and International Women’s Day


Review



Clinical Perspectives 129 Climate change: allergens and allergic diseases


Original Articles 135 EVOLVE: The Australian Rheumatology Association’s ‘top five’

list of investigations and interventions doctors and patients should question
















Brief Communications 194 Metformin-induced encephalopathy: the role of thiamine










Position Paper 210 Considerations for pre-transfusion immunohaematology

testing in patients receiving the anti-CD38 monoclonal antibody daratumumab for the treatment of multiple myeloma














Martin S. Knapp



228 Author reply






231 Author reply


233 Corrigendum

VOLUME 48 issue 2 February 2018

IMJ_48_2Cover.indd 1IMJ_48_2Cover.indd 1 02/02/18 1:21 PM02/02/18 1:21 PM