UvA-DARE (Digital Academic Repository) Development of ...€¦ · inappropriate antibiotic use is imperative.6 Antibiotic stewardship programs are developed to optimize antibiotic

UvA-DARE is a service provided by the library of the University of Amsterdam (http://dare.uva.nl)

UvA-DARE (Digital Academic Repository)

Development of quality indicators for appropriate antibiotic use in daily hospital practice

van den Bosch, C.M.A.

Link to publication

Citation for published version (APA):van den Bosch, C. M. A. (2016). Development of quality indicators for appropriate antibiotic use in daily hospitalpractice.

General rightsIt is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s),other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulationsIf you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, statingyour reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Askthe Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam,The Netherlands. You will be contacted as soon as possible.

Download date: 02 Dec 2020

https://dare.uva.nl/personal/pure/en/publications/development-of-quality-indicators-for-appropriate-antibiotic-use-in-daily-hospital-practice(9a54e316-5fc2-47ff-bf8c-2b84bfaf0f5b).html

Development of quality indicatorsfor appropriate antibiotic use

in daily hospital practice

Caroline M.A. van den Bosch

Caroline van d

en Bosch Q

uality indicators for appropriate antibiotic use

Development of Quality Indicators for

Appropriate Antibiotic Use in Daily Hospital Practice

Caroline Martine Alina van den Bosch

Development of Quality Indicators for


ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad van doctoraan de Universiteit van Amsterdamop gezag van de Rector Magnificus

prof. dr. D.C. van den Boomten overstaan van een door het College voor Promoties ingestelde commissie,

in het openbaar te verdedigen in de Agnietenkapelop vrijdag 22 januari 2016, te 12:00 uur

door

Caroline Martine Alina van den Bosch

geboren te HilversumPrinted by Mostert & Van Onderen, LeidenCover designed by Dimitri Valentijn - www.dimitrivalentijn.nl© 2015, Caroline van den BoschAll rights reserved. No parts of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, mechanically, by photocopy, by recording, or otherwise, without the prior permission from the author.

Contents

Chapter 1. General Introduction 7Chapter 2. An online national antimicrobial guide with locally

customized versions can promote appropriate antibiotic use in hospitals

17

Chapter 3. Development of quality indicators for antimicrobial treatment in adults with sepsis

35

Chapter 4. Quality indicators to measure appropriate antibiotic use in hospitalized adults

61

Chapter 5. Applicability of generic quality indicators for appropriate antibiotic use in daily hospital practice

89

Chapter 6. Appropriate antibiotic use reduces length of hospital stay 113

Chapter 7. General Discussion 145

Chapter 8. Appendices: 161

Nederlandse samenvatting 162

Dankwoord 174

Lijst van deelnemende centra 178

PhD portfolio 179

Curriculum Vitae 183

Promotiecommissie

Promotores: Prof. dr. S.E. Geerlings Universiteit van Amsterdam Prof. dr. J.M. Prins Universiteit van AmsterdamCopromotor: Prof. dr. M.E.J.L. Hulscher Radboud Universiteit Nijmegen

Overige leden: Prof. dr. L.P.H.J. Aarts Universiteit Leiden Prof. dr. M.A. Boermeester Universiteit van Amsterdam Dr. J.C.C. Braspenning Radboud Universiteit Nijmegen Prof. dr. M.D. de Jong Universiteit van Amsterdam Prof. dr. B.J. Kullberg Radboud Universiteit Nijmegen Prof. dr. R.A.A. Mathôt Universiteit van Amsterdam

Faculteit der Geneeskunde

Chapter

General introduction

1

Chapter 1

8


9

1Emergence of antibiotic resistanceThe breakthrough of antibiotics in the 1930s led to a dramatic decline in illness and death from infectious diseases. Antibiotics, often described as ‘wonder drugs’, fundamentally altered health care in the twentieth century by creating curable bacterial infectious diseases. Today antibiotics are indispensable in practically all health care systems.1 Achievements in modern medicine, such as major surgery, organ transplantation, treatment of preterm babies, and cancer chemotherapy would be impossible without an effective treatment for bacterial infections.2

However, the extensive use of antibiotics is also the most important driving force in the emergence of resistant microorganisms, threatening the effectiveness of these ‘wonder drugs’.3 Antibiotic consumption is high in human medicine, but antibiotic use for growth promotion and prevention of diseases in veterinary medicine, agriculture, aquaculture, and horticulture are also significant contributing factors.2 Worldwide, antibiotic consumption and antibiotic resistance are still on the rise, which, together with the steady decline in the discovery of new antibiotics, creates one of the greatest current threats to human health.3-6 With the arrival of untreatable strains of carbapenem-resistant Enterobacteriaceae, we are at the break of a post-antibiotic era.2, 7 Antibiotic stewardship programsThe reasons for resistance to antibiotics are complex, and also comprises human behavior at many levels of society.2 To help curbing antibiotic resistance in hospitals, better use of current agents is warranted and a decrease of inappropriate antibiotic use is imperative.6 Antibiotic stewardship programs are developed to optimize antibiotic selection, dosing, route, and duration of therapy, in order to maximize the chance of clinical cure or prevention of infection.8 At the same time, they aim to limit the unintended consequences of antibiotic use, such as the emergence of resistance, adverse drug events, and costs.8 Antibiotic stewardship programs have been shown to be effective and financially self-supporting.9-11 Multidisciplinary local stewardship teams are now established across the world, with the task to design programs in their own hospitals. In the Netherlands, the development of an Antibiotic-team (A-team) is mandatory in every hospital since 2014. This so-called A-team (antibiotic team) comprises infectious diseases specialists, microbiologists, hospital

pharmacists and quality assurance experts and focuses on authorization of antibiotics for special indications, measuring and monitoring antibiotic use, the organization of special improvement projects, and the provision of education, training, and consultations.12, 13 In order to set priorities and focus improvement, an essential tool for an effective stewardship program is the ability to measure the appropriateness of antibiotic use in individual patients.

Antibiotic guidelines Appropriate antibiotic use is described in the recommendations of guidelines. International, national and local guidelines on the management of various infections have been developed and are updated regularly. The Infectious Diseases Society of America (IDSA) and the European Society for Microbiology and Infectious Diseases (ESCMID) are important collaborations that develop guidelines for infectious diseases.14, 15 In the Netherlands, the SWAB, the Dutch Working Party on Antibiotic Policy, develops national antibiotic guidelines. Previous studies have shown that adherence to guidelines improves clinical outcome, is correlated with less development of resistance and lowers costs.16-19 The aim for guidelines is to support clinical decision making, and to increase effectiveness of antibiotic care by preventing under-, over- and misuse of antibiotic therapy. Moreover, adherence to these guidelines will result into a decrease of unwanted variability in practice.20 In summary, guidelines aim to improve the appropriateness of antibiotic therapy by bridging the gap between what is already known from scientific studies and what is performed in daily clinical practice.21

Nonetheless, not always are patients receiving care as recommended in guidelines: the publication and dissemination of guidelines do not directly ensure appropriate antibiotic use in daily clinical practice.21, 22 For inpatients receiving antibiotics, studies have shown that guideline adherence is often suboptimal, up to 50% of hospital antibiotic use is inappropriate, and adherence varies considerably between hospitals and between care providers.23-25

Measuring appropriate antibiotic useAfter developing guidelines, the next step to improve the appropriateness of antibiotic use is to gain insight into the quality of antibiotic use in daily clinical practice.26 Quality indicators (QIs) are measurable elements of practice

Chapter 1

10


11

1performance for which there is evidence or consensus that they can be used to assess and change the quality of antibiotic care provided.27 They consist of a numerator and denominator. The denominator describes the patients to whom the care should be applied to and the numerator the patients who have actually received the care as recommended.21 Table 1 shows an example of a QI; the quotient between the numerator and denominator is the indicator score. A well-known classification to categorize QIs is: structure -, process- and outcome indicators.28 Structure indicators measure requirements and attributes of settings in which antibiotic care occurs. Process indicators measure the process around antibiotic care at the patient level. Outcome indicators measure the effects of antibiotic care on the health status of patients and populations.

Table 1. Example of a quality indicator

Numerator Number of patients who started with empirical systemic antibi-otic therapy according to the national guideline

150

Denominator Total number of patients who started with empirical systemic antibiotic therapy

200

Indicator score 75%

Indicator developmentDuring the past decade quality indicators have been developed and implemented in many healthcare settings to measure, compare and improve appropriateness of care.29 Many governmental- and professional associations have developed QIs.30 To obtain the information needed to develop QIs, a systematic or non-systemic method can be used. Systematic methods rely on available scientific evidence complemented when necessary with expert opinion.30 The method used for the development of QIs in this thesis is the well-known RAND modified Delphi procedure. This consensus method is also a rigorous systematic process where QIs are developed based on scientific evidence combined with expert opinions.27, 31-33 There were no universally accepted requirements for using the Delphi technique, which led to considerable variability across studies in the characteristics and the reporting of the Delphi procedure.30 Boulkedid et al. developed a practical guidance on the use and reporting of the Delphi method in QI research.

When the study of this thesis started in 2011, a couple of indicator sets were already developed for antibiotic care in the hospital using a systematic method like the RAND modified Delphi procedure. More precisely, in previous studies we developed QIs for the treatment of the two most prevalent infections in the hospital: lower respiratory tract infections 33 and urinary tract infections.31 The Surviving Sepsis Campaign developed a care bundle for optimal start of treatment in the first 24 hours.34 The European Surveillance of Antimicrobial Consumption (ESAC) developed QIs to measure appropriate antibiotic use in outpatients in Europe.35 What was missing was a set of generic indicators covering all steps in the process of antibiotic use, applicable to all hospitalized adult patients with a suspected bacterial infection.

In 2011 we initiated the ‘Development of Reliable generic quality Indicators for the optimalisation of ANTibiotic use in the hospital (RIANT)’-study, with the aim to develop reliable generic quality indicators which define appropriate antibiotic use in the treatment of all bacterial infections in hospitalized adult patients. This set of QIs was intended to measure and monitor the various steps in the process of antibiotic use on the patient level, along the entire antibiotic pathway.

Indicator validationFor an optimal and reliable use of the developed QIs, their clinimetric properties have first to be tested in clinical practice.21 However, few study groups have tested their QIs in daily clinical practice before using them for example in improvement projects. Registration of data is different in every country and varies over time, which has an effect on applicability, validity and reliability of data collection. Therefore it is mandatory to locally test the clinimetric properties of the QIs, in order to discriminate between indicators that are feasible, valid and reliable in a specific setting and those that are not.21 There are several criteria to consider when assessing the QIs. These important clinimetric characteristics are described below. Measurability. Is defined as the availability of administrative data required to evaluate the indicator. Too much missing data may bias the results and therefore, the number of missing data should be reported.31, 33, 36 Applicability. An indicator should be applicable to a substantial proportion of the reviewed patient records.37

Chapter 1

12


13

1Inter-observer reliability. When two different registration employees are extracting data or data is collected by two different methods, this may affect the indicators scores by subjective interpretation of information in the medical records. Therefore, the level of agreement between employees or methods should be assessed after performing duplicate registration.38, 39 Room for improvement. QIs must be capable of detecting variability in quality of care between and within hospitals. High QI scores with little variation between hospitals make indicators less sensitive and therefore less useful in daily practice.27 Case mix stability. The QI scores may be influenced by characteristics of the underlying population (case mix).21 More specifically, patient factors such as co morbidity or severity of illness may influence the performance on process- and outcome indicators.39 This should be evaluated.

Outline of this thesisThis thesis describes the development and validation of QIs that can be used to assess and improve appropriateness of antibiotic use in the hospital. Secondly, it focuses on measuring antibiotic use in the Netherlands and its effect on clinical outcome. Each of the studies contributes to answering the question: What is appropriate antibiotic use and how can we measure it?

In 2006 the Dutch Working Party on Antibiotic Policy (SWAB) introduced an online national antimicrobial guide (SWAB-ID) for the treatment and prophylaxis of common infectious diseases in hospitals. Every hospital in the Netherlands was offered the opportunity to get a local copy of the national version, and to edit this version for local use. Customizing (national) guidelines for local use is necessary to meet local circumstances, resource constraints and barriers to implementation. By the end of 2012, approximately 40% of Dutch hospitals used a local, customized version of SWAB-ID. In Chapter 2 we compared two approaches, the local SWAB-ID based and local non-SWAB-ID based antimicrobial guides, and described the relative comprehensiveness and compliance with the national SWAB-ID antimicrobial guideline. To determine the impact of this online tool in providing comprehensive antibiotic treatment guidelines.The SWAB published an evidence-based guideline for antimicrobial use in hospitalized patients with sepsis in 2010. In Chapter 3 we used this guideline

to develop a set of guideline-based QIs to measure the appropriateness of antimicrobial use in hospitalized adult patients with sepsis. In Chapter 4 we answered the question ‘what is appropriate antibiotic use’ by performing a literature search and asking an international expert panel which potential QIs are the most important when considering good antibiotic therapy. We developed a set of generic evidence-based QIs to measure and monitor appropriate antibiotic use in hospitalized adult patients with a suspected bacterial infection, using a RAND modified Delphi method.

The RIANT-study was performed at 4 university and 18 non-university hospitals located throughout the Netherlands. PREZIES (prevention of nosocomial infections by surveillance) is a department of the RIVM (National Institute for Public Health and the Environment, the Netherlands) and organizes biannual point prevalence measurements primarily aimed at identifying nosocomial infections. This point prevalence measurement was used to include 1890 adult patients using antibiotics for a suspected bacterial infection. In Chapter 5 we validated the QIs we systematically developed in Chapter 3, by assessing their clinimetric characteristics in daily clinical practice in the 1890 patients treated with antibiotics for a suspected bacterial infection. Chapter 6 evaluates the association between appropriate antibiotic use for patients with a suspected bacterial infection and their length of hospital stay. Appropriate antibiotic use was defined by the previous developed and validated set of generic indicators. The study group was again the group of patients enrolled in the RIANT study.

In the general discussion, Chapter 7, the main results of this thesis are summarized and discussed, followed by final conclusions and implications for further research.

Chapter 1

14


15

1Reference List

1. Heddini A, ugent R. The path of least resistance. http://www.reactgroup.org/uploads/publications/react-publications/the-path-of-least-resistance.pdf.

2. Laxminarayan R, Duse A, Wattal C, et al. Antibiotic resistance-the need for global solutions. Lancet Infect Dis 2013; 13: 1057-98.

3. Centers for Disease Control and Prevention. Report 2011 (revision): A public health action plan to combat antimicrobial resistance. http://www.cdc.gov/drugresistance/pdf/public-health-action-plan-combat-antimicrobial-resistance.pdf.

4. NETHMAP 2013. Consumption of antimicrobial agents and antimicrobial resistance among medically important bacteria in the Netherlands. http://www.swab.nl/swab/cms3.nsf/uploads/ADFB2606CCFDF6E4C1257BDB0022F93F/$FILE/Nethmap_2013%20def_web.pdf.

5. Spellberg B, Blaser M, Guidos RJ, et al. Combating antimicrobial resistance: policy recommendations to save lives. Clin Infect Dis 2011; 52 Suppl 5: S397-S428.

6. World Health Organization. Report 2012: The envolving theat of antimicrobial resistance - Options for action. http://apps.who.int/iris/bitstream/10665/44812/1/9789241503181_eng.pdf

7. Centers for Disease Control and Prevention (CDC). Vital signs: carbapenem-resistant Enterobacteriaceae. MMWR Morb Mortal Wkly Rep 2013; 62: 165-70.

8. Dellit TH, Owens RC, McGowan JE, Jr., et al. Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America guidelines for developing an institutional program to enhance antimicrobial stewardship. Clin Infect Dis 2007; 44: 159-77.

9. Carling P, Fung T, Killion A, et al. Favorable impact of a multidisciplinary antibiotic management program conducted during 7 years. Infect Control Hosp Epidemiol 2003; 24: 699-706.

10. Lutters M, Harbarth S, Janssens JP, et al. Effect of a comprehensive, multidisciplinary, educational program on the use of antibiotics in a geriatric university hospital. J Am Geriatr Soc 2004; 52: 112-6.

11. Ruttimann S, Keck B, Hartmeier C, et al. Long-term antibiotic cost savings from a comprehensive intervention program in a medical department of a university-affiliated teaching hospital. Clin Infect Dis 2004; 38: 348-56.

12. LeBrun PPH. Antimicrobial stewardship. The Netherlands. http://www.hospitalpharmacyeurope.com/featured-articles/antimicrobial-stewardship.

13. SWAB Stichting Werkgroep Antibioticabeleid. De kwaliteit van het antibioticabeleid in Nederland. Advies aangaande het restrictief gebruik van antibiotica an het invoeren van Antibioticateams in de Nederlandse ziekenhuizen en in de Eerste lijn. http://www.swab.nl/swab/cms3.nsf/

uploads/5FD2BE2700E8B433C1257A680028D9F0/$FILE/visiedoc%20SWAB%20vs%2021%20junifinal.pdf.

14. IDSA Practice Guidelines. http://www.idsociety.org/IDSA_Practice_Guidelines/.

15. ESCMID guidelines. https://www.escmid.org/escmid_publications/medical_guidelines/escmid_guidelines/.

16. Arnold FW, LaJoie AS, Brock GN, et al. Improving outcomes in elderly patients with community-acquired pneumonia by adhering to national guidelines: Community-Acquired Pneumonia Organization International cohort study results. Arch Intern Med 2009; 169: 1515-24.

17. Butler CC, Dunstan F, Heginbothom M, et al. Containing antibiotic resistance: decreased antibiotic-resistant coliform urinary tract infections with reduction in antibiotic prescribing by general practices. Br J Gen Pract 2007; 57: 785-92.

18. Larson EL, Quiros D, Giblin T, et al. Relationship of antimicrobial control policies and hospital and infection control characteristics to antimicrobial resistance rates. Am J Crit Care 2007; 16: 110-20.

19. McCabe C, Kirchner C, Zhang H, et al. Guideline-concordant therapy and reduced mortality and length of stay in adults with community-acquired pneumonia: playing by the rules. Arch Intern Med 2009; 169: 1525-31.

20. Grol R. Successes and failures in the implementation of evidence-based guidelines for clinical practice. Med Care 2001; 39: II46-II54.

21. Wennekes L. Development and Validation of Quality Indicators for Cancer Care. Thesis. Radboud. 2013; 8-158.

22. Bero LA, Grilli R, Grimshaw JM, et al. Closing the gap between research and practice: an overview of systematic reviews of interventions to promote the implementation of research findings. The Cochrane Effective Practice and Organization of Care Review Group. BMJ 1998; 317: 465-8.

23. Davey P, Brown E, Fenelon L, et al. Interventions to improve antibiotic prescribing practices for hospital inpatients. Cochrane Database Syst Rev 2005: CD003543.

24. Hulscher ME, Grol RP, van der Meer JW. Antibiotic prescribing in hospitals: a social and behavioural scientific approach. Lancet Infect Dis 2010; 10: 167-75.

25. Zarb P, Amadeo B, Muller A, et al. Identification of targets for quality improvement in antimicrobial prescribing: the web-based ESAC Point Prevalence Survey 2009. J Antimicrob Chemother 2011; 66: 443-9.

26. Grol R, Grimshaw J. From best evidence to best practice: effective implementation of change in patients’ care. Lancet 2003; 362: 1225-30.

27. Campbell SM, Braspenning J, Hutchinson A, et al. Research methods used in developing and applying quality indicators in primary care. Qual Saf Health Care 2002; 11: 358-64.

28. Donabedian A. The quality of care. How can it be assessed? JAMA 1988; 260: 1743-8.

Chapter 1

16

29. Kotter T, Blozik E, Scherer M. Methods for the guideline-based development of quality indicators--a systematic review. Implement Sci 2012; 7: 21.

30. Boulkedid R, Abdoul H, Loustau M, et al. Using and reporting the Delphi method for selecting healthcare quality indicators: a systematic review. PLoS One 2011; 6: e20476.

31. Hermanides HS, Hulscher ME, Schouten JA, et al. Development of quality indicators for the antibiotic treatment of complicated urinary tract infections: a first step to measure and improve care. Clin Infect Dis 2008; 46: 703-11.

32. Mourad SM, Hermens RP, Nelen WL, et al. Guideline-based development of quality indicators for subfertility care. Hum Reprod 2007; 22: 2665-72.

33. Schouten JA, Hulscher ME, Wollersheim H, et al. Quality of antibiotic use for lower respiratory tract infections at hospitals: (how) can we measure it? Clin Infect Dis 2005; 41: 450-60.

34. Dellinger RP, Levy MM, Rhodes A, et al. Surviving sepsis campaign: international guidelines for management of severe sepsis and septic shock: 2012. Crit Care Med 2013; 41: 580-637.

35. Coenen S, Ferech M, Haaijer-Ruskamp FM, et al. European Surveillance of Antimicrobial Consumption (ESAC): quality indicators for outpatient antibiotic use in Europe. Qual Saf Health Care 2007; 16: 440-5.

36. Mourad SM, Nelen WL, Hermens RP, et al. Variation in subfertility care measured by guideline-based performance indicators. Hum Reprod 2008; 23: 2493-500.

37. Grol R, Baker R, Moss F. Quality improvement research: understanding the science of change in health care. Qual Saf Health Care 2002; 11: 110-1.

38. Landis JR, Koch GG. An application of hierarchical kappa-type statistics in the assessment of majority agreement among multiple observers. Biometrics 1977; 33: 363-74.

39. Rubin HR, Pronovost P, Diette GB. From a process of care to a measure: the development and testing of a quality indicator. Int J Qual Health Care 2001; 13: 489-96.

An online national antimicrobial guide with locally customized versions can promote

appropriate antibiotic use in hospitals

Emelie C. Schuts1, Caroline MA van den Bosch1, Inge C Gyssens2,3,4, Bart-Jan Kullberg2, Maurine A. Leverstein-van Hall5, Stephanie Natsch6, Fre Sebens7, Martha

B. Adams8, Richard Drew9, Jan M. Prins1

1Department of Internal Medicine, Division of Infectious Diseases, Center for Infection and Immunity Amsterdam (CINIMA), Academic Medical Center,

Meibergdreef 9, 1105 AZ Amsterdam, the Netherlands, 2Department of Internal Medicine, Radboud Center for Infectious Diseases, Radboud University Medical

Center, Nijmegen, the Netherlands, 3Department of Medical Microbiology and Infectious Diseases, Canisius Wilhelmina Ziekenhuis, Nijmegen, the Netherlands,

4Hasselt University, Hasselt, Belgium, 5Department of Medical Microbiology, Bronovo Hospital, Den Haag and Diaconessenhuis, Leiden, The Netherlands,

6Department of Pharmacy, Radboud University Medical Center, Nijmegen, The Netherlands, 7Department of Medical Microbiology and Infection Control, Deventer

Ziekenhuis, Deventer, The Netherlands, 8Duke University Medical Center, Durham, NC 27710, USA, 9Department of Medicine, Division of Infectious Diseases, Duke

University Medical Center, Durham, NC 27710, USA

A shortened version of this manuscript has been accepted for publication by the European Journal of Clinical Pharmacology

Chapter 2

Chapter 2

18

Customized local version of SWAB-ID

19

2

Abstract

Objective A local, evidence-based hospital antimicrobial guide is an essential element of a stewardship program. However, the lack of generally accepted professional standards for such guides results in variable quality between hospitals, and its compilation is time-consuming. In 2006 the Dutch Working Party on Antibiotic Policy (SWAB) developed an online national antimicrobial guide (SWAB-ID) for the prophylaxis and treatment of infectious diseases in hospitals. Every hospital in the Netherlands was offered the opportunity to edit this version for local use and distribute it through an independent website. By the end of 2012, approximately 40% of Dutch hospitals used a local, customized version of SWAB-ID. The purpose of our study was to determine the impact of this online tool in providing comprehensive local antibiotic treatment guidelines.

Methods We compared the local SWAB-ID based and non-SWAB-ID based antimicrobial guides for comprehensiveness and compliance with the national antimicrobial guidelines. Guides were scored on 199 pre-specified items to measure comprehensiveness and on 35 items to measure compliance.

Results Fifty antimicrobial guides (27 non-SWAB-ID and 23 SWAB-ID based), covering nearly every hospital in the Netherlands, were evaluated. The non-SWAB-ID based local guides were significantly less comprehensive (64% vs. 90%, p < 0.001) and less guideline-compliant according to the national guidelines (46% vs. 70%, p < 0.001).

Conclusions Use of a local version of a national antimicrobial guide significantly increased both comprehensiveness and guideline-compliance of local antimicrobial policies. A comprehensive national guideline with customized local versions may therefore contribute to the quality of local antimicrobial policy.

Introduction

The World Health Organisation (WHO) has signalled the emergence of antimicrobial resistance, along with the steady decline in the discovery of new antibiotics, as a major health threat for the coming decade. To help contain antimicrobial resistance, better use of current agents is essential and a decrease in inappropriate use of antibiotics is imperative.1

A major strategy being advocated as part of the efforts to control antimicrobial resistance is antimicrobial stewardship, which is an active multidisciplinary effort to improve the appropriateness of antibiotic prescribing.2, 3 Available literature indicates that an antimicrobial stewardship program can improve antibiotic prescribing with an associated decrease in resistance, without compromising short-term clinical outcomes.4-6 Besides promoting education, audits, restrictions and order sets, an essential requirement for an effective stewardship program is the presence of an evidence-based antimicrobial guide for the empirical therapy of infectious diseases, which takes into account local pathogens and their resistance patterns. However, there are no generally accepted professional standards for such guides, and the quality may vary significantly among hospitals. In addition, the compilation of a local antimicrobial guide is a time-consuming effort. The Dutch Working Party on Antibiotic Policy (SWAB ) was founded in 1996 on the initiative of the Dutch Society for Infectious Diseases , the Dutch Society for Medical Microbiology and the Dutch Association of Hospital Pharmacists. SWAB yearly reports on the use of antibiotics and trends in antimicrobial resistance in The Netherlands, in collaboration with the National Institute for Public Health and the Environment (RIVM)7, coordinates the surveillance of the use of antibiotics, and runs a guideline development program. In the Netherlands, each hospital has an antimicrobial guide. In 2006, the SWAB developed an online national antimicrobial guide (SWAB-ID) for the antimicrobial treatment and prophylaxis of common infectious diseases in hospitals. This guide also contains information on the most important characteristics of antimicrobial drugs. The recommendations on antimicrobial treatment are based on existing national evidence-based guidelines, where available, and include considerations of alternative(s) for patients with an allergy or contraindication to a first-line drug, therapeutic drug monitoring and treatment during pregnancy/lactation. Where no published guidelines are

Chapter 2

20


21

2

available, the recommendations are based on an inventory of the antimicrobial policies of the 12 Dutch centers with an infectious diseases or medical microbiology training program in 2005.8 In these cases, several equivalent alternatives are typically offered. The guide is updated the moment new national guidelines become available and the inventory of the antimicrobial policies of the 12 Dutch centers was updated at the end of 2008. The national antimicrobial guide can be accessed free of charge through the SWAB website (www.swabid.nl). In addition, to increase the quality of local antimicrobial guides, while at the same time reducing the local efforts required, every hospital antimicrobial formulary committee in the Netherlands has been offered the opportunity to obtain for a small fee a local, web-based copy of the national version, and to edit this version to local circumstances and resources. These local versions are hosted on an independent, localized website, maintained through the SWAB. The development of SWAB-ID and its customized local versions was done through a cooperative educational effort with CustomID (Duke University Medical Center, Durham, NC USA). In the ensuing years, 23 hospitals or groups of hospitals have obtained a local license, including 5 of the 8 university hospitals. There are also hospitals in the Netherlands who have not joined the SWAB-ID initiative and still fully design their own guide, either online, digital or on paper.In this study we sought to describe the relative comprehensiveness and compliance with national treatment guidelines between these two approaches, with the purpose to determine the impact of this online tool in providing comprehensive antibiotic treatment guidelines.

Methods

Hospital selectionIn May 2013 we approached all general and university hospitals in the Netherlands (n=91) through email or by phone, with the request to send us the current version of their institution’s antimicrobial guide. We excluded specialty hospitals not offering general care, e.g., orthopedic surgery only.

Data collectionWe scored all guides with regard to 234 pre-specified items, divided into two

groups. The first group, consisting of 199 items, evaluated the comprehensiveness of the antimicrobial guide, and included 192 items covering initial empiric therapy based on indication and seven general items (e.g., are there specific advices for pregnant women, and for patients with β-lactam allergy; Table 1; Supplementary Table 1). The second group, consisting of 35 frequently occurring infections, evaluated the compliance of the recommendations with the national recommendations provided by SWAB-ID.

Table 1 Comparison of comprehensiveness of SWAB-ID-based and non-SWAB-ID-based antimicrobial guides.

Recommendations for: SWAB-ID based local guides (n=23)

Non-SWAB-ID based local guides (n=27)

P-value

Empirical therapy (192 items) 90% 64% <0.01Pregnant women and lactation 100% 85% 0.08Children 100% 67% <0.01I.V. to oral switch 100% 52% <0.01Therapeutic Drug Monitoring 96% 70% 0.02β-Lactam allergy 100% 74% <0.01Adjusting to renal function 100% 85% 0.08Prophylaxis 100% 96% 0.54

Scoring comprehensivenessFor the comprehensiveness assessment, one point was granted for each item present in the local guide, the total score per hospital being the sum of all points. If a guide did not give an advice for a certain condition (e.g. tonsillitis), no points were scored. The maximum number of points that could be scored was therefore 199. The items in the questionnaire were based on the content of SWAB-ID version 1.7 (May 2013). Specific conditions that for many hospitals are not relevant, like tropical diseases, were excluded.

For the present paper we focused on empirical therapy. Therefore, local guidelines recommending consultation with an infectious disease physician or medical microbiologist, or requiring the therapy to be based on a culture result and not providing a specific recommendation, were considered to be inadequate for initial empirical therapy. Exceptions were diagnoses where also SWAB-ID recommends

Chapter 2

22


23

2

culture-based therapy or consultation of an expert, e.g. chronic osteomyelitis. For pathogen-specific infections (e.g., pneumonia due to Pseudomonas), recommendations for pneumonia due to groups of organisms (e.g., gram-negative rods) rather than the specific pathogen itself were considered inadequate. A recommendation referring to a protocol or another guideline was considered to be sufficient, since a recommendation on therapy was available for the user. Furthermore, if for a condition the recommendation was not to use antibiotics, this was considered to be sufficient.

Scoring complianceFor 35 infectious conditions (Table 2) each local recommendation was categorized as either ‘compliant with the national guide’, ‘non-compliant with the national guide’ or ‘not having a policy’. As it is based on the national guidelines, SWAB-ID version 1.7 was considered the Dutch standard and served as the reference. Only the therapy of first choice was scored. In case several alternatives were given with no clear preference, we scored all alternatives. For example, if a guide gave two recommendations on the treatment of a lung abscess that were consistent with the national guideline and one recommendation inconsistent with the national guideline, that hospital would obtain for that item 1 point on the compliance scale and 1 point on the non-compliance scale. If the local guide did not give any recommendation, because that condition was not mentioned in the guide, the hospital would get 1 point on the no-policy scale. When given the option we chose the antimicrobial recommendation for admitted patients. Furthermore, we focused on therapy recommendations for adults.

Several assumptions were made. If no age category was specified, we assumed the therapy was meant for adults. When the antimicrobial recommendation for a urinary tract infection in patients with a urinary catheter did not specify the length of presence of the catheter, we assumed this recommendation was for patients with a catheter in place for less than 10 days and we scored it accordingly.9 In case the therapy recommendation for pneumonia or sepsis did not specify whether it applied to community-acquired or hospital-acquired infections, we assumed it was meant for community-acquired infections. If there was no evident classification for pneumonia in mild, moderate or severe, we scored it as ‘not having a policy’.10

Table 2 Items used to analyze the compliance of the local antimicrobial guides with the national guidelines.

Infection1 Community-acquired cholangitis2 Cholecystitis3 Diverticulitis4 Liver abscess5 Peritonitis – Leakage/perforation6 Spontaneous bacterial peritonitis7 Community-acquired pneumonia (CAP) Category I8 CAP Category I – Penicillin allergy or inability to use doxycycline due to preg-

nancy/lactation9 CAP Category II

10 CAP Category II – Penicillin allergy11 CAP Category II – Aspiration 12 CAP Category II – Colonization Pseudomonas spp.13 CAP Category III14 CAP Category III – Post-influenza15 Lungabscess16 Pleural empyema17 Hospital-acquired pneumonia (HAP)18 HAP – Aspiration19 Cystitis – Diabetes Mellitus20 Cystitis – Pregnancy21 Urosepsis/Pyelonephritis22 Urinary tract infection – Urinary catheter in place ≤ 10 days23 Pyelonephritis – Pregnancy24 Community-acquired sepsis25 Nosocomial sepsis26 Community-acquired sepsis – Neutropenia27 Catheter-related bloodstream infection

Chapter 2

24


25

2

Statistical analysisHospitals were grouped according to their use of a customized local SWAB-ID guide. We assessed for each guide the comprehensiveness score: the percentage of the maximum score of 199 points, and compared the hospitals using local versions of the national SWAB-ID guide with the hospitals that use their own local antimicrobial guide. To evaluate compliance with the national guidelines, the local recommendations for 35 infections were categorized as previously described. For each of the 35 conditions we compared the distribution of scores of the hospitals using local versions of the national guide with those of the other hospitals. In both groups we added up for each item the points obtained in all guides on the compliant and non-compliant scales, and expressed compliant and non-compliant advices as percentage of the total number of advices per condition scored in SWAB-ID using respectively non-using hospitals. For reliability, we considered for each item a minimum of 5 points in both groups to be necessary to be retained in the analysis. Data were summarized using descriptive statistics. The comprehensiveness and compliance scores of the hospitals were compared between groups using the non-parametric Mann-Whitney U test. For all analyses SPSS 20.0 was used.

Results

We received a total of 50 antimicrobial guides, covering 85 of the 91 Dutch hospitals, because in a number of cases several hospitals shared the same guide. Twenty-three of these 50 guides were local versions of the national SWAB-ID guide and 27 were fully designed by the hospitals themselves. When known, the local antibiotic guides had a publishing date range between 2006 and 2013.

ComprehensivenessFigure 1 presents the comprehensiveness scores. The local non-SWAB-ID based antimicrobial guides (n=27) were significantly less comprehensive than the local guides based on the national SWAB-ID guide (n=23) (64% vs. 90% respectively, p < 0.001). Table 1 summarizes the results, and shows that the SWAB-ID based antimicrobial guides also scored higher on general policies, e.g. I.V. to oral switch.

Figure 1. Comprehensiveness scores (%) of the Dutch antimicrobial guides, for SWAB-ID based versus non-SWAB-ID based guides

Compliance with the national guidelinesFor 35 conditions all recommendations were retrieved and scored as ‘guideline-compliant’, ‘not guideline-compliant’, or ‘no policy’. Per condition we added up the scores for the SWAB-ID based and the non-SWAB-ID based local guides. Eight items had scores below 5 in either the SWAB or the non-SWAB group and were therefore removed from the analysis. A total of 27 items remained (Table 2).Figure 2 displays the compliance scores obtained by the SWAB-ID (Figure 2a) versus the non-SWAB-ID based local guides (Figure 2b), for each of the 27 items separately. Overall, for SWAB-ID based guides and non-SWAB-ID based guides, recommendations were guideline-compliant in 70% and 46% respectively (p < 0.001), non-guideline-compliant in 20% and 22% respectively and no recommendation was given in 10% and 32% of items respectively.

Chapter 2

26


27

2

Discussion

In 2006 the Dutch Working Party on Antibiotic Policy (SWAB) introduced an online national antimicrobial guide (SWAB-ID) for the treatment and prophylaxis of common infectious diseases in hospitals. Every hospital in the Netherlands was offered the opportunity to get a local copy of the national version, and to edit this version for local use. Customizing (national) guidelines for local use is necessary to meet local circumstances, resource constraints and barriers to implementation.11 By the end of 2012, approximately 40% of Dutch hospitals used a local, customized version of SWAB-ID. This study shows that hospitals using a local version of the national antimicrobial guide have a more comprehensive and more guideline-compliant antimicrobial policy than hospitals that have chosen to fully design their own guide. There are no obvious reasons why a hospital’s antimicrobial guide should deviate from the national antibiotic guidelines. In the Netherlands there are minimal ranges in local resistance rates7, which are not sufficient to explain the differences between policies in the antimicrobial guides. A limited inventory in three London hospitals also noticed significant discrepancies between the hospital antimicrobial guidelines and the recommendations in the literature, without a proper comment within the documents to justify these deviations.12 Adherence to evidence-based antimicrobial guidelines has been demonstrated to improve clinical outcome and to reduce the development of antimicrobial resistance.13-18 Previous studies have noted that resources and expertise required to develop a valid local guideline based on the latest evidence are mostly not available at a local level, so local adaption of a national guideline provides a cost-effective approach to a rigorous local guide development.11 Therefore, we think that this national guide with its local versions can contribute to appropriate antibiotic use in hospitals. Web-based tools are increasingly used for infectious disease surveillance, and to distribute guidelines and tools for antimicrobial stewardship programs.19-22 To our knowledge, the concept of an online national antimicrobial guide with local, customizable versions is unique. In addition, this study is the first to analyze the actual content of local antimicrobial guides and to compare local and national guidelines. While not included in the present study, the experience of the past years is that composition of a local antimicrobial guide is much less time-consuming when the hospitals use a local version of the national guide. Moreover, the

Figure 2a. Compliance of recommendations (%) with the national guideline recommendations, for 27 infectious conditions. SWAB-ID based guides.

Figure 2b. Compliance of recommendations (%) with the national guideline recommendations, for 27 infectious conditions. Non-SWAB-ID based guides

Chapter 2

28


29

2

local SWAB-ID using guides are more accurately kept up to date compared to non-SWAB-ID local guides, that were published between 2006 and 2012, and the recommendations more often trace back to the national evidence-based guidelines. Literature suggests antibiotic prescribing can improve by implementing an antimicrobial stewardship program.3-5 We believe this study proves that the use of a comprehensive national guideline like the SWAB-ID antimicrobial guide with local customizable versions is an important tool for improving the local quality of antimicrobial policy. We covered nearly every hospital in the Netherlands, and our questionnaire covered the usual range of antimicrobial policies to be found in an antimicrobial guide, but there are a number of limitations in our study. For a few infections we had to make assumptions to classify recommendations as guideline-compliant or not. Moreover, we used the national SWAB-ID antimicrobial guide as our standard and one could argue that not all policy by the SWAB is evidence-based, in which case an alternative policy is not necessarily inaccurate. As stated, in the Netherlands there are minimal ranges in local resistance rates,7 which are not sufficient to explain the differences between policies in the antimicrobial guides. Furthermore, the SWAB-ID recommendations on antimicrobial treatment are based on national evidence-based guidelines. Where no existing guideline is available the recommendation is based on expert opinion and consensus: an inventory of the antimicrobial policies of the 12 Dutch centers with an infectious diseases or medical microbiology training program.8 In these cases, all alternatives offered by these 12 local guides were considered adequate. Only recommendations that fell outside this range of options were scored non-compliant and it could therefore be debated if these deviations are plausible therapeutic strategies.Taken together, alternative options are difficult to justify. Sometimes the guidelines on which SWAB-ID is based were already a few years old, which could be a reason to deviate. One might also argue that the national SWAB-ID guide serving as reference was regularly updated (last update at the time of the inventory: April 2013), whilst the study was conducted at the same time, which is a clear disadvantage to hospitals not using SWAB-ID. However, this underscores our main point that SWAB-ID and its local version are web-based and easy to update, whereas the local versions were often several years old, and therefore in our opinion much more prone to being outdated.

In conclusion, the present study demonstrates that hospitals using a national guideline customized to the local environment have a more comprehensive and more guideline-compliant antibiotic policy than hospitals that have chosen to fully design their own guide. Future research should investigate to what extent the recommended policy in the antimicrobial guides translates into actual antibiotic prescription patterns in the hospitals in the Netherlands and in differences in outcomes of care.

Competing interestAll authors have completed the Unified Competing Interest form at http://www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare that they have no conflict of interest.The Dutch Working Party on Antibiotic Policy (SWAB) is financially supported by the National Institute for Public Health and the Environment (RIVM).

Chapter 2

30


31

2

Reference list

1 (2014) WHO Global Strategy for Containment of Antimicrobial Resistance. 2 Tamma PD, Cosgrove SE (2011) Antimicrobial stewardship. Infect Dis Clin

North Am 25 (1): 245-260 DOI 10.1016/j.idc.2010.11.0113 Dellit TH, Owens RC, McGowan JE, Jr., Gerding DN, Weinstein RA, Burke JP,

Huskins WC, Paterson DL, Fishman NO, Carpenter CF, Brennan PJ, Billeter M, Hooton TM (2007) Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America guidelines for developing an institutional program to enhance antimicrobial stewardship. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 44 (2): 159-177 DOI 10.1086/510393

4 Kaki R, Elligsen M, Walker S, Simor A, Palmay L, Daneman N (2011) Impact of antimicrobial stewardship in critical care: a systematic review. The Journal of antimicrobial chemotherapy 66 (6): 1223-1230 DOI 10.1093/jac/dkr137

5 Davey P, Brown E, Charani E, Fenelon L, Gould IM, Holmes A, Ramsay CR, Wiffen PJ, Wilcox M (2013) Interventions to improve antibiotic prescribing practices for hospital inpatients. The Cochrane database of systematic reviews 4: Cd003543 DOI 10.1002/14651858.CD003543.pub3

6 Wagner B, Filice GA, Drekonja D, Greer N, MacDonald R, Rutks I, Butler M, Wilt TJ (2014) Antimicrobial stewardship programs in inpatient hospital settings: a systematic review. Infection control and hospital epidemiology 35 (10): 1209-1228 DOI 10.1086/678057

7 (2014) NethMap 2014: Consumption of antimicrobial agents and antimicrobial resistance among medically important bacteria in The Netherlands in 2013.

8 van Vonderen MG, Gyssens IC, Hartwig NG, Kullberg BJ, Leverstein-van Hall MA, Natsch S, Prins JM (2006) [Optimalisation of the antibiotic policy in The Netherlands. XI. The national electronic antibiotic guide’SWAB-ID’ for use in hospitals]. Ned Tijdschr Geneeskd 150 (46): 2560-2564

9 Geerlings SE, van den Broek PJ, van Haarst EP, Vleming LJ, van Haaren KM, Janknegt R, Platenkamp GJ, Prins JM (2006) [Optimisation of the antibiotic policy in the Netherlands. X. The SWAB guideline for antimicrobial treatment of complicated urinary tract infections]. Ned Tijdschr Geneeskd 150 (43): 2370-2376

10 Wiersinga WJ, Bonten MJ, Boersma WG, Jonkers RE, Aleva RM, Kullberg BJ, Schouten JA, Degener JE, Janknegt R, Verheij TJ, Sachs AP, Prins JM (2012) SWAB/NVALT (Dutch Working Party on Antibiotic Policy and Dutch Association of Chest Physicians) guidelines on the management of community-acquired pneumonia in adults. Neth J Med 70 (2): 90-101

11 Eccles M, Clapp Z, Grimshaw J, Adams PC, Higgins B, Purves I, Russell I (1996) Developing valid guidelines: methodological and procedural issues from the

North of England Evidence Based Guideline Development Project. Qual Health Care 5 (1): 44-50

12 Al Mahdy H (2012) Quality assuring adult anti-microbial guidelines. Int J Health Care Qual Assur 25 (3): 226-231 DOI 10.1108/09526861211210448

13 Arnold FW, LaJoie AS, Brock GN, Peyrani P, Rello J, Menendez R, Lopardo G, Torres A, Rossi P, Ramirez JA (2009) Improving outcomes in elderly patients with community-acquired pneumonia by adhering to national guidelines: Community-Acquired Pneumonia Organization International cohort study results. Arch Intern Med 169 (16): 1515-1524 DOI 10.1001/archinternmed.2009.265

14 Asadi L, Eurich DT, Gamble JM, Minhas-Sandhu JK, Marrie TJ, Majumdar SR (2013) Impact of guideline-concordant antibiotics and macrolide/beta-lactam combinations in 3203 patients hospitalized with pneumonia: prospective cohort study. Clin Microbiol Infect 19 (3): 257-264 DOI 10.1111/j.1469-0691.2012.03783.x

15 Larson EL, Quiros D, Giblin T, Lin S (2007) Relationship of antimicrobial control policies and hospital and infection control characteristics to antimicrobial resistance rates. Am J Crit Care 16 (2): 110-120

16 Lee SS, Kim Y, Chung DR (2011) Impact of discordant empirical therapy on outcome of community-acquired bacteremic acute pyelonephritis. J Infect 62 (2): 159-164 DOI 10.1016/j.jinf.2010.10.009

17 Menendez R, Torres A, Reyes S, Zalacain R, Capelastegui A, Aspa J, Borderias L, Martin-Villasclaras JJ, Bello S, Alfageme I, de Castro FR, Rello J, Molinos L, Ruiz-Manzano J (2012) Initial management of pneumonia and sepsis: factors associated with improved outcome. Eur Respir J 39 (1): 156-162 DOI 10.1183/09031936.00188710

18 Spoorenberg V, Prins JM, Stobberingh EE, Hulscher ME, Geerlings SE (2013) Adequacy of an evidence-based treatment guideline for complicated urinary tract infections in the Netherlands and the effectiveness of guideline adherence. Eur J Clin Microbiol Infect Dis 32 (12): 1545-1556 DOI 10.1007/s10096-013-1909-6

19 Milinovich GJ, Hu W (2013) Web based surveillance systems could improve disease detection and the response to emerging disease events. BMJ 347: f4276 DOI 10.1136/bmj.f4276

20 Carneiro HA, Mylonakis E (2009) Google trends: a web-based tool for real-time surveillance of disease outbreaks. Clin Infect Dis 49 (10): 1557-1564 DOI 10.1086/630200

21 Troppy S, Haney G, Cocoros N, Cranston K, DeMaria A, Jr. (2014) Infectious disease surveillance in the 21st century: an integrated web-based surveillance and case management system. Public Health Rep 129 (2): 132-138

22 Pagani L, Gyssens IC, Huttner B, Nathwani D, Harbarth S (2009) Navigating the Web in search of resources on antimicrobial stewardship in health care institutions. Clin Infect Dis 48 (5): 626-632 DOI 10.1086/596762

Chapter 2

32


33

2

Supplementary Table 1. Items used to analyze the comprehensiveness of the local antimicrobial guides.

Recommendations for: Recommendations for:1 Pregnant women and lactation 101 Influenza2 Children 102 Isosporiasis3 I.V. to oral switch 103 Keratitis – Pathogen unknown4 Therapeutic Drug Monitoring 104 Keratitis – Candida albicans5 β-Lactam allergy 105 Herpes simplex keratitis6 Adjusting to renal function 106 Fungal keratitis7 Prophylaxis 107 Pertussis8 Acrodermatitis Chronica Atrophicans

(Lyme)108 Laryngitis

9 Adnexitis/salpingitis/PID 109 Laryngotracheobronchitis10 Amoebic liver abscess 110 Leptospirosis11 Animal bite 111 Lung abscess12 Arthritis – Lyme 112 P. malariae, P. vivax, P.ovale and P. falci-

parum13 Septic arthritis – Pathogen unknown 113 Puerperal mastitis14 Septic arthritis – Enterobacteriaceae 114 Acute mastoiditis15 Septic arthritis – Gonococci 115 Mediastinitis16 Septic arthritis – H. influenzae 116 Meningitis – Pathogen unknown17 Septic arthritis – Pseudomonas spp. 117 Meningitis – Candida albicans18 Septic arthritis – S. aureus 118 Meningitis – Cryptococcal19 Septic arthritis – Streptococcus 119 meningitis – E.coli20 Aspergillosis 120 meningitis – Group B streptococci21 Bacterial vaginosis (Gardnerella) 121 meningitis – H. influenzae22 Bartonellosis 122 meningitis – Listeria monocytogenes23 Blepharitis 123 meningitis – Meningococcus24 Borreliosis 124 meningitis – Pneumococcus25 Bronchitis 125 meningitis – S.aureus26 Bronchitis – Exacerbation COPD 126 meningitis – S.epidermidis27 Brucellosis 127 Microsporidiosis28 Burn wound 128 Ludwig’s angina29 Cutaneous candidiasis 129 MRSA carrier30 Esophageal candidiasis 130 Mycobacterium avium – Disseminated31 Oropharyngeal candidiasis 131 Mycobacterium kansasii32 Candidal vulvovaginitis 132 Necrotizing enterocolitis33 Disseminated candidiasis 133 Neuroborreliosis (Lyme)34 Renal candidiasis 134 Orbital cellulitis35 Cellulitis 135 Chronic osteomyelitis36 Chlamydia (including Lymphogranulo-

ma venereum)136 Acute osteomyelitis – Pathogen un-

known

37 Clostridium difficile 137 Acute osteomyelitis – group A strep-tococci

38 CMV colitis 138 Acute osteomyelitis – Pseudomonas spp.39 CMV radiculitis/myelitis 139 Acute osteomyelitis – S.aureus40 CMV retinitis 140 Acute osteomyelitis – Salmonella spp.41 CAP (Community-Acquired Pneumo-

nia) – Anaerobes141 Otitis externa, Otitis media, Malignant

Otitis Externa42 CAP – Chlamydia 142 Necrotizing pancreatitis43 CAP – H. influenzae 143 Parafaryngeal abscess /Retropharyn-

geal abscess44 CAP – Klebsiella pneumoniae 144 Parotitis45 CAP – Legionella 145 Peritonitis – Spontaneous bacterial

peritonitis46 CAP – Mycoplasma 146 Peritonitis – Peritoneal dialysis, Patho-

gen unknown47 CAP – Pneumococcus 147 Peritonsillar abscess48 CAP – Pseudomonas aeruginosa 148 Plague49 CAP – Q fever 149 Pityriasis versicolor50 CAP – S.aureus (not MRSA) 150 Pleural empyema51 Conjunctivitis – Pathogen unknown 151 Pneumocystis carinii pneumonia52 Conjunctivitis – Chlamydia trachomatis 152 Proctitis (man, STD)53 Cryptococcosis (cryptococcus) 153 Pyocystis54 Cryptosporidiosis 154 Rickettsiosis55 Dacryoadenitis/ Dacryocystitis 155 Scarlet fever56 Diabetic foot 156 Salmonella typhi – Typhoid fever57 Diphtheria 157 Salmonella typhi – Carrier58 Ehrlichiosis 158 Scabies59 Encephalitis – CMV 159 Pubic louse60 Encephalitis – Herpes simplex 160 Skin yeast infections61 Encephalitis – Varicella 161 Schistosomiasis62 Endocarditis 162 sepsis – Enterococcus63 Puerperal endometritis 163 sepsis – Listeria monocytogenes64 Endophthalmitis – Pathogen unknown 164 sespsis – Pseudomonas65 Endophthalmitis – Candida 165 sespsis – S.aureus66 Enterobiasis 166 Septic abortion67 Epididymo-orchitis 167 Sinusitis68 Epiglottitis 168 Staphylococcal scalded skin syndrome69 Erysipelas 169 Strongyloidiasis70 Erythema chronicum migrans (Lyme) 170 Syphilis71 Pharyngitis/ Cervical lymphadenitis 171 Tonsillitis72 Necrotizing fasciitis 172 Streptococcal toxic shock syndrome73 Furuncle 173 Cerebral toxoplasmosis74 Gastroenteritis – Pathogen unknown 174 Trichomonas vaginitis

Chapter 2

34

75 Gastroenteritis – Campylobacter spp. 175 Tuberculosis76 Gastroenteritis – Cryptosporidium 176 Ulcerative stomatitis77 Gastroenteritis – Cyclospora 177 Urethritis (man)78 Gastroenteritis – Dientamoeba fragilis 178 Urinary tract infection caused by kid-

ney stones79 Gastroenteritis – E. coli spp. 179 Varicella – Radiculitis/Myelitis80 Gastroenteritis – Entamoeba histolytica 180 Varicella - Chickenpox81 Gastroenteritis – Giardia lamblia 181 Dental root abscess82 Gastroenteritis – Isospora belli 182 Zygomycosis83 Gastroenteritis – Salmonella spp. (non

typhoidal)183 Cholangitis

84 Gastroenteritis – Shigella spp. 184 Cholecystisis85 Gastroenteritis – Vibrio cholerae 185 CAP mild – Pathogen unknown86 Gastroenteritis – Yersinia spp. 186 CAP moderate – Pathogen unknown87 Gonorrhea 187 CAP severe – Pathogen unknown88 Helicobacter pylori 188 Cystitis with diabetes mellitus89 Herpes genitalis 189 Cystitis during pregnancy90 Herpes simplex – Stomatitis 190 Diverticulitis91 Herpes simplex – mucocutaneous 191 Hospital-Acquired Pneumonia92 Herpes zoster 192 Liver abscess93 Herpes zoster ophthalmicus 193 Peritonitis (perforation)94 Cerebral abscess – onbekende verwek-

ker194 Pyelonephritis during pregnancy

95 Cerebral abscess – E.coli/Klebsiella 195 Sepsis – Pathogen unknown96 Cerebral abscess – Pseudomonas 196 Sepsis - Neutropenia97 Cerebral abscess – S. aureus 197 Urinary tract infection – urinary cath-

eter <10d98 Cerebral abscess – Streptococcus milleri 198 Urosepsis99 Hordeolum 199 Urinary tract infection – urinary cath-

eter >10d100 Impetigo

Development of quality indicators for antimicrobial treatment in adults with sepsis

Caroline MA van den Bosch1*, Marlies EJL Hulscher2, Stephanie Natsch3, Inge C Gyssens4,5,6, Jan M Prins1 and Suzanne E Geerlings1.

Dutch Sepsis QI expert panel

1 Department of Internal Medicine, division of Infectious Diseases, Center for Infection and Immunity Amsterdam (CINIMA) Academic Medical Center,

Meibergdreef 9, 1105, AZ Amsterdam Zuidoost, The Netherlands, 2 Scientific Institute for Quality of Healthcare (IQ healthcare), Radboud University Medical

Center, Nijmegen, The Netherlands, 3 Department of Clinical Pharmacy, Radboud University Medical Center, Nijmegen, The Netherlands, 4 Department of Internal

Medicine, Radboud University Medical Center, Nijmegen, The Netherlands, 5 Department of Medical Microbiology and Infectious Diseases, Canisius Wilhelmina

Ziekenhuis, Nijmegen, The Netherlands, 6 Hasselt University, Hasselt, Belgium

BMC Infect Dis 2014 juni 20;14:345.

Chapter 3

Chapter 3

36

Quality indicators for sepsis

37

3

Abstract

BackgroundOutcomes in patients with sepsis are better if initial empirical antimicrobial use is appropriate. Several studies have shown that adherence to guidelines dictating appropriate antimicrobial use positively influences clinical outcome, shortens length of hospital stay and contributes to the containment of antibiotic resistance.Quality indicators (QIs) can be systematically developed from these guidelines to define and measure appropriate antimicrobial use. We describe the development of a concise set of QIs to assess the appropriateness of antimicrobial use in adult patients with sepsis on a general medical ward or Intensive Care Unit (ICU).

MethodsA RAND-modified, five step Delphi procedure was used. A multidisciplinary panel of 14 experts appraised and prioritized 40 key recommendations from within the Dutch national guideline on antimicrobial use for adult hospitalized patients with sepsis (www.swab.nl/guidelines). A procedure to select QIs relevant to clinical outcome, antimicrobial resistance and costs was performed using two rounds of questionnaires with a face-to-face consensus meeting between the rounds over a period of three months.

ResultsThe procedure resulted in the selection of a final set of five QIs, namely: obtain cultures; prescribe empirical antimicrobial therapy according to the national guideline; start intravenous drug therapy; start antimicrobial treatment within one hour; and streamline antimicrobial therapy.

ConclusionThis systematic, stepwise method, which combined evidence and expert opinion, led to a concise and therefore feasible set of QIs for optimal antimicrobial use in hospitalized adult patients with sepsis. The next step will entail subjecting these quality indicators to an applicability test for their clinimetric properties and ultimately, using these QIs in quality-improvement projects. This information is crucial for antimicrobial stewardship teams to help set priorities and to focus improvement.

Background

Severe sepsis and septic shock are a substantial burden to health care, affecting millions of patients around the world each year1. The average cost of care for a patient with severe sepsis is about 22.000 USD2. It is often thought that severe sepsis is primarily seen on the intensive care unit (ICU). However, studies point out that the majority (50 – 68%) of patients with severe sepsis are admitted to a general medical ward3, with a mortality rate around 26 – 29.5%4,5. Since the diagnosis ‘severe sepsis’ is often poorly documented in the medical records by the treating clinicians, specific sepsis-targeted measures may not have been performed and antimicrobial use may have been inappropriate3.As the necessary first step in the improvement of appropriate use in patients with sepsis, guidelines have been developed that describe appropriate antimicrobial use in patients with sepsis admitted to a general medical ward or an Intensive Care Unit (ICU). Despite the availability of these guidelines, antimicrobials are used inappropriately: several studies show that inappropriate initial antimicrobial use in patients with severe sepsis or septic shock is associated with a reduction in survival6-9. As a second important step towards change and improvement of daily clinical care, the guideline-based development of quality indicators has been suggested10,11. Quality indicators (QIs) are measurable elements that can be used to gain insight into the appropriateness of the given antimicrobial treatment, which is important to set priorities and to focus improvement. The aim of our study was to develop a concise and therefore feasible set of QIs to measure and monitor the appropriateness of antimicrobial use in adults with sepsis admitted to a general medical ward and/or ICU.

Methods

Delphi surveyThe Dutch Working Party on Antibiotic Policy (SWAB) publishes evidence-based guidelines for antimicrobial use. We used the guideline for antimicrobial use in hospitalized patients with sepsis (published online in 2010) as a starting point for the development of a set of QIs12. This guideline covers antimicrobial use in all hospitalized adult patients with sepsis, except antimicrobial use in sepsis associated with indwelling intravascular devices that are not removed

Chapter 3

38


39

3

(tunnelled catheter or totally implantable vascular access devices) and therefore requires different therapy.We used a systematic approach -the RAND-modified Delphi method13,14- to develop a set of QIs in order to measure the appropriateness of antimicrobial treatment in adult patients with sepsis admitted to a general ward and/or ICU (Figure 1). For developing QIs by means of the Delphi method medical ethical approval was not required.

Extraction of guideline recommendationsOne infectious diseases physician and one quality-of-care specialist independently extracted key recommendations from the national guideline for antimicrobial use in hospitalized patients with sepsis. Disagreements were resolved by consensus. The selected recommendations were translated into potential QIs and each indicator was graded to determine its scientific soundness (level of evidence), using Tables 1 and 215.

Table 1 Grading system for methodological quality of individual studies15

Intervention Aetiology, prognosisA1 Systematic review of at least two independent A2-level studiesA2 Randomised Controlled Trial (RCT)

of sufficient methodological quality and power

Prospective cohort study with suffi-cient power and with adequate con-founding corrections

B Comparative Study lacking the same quality as mentioned at A2 (including patient-control and cohort studies)

Prospective cohort study lacking the same quality as mentioned at A2, retrospective cohort study or pa-tient-control study

C Non-comparative studyD Expert opinion

Table 2 Level of evidence of conclusions

Conclusions based on1 Study of level A1 or at least two independent studies of level A22 One study of level A2 or at least two independent studies of level B3 One study of level B or C4 Expert opinion

First questionnaire roundThe list of the potentially relevant QIs was converted into a written questionnaire and used for the RAND-modified Delphi method to achieve expert consensus on these QIs. The consensus procedure was performed between February and April 2011. All authors of the above mentioned national guideline, and an additional intensive care specialist and a hospital pharmacist were approached, and they all agreed to participate in the multidisciplinary expert panel. Our final expert panel consisted of four infectious diseases physicians (all working primarily outside the ICU), two medical microbiologists, two hospital pharmacists, three intensive care specialists, two hematologists and one general surgeon (14 experts).The questionnaire was sent by email to the experts, asking them to rate the QIs using the following criteria:

• ThepotentialQIleadstohealthgainforthepatient,lessbacterialresistanceor promotes efficiency of care;

• The potential QI is generalizable to all adult patients treated for sepsis with antimicrobial use;

• There is enough scientific evidence or expert consensus to justify the recommended care.

The expert panel was asked to rate the potential QIs using a 9-point Likert scale (with 1 denoting “definitely not appropriate care” and 9 denoting “definitely appropriate care”). The answer category ‘cannot assess’ was also available. The panel members (experts) were asked to add suggestions and comments regarding the potential QIs, and also to add additional potential QIs or topics for consideration.The results from the first questionnaire were analyzed using a standardized Microsoft Office Access-based consensus tool. Potential QIs rated with an overall median score of 8 or 9 without disagreement were considered to be face valid and reliable16, and were accepted as preliminary indicators. Disagreement was defined as the case in which less than 70% of the scores were in the top tertile (scores 7, 8, or 9)16. If there was disagreement and the median score was below 8, the potential indicator was rejected and not discussed during the consensus meeting. QIs with a median score of 7 without disagreement or a median score of 8 or 9 with disagreement were discussed during the consensus meeting.

Chapter 3

40


41

3

Expert panel meetingAll panel members were invited for a consensus meeting during which an overview of the first round ratings was provided. Goals of the consensus meeting were to achieve consensus on the QIs with a median score of 7 without disagreement or a median score of 8 or 9 with disagreement, and to rephrase accepted indicators using the comments from the panel in the first round of questionnaires. These QIs and the suggested new QIs by the panel members were discussed and reformulated when necessary.

Second questionnaire round, ranking procedureAfter the consensus meeting, all discussed, reformulated and added potential indicators were included in a second questionnaire. First, the panel was asked whether they agreed (yes or no) with the proposed indicators and their definitions. Redefined indicators were accepted if at least 70% of the experts agreed with the new formulation. Second, the panel was asked to prioritize the potential indicators by selecting the ‘top 5’ of most important indicators. For each number-one ranking by a panel member, we granted a potential QI five points, for each number-two ranking, we granted four points and so on. QIs receiving more than 15% of the maximum possible ranking points were considered to be the most relevant indicators.

Results

Extraction of guideline recommendationsKey recommendations were extracted from the national guideline independently by one infectious diseases physician (S.E.G.) and one quality-of-care specialist (M.E.L.J.H.). In consensus, 40 key recommendations were extracted from the national guideline and translated into potential QIs.Figure 1 shows the entire Delphi method as performed in the next steps.

First questionnaire roundThe 40 potential QIs were scored by 12 of the 14 panel members during the first questionnaire round (86% response rate). Twenty-two potential QIs had a high score (8 or 9) without disagreement and were accepted, and nine potential indicators were rejected, see Figure 1 and Table 3. The panel did not agree on

59

Figure 1. The step-wise RAND-modified Delphi method.

Figure 1 The step-wise RAND-modified Delphi method. The step-wise RAND-modified Delphi method. a. Accepted: the potential QI was selected for the next round because of an overall median score of 8 or 9, without disagreement. Disagreement was defined as the case in which less than 70% of the scores were in the top tertile (scores 7, 8, or 9). b. Discussion: the QI had a median score of 7 without disagreement or a median score of 8 or 9 with disagreement, and so it was discussed during the consensus meeting. c. Rejected: disagreement between panel members and the median was also lower than 8; the potential indicator was deselected and not discussed during the consensus meeting. d. Merged: multiple indicators were ‘rejected’ and merged into a composite, more generic indicator. e. Added: the indicator was proposed by one of the experts and was added to the initial set of indicators.

Step 3: expert panel meeting (start n = 32): 13 accepted, 2 rephrased, 17 were ‘rejected’ and mergedd into 4

new composite QIs, 1 new potential QIs was addede

Step 4: second questionnaire (start n = 20): ranking procedure to prioritize

6 QIs received the most points (2 QIs were merged into 1 QI after comments from the experts)

Step 5: Final set of 5 potential QIs

Step 2: first questionnaire round (start n = 40): 22 accepteda,

9 discussionb, 1 new proposed potential QI

Step 1: Extraction of key recommendations from the national SWAB guideline for antimicrobial therapy of hospitalized adult

patients with sepsis. n = 40

Rejected n = 14

Rejectedc n = 9

n = 40 potential QIs: n = 25 QIs on empirical therapy n = 9 QIs on modifying antimicrobial therapy/dose n = 3 QIs on monitoring antimicrobial therapy n = 2 QIs when to stop antimicrobial therapy n = 1 QI on iv-oral switching

Figure 1 The step-wise RAND-modified Delphi method. a. Accepted: the potential QI was selected for the next round because of an overall median score of 8 or 9, without disagreement. Disagreement was defined as the case in which less than 70% of the scores were in the top tertile (scores 7, 8, or 9). b. Discussion: the QI had a median score of 7 without disagreement or a median score of 8 or 9 with disagreement, and so it was discussed during the consensus meeting. c. Rejected: disagreement between panel members and the median was also lower than 8; the potential indicator was deselected and not discussed during the consensus meeting. d. Merged: multiple indicators were ‘rejected’ and merged into a composite, more generic indicator. e. Added: the indicator was proposed by one of the experts and was added to the initial set of indicators.

Chapter 3

42


43

3

nine potential QIs, they either had a median score of 7 without disagreement or a median score of 8 or 9 with disagreement. One new potential indicator regarding adapting the antimicrobial dose to renal function was proposed by one panel member. Results are shown in Table 3.

Expert panel meetingSix panel members (four infectious diseases specialists, one medical microbiologist and one intensive care specialist) were present during the consensus meeting (43%). The 22 accepted QIs with comments from the panel, the nine potential QIs with disagreement and the new indicator were discussed during the meeting.From these 32 QIs, this smaller panel accepted 13 potential QIs; two QIs were rephrased and one new potential QI was added. The other 17 QIs were rejected and merged into four new composite QIs. These potential QIs had a more generic formulation and contained: starting empirical antimicrobial therapy; duration of therapy; changing empirical therapy to pathogen-directed therapy; and harmonizing local guidelines with the national guideline (indicator 42 – 45, Table 3). This resulted in a total set of 20 potential QIs.

Second questionnaire round, ranking procedureAfter the consensus meeting, the resulting 20 potential indicators were sent to the entire panel for comments and approval. This questionnaire was scored by 13 of the 14 panel members (93% response rate). All 20 indicators were accepted, because in all cases 70% or more of the panel members agreed with the new content and rephrasing. Indicator 42 and 43 (local guidelines should correspond to the national guideline and prescribe according to the national guideline) were merged into one indicator because four out of the 13 panel members (31%) found them to be overlapping indicators. In the same questionnaire the panel members were also asked to rank the QIs. Out of this set of 19 indicators, five indicators received more than 15% of the maximum possible ranking points and were prioritized. They were found to be the most important QIs for antimicrobial care in adult patients with sepsis. The results of the second questionnaire are shown in Table 3 and Table 4 shows the final set of QIs. For the attendance list of the participation of the Delphi procedure and the development of the sepsis guideline, see Additional file 1: Table 5.

Qua

lity

indi

cato

rsLe

vel o

f su

ppor

ting

evid

ence

(see

Ta

ble 2

)

Firs

t que

stio

nnai

reC

onse

nsus

m

eetin

gSe

cond

que

stio

nnai

reM

edia

n sc

ore

% in

hi

ghes

t te

rtile

Con

clus

ion

Nr o

f exp

erts

pr

iorit

izin

g th

e QI

Tota

l sc

ore

Con

clus

ion

1. S

tart

antim

icro

bial

ther

apy

intra

veno

usly

in

adul

t pat

ient

s with

seps

is4

992

Acce

pted

aAc

cept

ed7

26Ac

cept

ed

2. S

tart

antim

icro

bial

ther

apy

as so

on as

pos

-sib

le, p

refe

rabl

y w

ithin

the fi

rst h

our i

n ad

ult

patie

nts w

ith se

vere

seps

is an

d se

ptic

shoc

k

29

100

Acce

pted

Acce

pted

1250

Acce

pted

3. B

efor

e sta

rtin

g an

timic

robi

al th

erap

y, at

le

ast t

wo

sets

of b

lood

cultu

res,

and

spec

imen

s fo

r cul

ture

from

susp

ecte

d sit

es o

f inf

ectio

n sh

ould

be t

aken

.

49

100

Acce

pted

Acce

pted

936

Acce

pted

4. F

or co

mm

unity

-acq

uire

d se

psis

with

out

neut

rope

nia a

nd w

ithou

t an

obvi

ous s

ite o

f in

fect

ion,

star

t a se

cond

or t

hird

gen

erat

ion

ceph

alos

porin

, or a

mox

icill

in an

d cl

avul

anic

ac

id +

an am

inog

lyco

side.

Dur

atio

n of

ther

a-py

: 7–1

0 da

ys.

*8

92Ac

cept

edM

erge

dd

into

num

-be

r 43/

44

5. F

or n

osoc

omia

l sep

sis w

ithou

t neu

trope

nia

and

with

no

obvi

ous s

ite o

f inf

ectio

n, st

art

pipe

raci

llin

with

tazo

bact

am, o

r a se

cond

or

third

gen

erat

ion

ceph

alos

porin

(exc

ept

cefta

zidi

me)

in co

mbi

natio

n w

ith ei

ther

an

amin

ogly

cosid

e or a

n an

ti-ps

eudo

mon

al fl

uo-

roqu

inol

one.

Dur

atio

n of

ther

apy:

7–1

0 da

ys.

*7

75D

iscus

si-on

bM

erge

d in

to n

um-

ber 4

3/44

Tabl

e 3

Resu

lts D

elphi

pro

cedu

re: fi

rst q

uesti

onna

ire, c

onse

nsus

mee

ting a

nd se

cond

que

stion

naire

Chapter 3

44


45

3

6. F

or co

mm

unity

-acq

uire

d or

nos

ocom

i-al

seps

is w

ith n

eutro

peni

a and

with

out a

n ob

viou

s site

of i

nfec

tion,

star

t pip

erac

illin

an

d ta

zoba

ctam

+/−

an am

inog

lyco

side o

r a

carb

apen

em w

ith an

ti-ps

eudo

mon

al ac

tivity

(i

mip

enem

/mer

open

em) a

s em

piric

al an

ti-ba

cter

ial r

egim

en. D

urat

ion

of th

erap

y: 7

–10

days

.

*7

75D

iscus

sion

Mer

ged

into

num

-be

r 43/

44

7. Th

e add

ition

of a

n am

inog

lyco

side t

o a

beta

-lact

am ag

ent i

n ad

ult p

atie

nts w

ith se

psis

is no

t rec

omm

ende

d, u

nles

s bas

ed o

n lo

cal

resis

tanc

e dat

a and

epid

emio

logy

(e.g

. risk

fac-

tors

for E

SBL)

a br

oad

spec

trum

of e

mpi

rical

th

erap

y ag

ains

t Gra

m-n

egat

ive p

atho

gens

is

need

ed.

*8

75Ac

cept

edAc

cept

ed3

6R

ejec

ted

8. G

lyco

pept

ides

shou

ld g

ener

ally

not

be p

art

of th

e em

piric

al an

tibac

teria

l reg

imen

in ad

ults

w

ith se

psis

(with

or w

ithou

t neu

trope

nia)

, un

less

pat

ient

s are

kno

wn

to b

e col

onise

d w

ith M

RSA

, or i

n pa

tient

s with

seve

re se

psis

and

neut

rope

nia w

ho re

ceiv

ed p

enic

illin

or

ceph

alos

porin

pro

phyl

axis.

*8

83Ac

cept

edAc

cept

ed0

0R

ejec

ted

9. F

or co

mm

unity

-acq

uire

d an

d no

soco

mia

l se

psis

and

prio

r use

of c

epha

losp

orin

s or

quin

olon

es w

ithin

30

days

bef

ore p

rese

ntat

i-on

, an

amin

ogly

cosid

e sho

uld

be ad

ded

or a

carb

apen

em w

ith an

tipse

udom

onal

activ

ity

shou

ld b

e sta

rted

. This

also

acco

unts

for a

dults

co

loni

sed

with

ESB

L-pr

oduc

ing

mic

ro-o

r-ga

nism

s and

for t

hose

adm

itted

to a

hosp

ital

with

hig

h pr

eval

ence

of E

SBL-

prod

ucin

g En

tero

bact

eria

ceae

. If p

reva

lenc

e is u

nkno

wn,

ris

k fa

ctor

s for

ESB

L sh

ould

be u

sed.

Risk

fa

ctor

s are

: a n

osoc

omia

l inf

ectio

n, p

rior u

se

of an

tibio

tics a

nd p

rese

nce o

f an

indw

ellin

g ur

inar

y ca

thet

er.

28

92Ac

cept

edAc

cept

ed2

4R

ejec

ted

10. E

mpi

rical

antif

unga

l the

rapy

may

be c

on-

sider

ed in

sele

cted

case

s: un

expl

aine

d se

psis

with

long

-term

ICU

stay

, sig

nific

ant C

andi

da

colo

nisa

tion,

and

clin

ical

risk

fact

ors s

uch

as

abdo

min

al su

rger

y, an

asto

mos

is le

akag

e, th

e pr

esen

ce o

f a ce

ntra

l ven

ous c

athe

ter a

nd th

e us

e of b

road

spec

trum

antib

iotic

s.

*8

73Ac

cept

edM

erge

d in

to n

um-

ber 4

3/44

11. F

or se

psis

with

a ho

spita

l-acq

uire

d pn

eum

onia

or a

vent

ilate

d-ac

quire

d pn

eu-

mon

ia, s

tart

amox

icill

in an

d cl

avul

anic

acid

+

an am

inog

lyco

side o

r cip

roflo

xaci

n, o

r the

co

mbi

natio

n of

a se

cond

/thi

rd g

ener

atio

n ce

phal

ospo

rin (e

xclu

ding

cefta

zidi

me)

with

an

amin

ogly

cosid

e or c

ipro

floxa

cin

or st

art p

iper

-ac

illin

with

tazo

bact

am. D

urat

ion

of th

erap

y:

max

imum

of 8

day

s.

17

67R

ejec

tedc

Chapter 3

46


47

3

12. F

or u

rose

psis,

star

t a se

cond

/thi

rd g

e-ne

ratio

n ce

phal

ospo

rin o

r the

com

bina

tion

of am

oxic

illin

and

gent

amic

in as

empi

rical

an

tibac

teria

l reg

imen

. Dur

atio

n of

ther

apy:

10

days

.

*7

92D

iscus

sion

Mer

ged

into

num

-be

r 43/

44

13. F

or u

rose

psis

and

an in

dwel

ling

urin

ary

cath

eter

, sta

rt a

seco

nd/t

hird

gen

erat

ion

cep-

halo

spor

in +

an am

inog

lyco

side o

r qui

nolo

ne

as em

piric

al an

tibac

teria

l reg

imen

.

*7

67R

ejec

ted

14. I

n ad

ults

with

uro

seps

is, g

lyco

pept

ides

sh

ould

be r

estr

icte

d to

thos

e sep

tic p

atie

nts

with

pre

viou

sly b

acte

riolo

gica

lly p

rove

n En

tero

cocc

us fa

eciu

m u

rinar

y tra

ct in

fect

ions

in

whi

ch en

tero

cocc

i are

susp

ecte

d to

be t

he

caus

ativ

e pat

hoge

ns.

*8

83Ac

cept

edM

erge

d in

to n

um-

ber 4

3d

15. F

or co

mm

unity

-acq

uire

d in

tra-a

bdom

inal

se

psis,

star

t a se

cond

/thi

rd g

ener

atio

n ce

p-ha

losp

orin

+ m

etro

nida

zole

+/−

an am

inog

-ly

cosid

e or a

mox

icill

in an

d cl

avul

anic

acid

+/

− an

amin

ogly

cosid

e. D

urat

ion

of th

erap

y:

5–7

days

.

28

92Ac

cept

edM

erge

d in

to n

um-

ber 4

3/44

16. F

or n

osoc

omia

l int

ra-a

bdom

inal

seps

is,

star

t a se

cond

/thi

rd g

ener

atio

n ce

phal

ospo

-rin

+ m

etro

nida

zole

+ an

amin

ogly

cosid

e or

amox

icill

in an

d cl

avul

anic

acid

+ an

amin

ogly

-co

side o

r pip

erac

illin

with

tazo

bact

am +

/− an

am

inog

lyco

side.

Dur

atio

n of

ther

apy:

5–7

da

ys.

27

75D

iscus

sion

Mer

ged

into

num

-be

r 43/

44

17. F

or co

mm

unity

-acq

uire

d se

psis

with

ch

olan

gitis

, sta

rt am

oxic

illin

+ an

amin

ogly

co-

side o

r am

oxic

illin

and

clav

ulan

ic ac

id +

/− an

am

inog

lyco

side.

Dur

atio

n of

ther

apy:

up

to 3

da

ys fo

llow

ing

adeq

uate

dra

inag

e.

*7

83D

iscus

sion

Mer

ged

into

num

-be

r 43/

44

18. F

or n

osoc

omia

l sep

sis w

ith ch

olan

gitis

, st

art a

mox

icill

in (w

ith o

r with

out c

lavul

anic

ac

id) +

an am

inog

lyco

side.

Dur

atio

n of

ther

a-py

: up

to 3

day

s fol

low

ing

adeq

uate

dra

inag

e.

*7

75D

iscus

sion

Mer

ged

into

num

-be

r 43/

44

19. F

or u

ncom

plic

ated

skin

and

skin

stru

ctur

e in

fect

ions

(SSS

I) w

ith se

psis,

star

t fluc

loxa

-ci

llin.

28

82Ac

cept

edM

erge

d in

to n

um-

ber 4

320

. For

com

mun

ity ac

quire

d co

mpl

icat

ed

SSSI

with

seps

is, st

art a

mox

icill

in an

d cl

avul

a-ni

c aci

d. D

urat

ion

of th

erap

y: 7

–10

days

.

*8

67D

iscus

sion

Mer

ged

into

num

-be

r 43/

4421

. For

nos

ocom

ial c

ompl

icat

ed S

SSI w

ith

seps

is, st

art a

mox

icill

in an

d cl

avul

anic

acid

+

an am

inog

lyco

side o

r pip

erac

illin

with

tazo

-ba

ctam

. Dur

atio

n of

ther

apy:

7–1

0 da

ys.

*8

75Ac

cept

edM

erge

d in

to n

um-

ber 4

3/44

22. F

or co

mm

unity

-acq

uire

d se

psis

and

necr

otisi

ng fa

sciit

is, st

art a

mox

icill

in an

d cl

avul

anic

acid

+ cl

inda

myc

in.

*6

50R

ejec

ted

23. F

or n

osoc

omia

l sep

sis an

d ne

crot

ising

fa

sciit

is, st

art a

mox

icill

in an

d cl

avul

anic

acid

+

an am

inog

lyco

side +

clin

dam

ycin

or p

iper

acil-

llin

with

tazo

bact

am +

/− an

amin

ogly

cosid

e +

clin

dam

ycin

.

*7

67R

ejec

ted

24. C

epha

losp

orin

s (+/

−met

roni

dazo

le) a

re

suita

ble a

ltern

ativ

es in

pat

ient

s with

non

-IgE

med

iate

d pe

nici

llin

rash

.

*7

67R

ejec

ted

Chapter 3

48


49

3

25. I

n ty

pe I

IgE

alle

rgic

reac

tions

to p

enic

il-lin

s, az

treon

am o

r cip

roflo

xaci

n +/

− an

ami-

nogl

ycos

ide i

n co

mbi

natio

n w

ith va

ncom

ycin

sh

ould

be c

hose

n.

*6

45R

ejec

ted

26. I

ndiv

idua

lizat

ion

of d

osin

g us

ing

the-

rape

utic

dru

g m

onito

ring

shou

ld b

e use

d w

hene

ver p

ossib

le in

adul

ts w

ith se

psis.

For

am

inog

lyco

sides

after

3 d

ays a

nd fo

r van

com

y-ci

n aft

er 5

day

s.

36

50R

ejec

ted

27. W

hen

star

ting

vanc

omyc

in th

erap

y, at

le

ast o

ne tr

ough

conc

entra

tion

(jus

t bef

ore

the f

ourt

h do

se) s

houl

d be

det

erm

ined

and

the c

once

ntra

tion

shou

ld b

e 15-

20 m

g/l.

37

50R

ejec

ted

28. F

requ

ent m

easu

ring

of va

ncom

ycin

trou

gh

conc

entra

tions

is re

com

men

ded

in p

atie

nts

with

an in

crea

sed

risk

of to

xici

ty o

r uns

tabl

e ki

dney

func

tion

and

> 5

days

of t

reat

men

t.

38

75Ac

cept

edR

ephr

ased

to

num

ber

47

29. W

ith p

rove

n Ps

eudo

mon

as b

acte

raem

ia,

com

bina

tion

ther

apy

shou

ld n

ot b

e pre

scri-

bed.

Dur

atio

n of

ther

apy

is 7

– 10

day

s.

27

70D

iscus

sion

Mer

ged

into

num

-be

r 44/

4530

. For

seps

is du

e to

met

hici

llin

susc

eptib

le

Stap

hylo

cocc

us au

reus

, sta

rt fl

uclo

xaci

llin.

28

100

Acce

pted

Acce

pted

13

Rej

ecte

d

31. M

icro

-org

anism

s with

MIC

s > 1

mg/

l su

ch as

Pse

udom

onas

aeru

gino

sa o

r pat

ient

s w

ith n

eutro

peni

a sho

uld

have

an in

trave

nous

ci

profl

oxac

in d

osag

e of 4

00 m

g tid

.

48

100

Acce

pted

Acce

pted

00

Rej

ecte

d

32. T

reat

men

t dur

atio

n sh

ould

be 1

4 da

ys fo

r se

psis

and

pneu

mon

ia d

ue to

S. a

ureu

s.4

989

Acce

pted

Mer

ged

into

num

-be

r 44

33. T

reat

men

t dur

atio

n sh

ould

be 1

4–21

day

s fo

r sep

sis an

d pn

eum

onia

due

to L

egio

nella

pn

eum

ophi

la, M

ycop

lasm

a pne

umon

iae o

r C

hlam

ydia

spp.

48

90Ac

cept

edM

erge

d in

to n

um-

ber 4

4

34. T

reat

men

t dur

atio

n sh

ould

be 1

4 da

ys

in an

unc

ompl

icat

ed S

taph

yloc

occu

s aur

eus

bact

erae

mia

.

49

91Ac

cept

edM

erge

d in

to n

um-

ber 4

435

. With

S. a

ureu

s bac

tera

emia

it is

impo

rtan

t to

sear

ch fo

r com

plic

atio

ns, t

his w

ill d

eter

-m

ine t

he d

urat

ion

of th

erap

y. C

ompl

icat

ions

ar

e: a

seco

ndar

y in

fect

ion

toge

ther

with

the

S. au

reus

bac

tera

emia

(lik

e an

endo

card

itis,

infe

cted

pro

sthe

sis, a

rthr

itis,

oste

omye

litis,

m

enin

gitis

, fas

ciiti

s, sp

leen

absc

ess)

48

67D

iscus

sion

Acce

pted

12

Rej

ecte

d

36. P

ersis

tenc

e of p

ositi

ve b

lood

cultu

res f

or

mor

e tha

n 72

hou

rs aft

er st

artin

g an

tibio

tics

shou

ld b

e con

sider

ed as

com

plic

ated

S. a

ureu

s ba

cter

aem

ia.

48

75Ac

cept

edAc

cept

ed1

1R

ejec

ted

37. W

ith se

psis

and

List

erio

sis, t

he d

urat

ion

of

ther

apy

shou

ld b

e 21

days

.4

767

Rej

ecte

d

38. A

fter c

linic

al re

cove

ry an

d w

hen

the

iden

tity

and

susc

eptib

ility

of t

he ca

usat

ive m

i-cr

o-or

gani

sm h

as b

een

dete

rmin

ed, a

switc

h to

ora

l age

nts w

ith h

igh

bioa

vaila

bilit

y sh

ould

be

mad

e.

28

91Ac

cept

edR

ephr

ased

to

num

ber

48

Chapter 3

50


51

3

39. E

mpi

rical

antim

icro

bial

ther

apy

for p

re-

sum

ed se

psis

shou

ld b

e disc

ontin

ued

in ca

se

of cl

inic

al im

prov

emen

t and

a la

ck o

f clin

ical

an

d m

icro

biol

ogic

al ev

iden

ce o

f inf

ectio

n.

Max

imum

dur

atio

n of

ther

apy

is 7

days

.

48

83Ac

cept

edAc

cept

ed2

5R

ejec

ted

40. D

iscon

tinue

bro

ad sp

ectr

um an

timic

ro-

bial

ther

apy

after

72

hour

s of c

linic

al st

abili

ty

in p

atie

nts w

ith p

ersis

ting

febr

ile n

eutro

pe-

nia t

hat s

how

no

clin

ical

or m

icro

biol

ogic

al

evid

ence

of i

nfec

tion.

Ora

l ant

imic

robi

al

prop

hyla

xis a

gain

st G

ram

-neg

ativ

e mic

ro-o

r-ga

nism

s sho

uld

be co

ntin

ued

until

reso

lutio

n of

neu

trope

nia.

29

91Ac

cept

edAc

cept

ed2

2R

ejec

ted

QIs

adde

d aft

er fi

rst q

uest

ionn

aire

:41

. Whe

n st

artin

g tre

atm

ent i

n ad

ults

with

se

psis,

dos

e and

dos

ing

inte

rval

of s

yste

mic

an

timic

robi

al th

erap

y sh

ould

be a

dapt

ed to

re

nal f

unct

ion.

4Ad

dede

Acce

pted

34

Rej

ecte

d

42. C

once

rnin

g em

piric

al th

erap

y fo

r adu

lt pa

tient

s with

seps

is, lo

cal g

uide

lines

shou

ld

corr

espo

nd to

the n

atio

nal g

uide

line,

but

shou

ld d

evia

te b

ased

on

loca

l res

istan

ce

patte

rns.

4Ad

ded

927

Acce

pted

an

d m

erge

d w

ith n

um-

ber 4

3

43. E

mpi

rical

antim

icro

bial

ther

apy

(onl

y ch

oice

of a

ntim

icro

bial

agen

t) in

all a

dult

patie

nts w

ith se

psis

shou

ld b

e pre

scrib

ed

acco

rdin

g to

the n

atio

nal g

uide

line.

*Ad

ded

27

Acce

pted

an

d m

erge

d w

ith n

um-

ber 4

2

*Bas

ed o

n av

aila

ble

Dut

ch e

pide

mio

logy

and

resis

tanc

e da

ta. a Ac

cept

ed: t

he p

oten

tial Q

I was

sele

cted

for t

he n

ext r

ound

bec

ause

of a

n ov

eral

l m

edia

n sc

ore

of 8

or 9

, with

out d

isagr

eem

ent.

Disa

gree

men

t was

defi

ned

as th

e ca

se in

whi

ch le

ss th

an 7

0% o

f the

scor

es w

ere

in th

e to

p te

rtile

(s

core

s 7, 8

, or 9

). b D

iscus

sion:

the

QI h

ad a

med

ian

scor

e of

7 w

ithou

t disa

gree

men

t or a

med

ian

scor

e of

8 o

r 9 w

ith d

isagr

eem

ent,

and

so it

was

di

scus

sed

durin

g th

e co

nsen

sus m

eetin

g. c R

ejec

ted:

disa

gree

men

t bet

wee

n pa

nel m

embe

rs a

nd th

e m

edia

n w

as a

lso lo

wer

than

8; t

he p

oten

tial

indi

cato

r was

des

elec

ted

and

not d

iscus

sed

durin

g the

cons

ensu

s mee

ting.

d Mer

ged:

mul

tiple

indi

cato

rs w

ere ‘

reje

cted

’ and

mer

ged

into

a co

mpo

site,

mor

e gen

eric

indi

cato

r. e Ad

ded:

the i

ndic

ator

was

pro

pose

d by

one

of t

he ex

pert

s and

was

adde

d to

the i

nitia

l set

of i

ndic

ator

s.

44. I

n al

l adu

lt pa

tient

s with

seps

is st

artin

g an

timic

robi

al th

erap

y th

e dur

atio

n of

ther

apy

shou

ld b

e pre

scrib

ed ac

cord

ing

to th

e nat

iona

l gu

idel

ine.

4Ad

ded

34

Rej

ecte

d

45. C

hang

e em

piric

al an

timic

robi

al th

erap

y to

pat

hoge

n-di

rect

ed th

erap

y if

cultu

re re

sults

be

com

e ava

ilabl

e.

3Ad

ded

715

Acce

pted

46. P

atie

nts w

ith a

S. au

reus

bac

tera

emia

sh

ould

hav

e a b

lood

cultu

re ta

ken

48 –

72

hour

s afte

r sta

rtin

g em

piric

al an

tibio

tic

ther

apy.

4Ad

ded

00

Rej

ecte

d

47. Th

erap

eutic

dru

g m

onito

ring

shou

ld b

e do

ne if

vanc

omyc

in o

r am

inog

lyco

sides

are

give

n >

48 h

ours

, acc

ordi

ng to

the l

ocal

gui

-de

line.

The v

anco

myc

in tr

ough

conc

entra

tion

shou

ld b

e 15-

20 m

g/l.

Res

ult

from

re

phra

sing

num

ber 2

8

11

Rej

ecte

d

48. A

fter c

linic

al re

cove

ry an

d w

hen

the

iden

tity

and

susc

eptib

ility

of t

he ca

usat

ive m

i-cr

o-or

gani

sm h

as b

een

dete

rmin

ed, a

switc

h to

ora

l age

nts w

ith h

igh

bioa

vaila

bilit

y sh

ould

be

mad

e. Ex

cept

ions

are:

S. a

ureu

s bac

tera

e-m

ia, l

iver

absc

ess,

empy

ema,

endo

card

itis,

men

ingi

tis an

d in

fect

ed p

rost

hetic

mat

eria

l.

Res

ult

from

re

phra

sing

num

ber 3

8

11

Rej

ecte

d

Chapter 3

52


53

3

Tabl

e 4

Fina

l set

of q

ualit

y ind

icato

rs to

mon

itor a

ntim

icrob

ial u

se in

hos

pita

lized

adu

lt pa

tient

s with

seps

is

Indi

ca-

tor n

um-

ber f

rom

Ta

ble 3

Qua

lity

indi

cato

rN

umer

ator

des

crip

tion

Den

omin

ator

des

crip

tion

All

patie

nts a

re: h

ospi

taliz

ed ad

ult p

atie

nts w

ith

seps

is, se

vere

seps

is or

sept

ic sh

ock,

whe

re sy

s-te

mic

antim

icro

bial

ther

apy

mus

t be s

tart

ed

All

patie

nts a

re: h

ospi

taliz

ed ad

ult p

atie

nts

with

seps

is, se

vere

seps

is or

sept

ic sh

ock,

w

here

syst

emic

antim

icro

bial

ther

apy

mus

t be

star

ted

All

patie

nts a

re: h

ospi

taliz

ed ad

ult

patie

nts w

ith se

psis,

seve

re se

psis

or

sept

ic sh

ock,

whe

re sy

stem

ic an

timi-

crob

ial t

hera

py m

ust b

e sta

rted

Num

ber

1.A

ntim

icro

bial

ther

apy

in ad

ult p

atie

nts w

ith

seps

is sh

ould

be s

tart

ed in

trave

nous

ly.N

umbe

r of p

atie

nts w

ho st

arte

d w

ith sy

ste-

mic

antim

icro

bial

ther

apy

intra

veno

usly.

Tota

l num

ber o

f pat

ient

s who

star

ted

with

syst

emic

antim

icro

bial

ther

apy.

Num

ber

2.A

ntim

icro

bial

ther

apy

shou

ld b

e sta

rted

as so

on

as p

ossib

le, p

refe

rabl

y w

ithin

the fi

rst h

our i

n ad

ult p

atie

nts w

ith se

vere

seps

is an

d se

ptic

sh

ock.

Num

ber o

f pat

ient

s with

seve

re se

psis

or

sept

ic sh

ock

who

star

ted

with

syst

emic

an

timic

robi

al th

erap

y w

ithin

the fi

rst h

our

after

the c

linic

al d

iagn

osis.

Tota

l num

ber o

f pat

ient

s with

seve

re

seps

is or

sept

ic sh

ock,

who

star

ted

with

syst

emic

antim

icro

bial

ther

apy.

Num

ber

3.Be

fore

star

ting

antim

icro

bial

ther

apy,

at le

ast

two

sets

of b

lood

cultu

res a

nd sp

ecim

ens f

or

cultu

re fr

om su

spec

ted

sites

of i

nfec

tion

shou

ld

be ta

ken.

Num

ber o

f pat

ient

s fro

m w

hom

at le

ast 2

bl

ood

cultu

res a

nd sp

ecim

ens f

or cu

lture

fro

m su

spec

ted

sites

of i

nfec

tion

wer

e ta-

ken

befo

re sy

stem

ic an

timic

robi

al th

erap

y w

as st

arte

d.

Tota

l num

ber o

f pat

ient

s who

star

ted

with

syst

emic

antim

icro

bial

ther

apy.

Num

ber

45.

Empi

ric sy

stem

ic an

timic

robi

al th

erap

y sh

ould

be

chan

ged

to p

atho

gen-

dire

cted

ther

apy

if cu

lture

resu

lts b

ecom

e ava

ilabl

e.

Num

ber o

f pat

ient

s with

a po

sitiv

e cul

ture

an

d em

piric

al sy

stem

ic an

timic

robi

al th

e-ra

py, w

hich

was

chan

ged

to p

atho

gen-

di-

rect

ed th

erap

y aft

er th

e res

ults

bec

ame

avai

labl

e.

Tota

l num

ber o

f pat

ient

s with

empi

-ric

al sy

stem

ic an

timic

robi

al th

erap

y w

hose

cultu

re b

ecam

e pos

itive

.

Num

ber

43 an

d nu

mbe

r 42

.

Empi

ric sy

stem

ic an

timic

robi

al th

erap

y (o

nly

choi

ce o

f ant

imic

robi

al ag

ent)

shou

ld b

e pre

s-cr

ibed

acco

rdin

g to

the n

atio

nal g

uide

line.

The

loca

l gui

delin

es sh

ould

corr

espo

nd to

the n

atio

-na

l gui

delin

e, bu

t sho

uld

devi

ate b

ased

on

loca

l re

sista

nce p

atter

ns.

Num

ber o

f pat

ient

s who

star

ted

with

empi

-ric

al sy

stem

ic an

timic

robi

al th

erap

y ac

cor-

ding

to th

e nat

iona

l gui

delin

e.

Tota

l num

ber o

f pat

ient

s who

star

ted

with

empi

rical

syst

emic

antim

icro

bial

th

erap

y (o

nly

choi

ce o

f ant

imic

robi

al

agen

t).

Num

ber o

f hos

pita

ls w

ith a

loca

l gui

delin

e th

at co

rres

pond

s with

the n

atio

nal g

uide

li-ne

or o

nly

devi

ates

bas

ed o

n lo

cal r

esist

an-

ce p

atter

ns.

Tota

l num

ber o

f hos

pita

ls w

ith a

loca

l gu

idel

ine.

Discussion

This systematic, stepwise method combining evidence and expert opinion generated a valid, concise and therefore feasible set of five QIs to measure appropriate antimicrobial use in adult patients with sepsis admitted to general medical wards and/or ICUs. One QI specifically applies to patients with severe sepsis or septic shock (Table 4, indicator 2).The issue of local resistance patterns and national versus local guideline recommendations for empirical treatment choices was extensively discussed during the Delphi consensus meeting. The national sepsis guideline underlines that hospitals can and should deviate from the recommendations based on local resistance patterns. We therefore favored to follow the national guidelines, but to guarantee the generalizability of the QIs another QI was added during the consensus meeting; local guidelines should correspond to the national guideline, but should deviate based on local resistance patterns (QI number 42).In a systematic review performed by McGregor et al. empiric or definitive antibiotic therapy was considered to be appropriate if the regimen exhibited in vitro activity against the isolated pathogen(s)17. We derived our key recommendations for appropriate antibiotic therapy from the national, evidence-based guideline for antimicrobial use in hospitalized patients with sepsis. This, according to the national and international experts, implies more than correct (empirical or definitive) antibiotic therapy alone: it defines correct antimicrobial use at patient level along the entire antibiotic pathway, from start (including appropriate diagnostics) to streamlining and discontinuing of antimicrobial therapy.The final set consists of independent QIs, that can be used to provide insight into the appropriateness of current antimicrobial use, to identify where there is room for improvement18. In a recent paper by our group19 we found that patients with urinary tract infections who in particular adherence to the total set of QIs, showed a significant dose–response relationship with a shorter length of hospital stay. This argues for application of the QIs in a bundle approach.This is the first study that specifically describes the development of QIs for the entire antimicrobial treatment of sepsis patients, also outside the ICU, via the modified Delphi technique. Several studies have described the use of quality measures for antimicrobial sepsis treatment20-25. However, some indicators were

Chapter 3

54


55

3

not systematically developed using a Delphi method21, and some only focused on optimal sepsis care on the ICU20,24 or focused on the start of treatment (first 24 hours) and not on the entire clinical course22,23,25.The results of our study show resemblance with the concise Surviving Sepsis Campaign Care Bundle originated from the guideline1,26. They also defined the optimal start of antimicrobial treatment and taking cultures as important parameters, only our panel members also defined streamlining as an important QI.Our study has several strengths. We used the systematic modified Delphi method, a common and validated technique in which scientific evidence is combined with expert opinion13,14,27,28. Boulkedid and colleagues recently reviewed its use and reporting, and formulated a practical guideline for using this RAND modified Delphi technique. Our procedure is consistent with their guideline29. Our panel was multidisciplinary, with 14 experts from 6 different specialties. Furthermore, the response rate of the first and second questionnaire (86% and 93%) was high, which increases the validity of the results.A limitation of this study is the national setting in which the QIs were developed, with a Dutch national expert panel and a Dutch guideline. This leads to the question of whether the results can be generalized to a wider international population. However, the guideline reviewed and graded the recent international literature and the QI development was performed by a multidisciplinary panel, in which several members have international experience and expertise on the topic.Another potential limitation was the attendance at the expert panel meeting, which was 43%. However an extensive summary concerning the results from the consensus meeting was sent to all panel members, as they were asked to give their final remarks and approval for the added and rephrased potential QIs. Since 93% returned the second questionnaire, we believe that an incomplete attendance did not undermine the validity of the results.

Conclusion

We describe the complete and precise development of a concise set of quality indicators for optimal antimicrobial use in hospitalized adult sepsis patients, by means of the Delphi method. This paper can be used as manual for others,

since transparency of healthcare becomes more important worldwide, and QIs give insight into the appropriateness of daily clinical care. At this moment, we are testing the applicability of this set of QIs in practice in 22 hospitals. After establishing their clinimetric properties we will analyze the association between adherence to the QIs and outcomes like duration of hospital stay. In the future, our guideline-based indicators can be used for national monitoring of antimicrobial use in hospitalized adults with sepsis or for quality improvement projects. This information is crucial for antimicrobial stewardship teams to help set priorities and to focus improvement.

AbbreviationsQIs, Quality indicators SWAB, Dutch working party on antibiotic policy ICU, Intensive care unit USD, United States dollar

Competing interestsThe authors declare that they have no competing interests.

AcknowledgmentsCollaborators: Dutch Sepsis QI expert panel. H.I. Bax, MD. E.F. Schippers, MD, PhD. S. van Assen, MD, PhD. C.W. Ang, MD, PhD. P. Sturm, MD, PhD. Y.G. van der Meer, PhD. Prof. M.A. Boermeester, MD. J.A. Schouten, MD, PhD. Prof. P. Pickkers, MD. J.J.W.M. Janssen, MD, PhD. Prof. N.M.A. Blijlevens, MD. N.P. Juffermans, MD, PhD.

FundsThis work was supported by Zon/MW, the Netherlands Organization for Health Research and Development, project number 205100003.

Chapter 3

56


57

3

References

1. Dellinger RP, Levy MM, Rhodes A, Annane D, Gerlach H, Opal SM, Sevransky JE, Sprung CL, Douglas IS, Jaeschke R, Osborn TM, Nunnally ME, Townsend SR, Reinhart K, Kleinpell RM, Angus DC, Deutschman CS, Machado FR, Rubenfeld GD, Webb SA, Beale RJ, Vincent JL, Moreno R, Surviving Sepsis Campaign Guidelines Committee including the Pediatric Subgroup: Surviving sepsis campaign: international guidelines for management of severe sepsis and septic shock: 2012. Crit Care Med 2013, 41:580–637.

2. Angus DC, Linde-Zwirble WT, Lidicker J, Clermont G, Carcillo J, Pinsky MR: Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care. Crit Care Med 2001, 29:1303–1310.

3. Rohde JM, Odden AJ, Bonham C, Kuhn L, Malani PN, Chen LM, Flanders SA, Iwashyna TJ: The epidemiology of acute organ system dysfunction from severe sepsis outside of the intensive care unit. J Hosp Med 2013, 8:243–247.

4. Esteban A, Frutos-Vivar F, Ferguson ND, Penuelas O, Lorente JA, Gordo F, Honrubia T, Algora A, Bustos A, Garcia G, Diaz-Regañón IR, de Luna RR: Sepsis incidence and outcome: contrasting the intensive care unit with the hospital ward. Crit Care Med 2007, 35:1284–1289.

5. Sundararajan V, Macisaac CM, Presneill JJ, Cade JF, Visvanathan K: Epidemiology of sepsis in Victoria, Australia. Crit Care Med 2005, 33:71–80.

6. Garnacho-Montero J, Garcia-Garmendia JL, Barrero-Almodovar A, Jimenez-Jimenez FJ, Perez-Paredes C, Ortiz-Leyba C: Impact of adequate empirical antibiotic therapy on the outcome of patients admitted to the intensive care unit with sepsis. Crit Care Med 2003, 31:2742–2751.

7. Kollef MH, Sherman G, Ward S, Fraser VJ: Inadequate antimicrobial treatment of infections: a risk factor for hospital mortality among critically ill patients. Chest 1999, 115:462–474.

8. Kumar A, Ellis P, Arabi Y, Roberts D, Light B, Parrillo JE, Dodek P, Wood G, Kumar A, Simon D, Peters C, Ahsan M, Chateau D, Cooperative Antimicrobial Therapy of Septic Shock Database Research Group: Initiation of inappropriate antimicrobial therapy results in a fivefold reduction of survival in human septic shock. Chest 2009, 136:1237–1248.

9. Menendez R, Torres A, Reyes S, Zalacain R, Capelastegui A, Aspa J, Borderias L, Martin-Villasclaras JJ, Bello S, Alfageme I, de Castro FR, Rello J, Molinos L, Ruiz-Manzano J: Initial management of pneumonia and sepsis: factors associated with improved outcome. Eur Respir J 2012, 39:156–162.

10. Grol R, Grimshaw J: From best evidence to best practice: effective implementation of change in patients’ care. Lancet 2003, 362:1225–1230.

11. Campbell SM, Braspenning J, Hutchinson A, Marshall MN: Research methods used in developing and applying quality indicators in primary care. BMJ 2003, 326:816–819.

12. SWAB guidelines for Antibacterial therapy of adult patients with sepsis. 2010. http://www.swab.nl/swab/cms3.nsf/uploads/65FB380648516FF2C125780F002C39E2/$FILE/swab_sepsis_guideline_december_2010.pdf.

13. Mourad SM, Hermens RP, Nelen WL, Braat DD, Grol RP, Kremer JA: Guideline-based development of quality indicators for subfertility care. Hum Reprod 2007, 22:2665–2672.

14. Stienen JJ, Tabbers MM, Benninga MA, Harmsen M, Ouwens MM: Development of quality indicators based on a multidisciplinary, evidence-based guideline on pediatric constipation. Eur J Pediatr 2011, 170(12):1513–1519.

15. CBO: Kwaliteitsinstituut voor de Gezondheidszorg CBO, handleiding voor werkgroepleden. [http://www.cbo.nl/themas/evidence-based-werken-richtlijnen-/projecten/richtlijnen]

16. Campbell SM, Cantrill JA, Roberts D: Prescribing indicators for UK general practice: Delphi consultation study. BMJ 2000, 321:425–428.

17. McGregor JC, Rich SE, Harris AD, Perencevich EN, Osih R, Lodise TP Jr, Miller RR, Furuno JP: A systematic review of the methods used to assess the association between appropriate antibiotic therapy and mortality in bacteremic patients. Clin Infect Dis 2007, 45:329–337.

18. Marwick C, Watts E, Evans J, Davey P: Quality of care in sepsis management: development and testing of measures for improvement. J Antimicrob Chemother 2007, 60:694–697.

19. Spoorenberg V, Hulscher ME, Akkermans RP, Prins JM, Geerlings SE: Appropriate antibiotic use for patients with urinary tract infections reduces length of hospital stay. Clin Infect Dis 2014, 58:164–169.

20. Berenholtz SM, Pronovost PJ, Ngo K, Barie PS, Hitt J, Kuti JL, Septimus E, Lawler N, Schilling L, Dorman T: Developing quality measures for sepsis care in the ICU. Jt Comm J Qual Patient Saf 2007, 33:559–568.

21. Diaz-Martin A, Martinez-Gonzalez ML, Ferrer R, Ortiz-Leyba C, Piacentini E, Lopez-Pueyo MJ, Martin-Loeches I, Levy MM, Artigas A, Garnacho-Montero J: Antibiotic prescription patterns in the empiric therapy of severe sepsis: combination of antimicrobials with different mechanisms of action reduces mortality. Crit Care 2012, 16:R223.

22. Levy MM, Dellinger RP, Townsend SR, Linde-Zwirble WT, Marshall JC, Bion J, Schorr C, Artigas A, Ramsay G, Beale R, Parker MM, Gerlach H, Reinhart K, Silva E, Harvey M, Regan S, Angus DC: The surviving sepsis campaign: results of an international guideline-based performance improvement program targeting severe sepsis. Intensive Care Med 2010, 36:222–231.

23. Nguyen HB, Corbett SW, Steele R, Banta J, Clark RT, Hayes SR, Edwards J, Cho TW, Wittlake WA: Implementation of a bundle of quality indicators for the early management of severe sepsis and septic shock is associated with decreased mortality. Crit Care Med 2007, 35:1105–1112.

Chapter 3

58


59

3

24. Pestana D, Espinosa E, Sanguesa-Molina JR, Ramos R, Perez-Fernandez E, Duque M, Martinez-Casanova E: Compliance with a sepsis bundle and its effect on intensive care unit mortality in surgical septic shock patients. J Trauma 2010, 69:1282–1287.

25. Schull MJ, Guttmann A, Leaver CA, Vermeulen M, Hatcher CM, Rowe BH, Zwarenstein M, Anderson GM: Prioritizing performance measurement for emergency department care: consensus on evidence-based quality of care indicators. CJEM 2011, 13:300–343.

26. Bochud PY, Bonten M, Marchetti O, Calandra T: Antimicrobial therapy for patients with severe sepsis and septic shock: an evidence-based review. Crit Care Med 2004, 32:S495–S512.

27. Hermanides HS, Hulscher ME, Schouten JA, Prins JM, Geerlings SE: Development of quality indicators for the antibiotic treatment of complicated urinary tract infections: a first step to measure and improve care. Clin Infect Dis 2008, 46:703–711.

28. Schouten JA, Hulscher ME, Wollersheim H, Braspennning J, Kullberg BJ, van der Meer JW, Grol RP: Quality of antibiotic use for lower respiratory tract infections at hospitals: (how) can we measure it? Clin Infect Dis 2005, 41:450–460.

29. Boulkedid R, Abdoul H, Loustau M, Sibony O, Alberti C: Using and reporting the Delphi method for selecting healthcare quality indicators: a systematic review. PLoS One 2011, 6:e20476.

Additional file 1:

Table 5 Participation Delphi procedure and development Sepsis guideline.

First question-naire round

Consensus meeting Second question-naire round

Preparatory committee sepsis guideline

Prof. I.C. Gyssens, MD, PhD(infec-tious diseases spe-cialist)

Prof. I.C. Gyssens, MD, PhD(infectious diseases specialist)

Prof. I.C. Gyssens, MD, PhD (infec-tious diseases spe-cialist)

Prof. I.C. Gyssens, MD, PhD (infec-tious diseases spe-cialist)

H.I. Bax, MD, PhD (infectious diseas-es specialist)

H.I. Bax, MD, PhD (infectious diseases specialist)



S. van Assen, MD, PhD (infectious diseases specialist)

S. van Assen, MD, PhD (infectious dis-eases specialist)

S. van Assen, MD, PhD(infectious diseases specialist)

S. van Assen, MD, PhD (infectious diseases specialist)

C.W. Ang, MD, PhD (medical mi-crobiologist)

C.W. Ang, MD, PhD (medical microbiol-ogist)



Prof. J.M. Prins, MD, PhD* (infectious diseases specialist)

E.F. Schippers, MD, PhD (infectious diseases specialist)

E.F. Schippers, MD, PhD (infectious diseases specialist)

N.P. Juffermans, MD, PhD (inten-sive care specialist)

N.P. Juffermans, MD, PhD (intensive care specialist)

N.P. Juffermans, MD, PhD (inten-sive care specialist)

Prof. M.A. Boer-meester, MD (sur-geon)



J.A. Schouten, MD, PhD (intensive care specialist)



Prof. P. Pickkers, MD (intensive care specialist)



J.J.W.M. Janssen, MD, PhD (hema-tologist)



Chapter 3

60

Prof. N.M.A. Bli-jlevens, MD (he-matologist)

Prof. N.M.A. Bli-jlevens, MD (hema-tologist)

Prof. N.M.A. Bli-jlevens, MD (hema-tologist)

Y.G. van der Meer, PhD (hospital pharmacist)

Y.G. van der Meer, PhD (hospital pharmacist)

P. Sturm, MD, PhD (medical microbi-ologist)

P. Sturm, MD, PhD (medical microbiol-ogist)

P. Sturm, MD, PhD (medical microbiol-ogist)

S. Natch, PhD (hospital pharma-cist)

*prof. Prins was a stand-in for dr. Schippers who was unable to attend the meeting on short notice.

Quality indicators to measure appropriate antibiotic use in hospitalized adults

Caroline M.A. van den Bosch1, Suzanne E. Geerlings1, Stephanie Natsch2, Jan M. Prins1 and Marlies E.J.L. Hulscher3

1 Department of Internal Medicine, Division of Infectious Diseases, Academic Medical Center, University of Amsterdam, the Netherlands, 2 Department of Clinical

Pharmacology, Radboud University Medical Center, Nijmegen, the Netherlands, 3 Departments Scientific Institute for Quality of Healthcare (IQ healthcare), Radboud

University Medical Center, Nijmegen, the Netherlands

Clin Infect Dis. 2015 Jan 15;60(2):281-91

Chapter 4

Chapter 4

62

Development of generic indicators

63

4

Abstract

Background An important requirement for an effective antibiotic stewardship program is the ability to measure appropriateness of antibiotic use. The aim of this study was to develop quality indicators (QIs) that can be used to measure appropriateness of antibiotic use in the treatment of all bacterial infections in hospitalized adult patients.

Methods A RAND-modified Delphi procedure was used to develop a set of QIs. Potential QIs were retrieved from literature. In two questionnaire mailings with an in-between face-to-face consensus meeting, an international multidisciplinary expert panel (17 experts) appraised and prioritized these potential QIs.

Results The literature search resulted in a list of 24 potential QIs. Nine QIs describing recommended care at patient level were selected: 1) take two blood cultures, 2) take cultures from suspected sites of infection, 3) prescribe empirical antibiotic therapy according to local guideline, 4) change empirical to pathogen directed therapy, 5) adapt antibiotic dosage to renal function, 6) switch from IV to oral, 7) document antibiotic plan, 8) perform therapeutic drug monitoring, 9) discontinue antibiotic therapy if infection is not confirmed. Two QIs describing recommended care at hospital level were also selected: 1) a local antibiotic guidelines should be present, and 2) these local guidelines should correspond to the national antibiotic guidelines.

Conclusion The selected QIs can be used in antibiotic stewardship programs to determine for which aspects of antibiotic use there is room for improvement. At this moment we are testing the clinimetric properties of these QIs in 1,800 hospitalized patients, in 22 Dutch hospitals.

Introduction

The World Health Organisation (WHO) signalled the emergence of antibiotic resistance, along with the steady decline in the discovery of new antibiotics, as a major health threat for the coming decade. To help control antibiotic resistance, better use of current agents is warranted and a decrease in inappropriate use of antibiotics is necessary.1 Antibiotic stewardship is an active interprofessional effort by multidisciplinary teams to optimize clinical outcome while minimizing unintended consequences of antibiotic use, including the emergence of resistance.2 Literature shows that stewardship programs can decrease incorrect antibiotic use and reduce health care costs without negatively influencing quality of care provided.2 An important requirement for an effective stewardship program in order to set priorities and focus improvement is the ability to measure the appropriateness of hospital antibiotic use.

Guidelines on the management of infections describe, by definition, appropriate antibiotic use.3 Adherence to such guidelines improves clinical outcome, is correlated with a lower rate of development of resistance to antibiotics, and lowers costs.4-8 Available guidelines and international literature can be used to systematically develop precise parameters, so-called quality indicators (QIs), to measure the appropriateness of antibiotic use.9-11 The European Surveillance of Antimicrobial Consumption (ESAC) developed QIs to measure appropriate outpatient antibiotic use in Europe.12 However, at this moment generic antibiotic use indicators-that is, indicators for measuring the appropriateness of antibiotic use in the treatment of all bacterial infections in hospitalized patients-are not available, but they are increasingly requested by policymakers. The aim of this study was to develop a set of generic indicators that can be used to assess the appropriateness of antibiotic use in the treatment of all bacterial infections in hospitalized adult patients.

Methods

We applied the RAND modified Delphi method to develop a set of QIs for appropriate antibiotic use in the treatment of all bacterial infections in

Chapter 4

64


65

4

hospitalized adult patients, with the exception of patients treated on the Intensive Care Unit (ICU)(figure 1).13, 14

Literature searchTo create an inventory of already available QIs we searched the databases of Pubmed and EMBASE to identify studies regarding the development or evaluation of QIs for antibiotic use in hospitalized adults. The search strategies are listed in figures 2a and b.

83

Figure 1. The RAND-modified Delphi procedure.

Expert panel meeting (start n = 25): 4 accepted, 4 rephrased, 4 were merged into 2 potential QIs, 1 new potential QI was added

Second questionnaire (start n = 11): ranking procedure to prioritize the potential QIs

Top-6: 1. Perform 2 blood cultures before starting AB treatment 2. Perform cultures from suspected sites of infection 3. Prescribe empirical therapy according to local guideline 4. Tailor AB treatment on the basis of culture results 5. AB booklet must be present in the hospital 6. Local guideline should correspond to the national guideline

328 potential QIs were extracted: 304 potential QI were excluded and

24 potential QIs remained.

First questionnaire round (start n = 24): 11 accepted, 4 disagreements, 10 newly proposed potential QIs

Literature search: 1574 articles, of which 29 contained potential QIs for appropriate antibiotic (AB) use in hospitalized

adults with an infection

Rejected n = 13

Rejected n = 9

Reasons for exclusion: - QIs not concerning antibiotic use n = 184 - QIs concerning prevention of infections n = 9 - QIs concerning antibiotic prophylaxis n = 10 - QIs which are not normative n = 32 - QIs concerning specific group of patients n = 12 - Double QIs n = 57

Figure 1. The RAND-modified Delphi procedure.

First, the abstracts were screened. Included were articles describing QIs for bacterial infections or antibiotic prescribing in hospitalized adult patients, excluding ICU patients. Potentially relevant publications were checked in full text format. Next, from these included publications, potentially relevant indicators regarding antibiotic prescribing/use were extracted, after which the exclusion criteria were applied (figure 1). QIs were excluded if they did not concern antibiotic use, were specified for a specific group of patients, concerned

quality indicator, health care [Mesh] OR

AND

anti-bacterial agents[Mesh] ORprocess indicator OR antibiotic ORclinical indicator OR anti bacterial ORprescribing indicator OR antibacterial ORperformance indicator OR anti-bacterial ORgeneric indicator OR anti-bacterial quality indicator OR agents[pharmacological action] ORoutcome indicator OR bacterial infections[Mesh] ORtransparency OR bacterial infection ORantibiotic bundle OR drug resistance, bacterial[Mesh] ORcare bundle OR antibiotic therapy ORindicator antibiotic prescribing

Figure 2a. Search strategy Medline Limits: humans, English, French, German, Italian, Spanish, Dutch

clinical indicator OR

AND

transparency ORcare bundle ORprocess indicator OR antibiotic ORprescribing indicator OR antibiotic agent ORperformance indicator OR antibiotic resistance ORgeneric indicator OR bacterial infectionoutcome indicator ORquality indicator ORindicator

Figure 2b. Search strategy EmbaseLimits: not animals, English, French, German, Italian, Spanish, Dutch, not case reports

Chapter 4

66


67

4

antibiotic prophylaxis, or were not normative. This process of excluding QIs was done by three reviewers (CvdB, SG and JP), who also determined the level of supporting evidence (table 1 and 2).

First questionnaire roundThe list of the potentially relevant QIs was converted into a written questionnaire and used for the RAND-modified Delphi method to achieve expert consensus on these QIs. We invited 20 experts from different countries and different specialties. All but 3 of the invited experts consented to participate in this survey. Ultimately, our international expert panel was composed of 17 members from the Netherlands, Spain, Belgium, Scotland, Croatia and Sweden, and consisted of 5 medical microbiologists, 4 infectious diseases specialists, 2 clinical hospital pharmacists, 2 general surgeons, 2 pulmonologists and 2 gynaecologists (see appendix 1).

Table 1. Methodological quality of individual studies

Intervention Aetiology, prognosisA1 Systematic review of at least two independent A2-level studiesA2 Randomized Controlled Trial (RCT)

of sufficient methodological quality and power

Prospective cohort study with suffi-cient power and with adequate con-founding corrections

B Comparative Study lacking the same quality as mentioned at A2 (includ-ing patient-control and cohort stud-ies)

Prospective cohort study lacking the same quality as mentioned at A2, retrospective cohort study or pa-tient-control study

C Non-comparative studyD Expert opinion

Table 2. Level of evidence of conclusions

Conclusions based on1 Study of level A1 or at least two independent studies of level A22 One study of level A2 or at least two independent studies of level B3 One study of level B or C4 Expert opinion

We asked the experts (panel members) to appraise the potential QIs while considering the following criteria:

• The recommended care leads to health gain for the patient, to less bacterial resistance or promotes efficiency of care;

• The recommended care is generalizable to all adult patients treated with antibiotics for a bacterial infection;

• There is sufficient scientific evidence or expert consensus to justify the recommended care.

To rate the degree with which the potential QI described appropriate antibiotic use (in accordance with these criteria), a Likert scale was used ranging from 1 (‘definitely not appropriate care’) to 9 (‘definitely appropriate care’), including an answer category ‘cannot assess’. The panel members could rephrase the potential indicator and could add new items and/or QIs. The results from the first questionnaire were analyzed using a standardized Microsoft Office Access-based consensus tool. QIs with a median score of 8 or 9 were accepted if there was no disagreement. Disagreement was defined as the case in which <70% of the scores were in the top tertile (scores 7, 8, or 9). If there was disagreement and the median score was ≤7, the QI was rejected. The QIs with a median score of 8 or 9 with disagreement or a median score of 7 without disagreement were discussed during the consensus meeting.15

Expert panel meetingFor pragmatic reasons, only Dutch panel members (n = 12) were invited for the expert panel meeting. The goal of the meeting was to present the results after the first round and to discuss the QIs with a median score of 8 or 9 with disagreement or a median score of 7 without disagreement. In addition, newly added potential QIs were discussed, and accepted QIs with comments from the experts were rephrased in consensus.

Second questionnaire round, ranking procedureAfter the consensus meeting, all of the accepted, added, and rephrased potential QIs were presented again in a questionnaire for final remarks, approval of the panel members, and prioritization of the potential QIs by asking the panel members to select a personal ‘top 5’ of most relevant QIs. An extensive

Chapter 4

68


69

4

summary with the results from the consensus meeting was sent to the panel members together with the second questionnaire. Rephrased indicators were accepted if at least 70% of the experts agreed with the new formulation. When an indicator was mentioned first in a panellist’s “top 5”, it was granted 5 points; the second was given 4 points, the third indicator was granted 3 points and so on. QIs receiving >15% of the maximum possible ranking points were considered to be the most important QIs for antibiotic care in all adult patients with a bacterial infection.

Results

Literature searchOf the 1574 identified articles regarding bacterial infections and/or antibiotic prescribing, 46 provided QIs, of which 29 articles described QIs for hospitalized adult patients, with the exception of patients treated in the ICU. From these 29 articles we derived 328 QIs, which also included five systematically developed, but at that moment no yet published QIs regarding antibiotic treatment in hospitalized adults with sepsis.16 See appendix 2 for these 29 articles. With 3 reviewers, we applied the predefined exclusion criteria and 304 QIs were excluded, mostly because they did not concern antibiotic use (184 QIs) or were doubles (57 QIs) (figure 1). This resulted in 24 potential generic QIs. These 24 potential QIs were put into a written questionnaire and sent to the 17 panel members (table 3, numbers 1 – 24).

First questionnaire roundThe consensus procedure was performed between May and October 2011. Sixteen members of the panel (all except 1 general surgeon) returned the first questionnaire (94% response). Eleven of the 24 initial indicators were accepted and 9 indicators were rejected (figure 1 and table 3). The panel members disagreed on 4 potential QIs and 10 new potential QIs were suggested (table 3, number 29, 33 - 41).

Expert panel meetingFive (29%) Dutch panel members (1 medical microbiologist, 2 infectious diseases specialists, 1 clinical hospital pharmacist and 1 general surgeon)

attended the consensus meeting. Discussed were the 11 accepted QIs with comments, the 4 with disagreement or a median score of 7, and the 10 newly proposed indicators from the first questionnaire round. Comments from the panel members regarding the first questionnaire were used to rephrase some of the accepted indicators. All 4 potential QIs requiring discussion and 9 of the 10 newly proposed QIs were rejected. From the 11 previously accepted indicators, 4 indicators were rephrased, 3 remained unchanged and another 4 indicators were merged into 2 indicators with similar content (table 3). One additional potential QI was added during the meeting.

Second questionnaire round, ranking procedureDuring the second questionnaire round, 11 potential QIs were presented to all panel members for final remarks and approval. All 17 questionnaires were returned (100% response) and no indicator was excluded, as ≥70% of the panelists agreed with each new formulation. The ranking of this entire set of indicators resulted in 6 QIs with the highest scores (table 3).

Final set of selected QIsTable 4 shows the entire set of 11 QIs representing the final, valid set of QIs that can be used to measure the appropriateness of antibiotic use in the treatment of all bacterial infections in hospitalized adult patients.

Chapter 4

70


71

4

Tabl

e 3.

Res

ults

Delp

hi p

roce

dure

: firs

t que

stion

naire

, con

sens

us m

eetin

g and

seco

nd q

uesti

onna

ire.

Qua

lity

indi

cato

rsLe

vel o

f su

ppor

ting

evid

ence

(see

ap

pend

ix 1

)

Firs

t que

stio

nnai

reC

onse

nsus

m

eetin

gSe

cond

que

stio

nnai

reQ

ualit

y in

dica

tors

firs

t que

stio

nnai

reM

edia

n%

in

high

est

tert

ile

Con

clus

ion

Num

-be

r of

expe

rts

Tota

l sc

ore

Con

clus

ion

1.In

hos

pita

lized

adul

ts w

ith a

susp

ecte

d ba

cter

ial

infe

ctio

n, em

piric

al th

erap

y sh

ould

be s

tart

ed in

tra-

veno

usly.

46

47R

ejec

tedc

2.In

hos

pita

lized

adul

ts w

ith a

susp

ecte

d ba

cter

ial

infe

ctio

n, em

piric

al th

erap

y sh

ould

be s

tart

ed as

so

on as

pos

sible

, pre

fera

bly

with

in th

e firs

t hou

r of

pres

enta

tion.

28

57D

iscus

sbR

ejec

ted

3.In

hos

pita

lized

adul

ts w

ith a

susp

ecte

d ba

cter

ial

infe

ctio

n, em

piric

al th

erap

y sh

ould

be s

tart

ed w

ith-

in 4

hou

rs aft

er cl

inic

al p

rese

ntat

ion.

2 (p

neum

o-ni

a)4

(UT

Is)

769

Rej

ecte

d

4.In

hos

pita

lized

adul

ts w

ith a

susp

ecte

d ba

cter

ial

infe

ctio

n, em

piric

al th

erap

y sh

ould

be s

tart

ed w

ith-

in 8

hou

rs aft

er ar

rival

in th

e em

erge

ncy

depa

rt-

men

t.

27

53R

ejec

ted

5.In

hos

pita

lized

adul

ts w

ith a

susp

ecte

d ba

cter

ial

infe

ctio

n, em

piric

al th

erap

y sh

ould

be a

dmin

ister

ed

whi

le th

e pat

ient

is in

the e

mer

genc

y de

part

men

t.

26

50R

ejec

ted

6.Be

fore

star

ting

syst

emic

antib

iotic

ther

apy

in h

os-

pita

lized

adul

ts w

ith a

susp

ecte

d ba

cter

ial i

nfec

tion,

at

leas

t tw

o se

ts o

f blo

od cu

lture

s sho

uld

be ta

ken.

2 (s

ever

e pn

eum

onia)

3 (s

epsis

)

880

Acce

pted

aAc

cept

ed7

15Ac

cept

ed

and

sele

ct-

ed fo

r top

6

7.Bl

ood

cultu

res b

efor

e sta

rt an

tibio

tics s

houl

d be

ob

tain

ed fr

om h

ospi

taliz

ed p

atie

nts w

ith a

susp

ect-

ed b

acte

rial i

nfec

tion

and

the c

linic

al in

dica

tion

liste

d he

re: I

CU

adm

issio

n, ca

vita

ry in

filtra

tes,

leuk

open

ia, a

ctiv

e alc

ohol

abus

e, ch

roni

c sev

ere

liver

dise

ase,

aspl

enia

or p

leur

al eff

usio

n.

28

53D

iscus

sR

ejec

ted

8.Be

fore

star

ting

antib

iotic

ther

apy

in h

ospi

taliz

ed

adul

ts w

ith a

susp

ecte

d ba

cter

ial i

nfec

tion,

spec

i-m

ens f

or cu

lture

from

susp

ecte

d sit

es o

f inf

ectio

n sh

ould

be t

aken

.

48

94Ac

cept

edR

ephr

ased

in

to n

um-

ber 2

6

9.Be

fore

star

ting

antib

iotic

ther

apy

in h

ospi

taliz

ed

adul

ts w

ith a

susp

ecte

d ba

cter

ial i

nfec

tion,

a G

ram

st

ain

of a

spec

imen

from

the s

uspe

cted

site

of i

nfec

-tio

n sh

ould

be p

erfo

rmed

.

45

27R

ejec

ted

10.In

hos

pita

lized

adul

ts w

ith a

susp

ecte

d ba

cter

ial

infe

ctio

n, em

piric

al th

erap

y sh

ould

be p

resc

ribed

ac

cord

ing

to th

e nat

iona

l gui

delin

e.

2 (i

ntra

-ab-

dom

inal

in-

fect

ions

)3

(pne

umo-

nia)

893

Acce

pted

Mer

gedd

into

indi

ca-

tor n

umbe

r 25

11.Th

e cho

ice o

f ini

tial a

ntib

iotic

s sho

uld

be em

-pi

rical

, but

shou

ld cl

early

be g

uide

d by

the c

linic

al

pict

ure a

nd th

e sen

sitiv

ity p

atter

ns o

f loc

al p

atho

-ge

ns.

28

73Ac

cept

edM

erge

d in

to in

dica

-to

r num

ber

2512

. Whe

n pr

escr

ibin

g em

piric

al th

erap

y in

hos

pi-

taliz

ed ad

ults

with

a su

spec

ted

bact

eria

l inf

ectio

n,

loca

l gui

delin

es sh

ould

corr

espo

nd to

the n

atio

n-al

gui

delin

e and

can

only

dev

iate

on

legi

timat

e gr

ound

s.

48

81Ac

cept

edR

ephr

ased

in

to in

dica

-to

r num

ber

32

Chapter 4

72


73

4

13.In

hos

pita

lized

adul

ts w

ith a

susp

ecte

d ba

cter

ial

infe

ctio

n, em

piric

al tr

eatm

ent w

ith fl

uoro

quin

olo-

nes s

houl

d on

ly b

e use

d if

oral

ther

apy

is gi

ven

or

in ca

se o

f ana

phyl

axis

rela

ted

to b

eta-

lact

am an

tibi-

otic

s.

36

33R

ejec

ted

14.In

hos

pita

lized

adul

ts w

ith a

susp

ecte

d ba

cter

ial

infe

ctio

n, d

ose a

nd d

osin

g in

terv

al o

f ant

ibio

tics

shou

ld b

e ada

pted

to re

nal f

unct

ion.

48

88Ac

cept

edAc

cept

ed2

3Ac

cept

ed

15.In

hos

pita

lized

adul

ts w

ith a

susp

ecte

d ba

cter

ial

infe

ctio

n, em

piric

al an

tibio

tics s

houl

d be

chan

ged

to p

atho

gen-

dire

cted

ther

apy

as so

on as

cultu

re

resu

lts b

ecom

e ava

ilabl

e.

3 (a

ll in

fec-

tions

)4

(UT

Is an

d pn

eum

onia)

988

Acce

pted

Acce

pted

1123

Acce

pted

an

d se

lect

-ed

for t

op 6

16.In

hos

pita

lized

adul

ts w

ith a

bact

eria

l inf

ectio

n,

antib

iotic

ther

apy

shou

ld b

e sw

itche

d fro

m in

trave

-no

us to

ora

l ant

ibio

tic th

erap

y aft

er 4

8 -7

2 ho

urs o

n th

e bas

is of

the c

linic

al co

nditi

on.

1(pn

eum

o-ni

a)2

(all

infe

c-tio

ns)

3 (U

TIs

)

880

Acce

pted

Rep

hras

ed

into

indi

ca-

tor n

umbe

r 27

17.H

ospi

taliz

ed a

dults

with

a b

acte

rial

infe

ctio

n sh

ould

be

switc

hed

from

intra

veno

us to

ora

l ant

ibi-

otic

s with

in 2

4 ho

urs o

f bei

ng ca

ndid

ates

for s

witc

h th

erap

y. Cr

iteria

for s

witch

ing:

1. c

linica

l sym

ptom

s are

impr

ov-

ing,

2. p

atien

t is a

febril

e for

at l

east

8 ho

urs,

3. th

e whi

te

bloo

d ce

ll co

unt i

s nor

mal

izin

g, 4

. ora

l int

ake a

nd ga

s-tro

inte

stina

l abs

orpt

ion

are a

dequ

ate.

27

56R

ejec

ted

18.H

ospi

taliz

ed a

dults

with

a b

acte

rial

infe

ctio

n sh

ould

be

switc

hed

from

intra

veno

us to

ora

l ant

i-bi

otic

s w

hen

the

patie

nt m

eets

all

of th

e fo

llow

ing

crite

ria: a

clin

ical

ly im

prov

ing

cond

ition

, hem

ody-

nam

ic st

abili

ty, a

nd to

lera

nce

of o

ral m

edic

atio

n or

fo

od an

d flu

ids.

28

67D

iscus

sR

ejec

ted

19.W

hen

antib

iotic

ther

apy

was

star

ted

in a

hosp

i-ta

lized

adul

t, th

e act

ual l

engt

h of

trea

tmen

t sho

uld

be in

acco

rdan

ce w

ith th

e len

gth

men

tione

d in

the

patie

nt’s

med

ical

file

.

45

15re

ject

ed

20. I

f the

pre

sent

ing

clin

ical

synd

rom

e in

a hos

pi-

taliz

ed ad

ult i

s det

erm

ined

to b

e due

to a

non-

infe

c-tio

us ca

use,

antib

iotic

ther

apy

shou

ld b

e sto

pped

pr

ompt

ly.

29

88Ac

cept

edR

ephr

ased

in

to in

dica

-to

r num

ber

3021

. In

a hos

pita

lized

adul

t with

a su

spec

ted

bact

eri-

al in

fect

ion

who

was

initi

ally

star

ted

on in

trave

nous

(i

v) an

tibio

tic th

erap

y, th

e pos

sibili

ty o

f iv-

oral

sw

itch

shou

ld b

e doc

umen

ted

in ca

se n

otes

.

46

50R

ejec

ted

22.In

a ho

spita

lized

adul

t with

a su

spec

ted

bact

eria

l in

fect

ion,

the i

ndic

atio

n to

star

t ant

ibio

tics s

houl

d be

doc

umen

ted

in ca

se n

otes

.

48

86Ac

cept

edM

erge

d in

to in

dica

-to

r num

ber

2823

.In a

hosp

italiz

ed ad

ult w

ith a

susp

ecte

d ba

cter

ial

infe

ctio

n w

ho w

as st

arte

d on

antib

iotic

ther

apy,

an an

tibio

tic p

lan

(nam

e, do

se, r

oute

, int

erva

l of

adm

inist

ratio

n an

d pl

anne

d du

ratio

n) sh

ould

be

docu

men

ted

in ca

se n

otes

.

48

79Ac

cept

edM

erge

d in

to in

dica

-to

r num

ber

28

24.In

a ho

spita

lized

adul

t with

a su

spec

ted

bact

eria

l in

fect

ion

who

was

star

ted

on an

tibio

tic th

erap

y, a

revi

ew o

f the

dia

gnos

is sh

ould

be d

ocum

ente

d in

ca

se n

otes

.

47

77D

iscus

sR

ejec

ted

Chapter 4

74


75

4

25.In

hos

pita

lized

adul

ts w

ith a

susp

ecte

d ba

cter

ial

infe

ctio

n, em

piric

al sy

stem

ic an

tibio

tic tr

eatm

ent

shou

ld b

e pre

scrib

ed ac

cord

ing

to th

e loc

al g

uide

-lin

e. *if

loca

l gui

delin

es a

re m

issin

g, p

resc

ribe a

ccor

ding

to

natio

nal g

uide

line.

If na

tiona

l gui

delin

es a

re a

lso m

iss-

ing,

pre

scrib

e acc

ordi

ng to

inte

rnat

iona

l gui

delin

e.

Res

ult f

rom

m

ergi

ng

indi

cato

rs

nr 1

0 an

d nu

mbe

r 11

1325

Acce

pted

an

d se

lect

-ed

for t

op 6

26.W

hen

star

ting

syst

emic

antib

iotic

ther

apy

in

hosp

italiz

ed ad

ults

with

a su

spec

ted

bact

eria

l in-

fect

ion,

spec

imen

s for

cultu

re fr

om su

spec

ted

sites

of

infe

ctio

n sh

ould

be t

aken

as so

on as

pos

sible

, pr

efer

ably

bef

ore a

ntib

iotic

s are

star

ted.

*Cul

ture

s sh

ould

be t

aken

unt

il m

axim

al 2

4 ho

urs a

fter a

ntib

iot-

ics a

re st

arted

.

Res

ult f

rom

re

phra

sing

indi

cato

r nu

mbe

r 8

613

Acce

pted

an

d se

lect

-ed

for t

op 6

27.In

hos

pita

lized

adul

ts w

ith a

bact

eria

l inf

ectio

n,

syst

emic

antib

iotic

ther

apy

shou

ld b

e sw

itche

d fro

m in

trave

nous

to o

ral a

ntib

iotic

ther

apy

with

in

48 -7

2 ho

urs o

n th

e bas

is of

the c

linic

al co

nditi

on

and

whe

n or

al tr

eatm

ent i

s ade

quat

e. *A

dequ

ate =

1. w

hen

antib

iotic

is a

vaila

ble o

rally

2.

Whe

n or

al in

take

and

gastr

oint

estin

al a

bsor

ptio

n ar

e ad

equa

te. 3

. Ade

quat

e in

term

s of d

iagn

osis

(exc

ep-

tions

like

endo

card

itis,

men

ingi

tis).

Res

ult f

rom

re

phra

sing

indi

cato

r nu

mbe

r16

34

Acce

pted

28.In

a ho

spita

lized

adul

t with

syst

emic

antib

iotic

th

erap

y, an

antib

iotic

pla

n sh

ould

be d

ocum

ente

d in

the c

ase n

otes

at th

e sta

rt o

f tre

atm

ent (

indi

ca-

tion,

nam

e, do

ses,

rout

e, in

terv

al o

f adm

inist

ratio

n).

Res

ult f

rom

m

ergi

ng

indi

cato

rs

num

ber 2

2 an

d 23

25

Acce

pted

29.Th

erap

eutic

dru

g m

onito

ring

shou

ld b

e per

-fo

rmed

whe

n th

e tre

atm

ent d

urat

ion

is lo

nger

than

3

days

for a

min

ogly

cosid

es an

d 5

days

for v

anco

-m

ycin

.

Adde

deAc

cept

ed as

ne

w in

di-

cato

r

11

Acce

pted

30.E

mpi

rical

antib

iotic

ther

apy

for p

resu

med

bac

-te

rial i

nfec

tion

shou

ld b

e disc

ontin

ued

base

d on

the

lack

of c

linic

al- a

nd/o

r mic

robi

olog

ical

evid

ence

of

infe

ctio

n. Th

e max

imum

dur

atio

n of

empi

rical

sy

stem

ic an

tibio

tic tr

eatm

ent s

houl

d be

7 d

ays.

Res

ult f

rom

re

phra

sing

indi

cato

r nu

mbe

r 20

37

Acce

pted

Proc

ess i

ndic

ator

s:31

.A cu

rren

t loc

al an

tibio

tic g

uide

line s

houl

d be

pr

esen

t in

the h

ospi

tal a

nd an

eval

uatio

n w

heth

-er

an u

pdat

e sho

uld

be co

nsid

ered

shou

ld b

e do

ne:

Adde

deAc

cept

ed

and

sele

ct-

ed fo

r top

6

-

ever

y ye

ar?

2

- ev

ery

two

year

s?7

-

ever

y th

ree y

ears

?7

-

ever

y fiv

e yea

rs?

132

.Loc

al an

tibio

tic g

uide

lines

shou

ld co

rres

pond

to

the n

atio

nal a

ntib

iotic

gui

delin

es, b

ut sh

ould

de

viat

e bas

ed o

n lo

cal r

esist

ance

patt

erns

.

Res

ult f

rom

re

phra

sing

indi

cato

r nu

mbe

r 12

Acce

pted

an

d se

lect

-ed

for t

op 6

33. W

hen

not t

o tre

at (a

sym

ptom

atic

UT

I, M

RSA

in

sput

um, c

onta

min

atio

n et

c.)

Sugg

este

d ne

w to

pic/

QI

Rej

ecte

d

34. D

ose o

f ant

ibio

tic th

erap

ySu

gges

ted

new

topi

c/Q

I

Rej

ecte

d

35. P

resc

ribin

g ac

cord

ing

to P

k/Pd

prin

cipl

esSu

gges

ted

new

topi

c/Q

I

Rej

ecte

d

Chapter 4

76


77

4

36. A

ppro

pria

te m

icro

biol

ogic

al sp

ecim

ens a

lso fo

r vi

ruse

sSu

gges

ted

new

topi

c/Q

I

Rej

ecte

d

37. R

apid

urin

e ant

igen

test

in se

vere

pne

umo-

nia

Sugg

este

d ne

w to

pic/

QI

Rej

ecte

d

38. D

ocum

entin

g cl

inic

al o

utco

mes

and

treat

men

t fa

ilure

sSu

gges

ted

new

topi

c/Q

I

Rej

ecte

d

39. D

ocum

entin

g se

verit

y of

seps

is in

case

not

es at

st

art o

f tre

atm

ent

Sugg

este

d ne

w to

pic/

QI

Rej

ecte

d

40. B

road

spec

trum

IV th

erap

y fo

r pat

ient

s with

se

vere

seps

isSu

gges

ted

new

topi

c/Q

I

Rej

ecte

d

41. L

engt

h of

antib

iotic

trea

tmen

t Su

gges

ted

new

topi

c/Q

I

Rej

ecte

d

Abb

revi

atio

ns: I

CU

, inte

nsiv

e car

e uni

t; IV

, intra

veno

us; M

RSA

, met

hici

llin-

resis

tant

Sta

phyl

ococ

cus a

ureu

s; PD

, pha

rmac

odyn

amic

; PK

, pha

rmac

okin

etic

; QI,

qual

ity in

dica

tor;

UT

I, ur

inar

y tra

ct in

fect

ion.

a. A

ccep

ted:

the

pote

ntia

l QI

was

sel

ecte

d fo

r th

e ne

xt

roun

d be

caus

e of

an

over

all m

edia

n sc

ore

of 8

or 9

, with

out d

isagr

eem

ent.

Disa

gree

men

t was

defi

ned

as th

e ca

se in

whi

ch le

ss th

an

70%

of t

he sc

ores

wer

e in

the

top

tert

ile (s

core

s 7, 8

, or 9

). b.

Disc

ussio

n: th

e Q

I had

a m

edia

n sc

ore

of 7

with

out d

isagr

eem

ent o

r a

med

ian

scor

e of

8 o

r 9 w

ith d

isagr

eem

ent,

and

so it

was

disc

usse

d du

ring

the

cons

ensu

s mee

ting.

c. R

ejec

ted:

disa

gree

men

t bet

wee

n pa

nel m

embe

rs an

d th

e m

edia

n w

as al

so lo

wer

than

8; t

he p

oten

tial i

ndic

ator

was

des

elec

ted

and

not d

iscus

sed

durin

g th

e co

nsen

sus

mee

ting.

d. M

erge

d: m

ultip

le in

dica

tors

wer

e ‘re

ject

ed’ a

nd m

erge

d in

to a

com

posit

e, m

ore g

ener

ic in

dica

tor.

e. A

dded

: the

indi

cato

r w

as p

ropo

sed

by o

ne o

f the

exp

erts

and

was

add

ed to

the

initi

al se

t of i

ndic

ator

s. Su

ppor

ting

evid

ence

var

ied

som

etim

es b

etw

een

popu

latio

ns (a

dults

with

pne

umon

ia o

r with

seps

is or

with

com

plic

ated

urin

ary

tract

infe

ctio

ns)

Tabl

e 4.

Fin

al li

st of

qua

lity

indi

cato

rs to

mon

itor

antib

iotic

use

for

all b

acte

rial i

nfec

tions

in h

ospi

taliz

ed a

dult

patie

nts

on n

on-IC

U

depa

rtm

ents.

Num

ber

from

tabl

e 3

Qua

lity

indi

cato

r

All

patie

nts a

re: a

dults

, adm

itted

on

a no

n-IC

U d

epar

tmen

t with

> 2

4h o

f sy

stem

ic a

ntib

iotic

s bec

ause

of a

sus-

pect

ed b

acte

rial

infe

ctio

n

Num

erat

or d

escr

iptio

n

All

patie

nts a

re: a

dults

, adm

it-te

d on

a n

on-I

CU

dep

artm

ent

with

> 2

4h o

f sys

tem

ic a

ntib

iot-

ics b

ecau

se o

f a su

spec

ted

bact

e-ri

al in

fect

ion

Den

omin

ator

des

crip

tion

All

patie

nts a

re: a

dults

, adm

itted

on

a n

on-I

CU

dep

artm

ent w

ith >

24

h of

syst

emic

ant

ibio

tics b

ecau

se

of a

susp

ecte

d ba

cter

ial i

nfec

tion

Num

ber 2

5Em

piric

al sy

stem

ic an

tibio

tic th

erap

y sh

ould

be p

resc

ribed

acco

rdin

g to

the

loca

l gui

delin

e. (if

loca

l gui

delin

es a

re m

issin

g, p

resc

ribe

acco

rdin

g to n

atio

nal g

uide

line.

If na

tiona

l gu

ideli

nes a

re a

lso m

issin

g, p

resc

ribe a

ccor

d-in

g to i

nter

natio

nal g

uide

line)

.

Num

ber o

f pat

ient

s who

star

ted

with

empi

rical

syst

emic

antib

iotic

th

erap

y ac

cord

ing

to th

e loc

al

guid

elin

e

Tota

l num

ber o

f pat

ient

s who

star

ted

with

empi

rical

syst

emic

antib

iotic

th

erap

y

Num

ber 6

Befo

re st

artin

g sy

stem

ic an

tibio

tic th

erap

y at

leas

t tw

o se

ts o

f blo

od cu

lture

s sho

uld

be ta

ken.

Num

ber o

f pat

ient

s in

who

m at

le

ast 2

sets

of b

lood

cultu

res w

ere

take

n be

fore

syst

emic

antib

iotic

th

erap

y w

as st

arte

d

Tota

l num

ber o

f pat

ient

s who

star

ted

with

syst

emic

antib

iotic

ther

apy

Num

ber 2

6W

hen

star

ting

syst

emic

antib

iotic

ther

apy

spec

imen

s for

cultu

re fr

om su

spec

ted

sites

of

infe

ctio

n sh

ould

be t

aken

as so

on as

po

ssib

le, p

refe

rabl

y be

fore

antib

iotic

s are

st

arte

d.

*Cul

ture

s sho

uld

be ta

ken

until

max

imal

24

hour

s afte

r ant

ibio

tics a

re st

arted

.

Num

ber o

f pat

ient

s in

who

m cu

l-tu

res f

rom

susp

ecte

d sit

es o

f inf

ec-

tion

wer

e tak

en w

ithin

24

hour

s aft

er th

e sys

tem

ic an

tibio

tics w

ere

star

ted

Tota

l num

ber o

f pat

ient

s who

star

ted

with

syst

emic

antib

iotic

ther

apy

Chapter 4

78


79

4

Num

ber 1

5Em

piric

al an

tibio

tic th

erap

y sh

ould

be

chan

ged

to p

atho

gen-

dire

cted

ther

apy

if cu

lture

resu

lts b

ecom

e ava

ilabl

e

Num

ber o

f pat

ient

s with

empi

rical

sy

stem

ic an

tibio

tics w

hose

cultu

re

beca

me p

ositi

ve an

d ch

angi

ng to

pa

thog

en- d

irect

ed th

erap

y w

as

done

corr

ectly

Tota

l num

ber o

f pat

ient

s with

empi

ri-ca

l sys

tem

ic an

tibio

tics w

hose

cultu

re

beca

me p

ositi

ve

Num

ber 1

4D

ose a

nd d

osin

g in

terv

al o

f sys

tem

ic

antib

iotic

s sho

uld

be ad

apte

d to

rena

l fu

nctio

n

Num

ber o

f pat

ient

s with

a co

m-

prom

ised

rena

l fun

ctio

n w

ith a

dosin

g re

gim

en ad

just

ed to

rena

l fu

nctio

n (d

efine

d as

an

estim

ated

Glo

mer

-ul

ar F

iltra

tion

Rate

< 5

0 m

l/m

in/1

,73m

2 )

Tota

l num

ber o

f pat

ient

s who

star

ted

with

syst

emic

antib

iotic

ther

apy

and

who

had

a co

mpr

omise

d re

nal f

unc-

tion.

*d

efine

d as

an

estim

ated

Glo

mer

ular

Fi

ltrat

ion

Rate

< 5

0 m

l/m

in/1

,73m

2 )

Num

ber 2

7Sy

stem

ic an

tibio

tic th

erap

y sh

ould

be

switc

hed

from

intra

veno

us to

ora

l ant

ibi-

otic

ther

apy

with

in 4

8 -7

2 ho

urs o

n th

e ba

sis o

f the

clin

ical

cond

ition

and

whe

n or

al tr

eatm

ent i

s ade

quat

e. *A

dequ

ate =

1. w

hen

antib

iotic

is a

vaila

ble

oral

ly 2.

Whe

n or

al in

take

and

gastr

oint

esti-

nal a

bsor

ptio

n ar

e ade

quat

e. 3.

Ade

quat

e in

term

s of d

iagn

osis

(exc

eptio

ns li

ke en

doca

r-di

tis, m

enin

gitis

)

Num

ber o

f pat

ient

s with

intra

-ve

nous

antib

iotic

s for

48-

72h,

in

who

m ch

angi

ng to

ora

l ant

ibio

tic

ther

apy

on th

e bas

is of

clin

ical

co

nditi

on w

as d

one

Tota

l num

ber o

f pat

ient

s with

intra

ve-

nous

antib

iotic

s for

48-

72h,

in w

hom

ch

angi

ng to

ora

l ant

ibio

tic th

erap

y on

th

e bas

is of

the c

linic

al co

nditi

on w

as

indi

cate

d

Num

ber 2

8A

n an

tibio

tic p

lan

shou

ld b

e doc

umen

ted

in th

e cas

e not

es at

the s

tart

of s

yste

mic

an

tibio

tic tr

eatm

ent (

antib

iotic

pla

n is

indi

catio

n, n

ame,

dose

s, ro

ute a

nd in

terv

al of

ad

min

istra

tion)

Num

ber o

f pat

ient

s who

star

ted

with

syst

emic

antib

iotic

ther

apy

for w

hom

an an

tibio

tic p

lan

was

do

cum

ente

d in

the c

ase n

otes

Tota

l num

ber o

f pat

ient

s who

star

ted

with

syst

emic

antib

iotic

ther

apy

Num

ber 2

9Th

erap

eutic

dru

g m

onito

ring

shou

ld b

e pe

rform

ed w

hen

the t

reat

men

t dur

atio

n is

> 3

days

for a

min

ogly

cosid

es an

d >5

day

s fo

r van

com

ycin

Num

ber o

f the

se p

atie

nts w

ith at

le

ast o

ne se

rum

dru

g le

vel m

ea-

sure

men

t

Tota

l num

ber o

f pat

ient

s who

re-

ceiv

ed am

inog

lyco

sides

for >

3 d

ays

and/

or va

ncom

ycin

for >

5 d

ays

Num

ber 3

0Em

piric

al an

tibio

tic th

erap

y fo

r pre

sum

ed

bact

eria

l inf

ectio

n sh

ould

be d

iscon

tinue

d ba

sed

on th

e lac

k of

clin

ical

and/

or m

i-cr

obio

logi

cal e

vide

nce o

f inf

ectio

n. Th

e m

axim

um d

urat

ion

of em

piric

al sy

stem

ic

antib

iotic

trea

tmen

t sho

uld

be 7

day

s

Num

ber o

f pat

ient

s who

se em

pir-

ical

antib

iotic

ther

apy

was

disc

on-

tinue

d w

ithin

7 d

ays,

beca

use o

f la

ck o

f clin

ical

and/

or m

icro

bio-

logi

cal e

vide

nce o

f inf

ectio

n

Tota

l num

ber o

f pat

ient

s who

star

ted

empi

rical

syst

emic

antib

iotic

ther

apy,

but l

acke

d cl

inic

al an

d/or

mic

robi

o-lo

gica

l evi

denc

e of i

nfec

tion

Num

ber 3

1A

curr

ent l

ocal

antib

iotic

gui

delin

e sho

uld

be p

rese

nt in

the h

ospi

tal a

nd an

eval

ua-

tion

whe

ther

an u

pdat

e sho

uld

be co

nsid

-er

ed sh

ould

be d

one e

very

3 ye

ars

Num

ber 3

2Lo

cal a

ntib

iotic

gui

delin

es sh

ould

cor-

resp

ond

to th

e nat

iona

l ant

ibio

tic g

uide

-lin

es, b

ut sh

ould

dev

iate

bas

ed o

n lo

cal

resis

tanc

e patt

erns

Chapter 4

80


81

4

Discussion

To our knowledge this is the first study that used the RAND modified Delphi method to systematically develop a concise set of generic QIs defining appro-priate antibiotic use in the treatment of all bacterial infections in adult patients hospitalized at non-ICU departments. Antibiotic stewardship programmes are increasingly being implemented in hospitals to optimize antibiotic use. Most important in these programmes tends to be the provision of guidelines and instructions for prescribers, but this alone will not be sufficient to bring about change and improvement of antibiotic use.17 Our set of generic QIs provides important parameters that can be used to measure the various steps in the process of antibiotic use on patient level –as described with our QIs– along the entire antibiotic pathway. These QIs enable stewardship teams to determine for which steps along the antibiotic pathway there is room for improvement, and to set priorities for targeted improvement actions in their specific hospital. The effectiveness of these actions can, again, be measured using the QIs. In this manner a quality system can be introduced in hospitals to continuously self-monitor and improve the appropriateness of antibiotic use. Of course, our QIs can also be used by groups of hospitals for benchmarking inpatient hospital QI performance to further improve antibiotic use.The generic set contains 11 QIs describing appropriate antibiotic use, from start to discontinuation of antibiotics. All indicators received a high score in the first questionnaire round. We also asked the panel members to rank this complete set, to see if there was a hierarchy within this set of QIs. The results show that taking cultures, prescribing empirical therapy according to the guideline, and streamlining antibiotic therapy received the highest scores.

Our study has several strengths. First, the set of QIs was specifically designed for hospitalized patients. The European Surveillance of Antimicrobial Consumption (ESAC) developed QIs to measure appropriate outpatient antibiotic use in Europe.12, 18 However, this set was not designed to measure the appropriateness of antibiotic use in individual patients. Second, we used the Delphi procedure, where scientific evidence is combined with expert opinion, which is well known and described in other studies.10, 11, 19 The application of this systematic and rigorous consensus method for indicator development resulted

in indicators with high content validity. Recently 2 reviews were published on methods for developing QIs, and the use and reporting of the Delphi method. Both reviews reported a substantial variety among studies.20, 21 Boulkedid and colleagues developed practical guidelines for using the RAND modified Delphi technique and our procedure is consistent with these guidelines.20 Another strength was the multidisciplinary expert panel, which was an international panel in which all the main specialities involved in antibiotic treatment were represented. This resulted in a diversity of practices and opinions, which strengthens the results of the Delphi procedure. In addition, both the scientific literature search and the expert panel were international. We therefore believe these QIs represent a valid set that can be used internationally.

This study also has some limitations. Twenty-nine per cent of the experts attended the panel meeting. All attendees were Dutch because of logistical reasons. Nevertheless, the response rate of the first and second questionnaire was respectively 94% and 100%, which is very high. An extensive summary with regard to the results from the consensus meeting was sent to all panel members, and they were asked to give their final remarks and approval for the added and rephrased potential QIs. Because the entire panel returned the second questionnaire, we believe that an incomplete attendance did not undermine the validity of the results. Another potential limitation was that none of the QIs had “grade 1” evidence (table 1-3). This is, however, exactly the reason why we used the Delphi method, as it systematically combines evidence and consensus of experts, which enables the assessment of a broader range of topics than would otherwise be possible.

In conclusion, the applicability of QIs should always be tested in practice first, as registration of data is different in every country, which affects the feasibility, validity and reliability of data collection.11 Also, within a country, registration may vary between and sometimes even within clinical settings. We therefore strongly advise to first test the clinimetric properties of the QIs to discriminate between indicators that are feasible, valid and reliable in a specific setting and those that or not. Such a test will also facilitate acceptance of the measures. For example, at this moment we are testing the clinimetric properties of our QIs in approximately 1,800 hospitalized patients, in 22 Dutch hospitals. Similarly, the feasibility, validity and reliability of the QIs should be tested in other countries/

Chapter 4

82


83

4

states, to check whether our antibiotic stewardship QIs are also applicable and comparable internationally.

FundingThis work was supported by the Netherlands Organisation for Health Research and Development (Zon/MW) [grant number 205100003].

Conflicts of interestCaroline M.A. van den Bosch, Suzanne E. Geerlings and Marlies E.J.L. Hulscher, all received the above mentioned research grant from the Netherlands Organisation for Health Research and Development (Zon/MW) [grant number 205100003].

AcknowledgementsProf. dr. A. Tambic Andrasevic, medical microbiologist, CroatiaProf. dr. J. Rodríguez-Baño, medical microbiologist, SpainDr. M. Sundqvist, medical microbiologist, SwedenProf. dr. P.G. Davey, infectious diseases specialist, ScotlandProf. dr. L. Willems, clinical hospital pharmacist, BelgiumProf. dr. I.C. Gyssens, infectious diseases specialist, Belgium and the NetherlandsProf. dr. C.J.M. de Groot, gynaecologist, the NetherlandsDr. A. Timmermans, gynaecologist, the NetherlandsProf. dr. M.A. Boermeester, general surgeon, the NetherlandsDr. M.J.T.F.D. Vrancken Peeters, general surgeon, the NetherlandsProf. dr. M.J.M. Bonten, medical microbiologist, the NetherlandsDr. L. Spanjaard, medical microbiologist, the NetherlandsDr. P. Bresser, pulmonologist, the NetherlandsDr. W.G. Boersma, pulmonologist, the Netherlands

Reference List

(1) WHO Global Strategy for Containment of Antimicrobial Resistance. 2011.(2) Dellit TH, Owens RC, McGowan JE, Jr., et al. Infectious Diseases Society of

America and the Society for Healthcare Epidemiology of America guidelines for developing an institutional program to enhance antimicrobial stewardship. Clin Infect Dis 2007 Jan 15; 44(2):159-77.

(3) Woolf SH, Grol R, Hutchinson A, Eccles M, Grimshaw J. Clinical guidelines: potential benefits, limitations, and harms of clinical guidelines. BMJ 1999 Feb 20; 318(7182):527-30.

(4) Arnold FW, LaJoie AS, Brock GN, et al. Improving outcomes in elderly patients with community-acquired pneumonia by adhering to national guidelines: Community-Acquired Pneumonia Organization International cohort study results. Arch Intern Med 2009 Sep 14; 169(16):1515-24.

(5) Asadi L, Eurich DT, Gamble JM, Minhas-Sandhu JK, Marrie TJ, Majumdar SR. Impact of guideline-concordant antibiotics and macrolide/beta-lactam combinations in 3203 patients hospitalized with pneumonia: prospective cohort study. Clin Microbiol Infect 2013 Mar; 19(3):257-64.

(6) Larson EL, Quiros D, Giblin T, Lin S. Relationship of antimicrobial control policies and hospital and infection control characteristics to antimicrobial resistance rates. Am J Crit Care 2007 Mar; 16(2):110-20.

(7) Lee SS, Kim Y, Chung DR. Impact of discordant empirical therapy on outcome of community-acquired bacteremic acute pyelonephritis. J Infect 2011 Feb; 62(2):159-64.

(8) Menendez R, Torres A, Reyes S, et al. Initial management of pneumonia and sepsis: factors associated with improved outcome. Eur Respir J 2012 Jan; 39(1):156-62.

(9) Campbell SM, Braspenning J, Hutchinson A, Marshall MN. Research methods used in developing and applying quality indicators in primary care. BMJ 2003 Apr 12; 326(7393):816-9.

(10) Hermanides HS, Hulscher ME, Schouten JA, Prins JM, Geerlings SE. Development of quality indicators for the antibiotic treatment of complicated urinary tract infections: a first step to measure and improve care. Clin Infect Dis 2008 Mar 1; 46(5):703-11.

(11) Schouten JA, Hulscher ME, Wollersheim H, et al. Quality of antibiotic use for lower respiratory tract infections at hospitals: (how) can we measure it? Clin Infect Dis 2005 Aug 15; 41(4):450-60.

(12) Coenen S, Ferech M, Haaijer-Ruskamp FM, et al. European Surveillance of Antimicrobial Consumption (ESAC): quality indicators for outpatient antibiotic use in Europe. Qual Saf Health Care 2007 Dec; 16(6):440-5.

Chapter 4

84


85

4

(13) Campbell SM, Braspenning J, Hutchinson A, Marshall M. Research methods used in developing and applying quality indicators in primary care. Qual Saf Health Care 2002 Dec; 11(4):358-64.

(14) Fitch K, Bernstein SJ, Aguilar MS, et al. RAND/UCLA Appropriateness method user’s manual. 2001.

(15) Campbell SM, Cantrill JA, Roberts D. Prescribing indicators for UK general practice: Delphi consultation study. BMJ 2000 Aug 12; 321(7258):425-8.

(16) van den Bosch CM, Hulscher ME, Natsch S, Gyssens IC, Prins JM, Geerlings SE. Development of quality indicators for antimicrobial treatment in adults with sepsis. BMC Infect Dis 2014 Jun 20; 14(1):345.

(17) Charani E, Cooke J, Holmes A. Antibiotic stewardship programmes--what’s missing? J Antimicrob Chemother 2010 Nov; 65(11):2275-7.

(18) Adriaenssens N, Coenen S, Versporten A, Muller A, Vankerckhoven V, Goossens H. European Surveillance of Antimicrobial Consumption (ESAC): quality appraisal of antibiotic use in Europe. J Antimicrob Chemother 2011 Dec; 66 Suppl 6:vi71-vi77.

(19) van Hulst LT, Fransen J, den Broeder AA, Grol R, van Riel PL, Hulscher ME. Development of quality indicators for monitoring of the disease course in rheumatoid arthritis. Ann Rheum Dis 2009 Dec; 68(12):1805-10.

(20) Boulkedid R, Abdoul H, Loustau M, Sibony O, Alberti C. Using and reporting the Delphi method for selecting healthcare quality indicators: a systematic review. PLoS One 2011; 6(6):e20476.

(21) Kotter T, Blozik E, Scherer M. Methods for the guideline-based development of quality indicators--a systematic review. Implement Sci 2012; 7:21.

Appendix 1. Panel members

Prof. dr. A. Tambic Andrasevic, medical microbiologist, CroatiaProf. dr. J. Rodríguez-Baño, medical microbiologist, SpainDr. M. Sundqvist, medical microbiologist, SwedenProf. dr. P.G. Davey, infectious diseases specialist, ScotlandProf. dr. L. Willems, clinical hospital pharmacist, BelgiumProf. dr. I.C. Gyssens, infectious diseases specialist, Belgium and the NetherlandsProf. dr. C.J.M. de Groot, gynaecologist, the NetherlandsDr. A. Timmermans, gynaecologist, the NetherlandsProf. dr. M.A. Boermeester, general surgeon, the NetherlandsDr. M.J.T.F.D. Vrancken Peeters, general surgeon, the NetherlandsProf. dr. M.J.M. Bonten, medical microbiologist, the NetherlandsDr. L. Spanjaard, medical microbiologist, the NetherlandsDr. P. Bresser, pulmonologist, the NetherlandsDr. W.G. Boersma, pulmonologist, the NetherlandsDr. S. Natsch, clinical hospital pharmacist, the NetherlandsProf. dr. J.M. Prins, infectious diseases specialist, the NetherlandsDr. S.E. Geerlings, infectious diseases specialist, the Netherlands

Chapter 4

86


87

4

Appendix 2. Articles found in literature search

1. Ballard DJ (2003) Indicators to improve clinical quality across an integrated health care system. Int J Qual Health Care 15 Suppl 1: i13-i23.

2. Chalker J (2001) Improving antibiotic prescribing in Hai Phong Province, Viet Nam: the “antibiotic-dose” indicator. Bull World Health Organ 79: 313-320. S0042-96862001000400008 [pii].

3. Ansari F, Erntell M, Goossens H, Davey P (2009) The European surveillance of antimicrobial consumption (ESAC) point-prevalence survey of antibacterial use in 20 European hospitals in 2006. Clin Infect Dis 49: 1496-1504. 10.1086/644617 [doi].

4. Apostolopoulou E, Zaxos N, Georgoudi A, Tsakona M, Nikoloudi P (2010) Adherence with guidelines of perioperative antibiotic prophylaxis and cost among women undergoing cesarean section. Review of Clinical Pharmacology and Pharmacokinetics 24: 19-24.

5. Berenholtz SM, Dorman T, Ngo K, Pronovost PJ (2002) Qualitative review of intensive care unit quality indicators. J Crit Care 17: 1-12. S0883944102500137 [pii].

6. Berenholtz SM, Pronovost PJ, Ngo K, Barie PS, Hitt J, Kuti JL, Septimus E, Lawler N, Schilling L, Dorman T (2007) Developing quality measures for sepsis care in the ICU. Jt Comm J Qual Patient Saf 33: 559-568.

7. Capelastegui A, Espana PP, Quintana JM, Gorordo I, Ortega M, Idoiaga I, Bilbao A (2004) Improvement of process-of-care and outcomes after implementing a guideline for the management of community-acquired pneumonia: a controlled before-and-after design study. Clin Infect Dis 39: 955-963. CID33239 [pii];10.1086/423960 [doi].

8. Challenor R, Warwick Z (2011) Developing quality indicators for genitourinary medicine services with patient and public involvement: an audit of performance against proposed local measures. Int J STD AIDS 22: 604-607. 22/10/604 [pii];10.1258/ijsa.2009.009321 [doi].

9. Chu LA, Bratzler DW, Lewis RJ, Murray C, Moore L, Shook C, Weingarten SR (2003) Improving the quality of care for patients with pneumonia in very small hospitals. Arch Intern Med 163: 326-332. ioi20006 [pii].

10. Curtis C, Marriott J, Langley C (2004) Development of a prescribing indicator for objective quantification of antibiotic usage in secondary care. J Antimicrob Chemother 54: 529-533. 10.1093/jac/dkh362 [doi];dkh362 [pii].

11. Dellinger RP, Carlet JM, Masur H, Gerlach H, Calandra T, Cohen J, Gea-Banacloche J, Keh D, Marshall JC, Parker MM, Ramsay G, Zimmerman JL, Vincent JL, Levy MM (2004) Surviving Sepsis Campaign guidelines for management of severe sepsis and septic shock. Crit Care Med 32: 858-873. 00003246-200403000-00038 [pii].

12. Hermanides HS, Hulscher ME, Schouten JA, Prins JM, Geerlings SE (2008) Development of quality indicators for the antibiotic treatment of complicated urinary tract infections: a first step to measure and improve care. Clin Infect Dis 46: 703-711. 10.1086/527384 [doi].

13. Mandy B, Koutny E, Cornette C, Woronoff-Lemsi MC, Talon D (2004) Methodological validation of monitoring indicators of antibiotics use in hospitals. Pharm World Sci 26: 90-95.

14. Morris AC, Hay AW, Swann DG, Everingham K, McCulloch C, McNulty J, Brooks O, Laurenson IF, Cook B, Walsh TS (2011) Reducing ventilator-associated pneumonia in intensive care: impact of implementing a care bundle. Crit Care Med 39: 2218-2224. 10.1097/CCM.0b013e3182227d52 [doi].

15. Mylotte JM, Weislo P (2000) Antibiotic use and cost indicators at a rural hospital: a pilot project. Am J Infect Control 28: 415-420. S0196-6553(00)36410-0 [pii];10.1067/mic.2000.109910 [doi].

16. Nathwani D, Rubinstein E, Barlow G, Davey P (2001) Do guidelines for community-acquired pneumonia improve the cost-effectiveness of hospital care? Clin Infect Dis 32: 728-741. CID000552 [pii];10.1086/319216 [doi].

17. Nguyen HB, Corbett SW, Steele R, Banta J, Clark RT, Hayes SR, Edwards J, Cho TW, Wittlake WA (2007) Implementation of a bundle of quality indicators for the early management of severe sepsis and septic shock is associated with decreased mortality. Crit Care Med 35: 1105-1112. 10.1097/01.CCM.0000259463.33848.3D [doi].

18. Pestana D, Espinosa E, Sanguesa-Molina JR, Ramos R, Perez-Fernandez E, Duque M, Martinez-Casanova E (2010) Compliance with a sepsis bundle and its effect on intensive care unit mortality in surgical septic shock patients. J Trauma 69: 1282-1287. 10.1097/TA.0b013e3181c4539f [doi].

19. Poulton B (2006) Advances in the management of sepsis: the randomised controlled trials behind the Surviving Sepsis Campaign recommendations. Int J Antimicrob Agents 27: 97-101. S0924-8579(05)00327-4 [pii];10.1016/j.ijantimicag.2005.11.003 [doi].

20. Pulcini C, Defres S, Aggarwal I, Nathwani D, Davey P (2008) Design of a ‘day 3 bundle’ to improve the reassessment of inpatient empirical antibiotic prescriptions. J Antimicrob Chemother 61: 1384-1388. dkn113 [pii];10.1093/jac/dkn113 [doi].

21. Ramirez JA (2005) Worldwide perspective of the quality of care provided to hospitalized patients with community-acquired pneumonia: results from the CAPO international cohort study. Semin Respir Crit Care Med 26: 543-552. 10.1055/s-2005-925521 [doi].

22. Rhew DC (2001) Quality indicators for the management of pneumonia in vulnerable elders. Ann Intern Med 135: 736-743. 200110161-00013 [pii].

Chapter 4

88

23. Schouten JA, Hulscher ME, Wollersheim H, Braspennning J, Kullberg BJ, van der Meer JW, Grol RP (2005) Quality of antibiotic use for lower respiratory tract infections at hospitals: (how) can we measure it? Clin Infect Dis 41: 450-460. CID35866 [pii];10.1086/431983 [doi].

24. Sheddon J, Nathwani D, Patton A, Malcolm W, Watson E (2011) Development of national prescribing indicators for antimicrobials to support reduction in Clostridium difficile infection. Clin Microbiol Infect Conference: 21st ECCMID/27th ICC Milan Italy 17: S7.

25. Shorr AF, Owens RC, Jr. (2009) Guidelines and quality for community-acquired pneumonia: measures from the Joint Commission and the Centers for Medicare and Medicaid Services. Am J Health Syst Pharm 66: S2-S7. 66/12_Supplement_4/S2 [pii];10.2146/090087a [doi].

26. Toth NR, Chambers RM, Davis SL (2010) Implementation of a care bundle for antimicrobial stewardship. Am J Health Syst Pharm 67: 746-749. 67/9/746 [pii];10.2146/ajhp090259 [doi].

27. Zarb P, Ansari F, Muller A, Vankerckhoven V, Davey PG, Goossens H (2011) Drug utilization 75% (DU75%) in 17 European hospitals (2000-2005): results from the ESAC-2 Hospital Care Sub Project. Curr Clin Pharmacol 6: 62-70. BSP/CCP/E-Pub/0032 [pii].

28. Zarb P, Goossens H (2011) European Surveillance of Antimicrobial Consumption (ESAC): Value of a Point-Prevalence Survey of Antimicrobial Use Across Europe. Drugs 71: 745-755. 5 [pii];10.2165/11591180-000000000-00000 [doi].

29. Zarb P, Amadeo B, Muller A, Drapier N, Vankerckhoven V, Davey P, Goossens H (2011) Identification of targets for quality improvement in antimicrobial prescribing: the web-based ESAC Point Prevalence Survey 2009. J Antimicrob Chemother 66: 443-449. dkq430 [pii];10.1093/jac/dkq430 [doi].

Added: van den Bosch CM, Hulscher ME, Natsch S, Gyssens IC, Prins JM, Geerlings SE

(2014) Development of quality indicators for antimicrobial treatment in adults with sepsis. BMC Infect Dis. 2014 Jun 20; 14:345

Chapter 5Applicability of Generic Quality Indicators for


Caroline M.A. van den Bosch1, Marlies E.J.L. Hulscher2, Stephanie Natsch3, Jan Wille4, Jan M. Prins1 and Suzanne E. Geerlings1

1Department of Internal Medicine, Division of Infectious Diseases, Academic Medical Centre, University of Amsterdam, Amsterdam, the Netherlands,

2Department of Scientific Institute for Quality of Healthcare (IQ healthcare), Radboud University Medical Center, Nijmegen, the Netherlands, 3Department of Pharmacy, Radboud University Medical Centre, Nijmegen, the Netherlands,

4Department of Centre for Infectious Diseases Epidemiology and Surveillance, The National Institute for Public Health and the Environment (RIVM), Bilthoven, The

Netherlands

Submitted

Chapter 5

90

Applicability of generic indicators

91

5

Abstract

ObjectiveThe ability to monitor the appropriateness of hospital antibiotic use is a key element of an effective antibiotic stewardship program. A set of 11 generic quality indicators (QIs) was previously developed to assess the quality of antibiotic use in hospitalized adults treated for a bacterial infection. The primary aim of the current study was to assess the clinimetric properties of these QIs (nine process and two structure indicators) in daily clinical practice.

Methods QIs were developed following a RAND-modified Delphi procedure. An observational multicenter study included 1890 in patients treated with antibiotics for a suspected bacterial infection. In this cohort we tested the measurability, applicability, reliability, room for improvement and case mix stability of these QIs.

Results Low applicability (≤ 10%) was found for the QIs ‘therapeutic drug monitoring’, ‘adapting antibiotics to renal function’ and ‘discontinue empirical therapy in case of lack of clinical and/or microbiological evidence of infection’. For the latter, we also found a low inter-observer agreement (kappa < 0.4). One QI showed low improvement potential. The remaining seven QIs had sound clinimetric properties. Case-mix correction was necessary for most process QIs. For all QIs, we found ample room for improvement and large variation between hospitals.

Conclusion Establishing the clinimetric properties was essential, as four of the eleven previously selected QIs showed unsatisfactory properties in this practice test. Since the quality of antibiotic use and the process of documenting data is changing over time and may vary per country, QIs should always be tested in practice first.

Introduction

Around 30-40% of patients do not receive care based on available scientific evidence.1;2 For antibiotic treatment this means unnecessary or inaccurate use of antibiotics, which can have negative consequences for patient outcome and the development of antimicrobial resistance.3;4 To curb the development of antibiotic resistance, better use of current agents and a decrease of inappropriate antibiotic use are necessary.5 Studies show that antibiotic stewardship programs can improve antibiotic use and reduce healthcare costs without negatively influencing the quality of care provided.4 One of the key elements of an effective stewardship program is its ability to monitor the quality of hospital antibiotic use, with the aim to set priorities and focus improvement. Monitoring the appropriateness of hospital antibiotic use can be accomplished using quality indicators (QIs).6 Quality indicators (QIs) function as measurable elements for which there is evidence or consensus that they can be used to assess the appropriateness of daily antibiotic care provided.6;7 QIs can be developed using recommendations and already available QIs that are systematically extracted from international guidelines and literature.6-9 After developing QIs, assessing their feasibility in daily practice is essential before using them to measure appropriateness of antibiotic use. This can be done by testing important clinimetric characteristics, like inter-observer reliability.8;9

QIs measuring antibiotic use in hospitals have been described in literature before, but for none of these measures the clinimetric properties were tested in practice.10-16 We reported previously on the systematic development, using the RAND-modified Delphi procedure, of generic indicators for measuring the appropriateness of antibiotic use in hospitalized adults treated for a suspected bacterial infection, see table 1.17 The primary aim of this current study was to assess the clinimetric properties of these generic QIs in a large number of inpatients in daily clinical practice.

Chapter 5

92


93

5

Tabl

e 1.

List

of g

ener

ic qu

ality

indi

cato

rs to

mon

itor

antib

iotic

use

for

all b

acte

rial i

nfec

tions

in h

ospi

taliz

ed a

dult

patie

nts o

n no

n-IC

U

depa

rtm

ents.

Num

-be

r Q

ualit

y in

dica

tor

All

patie

nts a

re: a

dults

, ad

mitt

ed o

n a

non-

ICU

de

part

men

t, us

ing

syst

emic

an

tibio

tics f

or >

24h

be-

caus

e of

a su

spec

ted

bact

e-ri

al in

fect

ion

Num

erat

or d

escr

iptio

n D

enom

inat

or d

escr

iptio

nC

omm

ent

1Em

piric

al sy

stem

ic an

tibio

tic

ther

apy

shou

ld b

e pre

scrib

ed

acco

rdin

g to

the n

atio

nal

guid

elin

e

Num

ber o

f pat

ient

s who

st

arte

d w

ith em

piric

al sy

s-te

mic

antib

iotic

ther

apy

acco

rdin

g to

the n

atio

nal

guid

elin

e

Tota

l num

ber o

f pat

ient

s who

st

arte

d w

ith em

piric

al sy

s-te

mic

antib

iotic

ther

apy

2Be

fore

star

ting

syst

emic

an-

tibio

tic th

erap

y at

leas

t tw

o se

ts o

f blo

od cu

lture

s sho

uld

be ta

ken

Num

ber o

f pat

ient

s in

who

m

at le

ast 2

sets

of b

lood

cul-

ture

s wer

e tak

en b

efor

e sys

-te

mic

antib

iotic

ther

apy

was

st

arte

d

Tota

l num

ber o

f pat

ient

s who

st

arte

d w

ith sy

stem

ic an

tibi-

otic

ther

apy

Bloo

d cu

lture

s sho

uld

be

perfo

rmed

bet

wee

n on

e wee

k be

fore

star

t of t

reat

men

t and

ac

tual

star

t of t

reat

men

t

3W

hen

star

ting

syst

emic

anti-

biot

ic th

erap

y sp

ecim

ens f

or

cultu

re fr

om su

spec

ted

sites

of

infe

ctio

n sh

ould

be t

aken

as

soon

as p

ossib

le, p

refe

rabl

y be

fore

antib

iotic

s are

star

ted.

Num

ber o

f pat

ient

s in

who

m

cultu

res f

rom

susp

ecte

d sit

es

of in

fect

ion

wer

e tak

en w

ith-

in 2

4 ho

urs a

fter t

he sy

stem

ic

antib

iotic

s wer

e sta

rted

Tota

l num

ber o

f pat

ient

s who

st

arte

d w

ith sy

stem

ic an

tibi-

otic

ther

apy

Cul

ture

s sho

uld

be ta

ken

betw

een

two

wee

ks b

efor

e st

art o

f tre

atm

ent a

nd o

ne

day

after

star

t of t

reat

men

t

4A

n an

tibio

tic p

lan

shou

ld b

e do

cum

ente

d in

the c

ase n

otes

at

the s

tart

of s

yste

mic

antib

i-ot

ic th

erap

y

Num

ber o

f pat

ient

s for

w

hom

an an

tibio

tic p

lan

was

do

cum

ente

d in

the c

ase n

otes

Tota

l num

ber o

f pat

ient

s who

st

arte

d w

ith sy

stem

ic an

tibi-

otic

ther

apy

Ant

ibio

tic p

lan

incl

udes

in-

dica

tion,

and

nam

e, do

sage

, ro

ute a

nd in

terv

al o

f adm

inis-

tratio

n of

the a

ntib

iotic

5Sy

stem

ic an

tibio

tic th

erap

y sh

ould

be s

witc

hed

from

in

trave

nous

to o

ral a

ntib

iotic

th

erap

y w

ithin

48

-72

hour

s on

the b

asis

of th

e clin

ical

co

nditi

on an

d w

hen

oral

tre

atm

ent i

s ade

quat

e

Num

ber o

f pat

ient

s with

in

trave

nous

antib

iotic

s for

48

-72h

, in

who

m ch

angi

ng to

or

al an

tibio

tic th

erap

y on

the

basis

of c

linic

al co

nditi

ons

was

don

e

Tota

l num

ber o

f pat

ient

s with

in

trave

nous

antib

iotic

s for

48

-72h

, in

who

m ch

angi

ng to

or

al an

tibio

tic th

erap

y on

the

basis

of t

he cl

inic

al co

nditi

on

was

indi

cate

d

Defi

nitio

ns: s

ee B

ox 1

6Em

piric

al an

tibio

tic th

erap

y sh

ould

be c

hang

ed to

pat

ho-

gen-

dire

cted

ther

apy

if cu

l-tu

re re

sults

bec

ome a

vaila

ble

Num

ber o

f pat

ient

s with

em-

piric

al th

erap

y w

hose

cultu

re

beca

me p

ositi

ve an

d ch

ang-

ing

to p

atho

gen-

dire

cted

th

erap

y w

as d

one c

orre

ctly

Tota

l num

ber o

f pat

ient

s with

em

piric

al sy

stem

ic an

tibio

t-ic

s, w

hose

cultu

re b

ecam

e po

sitiv

e.

Ther

apy

is pa

thog

en-d

irect

ed

if it

is in

acco

rdan

ce w

ith

resis

tanc

e patt

ern.

If p

ossib

le,

antib

iotic

s sho

uld

be ch

osen

fro

m th

e gro

up o

f nar

row

sp

ectr

um an

tibio

tics (

Box 2

)7

Dos

e and

dos

ing

inte

rval

of

syst

emic

antib

iotic

ther

apy

shou

ld b

e ada

pted

to re

nal

func

tion

Num

ber o

f pat

ient

s with

a co

mpr

omise

d re

nal f

unct

ion

with

a do

sing

regi

men

adju

st-

ed to

rena

l fun

ctio

n.

Tota

l num

ber o

f pat

ient

s who

st

arte

d w

ith sy

stem

ic an

ti-bi

otic

ther

apy

whi

ch sh

ould

be

dos

ed ac

cord

ing

to re

nal

func

tion,

and

who

had

a un

-kn

own

or co

mpr

omise

d re

nal

func

tion

Com

prom

ised

rena

l fun

c-tio

n de

fined

as an

estim

ated

G

lom

erul

ar F

iltra

tion

Rat

e <

50m

l/m

in/1

.73m

2

Chapter 5

94


95

5

8Th

erap

eutic

dru

g m

onito

ring

shou

ld b

e per

form

ed w

hen

the t

hera

py d

urat

ion

is >3

da

ys fo

r am

inog

lyco

sides

and

>5 d

ays f

or va

ncom

ycin

Num

ber o

f the

se p

atie

nts

with

at le

ast o

ne se

rum

dru

g le

vel m

easu

rem

ent

Tota

l num

ber o

f pat

ient

s who

re

ceiv

ed am

inog

lyco

sides

for

>3 d

ays a

nd/o

r van

com

ycin

fo

r >5

days

9Em

piric

al an

tibio

tic th

erap

y fo

r pre

sum

ed b

acte

rial i

nfec

-tio

n sh

ould

be d

iscon

tinue

d ba

sed

on th

e lac

k of

clin

ical

an

d/or

mic

robi

olog

ical

evi-

denc

e of i

nfec

tion.

The m

ax-

imum

dur

atio

n of

empi

rical

sy

stem

ic an

tibio

tic tr

eatm

ent

shou

ld b

e 7 d

ays

Num

ber o

f pat

ient

s who

se

empi

rical

antib

iotic

ther

apy

was

disc

ontin

ued

with

in 7

da

ys

Tota

l num

ber o

f pat

ient

s who

st

arte

d em

piric

al sy

stem

ic

antib

iotic

ther

apy,

but l

acke

d cl

inic

al an

d/or

mic

robi

olog

i-ca

l evi

denc

e of i

nfec

tion

Defi

nitio

ns:

see B

ox 3

10A

curr

ent l

ocal

antib

iotic

gu

idel

ine s

houl

d be

pre

sent

in

the h

ospi

tal a

nd an

eval

-ua

tion

whe

ther

an u

pdat

e sh

ould

be c

onsid

ered

shou

ld

be d

one e

very

3 ye

ars

11Lo

cal a

ntib

iotic

gui

delin

es

shou

ld co

rres

pond

to th

e na

tiona

l ant

ibio

tic g

uide

lines

, bu

t sho

uld

devi

ate b

ased

on

loca

l res

istan

ce p

atter

ns

Methods

Setting, study population and data collectionThis was an observational multicenter study where we measured the applicability of the developed QIs in four university and 18 non-university (medium-sized and small) hospitals. To include a representative patient group, we identified in each hospital on non-Intensive Care Unit (ICU) departments, all adult inpatients using antibiotics for >24 hours for a suspected hospital- or community-acquired bacterial infection on the day of a point prevalence measurement. This point prevalence measurement was performed biannually by PREZIES (prevention of nosocomial infections by surveillance), a department of the RIVM (National Institute for Public Health and the Environment, the Netherlands). Patients were included in October 2011, March 2012 and October 2012. Of all included patients clinical and laboratory data of the entire antibiotic pathway, from start of antibiotics till discharge, were extracted retrospectively from medical and nursing records, and medication charts. Patients were excluded if antibiotics were used as prophylaxis, or when antibiotic treatment was started on the ICU department or in another hospital, see figure 1. All data needed to compute the QIs or to assess potential determinants needed for case mix correction (see below) were extracted by trained research nurses and the study coordinator in a uniform way and entered in the database anonymously. The ethics committee assessed the study and concluded that it was deemed exempt from their approval.

Quality indicators and definitions The set of 11 QIs for appropriate antibiotic use of all bacterial infections in hospitalized adult patients admitted on a non-ICU department (table 1) were developed using the systematic RAND modified Delphi method.17 QIs one to nine are process indicators, measuring the process around appropriate antibiotic use at the patient level. QIs 10 and 11 are structure indicators, which measure requirements for appropriate antibiotic use at the hospital level. The QIs with their numerators and denominators are described in detail in Table 1. For some of the QIs, we had to develop more extensive working definitions:QI.1 Prescribe empirical antibiotic therapy according to the national guideline. We defined empirical therapy as the antibiotic (combination) prescribed on the day of diagnosis, before identification of a causative pathogen. If

Chapter 5

96


97

5

initial therapy was directed at a previously cultured pathogen, this was not called ‘empirical therapy’ and these patients were excluded for this indicator. Prescribing in accordance with the guideline was evaluated in relation to the at that moment current Dutch national guideline, http://www.swab.nl/guidelines. Individual patient characteristics, in particular allergies, pregnancy, and previous extended-spectrum β-lactamase (ESBL) infection were taken into account when computing the algorithms for appropriate empirical therapy.

Figure 1. Patients Identified During the Point Prevalence Measurements

116

Figure 1. Patients Identified During the Point Prevalence Measurements

9395 patients

No antibiotics n = 5675 (68%)

Antibiotics < 24 hours n = 218 (3%)

Adult patients on non-ICU departments n = 8314 (88%)

Children (< 18 years old) n = 733 (8%)

Patients on the Intensive Care Unit (ICU) n = 348 (4%)

Antibiotic prophylaxis or other reasons n = 298 (12%)

Antibiotic started on ICU department n = 109 (5%)

Antibiotic started at another hospital n = 69 (3%)

Medical chart could not be located n= 55 (2%)

Antibiotic started for suspected bacterial infection n = 1890 (78%)

Included in the study

Antibiotics > 24 hours n = 2421 (29%)

Only patients with one or two (possible) diagnoses covered by a guideline were included. Patients with more than two (possible) infections or with a (possible) infection not covered by a guideline were excluded. In case two diagnosed or possible infections affected a single organ system (e.g. upper and lower respiratory tract infection (LRTI)), it usually concerned a differential diagnosis, and empiric antibiotic therapy was considered correct if covered by one of the two applicable guidelines. If the two (possible) infections affected different organ systems (e.g. LRTI and urinary tract infection), antibiotic therapy was only considered correct if both diagnoses were covered according to their respective guidelines. Two infectious diseases specialists assessed the algorithms regarding guideline adherent antibiotic therapy for combinations of diagnoses. QI.5 Antibiotic therapy should be switched from intravenous to oral therapy within 48 – 72 hours.This QI was applicable to the subgroup of patients who started with intravenous (iv) antibiotic treatment and fulfilled the criteria for safe early switch, see Box 1.18 Clinical stability at 48 – 72 hours of antibiotic treatment was also considered to have been present when temperature or blood pressure was not mentioned in the medical records or white blood cells (WBC) had not been determined, because we assumed that if these parameters were abnormal, they would have been documented. QI.6 Empirical antibiotic therapy should be changed to pathogen-directed therapy if culture results become available.This QI applied to patients with a positive culture (denominator), and it measured the proportion of patients (numerator) that received pathogen-directed antibiotic treatment after culture results became positive. We estimated conservatively that in general at day 5 culture results are available. Therefore, the antibiotic (combination) given at day 5 was considered the final therapy. This therapy was considered pathogen-directed when it was in accordance with the resistance pattern of the cultured microorganism, regardless whether antibiotic therapy was changed or not. If possible, antibiotics should be chosen from a group of recommended (‘more’ narrow) antibiotics (Box 2).10 QI.7 Dose and dosing interval of antibiotic therapy should be adapted to renal function.Creatinine clearance was measured using the Modification of Diet in Renal Disease (MDRD) formula. Antibiotics were grouped according to necessity to

Chapter 5

98


99

5

Box 1.

QI 5 Criteria for safe switch within 2-3 days18

1. Hemodynamic stable 2. No fever3. Normal WBC count (4 – 12 x109/L) 4. Patient must be able to take oral medication5. Patient must have a functioning gastrointestinal tract, without signs of malab-

sorption6. Safe switch not possible in case of: meningitis, intracranial abscess, endocarditis,

mediastinitis, Legionella pneumonia and exacerbations of cystic fibrosis, inade-quately drained abscess and empyema, severe soft tissue infection such as group A streptococcal infections, infection of foreign bodies, Staphylococcus aureus or Pseudomonas aeruginosa bacteraemia, liver abscess and empyema, osteomyelitis and arthritis, chemotherapy related neutropenia.

Box 2.

QI 6 Recommended (more narrow) antibiotics10 Penicillin CiprofloxacinPhenethicillin NorfloxacinAmoxicillin OfloxacinFlucloxacillin DoxycyclineCephalexin TetracyclineCephalothin MinocyclineCefazolin Nitrofurantoin Erythromycin FosfomycinCo-trimoxazole MetronidazoleTrimethoprim

Box 3.

QI 9 Criteria for discontinuing empirical antibiotic therapy1. Antibiotic therapy was started empirical2. Blood cultures and cultures from suspected site of infection have been per-

formed3. All cultures remained negative4. Haemodynamic stable 2 – 3 days after start treatment5. WBC count normal (4 – 12 x109/L) 2 – 3 days after start treatment6. No fever at day 6 or 77. No antibiotics in the 30 days before start of empirical treatment

adapt the dosing regimen in case of a compromised renal function: no adaption to renal function necessary, or adaption necessary with estimated Glomerular Filtration Rate (eGFR) 30-50 mL/min/1.72m2, 10-30 mL/min/1.72m2, or <10 mL/min/1.72m2. In patients using an antibiotic that should be dosed according to renal function, the plasma creatinine should therefore be checked sometime between 3 days before and 2 days after the start of antibiotics. QI.11 Local antibiotic guidelines should correspond to the national antibiotic guidelines, but should deviate based on local resistance patterns.We compared the local guidelines with the national guidelines in the Netherlands (http://www.swab.nl/guidelines). Since between various regions in the Netherlands minimal differences in local resistance rates exist, deviation based on local resistance patterns is in our setting not an issue.19 For the same reason we evaluated in QI.1 empirical antibiotic therapy according to the national guidelines, instead of the local guideline.9

Assessment of the clinimetric properties of the Quality indicators The denominator and numerator of the QIs were translated into computerized algorithms in SPSS syntax, in which data from clinical patient records were entered, to determine for every QI whether a patient qualified for the denominator and to estimate the numerator. Measurability is defined as the availability of administrative data required to evaluate the indicator.8;9;20 An indicator was considered measurable if data necessary to score the indicator could be abstracted from the available data for >75% of the cases.21

Applicability. A QI should be applicable to at least 10% of the reviewed patient records (with a minimum of 10 patients per hospital).22 Inter-observer reliability (only estimated in process QIs). A second investigator rated 50 records from different departments in four medical centres (two university- and two non-university hospitals, two with and two without Electronic Patient Record System [EPRS]). The percentage of agreement on the level of indicator outcome was calculated and expressed in Kappa (k) coefficients.6;23 Scores of k >0·6 were considered to be good, k values <0·4 unreliable. Potential room for improvement measures the sensitivity of a potential indicator to detect variability in quality of care between and within hospitals. It is expressed as 100% minus the performance score, with performance

Chapter 5

100


101

5

expressing the percentage of adherence to an indicator.21 High QI scores with little variation between hospitals make indicators less sensitive and therefore less useful in daily practice, so the improvement potential was considered ‘low’ if the performance scores of all hospitals were >85%. Case mix stability (only estimated in process QIs). For indicators that showed sound clinimetric properties, case-mix stability was analysed, to determine whether an indicator can monitor quality in a specific hospital over time, and to compare hospitals of different sizes and settings. The effects on the indicator scores of the following potential determinants were studied: sex, age, community- versus hospital-acquired infection, sepsis (SIRS criteria ≥2) versus no sepsis,24 severity of illness (MEWS score),25;26 co morbidity (Charlson Co morbidity Index),27 prior use of antibiotics and iv or oral start of antibiotics. The need for case-mix correction did not lead to elimination of the indicator.

Correlation between performance scores of the Quality indicators We finally examined the relationships between the performance scores of the process QIs, with the ultimate aim of reducing the number of QIs, and therefore the required data collection necessary to assess ‘appropriate antibiotic use’. A correlation coefficient > 0·4 was considered to represent a moderate correlation between QIs.

AnalysisCollected data were entered in a database using the Statistical Package for the Social Sciences (SPSS 20·0 for Windows®, SPSS Inc., Chicago, IL, USA). Descriptive analyses were performed for each indicator. Performance was expressed as percentage of adherence to an indicator. Case-mix stability was analysed by a multilevel (mixed model) analysis. With these analyses we take into account the hierarchical structure: patients nested within hospitals. We performed a mixed model with a random intercept and all other variables fixed.8;9 For examining the relationship between the process QIs, we used the bivariate Spearman correlation coefficient.

Role of the funding sourceThe design and conduct of the study, the collection, analysis and interpretation of the data were funded by the Netherlands Organisation for Health Research and Development (ZonMW) [grant number 205100003].

The funders had no role in the preparation, review, or approval of the manuscript, and decision to submit the manuscript for publication.

Results

PatientsThe QIs were tested in a sample of 1890 adult patients, admitted to non-ICU departments of 22 hospitals in the Netherlands and treated for a suspected bacterial infection. Of the 1890 patients, 51·5% were men, and 73·6% of the suspected bacterial infections were community-acquired. Further baseline characteristics are shown in table 2.

Clinimetric properties of the QIs Measurability: Some indicators had more missing values than others, but for all QIs measurability was good (Table 3).Applicability: QI.7, adapting antibiotics to renal function, applied to only 8·7% of patients, QI.8, on Therapeutic Drug Monitoring, applied to 5·8% of patients and QI.9, regarding discontinuing empirical antibiotic therapy, applied to 8·6%. Inter-observer reliability: QI.9, discontinuing empirical antibiotic therapy, had a kappa of 0·24, indicating very low inter-observer reliability, partly because the impossibility to design a good algorithm for ‘lack of clinical evidence of infection’ left it subject to personal interpretation. All other indicators scored kappa’s >0·6, suggesting good reliability. Potential room for improvement: Structure QI.10 states that a current local guide should be present in the hospital. Ninety-five percent of the hospitals had a local, recent guide, which leaves little room for improvement. Case-mix stability: Multilevel analysis showed that in five of the six QIs one or more determinants significantly influenced indicator scores. The most important variables that influenced indicator scores were hospital- versus community-acquired infection, co morbidity, severity of illness and start of i.v. versus oral antibiotic treatment. The required corrections per QI, and all other clinimetric properties are shown in Table 3.

Chapter 5

102


103

5

Correlation between QIs The highest correlation coefficient, 0·17, was found between QI.5 (IV-oral switching) and QI.6 (streamlining) (data not shown). So, no strong correlations between QIs were found and reducing the number of QIs is therefore not possible.

Table 2. Baseline characteristics of the patients (n = 1890)

Sex Male 974 (51.5)* Female 916 (48.5)Age, mean (SD) 69·8 (17)Infection Community-acquired 1391 (74) Hospital-acquired 499 (26)Diagnosis Lower respiratory tract infection 311 (17) Urinary tract infection Skin and soft tissue infection

265 (14)177 (9)

Intra-abdominal infection Upper respiratory tract infection

174 (9)152 (8)

Sepsis (mostly with no obvious focus of infection)

53 (3)

2 diagnoses possible at start of antibiotic treatment 275 (15) Other 483 (25)SIRS criteria24 ≥ 2 (sepsis) 774 (41) < 2 (no sepsis) 1116 (59)Antibiotics Started intravenously 1399 (74)† Started orally 489 (26)Received antibiotics before start of (empirical) treatment No 1180 (62) Yes 710 (38)

*Numbers are n(%) unless otherwise indicated† In 2 patients data on route of administration was missing

Tabl

e 3.

Clin

imet

ric p

rope

rties

of th

e qua

lity i

ndica

tors

IND

ICAT

OR

Mea

sura

bil-

ity;

mis

sing

s %

(n)

App

li-ca

bilit

y %

(n)

Inte

robs

erve

r re

liabi

lity

(Kap

pa)

Impr

ovem

ent

pote

ntia

l (%

) (1

00-p

erfo

rman

ce)

Cas

e m

ix

stab

ility

1.Pr

escr

ibe

empi

rica

l ant

ibio

tic th

erap

y ac

cord

ing

to g

uide

line

1·1

(20)

73

(136

1)0·

75

59N

oA, B

, C

2.B

efor

e st

artin

g an

tibio

tic th

erap

y tw

o se

ts o

f blo

od c

ultu

res

shou

ld b

e ta

ken

010

0 (1

890)

0·74

64N

oB,D

,E,F

3.W

hen

star

ting

antib

iotic

ther

apy

cultu

res s

houl

d be

take

n fr

om su

spec

ted

site

s of i

nfec

tion

0·9

(17)

65

(1

217)

0·74

51

NoA

,C,D

4.A

n an

tibio

tic p

lan

shou

ld b

e do

cum

ente

d in

the

case

not

es a

t th

e st

art o

f ant

ibio

tic th

erap

y0

100

(189

0)0·

7739

NoC

,D,F

5.A

ntib

iotic

ther

apy

shou

ld b

e sw

itche

d fr

om in

trav

enou

s to

oral

ther

apy

with

in 4

8 –

72 h

ours

1 (1

9)23

(4

22)

0·90

68

NoC

6.Em

piri

cal a

ntib

iotic

ther

apy

shou

ld b

e ch

ange

d to

pat

ho-

gen-

dire

cted

ther

apy

if cu

lture

resu

lts b

ecom

e av

aila

ble

0·2

(3)

24

(453

)0·

94

50Ye

s

7.D

ose a

nd d

osin

g in

terv

al o

f ant

ibio

tic th

erap

y sh

ould

be a

dapt

ed

to re

nal f

unct

ion

09

(165

)0·

87

62-

8.Ther

apeu

tic d

rug

mon

itorin

g sh

ould

be p

erfo

rmed

whe

n th

e th

erap

y du

ratio

n is

long

er th

an 3

day

s for

amin

ogly

cosid

es an

d 5

days

for v

anco

myc

in

0·1

(2)

6 (1

12)

0·79

21

-

9.Em

piric

al th

erap

y sh

ould

be d

iscon

tinue

d in

case

of l

ack

of cl

ini-

cal a

nd/o

r mic

robi

olog

ical

evid

ence

of i

nfec

tion.

Max

imum

dur

a-tio

n of

ther

apy

is 7

days

0·2

(3)

9 (1

62)

0·24

67

-

10.A

curr

ent l

ocal

antib

iotic

gui

delin

e sho

uld

be p

rese

nt in

the

hosp

ital,

and

an u

pdat

e sho

uld

be d

one e

very

3 ye

ars

14 (3

)10

0 (1

9)-

5-

11.L

ocal

ant

ibio

tic g

uide

lines

shou

ld co

rres

pond

to th

e na

-tio

nal a

ntib

iotic

gui

delin

es, b

ut sh

ould

dev

iate

bas

ed o

n lo

cal

resi

stan

ce p

atter

ns

9 (2

)10

0 (2

0)-

100

-

A Cor

rect

ion

requ

ired

for (

A) ag

e, (B

) com

orbi

dity

(Cha

rlson

Inde

x), (

C) h

ospi

tal-

vs. c

omm

unity

-acq

uire

d in

fect

ion,

(D) s

ever

ity o

f ill

ness

(MEW

S sc

ore)

, (E)

seps

is ve

rsus

no

seps

is, (F

) sta

rt an

tibio

tic tr

eatm

ent I

V ve

rsus

ora

lly.

Chapter 5

104


105

5

Table 4. Performance of the validated QIs in the 22 Dutch hospitals.

INDICATOR n Performance % (range)

1.Prescribe empirical antibiotic therapy according to guideline

563 41 (24 – 58)

2.Before starting antibiotic therapy two sets of blood cultures should be taken

674 36 (9 – 59)

3.When starting antibiotic therapy cultures should be taken from suspected sites of infection

595 49 (33 – 73)

4.An antibiotic plan should be documented in the case notes at the start of antibiotic therapy

1145 61 (23 – 98)

5.Antibiotic therapy should be switched from intrave-nous to oral therapy within 48 – 72 hours

134 32 (5 – 50)

6.Empirical antibiotic therapy should be changed to pathogen-directed therapy if culture results become available

228 50 (21 – 85)

11.Local antibiotic guidelines should correspond to the national antibiotic guidelines, but should deviate based on local resistance patterns

0 0 (63 – 94*)

*not one hospital had local guidelines that corresponded to the national guidelines completely, so performance was 0%, but the % of overlap between the local and the national guidelines ranged from 63 – 94%, with a mean of 80%.

Performance scoresThe mean performance scores for each QI are shown in table 4. The highest score (61%) was found for QI.4 (antibiotic plan should be documented). QI.11 (local guidelines should correspond to the national guidelines) had the lowest score: 0%. The mean performance score for all QIs together was 38%. For each QI we established a large variation between hospitals.

Discussion

We demonstrated the importance of testing the clinimetric properties of previously developed QIs before using them,17 for example in antibiotic stewardship programs. In this practice test in 22 hospitals in the Netherlands, seven of the eleven generic QIs showed sound clinimetric properties. We also demonstrated that there was no relevant correlation between the process QIs, precluding a reduction in total number of QIs and data required. This study also showed low adherence to the QIs, with a wide performance range across hospitals. Four indicators scored below 50% adherence. This means a large room for improvement concerning the quality of antibiotic treatment in Dutch hospitals. Although some QIs showed no sound clinimetric properties, this does not imply that these QIs are not important when assessing appropriate antibiotic use. However, they should not be used as generic QIs, but for instance in targeted audits or in specific settings (QI.7 renal function and QI.8 TDM) or after being redefined (QI.9 discontinue empiric therapy).

Our study has several strengths. To our knowledge, this is the first study to develop and test the clinimetric properties of a set of generic QIs describing appropriate antibiotic treatment for all bacterial infections in the hospital, along the entire antibiotic pathway from start to discontinuing antibiotics, on an individual patient level. Previously other studies have developed and used QIs for antibiotic use in the hospital, such as the ESAC group, but those QIs were not tested in clinical practice and e.g. inter-observer reliability was not measured.10;11;13;15;16 Testing the internal validity of data collected, and therefore knowing your data was uniformly collected is very important. The QIs of Thern et al. showed some overlap with our set, however most their indicators applied to specific patient groups, e.g. only to patients with community acquired infections.14 Our set of QIs is generic and therefore applies to all adult inpatients treated with antibiotics. Pulcini et al. developed QIs specifically for the reassessment of empirical antibiotic use on day 3, instead of the whole antibiotic pathway as measured in our study.28 Secondly, the RAND modified Delphi procedure was used for the QI development and many international experts participated in this procedure.17 This systematic and rigorous consensus method for indicator development,

Chapter 5

106


107

5

together with an applicability test in nearly 1900 patients, results in robust indicators.7-9;17;20;29 Thirdly, all QIs showed high inter-observer reliability, except for QI.9. The overall high kappa was probably because the data were collected in a uniform and standardized way by the investigators. The interpretation was also uniform using the computerized algorithms in SPSS syntax to qualify a patient for the denominator and estimating the numerator.

This study also has potential limitations. One potential limitation could be that the QIs were developed and tested in a Dutch setting, so the results may not automatically translate to other countries. The clinimetric properties can be different for other countries. However, since the QIs were developed using international studies and with an international expert panel, we believe these indicators represent a theoretically sound set that can be used internationally.17 Nevertheless, feasibility should always be tested in practice first, because registration of data may vary between countries and sometimes even within clinical settings. Another limitation with assessing the QI ‘prescribe according to national guideline’, was that some guideline recommendations may have been outdated. Therefore, although guidelines should be reassessed every three years,30 in some cases empirical treatment may have been scored as not correct, while in reality it was correct according to the most recent insights. We also recommend that QI ‘prescribe according to national guideline’ should be applicable to patients with one diagnosis only, because constructing algorithms regarding guideline-adherent therapy for all possible combinations of diagnoses was time-consuming and difficult.

A third limitation was the retrospective, observational design of the study, which makes it impossible to collect all relevant data. For example, not all data that influenced the physicians’ choices may have been documented properly. In addition, most Dutch hospitals still do not have a systematic and robust registration system, therefore collecting data is a time-consuming activity. The expectation is that the current introduction of EPRS may solve part of this problem over time.9 Since collecting data for process QIs is time-consuming, especially for QI.1, we explored what the consequences would be when certain patient characteristics (allergies, pregnancy, and previous ESBL infection)

were not taken into account when computing the algorithm for QI.1 ‘prescribe empirical antibiotic therapy according to guideline’. Excluding these parameters only marginally decreased the mean performance rate: from 41·4% to 40·3%. Therefore, omitting these parameters when assessing QI.1 could be considered.

In summary, we evaluated a set of generic and robust QIs to measure and monitor the appropriateness of antibiotic use in the treatment of all bacterial infections in hospitalized adult patients. A practice test in the Netherlands demonstrated that seven out of the eleven QIs had sound clinimetric properties, and ample room for improvement with large variation between hospitals. Since the process of collecting data is changing over time and varies per country, theoretically sound QIs should always be tested in practice first. Future studies can use this article as a guide for testing the clinimetric properties of their QIs or use this reliable set as a tool to monitor the quality of hospital antibiotic use and to evaluate strategies to improve it in antibiotic stewardship programs.

FundingThis work was supported by the Netherlands Organisation for Health Research and Development (ZonMW) [grant number 205100003].

Transparency declarationsNone declared by all authors

Chapter 5

108


109

5

- all participating hospitals: Academic Medical Centre, Amsterdam Rijnstate Hospital, ArnhemRadboud University Medical Centre, NijmegenUniversity Medical Centre Utrecht, UtrechtMaxima Medical Centre, VeldhovenGelderse Vallei, EdeWestfriesgasthuis, HoornHospital Rivierenland, TielDiakonessenhuis, UtrechtLeiden University Medical Centre, LeidenGemini Hospital, Den HelderOrbis Medical Centre, SittardHospital Bernhoven, UdeMedical Centre Haaglanden, Den HaagJeroen Bosch Hospital, Den BoschVlietland Hospital, SchiedamAtrium Medical Centre Parkstad, HeerlenFranciscus Hospital, RoosendaalLangeLand Hospital, ZoetermeerGroene Hart Hospital, GoudaLievensberg Hospital, Bergen op ZoomMedisch Spectrum Twente, Enschede

Reference List

(1) McGlynn EA, Asch SM, Adams J, Keesey J, Hicks J, DeCristofaro A et al. The quality of health care delivered to adults in the United States. N Engl J Med 2003; 348(26):2635-2645.

(2) Schuster MA, McGlynn EA, Brook RH. How good is the quality of health care in the United States? Milbank Q 1998; 76(4):517-63, 509.

(3) Braykov NP, Morgan DJ, Schweizer ML, Uslan DZ, Kelesidis T, Weisenberg SA et al. Assessment of empirical antibiotic therapy optimisation in six hospitals: an observational cohort study. Lancet Infect Dis 2014; 14(12):1220-1227.

(4) Dellit TH, Owens RC, McGowan JE, Jr., Gerding DN, Weinstein RA, Burke JP et al. Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America guidelines for developing an institutional program to enhance antimicrobial stewardship. Clin Infect Dis 2007; 44(2):159-177.

(5) WHO Global Strategy for Containment of Antimicrobial Resistance. 2011. 1-11-2010. Ref Type: Online Source

(6) Rubin HR, Pronovost P, Diette GB. From a process of care to a measure: the development and testing of a quality indicator. Int J Qual Health Care 2001; 13(6):489-496.

(7) Campbell SM, Braspenning J, Hutchinson A, Marshall MN. Research methods used in developing and applying quality indicators in primary care. BMJ 2003; 326(7393):816-819.

(8) Hermanides HS, Hulscher ME, Schouten JA, Prins JM, Geerlings SE. Development of quality indicators for the antibiotic treatment of complicated urinary tract infections: a first step to measure and improve care. Clin Infect Dis 2008; 46(5):703-711.

(9) Schouten JA, Hulscher ME, Wollersheim H, Braspennning J, Kullberg BJ, van der Meer JW et al. Quality of antibiotic use for lower respiratory tract infections at hospitals: (how) can we measure it? Clin Infect Dis 2005; 41(4):450-460.

(10) Adriaenssens N, Coenen S, Versporten A, Muller A, Vankerckhoven V, Goossens H. European Surveillance of Antimicrobial Consumption (ESAC): quality appraisal of antibiotic use in Europe. J Antimicrob Chemother 2011; 66 Suppl 6:vi71-vi77.

(11) Aldeyab MA, Kearney MP, McElnay JC, Magee FA, Conlon G, MacIntyre J et al. A point prevalence survey of antibiotic use in four acute-care teaching hospitals utilizing the European Surveillance of Antimicrobial Consumption (ESAC) audit tool. Epidemiol Infect 2012; 140(9):1714-1720.

(12) Allerberger F, Lechner A, Wechsler-Fordos A, Gareis R. Optimization of antibiotic use in hospitals--antimicrobial stewardship and the EU project ABS international. Chemotherapy 2008; 54(4):260-267.

Chapter 5

110


111

5

(13) Ansari F, Erntell M, Goossens H, Davey P. The European surveillance of antimicrobial consumption (ESAC) point-prevalence survey of antibacterial use in 20 European hospitals in 2006. Clin Infect Dis 2009; 49(10):1496-1504.

(14) Thern J, de WK, Strauss R, Steib-Bauert M, Weber N, Kern WV. Selection of hospital antimicrobial prescribing quality indicators: a consensus among German antibiotic stewardship (ABS) networkers. Infection 2014; 42(2):351-362.

(15) Zarb P, Ansari F, Muller A, Vankerckhoven V, Davey PG, Goossens H. Drug utilization 75% (DU75%) in 17 European hospitals (2000-2005): results from the ESAC-2 Hospital Care Sub Project. Curr Clin Pharmacol 2011; 6(1):62-70.

(16) Zarb P, Amadeo B, Muller A, Drapier N, Vankerckhoven V, Davey P et al. Identification of targets for quality improvement in antimicrobial prescribing: the web-based ESAC Point Prevalence Survey 2009. J Antimicrob Chemother 2011; 66(2):443-449.

(17) van den Bosch CM, Geerlings SE, Natsch S, Prins JM, Hulscher ME. Quality indicators to measure appropriate antibiotic use in hospitalized adults. Clin Infect Dis 2015; 60(2):281-291.

(18) Sevinc F, Prins JM, Koopmans RP, Langendijk PN, Bossuyt PM, Dankert J et al. Early switch from intravenous to oral antibiotics: guidelines and implementation in a large teaching hospital. J Antimicrob Chemother 1999; 43(4):601-606.

(19) de Greeff SC, Mouton JW. Nethmap 2013. Consumption of antimicrobial agents and antimicrobial resistance among medically important bacteria in the Netherlands. SWAB (Stichting Werkgroep Antibioticabeleid) [2013 Available from: URL:http://www.rivm.nl/dsresource?objectid=rivmp:216911&type=org&disposition=inline&ns_nc=1

(20) Mourad SM, Nelen WL, Hermens RP, Bancsi LF, Braat DD, Zielhuis GA et al. Variation in subfertility care measured by guideline-based performance indicators. Hum Reprod 2008; 23(11):2493-2500.

(21) Campbell SM, Braspenning J, Hutchinson A, Marshall M. Research methods used in developing and applying quality indicators in primary care. Qual Saf Health Care 2002; 11(4):358-364.

(22) Grol R, Baker R, Moss F. Quality improvement research: understanding the science of change in health care. Qual Saf Health Care 2002; 11(2):110-111.

(23) Landis JR, Koch GG. An application of hierarchical kappa-type statistics in the assessment of majority agreement among multiple observers. Biometrics 1977; 33(2):363-374.

(24) Dellinger RP, Levy MM, Rhodes A, Annane D, Gerlach H, Opal SM et al. Surviving sepsis campaign: international guidelines for management of severe sepsis and septic shock: 2012. Crit Care Med 2013; 41(2):580-637.

(25) Paterson R, MacLeod DC, Thetford D, Beattie A, Graham C, Lam S et al. Prediction of in-hospital mortality and length of stay using an early warning scoring system: clinical audit. Clin Med 2006; 6(3):281-284.

(26) Subbe CP, Kruger M, Rutherford P, Gemmel L. Validation of a modified Early Warning Score in medical admissions. QJM 2001; 94(10):521-526.

(27) Sundararajan V, Henderson T, Perry C, Muggivan A, Quan H, Ghali WA. New ICD-10 version of the Charlson comorbidity index predicted in-hospital mortality. J Clin Epidemiol 2004; 57(12):1288-1294.

(28) Pulcini C, Defres S, Aggarwal I, Nathwani D, Davey P. Design of a ‘day 3 bundle’ to improve the reassessment of inpatient empirical antibiotic prescriptions. J Antimicrob Chemother 2008; 61(6):1384-1388.

(29) van den Bosch CM, Hulscher ME, Natsch S, Gyssens IC, Prins JM, Geerlings SE. Development of quality indicators for antimicrobial treatment in adults with sepsis. BMC Infect Dis 2014; 14(1):345.

(30) Shekelle PG, Ortiz E, Rhodes S, Morton SC, Eccles MP, Grimshaw JM et al. Validity of the Agency for Healthcare Research and Quality clinical practice guidelines: how quickly do guidelines become outdated? JAMA 2001; 286(12):1461-1467.

Appropriate antibiotic use reduces length of hospital stay

Caroline M.A. van den Bosch1, Marlies E.J.L. Hulscher2, Reinier P. Akkermans2, Jan Wille3, Suzanne E. Geerlings1 and Jan M. Prins1

1Department of Internal Medicine, Division of Infectious Diseases, Academic Medical Center, University of Amsterdam, the Netherlands, 2IQ healthcare, Radboud

university medical center, Radboud Institute for Health Sciences, Nijmegen, The Netherlands, 3Centre for Infectious Diseases, Epidemiology and Surveillance, The

National Institute for Public Health and the Environment (RIVM), Bilthoven, The Netherlands

Submitted

Chapter 6

Chapter 6

114

Length of hospital stay

115

6

Abstract

BackgroundTo define appropriate antibiotic use in hospitalized adults treated for a bacterial infection, we previously developed and validated a set of six generic quality indicators (QIs) covering all steps in the process of antibiotic use. We assessed the association between appropriate antibiotic use, defined by these QIs, and length of hospital stay (LOS).

MethodsAn observational multicenter study in 22 hospitals in the Netherlands included 1890 patients using antibiotics for a suspected bacterial infection. Performance scores were calculated for all QIs separately (appropriate or not), and a sum score described performance on the total set of QIs (performance on all QIs, divided by the number of QIs that applied to that patient). We divided the sum scores into two groups: low sum scores (0 – 49%) versus high (50 – 100%). Multilevel analyses, correcting for confounders, were used to correlate QI performance (single and combined) with (log-transformed) LOS and in-hospital mortality.

ResultsThe only single QI associated with shorter LOS was appropriate i.v.-oral switch (geometric means: 6.5 vs. 11.2 days; P <0.001). A high sum score was associated with a shorter LOS (10.1 vs. 11.2 days; P =0·002), also in the subgroup community-acquired infections (9.7 vs. 10.9 days; P =0·007). We found no association between performance on QIs and in-hospital mortality or readmission rate.

ConclusionAppropriate antibiotic use, defined by validated process QIs, in hospitalized patients with a suspected bacterial infection appears to reduce length of stay with one day and therefore positively contributes to patient outcome and healthcare costs.

Background

There is an increasing interest in appropriate antibiotic usage and more importantly how we can improve it in daily clinical care.1-3 A growing body of evidence illustrates that Antibiotic Stewardship Programs (ASP) can both optimize treatment of infections and reduce antibiotic resistance.4, 5 An ASP requires time, efforts and costs, but for patients with lower respiratory tract infections (lower RTI), complicated urinary tract infections (UTI) or sepsis, appropriate antibiotic use has been associated with a lower mortality rate, a decreased admission rate to the intensive care unit (ICU) and a decreased length of hospital stay.6-8 In most of these studies guideline-concordant empirical therapy was used to define appropriate antibiotic use.6, 7 However, appropriate antibiotic use requires more than appropriate empirical therapy alone. In earlier studies we developed and validated a comprehensive set of quality indicators (QIs) to define appropriate antibiotic use in patients with urinary tract infections.9 For patients with a complicated UTI appropriate antibiotic use, defined by this comprehensive set of UTI-specific QIs, reduced length of hospital stay.8 Length of hospital stay (LOS) is a relevant outcome measure, because it reflects the recovering time of patients and defines hospital costs.10 To be able to evaluate the quality of antibiotic use covering all steps in the process of antibiotic treatment and in all hospitalized adult patients with a suspected bacterial infection, we recently developed a set of generic QIs.11 This set was tested in daily hospital practice and six QIs, which measured various steps in the process of antibiotic use on the patient level, turned out to have sound clinimetric properties.12 These QIs were: Prescribe empirical treatment in accordance with the guideline; Take two blood cultures before start of treatment; Take a culture from the suspected site of infection; Document an antibiotic plan in the case notes; Switch from intravenous to oral therapy within 72 hours; and Tailor antibiotic treatment to pathogen-directed therapy on the basis of the culture results.The aim of the current study was to evaluate, in 1890 hospitalized adult patients with a broad range of (suspected) bacterial infections, the association between appropriate antibiotic use as defined by this set of generic QIs, and LOS. Secondary outcomes were in-hospital mortality, ICU admission during hospital stay and readmission within 30 days.

Chapter 6

116


117

6

Methods

Setting, study population and data collectionWe conducted a prospective observational study in 4 university and 18 non-university (medium-sized and small) hospitals. Details of this study have been described earlier.11, 12 In short, we identified in each hospital, in a point prevalence measurement on non-Intensive Care Unit (ICU) departments, all adult inpatients using antibiotics for >24 hours for a suspected hospital- or community-acquired bacterial infection. This point prevalence measurement primarily aimed at identifying nosocomial infections.13 Patients were included in October 2011, March 2012 and October 2012. Clinical and laboratory data were, retrospectively, extracted from medical and nursing records and medication charts. Co morbidity was measured by means of the Charlson Co morbidity Index,14 and severity of illness was defined as the MEWS (Modified Early Warning Score) at the start of treatment.15 Patients were excluded if antibiotics were used as prophylaxis, or when antibiotic treatment was started on the ICU department or in another hospital, see figure 1. All data needed to compute the QIs or to assess potential confounders were extracted in a uniform way and entered in the database anonymously. The ethics committee assessed the study and concluded that it was deemed exempt from their approval.

Quality indicators and definitionsThe development and validation of the QIs has been described earlier.11, 12 The final set of six validated, generic QIs defining appropriate antibiotic use and covering various steps in the process at the patient level is described in detail in Table 1. For some of the QIs, we had to develop the following extensive working definitions and algorithms to compute appropriateness scores:QI.1 Prescribe empirical antibiotic therapy according to the national guideline. We defined empirical therapy as the antibiotic (combination) prescribed on the day of diagnosis, before identification of a causative pathogen. If initial therapy was directed at a previously cultured pathogen, this was not called ‘empirical therapy’ and these patients were excluded for this indicator. We used the at that moment current Dutch national guidelines (http://www.swab.nl/guidelines and www.swabid.nl) to evaluate whether the prescribed empirical regimen was in accordance with the guideline. Individual patient characteristics, in particular allergies, pregnancy, and previous extended-spectrum β-lactamase

(ESBL) infection were taken into account when computing the algorithms for appropriate empirical therapy. Only patients with one or two (possible) diagnoses covered by a guideline were included. Patients with more than two (possible) infections, or with a (possible) infection not covered by a guideline were excluded. In case two diagnosed or possible infections affected a single organ system (e.g. upper and lower RTI), it usually concerned a differential diagnosis, and antibiotic therapy was considered correct if covered by one of the two applicable guidelines. If the two (possible) infections affected different organ systems (e.g. LRTI and urinary tract infection), antibiotic therapy was only considered correct if both diagnoses were covered according to their respective guidelines. QI.5 Antibiotic therapy should be switched from intravenous to oral therapy within 48 – 72 hours.This QI was applicable to the subgroup of patients who started with i.v. antibiotic treatment and fulfilled the criteria for safe early switch, see Box 1.16 Clinical stability at 48 – 72 hours of antibiotic treatment was also considered to have been present when temperature or blood pressure was not mentioned in the medical records or WBC had not been determined, because we assumed that if these parameters were abnormal, they would have been documented. QI.6 Empirical antibiotic therapy should be changed to pathogen-directed therapy if culture results become available.This QI applied to patients with a positive culture result (denominator), and it measured the proportion of patients (numerator) that received pathogen-directed antibiotic treatment after culture results became positive. We estimated conservatively that in general at day 5 culture results are available. Therefore, the antibiotic (combination) given at day 5 was considered the final therapy. This therapy was considered pathogen-directed when it was in accordance with the resistance pattern of the cultured microorganism, regardless whether antibiotic therapy was changed or not. Where possible, antibiotics should be chosen from the group of narrow spectrum antibiotics (Box 2).1

QI sum scoreThis sum score was defined as the overall performance on all QIs (appropriate =1 and inappropriate =0) for a patient, divided by the number of QIs that applied to that specific patient. For example, i.v.-oral switch was not possible in all patients. As QI.2, taking blood cultures, and QI.4, documenting the antibiotic plan, applied to all patients, the minimum number of applicable QIs

Chapter 6

118


119

6

Box 1.

QI 5 Criteria for safe switch within 2-3 days16

1. Hemodynamic stable 2. No fever3. Normal WBC count (4 – 12 x109/L) 4. Patient must be able to take oral medication5. Patient must have a functioning gastrointestinal tract, without signs of malab-

sorption6. Safe switch not possible in case of: meningitis, intracranial abscess, endocarditis,

mediastinitis, Legionella pneumonia and exacerbations of cystic fibrosis, inade-quately drained abscess and empyema, severe soft tissue infection such as group A streptococcal infections, infection of foreign bodies, Staphylococcus aureus or Pseudomonas aeruginosa bacteraemia, liver abscess and empyema, osteomyelitis and arthritis, chemotherapy related neutropenia.

Box 2.

QI 6: Narrow spectrum antibiotics1 Penicillin CiprofloxacinPhenethicillin NorfloxacinAmoxicillin OfloxacinFlucloxacillin DoxycyclineCephalexin TetracyclineCephalothin MinocyclineCefazolin Nitrofurantoin Erythromycin FosfomycinCo-trimoxazole MetronidazoleTrimethoprim

was two, and the maximum number six. We divided the sum scores into two groups: low sum scores (0 – 49%) versus high scores (50 – 100%). Definitions Length of hospital stay (LOS) in community-acquired infections was defined as the number of days between admission and discharge. Hospital-acquired infections were defined by the start of antibiotic treatment because of a suspected infection at least 48 hour after admission. In these patients LOS was defined as the number of days between start of antibiotic treatment and discharge. Readmission was defined as an admission within 30 days after discharge, with as primary diagnosis an infection in the same organ system as before discharge.

AnalysisCollected data were entered in a database using the Statistical Package for the Social Sciences (SPSS 20·0 for Windows®, SPSS Inc., Chicago, IL, USA). Algorithms were created in SPSS for every QI to determine whether a patient met the criteria for the denominator and numerator. To investigate the relationship between LOS and the performance on the single QIs and the overall performance on the set of QIs (i.e. the sum score), we used multilevel (linear mixed model) analyses, because of the hierarchical structure of this study (patients nested within hospitals). We used a model with random intercept and all other variables fixed. We adjusted for the quantity of the QIs in the sum score and conducted sub-analyses for patients with community-acquired versus hospital-acquired infections. LOS was log-transformed before analyses were done to satisfy normality assumptions, and LOS was back-transformed for presentation as geometric mean (95% confidence interval), and estimated difference in percent.17 Secondary outcomes (in-hospital mortality, ICU admission and readmission within 30 days) were also analyzed using multilevel (linear mixed model) analyses. The following variables were included as potential confounder for the relationship between QI scores and outcome: sex, age, community versus hospital-acquired infections, type of diagnosis (divided in seven groups: UTI, RTI, skin and soft tissue infections, intra-abdominal infections, other infections, two possible diagnoses at start of antibiotic treatment and more than two possible diagnoses/diagnosis not covered by guideline), severity of illness (MEWS score), co morbidity (Charlson Co morbidity Index), prior

Chapter 6

120


121

6

use of antibiotics (last 30 days before start of treatment), start of antibiotics intravenously or orally, and type of hospital (university hospital, large (more than 500 beds) or small (less than 500 beds) non-university hospital). We tested with multivariate multilevel (linear mixed) analyses if these variables had an effect (p < 0.10) on LOS. If so, confounding might be present and in that case we adjusted for this variable. A value of P < 0.05 was considered statistically significant.

Results

Study populationThe patient population consisted of 1890 in patients treated with antibiotics for a suspected bacterial infection (Figure 1). 1391 (74%) were community-acquired infections. The mean patient age was 69·8 years, 51.5% were male and 41% of patients fulfilled the sepsis criteria (≥ 2 SIRS criteria). 27% of patients had a MEWS score ≥3. For further baseline characteristics, see Table 2. Not every indicator applied to all included patients, so the sample size of the QIs varied (Table 3a and b). The percentages appropriate antibiotic use as defined by the six QIs varied between 32% (i.v.-oral switch) and 61% (documenting antibiotic plan).

Quality indicators and LOSThe geometric mean for LOS was 10.2 (95% CI: 3 – 42) days for community-acquired infections and 11.9 (95% CI: 3 - 53) days for hospital-acquired infections. Variables with an effect on LOS were community- versus hospital-acquired infections, type of diagnosis, co morbidity (Charlson Co morbidity Index), severity of illness (MEWS score), prior use of antibiotics (last 30 days before start of treatment), start antibiotics IV or orally, age and sex. In the subgroups not all variables had an effect on LOS, see table 4.Associations between appropriate antibiotic use (as described by the six QIs) and LOS are listed in Table 3a and b. A significant association was demonstrated between an early i.v.-oral switch and a shorter LOS (6.5 vs. 11.2 days; P <0.001), also in both subgroups separately. Prescribing empirical therapy in accordance with the national guideline had a significant association with a shorter LOS in the subgroup of community-acquired infections (8.8 vs. 10.5 days, P = 0.04) and

streamlining empiric therapy to pathogen-directed therapy in the subgroup of hospital-acquired infections (10.8 vs. 15.6 days, P = 0.002). Performing blood cultures or a culture from the suspected site of infection and documenting an antibiotic plan were not significantly associated with LOS.

Performance on the total set of QIs and LOSIn many patients only 2, 3, 4 or 5 QIs applied; the QI regarding i.v.-oral switch involved only 22% of the patients and streamlining antibiotic therapy 21% of the patients. The proportion appropriate use in each patient was associated with LOS: 10.1 days for patients with a high sum score versus 11.2 days for those with a low score (P = 0.002). This also applied to the subgroup community-acquired infections (9.7 vs. 10.9 days; P = 0.007). The subgroup hospital-acquired infections (11.6 vs. 12.2 days; P = 0.08) was smaller and although patients with a high sum score had a shorter length of hospital stay, the difference was not significant, see table 3a and 3b.

Quality indicators and secondary outcomesAppropriate antibiotic use as defined by the six QIs was not significantly associated with lower in-hospital mortality, ICU admissions or readmission within 30 days after discharge, see appendix 1.

Chapter 6

122


123

6

Table 1. List of generic quality indicators to monitor antibiotic use in hospitalized adult patients on non-ICU departments.

Num-ber

Quality indicator

All patients are: adults, admitted on a non-ICU department, using systemic antibi-otics for > 24h because of a sus-pected bacterial infection

Numerator de-scription

Denominator description

Comment

1 Empirical systemic antibiotic therapy should be pre-scribed according to the national guideline

Number of patients who started with empirical systemic antibiotic therapy according to the national guideline

Total number of patients who start-ed with empirical systemic antibiotic therapy

2 Before starting systemic antibiotic therapy at least two sets of blood cultures should be taken

Number of patients in whom at least 2 sets of blood cultures were taken before systemic antibiotic therapy was started

Total number of patients who start-ed with systemic antibiotic therapy

Blood cultures should be per-formed between one week before start of treatment and actual start of treatment

3 When starting systemic antibiotic therapy specimens for culture should be taken from suspected sites of infection

Number of patients in whom cultures from suspected sites of infection were taken within 24 hours after starting systemic antibiotics


Cultures should be taken between two weeks before start of treatment and one day after start of treatment

4 An antibiotic plan should be docu-mented in the case notes at the start of systemic antibiotic therapy

Number of patients for whom an an-tibiotic plan was documented in the case notes


Antibiotic plan is indication, and name, dosage, route and interval of administration of the antibiotic

5 Systemic antibiotic therapy should be switched from intravenous to oral antibiotic therapy within 48 -72 hours on the basis of the clinical condition and when oral treatment is ade-quate

Number of patients with intravenous antibiotics for 48-72h who changed to oral therapy on the basis of the clinical condition

Total number of patients with intravenous anti-biotics for 48-72h in whom changing to oral therapy was indicated on the basis of the clinical condition

Definitions: see

Box 1

6 Empirical antibiot-ic therapy should be changed to pathogen-directed therapy if culture results become available

Number of patients whose culture became positive and who correct-ly changed to pathogen-directed therapy

Total number of patients who start-ed with empirical systemic antibiot-ics, whose culture became positive

Therapy is patho-gen-directed if it is in accordance with resistance pattern. If possible, antibiotics should be chosen from the group of narrow spectrum antibiotic (Box 2)

Chapter 6

124


125

6

Table 2. Baseline characteristics of the participating patients

Characteristics Total (n=1890)

Community acquired infections(n=1391)

Hospital acquired infections (n=499)

Sex, male 974 (51.5)* 717 (51.5)* 257 (51.5)*Age, mean (SD) 69.8 (17) 69.2 (17) 71.5 (16)Infection, Community-acquired versus Hospital-acquired infection

1391 (74) 499 (26)

DiagnosisRespiratory tract infection 463 (24) 349 (25) 114 (23)Urinary tract infection 265 (14) 176 (13) 89 (18)Skin and soft tissue infection 177 (9) 130 (9) 47 (9)Intra-abdominal infection 122 (7) 92 (7) 30 (6)Other infections 184 (10) 135 (10) 49 (10)Two diagnoses possible at start of antibi-otic treatment

275 (15) 223 (16) 52 (10)

more than two possible diagnoses/ diag-nosis not covered by guideline

404 (21) 286 (20) 118 (24)

SIRS criteria22 ≥ 2 (sepsis) 774 (41) 596 (43) 178 (36) < 2 (no sepsis) 1116 (59) 795 (57) 321 (64)Severity of illness (MEWS score) 0 451 (24) 301 (22) 150 (30) 1 666 (35) 491 (35) 175 (35) 2 272 (15) 206 (15) 66 (13) 3 194 (10) 150 (11) 44 (9) 4 136 (7) 103 (7) 33 (6) 5 74 (4) 60 (4) 14 (3) 6 52 (3) 39 (3) 13 (3) ≥ 7 45 (2) 41 (3) 4 (1)Comorbidity (Charlson comorbidity index) 0 443 (23) 348 (25) 95 (19) 1 406 (22) 321 (23) 85 (17) 2 444 (23) 303 (22) 141 (28) 3 329 (17) 226 (16) 103 (20) 4 146 (8) 111 (8) 35 (7) 5 71 (4) 47 (3) 24 (5) 6 36 (2) 24 (2) 12 (3) ≥ 7 15 (1) 11 (1) 4 (1)

Received antibiotics before start of (em-piric) treatment No 1180 (62) 891 (64) 289 (58) Yes 710 (38) 500 (36) 210 (42)Antibiotics Started empirical** 1732 (92) 1303 (94) 429 (86) Started based on cultures 155 (8) 86 (6) 69 (14) Started intravenously (IV)*** 1399 (74) 1075 (77) 324 (65) Started orally 489 (26) 314 (23) 175 (35)

* Numbers are n (%), unless otherwise indicated.**In 3 patients it was unknown whether antibiotics were started empirical or based on culture.*** In 2 patients it was unknown whether antibiotics were started IV or orally.

Chapter 6

126

Tabl

e 3a

. Asso

ciatio

n be

twee

n Q

Is an

d LO

S

Qua

lity

indi

cato

r

All

patie

nts (

n =

1890

)A

ppro

pria

te a

ntib

i-ot

ic tr

eatm

ent

Inap

prop

riat

e an

ti-bi

otic

trea

tmen

tN

(%)

LOS,

day

s*

N (%

)LO

S, d

ays*

D

iffer

ence

(95%

CI)

†%

LO

S**

p-va

lue†

Empi

rica

l the

rapy

in a

ccor

danc

e w

ith n

atio

n-al

gui

delin

es (n

= 1

361)

563

(41)

9.3

(3 –

41)

10.2

79

8 (5

9)10

.7 (3

- 42

)10

.7

0.96

(0.8

7 –

1.05

) -4

%0.

33

Blo

od c

ultu

res t

aken

(n =

189

0)67

4 (3

6)11

(3 –

42)

10.7

1216

(6

4)10

.4 (3

- 47

)10

.70.

99 (0

.93

– 1.

07)

-0.1

%0.

99

Cul

ture

take

n of

susp

ecte

d si

te o

f inf

ectio

n (n

=

1217

) 59

5 (4

9)10

.5 (3

- 47

)12

.662

2 (5

1)10

.5 (3

- 43

)12

1.05

(0.9

5 –

1.15

)+5

%0.

32

Doc

umen

tatio

n of

ant

ibio

tic p

lan

(n =

189

0)11

45

(61)

10.5

(3 –

44)

10.6

745

(39)

10.7

(3 -

52)

10.9

0.98

(0.9

– 1

.05)

-2%

0.43

Earl

y sw

itch

if fu

lfilli

ng c

rite

ria

for s

afe

IV-

oral

switc

h (n

= 4

22)

134

(32)

6.5

(2 -

27)

6.9

288

(68)

11.2

(4 –

44)

10.9

0.

63 (0

.55

– 0.

72)

-37%

0.00

0

Path

ogen

-dir

ecte

d an

tibio

tic th

erap

y if

cul-

ture

bec

omes

pos

itive

(n =

453

)22

8 (5

0)12

.1 (3

- 46

)13

225

(50)

13.3

(3 -

65)

13.7

0.96

(0.8

– 1

.1)

-4%

0.49

SUM

scor

e (1

890)

Inap

prop

riat

e =0

-49%

App

ropr

iate

=50

-100

%

1057

(5

6)10

.1 (3

- 42

)10

.183

3 (4

4)11

.2 (3

- 51

)11

.40.

89 (0

.8 –

0.9

5)-1

1%0.

002

* geo

met

ric m

ean

(95%

CI)

, and

geo

met

ric m

ean

after

adj

ustin

g fo

r pos

sible

conf

ound

ers

† D

iffer

ence

(app

ropr

iate

vs. in

appr

opria

te), e

xpre

ssed

as th

e rat

io of

the g

eom

etric

mea

n of

LO

S fo

r app

ropr

iate

ant

ibio

tic tr

eatm

ent o

ver t

he

geom

etric

mea

n of

LO

S fo

r ina

ppro

pria

te a

ntib

iotic

trea

tmen

t. A

djus

ted

for c

onfo

unde

rs: w

ithin

the g

roup

of a

ll in

fectio

ns w

e adj

usted

for:

age,

com

mun

ity- v

ersu

s hos

pita

l-acq

uire

d in

fectio

n, ty

pe o

f dia

gnos

is, se

verit

y of i

llnes

s (M

EWS

scor

e), c

omor

bidi

ty (C

harls

on C

omor

bidi

ty

Inde

x), p

rior u

se of

ant

ibio

tics,

and

start

of a

ntib

iotic

s int

rave

nous

ly or

oral

ly.**

Est

imat

ed d

iffer

ence

in L

OS

(%) w

hen

antib

iotic

trea

tmen

t is a

ppro

pria

te vs

. inap

prop

riate

(fo

r exa

mpl

e: d

iffer

ence

= 0

.96,

so th

ere i

s a

4% re

ducti

on on

LO

S wh

en em

piric

al th

erap

y is i

n ac

cord

ance

with

guid

eline

).

Tabl

e 3b

. Asso

ciatio

n be

twee

n Q

Is an

d LO

S in

subg

roup

s

Qua

lity

indi

cato

r

Patie

nts w

ith co

mm

unit

y-ac

quir

ed in

fect

ion

(n =

139

1)Pa

tient

s with

hos

pita

l-acq

uire

d in

fect

ion

(n =

499

)A

ppro

pria

te a

nti-

biot

ic tr

eatm

ent

Inap

prop

riat

e an

tibio

tic

trea

tmen

t

App

ropr

iate

an

tibio

tic

trea

tmen

t

Inap

prop

ri-

ate

antib

iotic

tr

eatm

ent

N (%

)LO

S,

days

* N

(%)

LOS,

da

ys*

Diff

eren

ce

(95%

CI)

†%

LO

S **

p-va

lue†

N (%

)LO

S,

days

* N (%

)LO

S,

days

* D

iffer

ence

(9

5% C

I)†

%

LOS*

*p-

valu

e†Em

piri

cal t

hera

py in

acc

or-

danc

e w

ith n

atio

nal g

uide

lines

46

2 (4

5)8.

8 (3

-36)

9.4

571

(55)

10.5

(3

-40)

10.4

0.91

(0.8

-0.9

6)-9

%0.

0410

1 (3

1)12

.4

(3-5

4)14

227

(69)

11.3

(2

-49)

12.4

1.1

(0.9

-1.4

)+1

0%0.

21

Blo

od c

ultu

res t

aken

517

(37)

10.5

(3

-39)

10.2

874

(63)

10

(3-4

4)10

.2

1.0

(0.9

-1.1

)+0

.1%

0.99

157

(31)

12.8

(4

-53)

12.3

342

(69)

11.5

(3

-54)

12.7

0.97

(0

.8-1

.1)

-3%

0.69

Cul

ture

take

n of

susp

ecte

d si

te

of in

fect

ion

411

(46)

10.2

(3

-41)

12.2

487

(54)

10.2

(3

-42)

11.2

1.1

(0.9

8-1.

2)

+10%

0.1

184

(58)

11.1

(2

-50)

14.2

135

(42)

11.7

(3

-49)

14.3

0.99

(0

.8-1

.2)

-1%

0.97

Doc

umen

tatio

n of

ant

ibio

tic

plan

87

6 (6

3)10

.1

(3-3

9)10

515

(37)

10.4

(3

-44)

10.4

0.95

(0

.9-1

.05)

-5%

0.37

269

(54)

12.2

(3

-47)

12.7

230

(46)

11.5

(2

-57)

12.5

1.01

(0.9

-1.2

)+1

%0.

88

Earl

y sw

itch

if fu

lfilli

ng c

rite

-ri

a fo

r saf

e IV

-ora

l sw

itch

115

(34)

6.4

(2-2

8)6.

5

220

(66)

10.3

(4

-41)

10.1

0.65

(0.6

-0.7

6)-3

5%0.

000

19

(22)

7.7

(2-)

8.2

68 (78)

14.4

(4

-56)

15.1

0.55

(0

.4-0

.8)

-45%

0.00

4

Path

ogen

-dir

ecte

d an

tibio

tic

ther

apy

if cu

lture

bec

omes

po

sitiv

e

153

(49)

12.7

(4

-48)

13.7

159

(51)

12.4

(3

-58)

12.5

1.1

(0.9

-1.3

)+1

0%0.

2675

(5

3)10

.8

(2-4

8)12

.4

66 (47)

15.6

(4

-77)

19.6

0.63

(0.5

-0.8

)-3

7%0.

002

SUM

scor

e In

appr

opri

ate

=0-4

9%A

ppro

pria

te =

50-1

00%

804

(58)

9.7

(3-3

7)9.

6

587

(42)

10.9

(3

-48)

10.9

0.89

(0

.8-0

.97)

-11%

0.00

725

3 (5

1)11

.6

(3-4

7)11

.8

246

(49)

12.2

(3

-59)

13.4

0.87

(0

.8-1

)-1

3%0.

08

* geo

met

ric m

ean

(95%

CI)

, and

geo

met

ric m

ean

after

adj

ustin

g fo

r pos

sible

conf

ound

ers

† D

iffer

ence

(app

ropr

iate

vs. in

appr

opria

te), e

xpre

ssed

as th

e rat

io of

the g

eom

etric

mea

n of

LO

S fo

r app

ropr

iate

antib

iotic

trea

tmen

t ove

r the

geom

etric

mea

n of

LO

S fo

r ina

ppro

pria

te

antib

iotic

trea

tmen

t. A

djus

ted

for

conf

ound

ers;

with

in th

e gro

up o

f com

mun

ity- a

cqui

red

infec

tions

we a

djus

ted fo

r: ag

e, ty

pe o

f dia

gnos

is, co

mor

bidi

ty (

Char

lson

Com

orbi

dity

In

dex)

and

star

t of a

ntib

iotic

s int

rave

nous

ly or

ora

lly. W

ithin

the g

roup

of h

ospi

tal-a

cqui

red

infec

tions

we a

djus

ted fo

r: ty

pe o

f dia

gnos

is, se

verit

y of

illn

ess (

MEW

S), c

omor

bidi

ty

(Cha

rlson

Com

orbi

dity

Inde

x), p

rior u

se of

ant

ibio

tics,

and

start

of a

ntib

iotic

s int

rave

nous

ly or

oral

ly.

** E

stim

ated

diff

eren

ce in

LO

S (%

) whe

n an

tibio

tic tr

eatm

ent i

s app

ropr

iate

vs. i

napp

ropr

iate

(fo

r exa

mpl

e: d

iffer

ence

= 0

.91,

so th

ere i

s a 9

% re

ducti

on o

n LO

S wh

en em

piric

al

ther

apy i

s in

acco

rdan

ce w

ith gu

ideli

ne).

Chapter 6

128


129

6

Tabl

e 4.

Con

foun

ders

for L

OS

Vari

able

sTo

tal (

n=18

90)

Com

mun

ity a

cqui

red

infe

ctio

ns(n

=139

1)H

ospi

tal a

cqui

red

infe

ctio

ns

(n=4

99)

Estim

ate (

95%

CI)

PEs

timat

e (95

% C

I)P

Estim

ate (

95%

CI)

PA

ge0.

003

(0.0

02 –

0.

004)

0.00

00.

004

(0.0

03 –

0.0

05)

0.00

00.

0008

(-0.

003

– 0.

001)

0.49

Com

mun

ity-

ver

sus h

ospi

tal-

acqu

ired

infe

ctio

n0.

07 (0

.03

– 0.

1)0.

000

Type

of d

iagn

osis

(div

ided

in

seve

n gr

oups

)*ov

eral

l tes

t0.

000

over

all t

est

0.00

0ov

eral

l tes

t0.

000

Seve

rity

of i

llnes

s (M

EWS)

0.04

(0.0

04 –

0.0

8)0.

030.

03 (-

0.01

– 0

.07)

0.20

0.08

(-0.

002

– 0.

2)0.

06C

omor

bidi

ty (C

harl

son

Com

orbi

dity

Inde

x)0.

02 (0

.01

– 0.

03)

0.00

00.

02 (0

.006

– 0

.03)

0.00

40.

02 (-

0.01

– 0

.04)

0.07

Prio

r use

of a

ntib

iotic

s (la

st 3

0 da

ys b

efor

e st

art o

f tre

atm

ent)

0.04

(0.0

1 –

0.08

)0.

010.

006

(-0.

32 –

0.0

4)0.

760.

11 (0

.05

– 0.

2)0.

001

Star

t AB

ora

lly o

r IV

0.08

(0.0

4 –

0.1)

0.00

00.

05 (0

.009

– 0

.1)

0.02

0.1

(0.0

3 –

0.2)

0.00

7Se

x-0

.03

(-0.

06 –

0.0

04)

0.09

-0.0

1 (-

0.05

– 0

.02)

0.45

-0.0

5 (-

0.12

– 0

.02)

0.13

* UT

I, RT

I, sk

in a

nd so

ft tis

sue

infe

ctio

ns, i

ntra

-abd

omin

al in

fect

ions

, oth

er in

fect

ions

, tw

o po

ssib

le d

iagn

oses

at s

tart

of a

ntib

iotic

tre

atm

ent a

nd m

ore t

han

two

poss

ible

dia

gnos

es/d

iagn

osis

not c

over

ed b

y gu

idel

ine

Discussion

In a cohort of 1890 hospitalized patients treated with antibiotics for a suspected bacterial infection, an early i.v.-oral switch and a higher proportion of appropriate antibiotic use (defined by the performance on the total set of QIs) appears to be associated with a shorter length of stay of one day. We found no association between performance on QIs and in-hospital mortality, ICU admission or readmission rate. To our knowledge, this study is the first to evaluate the relationship between appropriate antibiotic use, defined by a set of validated generic QIs, and LOS in hospitalized patients treated with antibiotics for a variety of suspected bacterial infections.

The major strength of this study is the unselected and large patient population, from many different hospitals and departments. Our study population comprises the everyday clinical case-mix of hospitalized patients. Since community-acquired and hospital-acquired infections define different patients groups, especially concerning length of stay, we adjusted for this variable in all analyses, and also performed sub analyses with only community-acquired infections. Another strength was the use of validated generic QIs,11, 12 which measure the various steps in the process of antibiotic use on the patient level. After collection of the data, algorithms were created in SPSS to score the QIs, thereby excluding personal interpretations or preferences.11, 12 Finally, length of stay is a continuous variable applicable to all patients and should in theory be sensitive to differences in quality of care, but LOS is subjected to numerous confounders.18 We therefore controlled for potential confounders which might have impact on LOS, and since outliers in LOS are common, LOS was log-transformed before analyses were done to satisfy normality assumptions. Adjusting for confounders did not alter our conclusions.

Our study has some limitations. The design is observational and although we controlled for known confounders,6, 18 we might not have controlled for all variables associated with LOS. Next, data were collected retrospectively, and it is not possible to collect all clinically relevant data. For example, not all data that affected the physicians’ choices may have been documented properly, which may lead to wrong conclusions when classifying antibiotic use as appropriate

Chapter 6

130


131

6

or inappropriate. However, we constructed the QIs to be based on objective, retrievable data as much as possible.11, 12 Finally, dosing of antibiotics was not evaluated, and selection of appropriate, but under dosed antibiotics may still lead to clinical failure and longer LOS.

An early i.v.-oral switch has been shown to be safe and is associated with significant reduced LOS.8, 19 Our study and previous studies point out that it is still not routine clinical care.8, 19 Even after 5 days only 57% of the patients who met the safe switch criteria within 72 hours were switched to oral antibiotics, with a wide range between hospitals. An early switch was the only single QI associated with a shorter LOS, but this does not explain the association between the sum score and LOS, as the QI regarding i.v.-oral switch applied to only 22% of patients. We only found an association between empirical antibiotic treatment according to (national) guidelines and LOS in patients with community-acquired infections, in accordance with previous studies.6, 20 At the same time there are studies that did not find such an association.8, 18 Appropriate empiric therapy is effective for increasing coverage rates without prescribing unnecessarily broad regimens,3 but if inappropriate empiric treatment is unnecessary broad, this may not have a direct negative effect on LOS. Another explanation why guideline-adherent empirical therapy does not affect LOS might be that in patients with lower severity of illness adequate antibiotic therapy may be less crucial for clinical outcomes.21 In our patient sample only 40% met the criteria for sepsis (SIRS criteria).22 No associations were found between performing blood cultures or a culture from the suspected site of infection, or, for community-acquired infections, streamlining of antibiotic therapy, and LOS. It appears that influencing LOS by performing appropriate microbiological investigations is difficult, since only a few studies found a positive effect on clinical outcome.23, 24 However, microbiological investigations not only allow for i.v.-oral switching, but also for switching of empirical to pathogen-directed (streamlined) therapy,25-27 which has been associated with a reduction in antibiotic use and costs.25, 28 Further research is needed to evaluate the usefulness of streamlining in terms of patient outcome.29, 30 Documenting an antibiotic plan in case notes, while potentially contributing to appropriate antibiotic use,31 was not associated with LOS, in accordance with previous studies.31, 32

The most important outcome of this study was that not adherence to one single QI, but that adherence to a combination of concomitant processes of care resulted in a decreased duration of hospital stay. It has been demonstrated before that in particular adherence to bundles of care improve patient outcome and might be a surrogate marker for a better global quality of care.8, 22, 33, 34

In conclusion, we found that appropriate antibiotic use, as defined by the performance on the total set of QIs, in hospitalized patients with a suspected community-acquired infection appears to reduce length of stay with one day. Illustrating links between processes of antibiotic care and outcome is essential to convince clinicians of the importance of appropriate prescribing and antibiotic stewardship program compliance. In addition, initiating stewardship activities that aim at improving compliance with these validated generic QIs might be an effective means to promote such appropriate care.

Declaration of interestsNone declared by all authors

Acknowledgements - all participating hospitals: Academic Medical Center, Amsterdam Rijnstate Hospital, ArnhemRadboud University Medical Center, NijmegenUniversity Medical Center Utrecht, UtrechtMaxima Medical Center, VeldhovenGelderse Vallei, EdeWestfriesgasthuis, HoornHospital Rivierenland, TielDiakonessenhuis, UtrechtLeiden University Medical Center, LeidenGemini Hospital, Den HelderOrbis Medical Center, SittardHospital Bernhoven, UdeMedical Center Haaglanden, Den HaagJeroen Bosch Hospital, Den BoschVlietland Hospital, Schiedam

Chapter 6

132


133

6

Atrium Medical Center Parkstad, HeerlenFranciscus Hospital, RoosendaalLangeLand Hospital, ZoetermeerGroene Hart Hospital, GoudaLievensberg Hospital, Bergen op ZoomMedisch Spectrum Twente, Enschede

Reference List

1. Adriaenssens N, Coenen S, Versporten A, et al. European Surveillance of Antimicrobial Consumption (ESAC): quality appraisal of antibiotic use in Europe. J Antimicrob Chemother 2011; 66 Suppl 6: vi71-vi77.

2. Laxminarayan R, Duse A, Wattal C, et al. Antibiotic resistance-the need for global solutions. Lancet Infect Dis 2013; 13: 1057-98.

3. Spoorenberg V, Prins JM, Stobberingh EE, et al. Adequacy of an evidence-based treatment guideline for complicated urinary tract infections in the Netherlands and the effectiveness of guideline adherence. Eur J Clin Microbiol Infect Dis 2013; 32: 1545-56.


5. DiazGranados CA. Prospective audit for antimicrobial stewardship in intensive care: impact on resistance and clinical outcomes. Am J Infect Control 2012; 40: 526-9.


7. Shorr AF, Micek ST, Welch EC, et al. Inappropriate antibiotic therapy in Gram-negative sepsis increases hospital length of stay. Crit Care Med 2011; 39: 46-51.

8. Spoorenberg V, Hulscher ME, Akkermans RP, et al. Appropriate antibiotic use for patients with urinary tract infections reduces length of hospital stay. Clin Infect Dis 2014; 58: 164-9.

9. Hermanides HS, Hulscher ME, Schouten JA, et al. Development of quality indicators for the antibiotic treatment of complicated urinary tract infections: a first step to measure and improve care. Clin Infect Dis 2008; 46: 703-11.

10. Raut M, Schein J, Mody S, et al. Estimating the economic impact of a half-day reduction in length of hospital stay among patients with community-acquired pneumonia in the US. Curr Med Res Opin 2009; 25: 2151-7.

11. van den Bosch CM, Geerlings SE, Natsch S, et al. Quality indicators to measure appropriate antibiotic use in hospitalized adults. Clin Infect Dis 2015; 60: 281-91.

12. van den Bosch CMA, Hulscher MEJL, Natsch S, et al. Quality Indicators for Monitoring Appropriate Antibiotic Use in Hospitals: an Important Tool for Antibiotic Stewardship. http://www.abstractsonline.co m / Pl an / Vi e wA b st rac t . a s px ? m I D = 3 5 2 9 & sK e y = 8 3 8 cd 3 3 9 - 5 0 f 1 -4 e a 5 - 8 0 f 7 - 9 6 7 9 8 0 f 0 5 2 2 9 & c K e y = 2 a 1 3 f d b c - 4 8 c 7 - 4 1 7 6 - b d 9 3 -60b65561e137&mKey=5d6b1802-e453-486b-bcbb-b11d1182d8bb.

Chapter 6

134


135

6

13. van der Kooi TI, Mannien J, Wille JC, et al. Prevalence of nosocomial infections in The Netherlands, 2007-2008: results of the first four national studies. J Hosp Infect 2010; 75: 168-72.

14. Sundararajan V, Henderson T, Perry C, et al. New ICD-10 version of the Charlson comorbidity index predicted in-hospital mortality. J Clin Epidemiol 2004; 57: 1288-94.

15. Paterson R, MacLeod DC, Thetford D, et al. Prediction of in-hospital mortality and length of stay using an early warning scoring system: clinical audit. Clin Med 2006; 6: 281-4.

16. Sevinc F, Prins JM, Koopmans RP, et al. Early switch from intravenous to oral antibiotics: guidelines and implementation in a large teaching hospital. J Antimicrob Chemother 1999; 43: 601-6.

17. Edwards MS, Andrews JS, Edwards AF, et al. Results of endovascular aortic aneurysm repair with general, regional, and local/monitored anesthesia care in the American College of Surgeons National Surgical Quality Improvement Program database. J Vasc Surg 2011; 54: 1273-82.

18. Thom KA, Schweizer ML, Osih RB, et al. Impact of empiric antimicrobial therapy on outcomes in patients with Escherichia coli and Klebsiella pneumoniae bacteremia: a cohort study. BMC Infect Dis 2008; 8: 116.

19. Athanassa Z, Makris G, Dimopoulos G, et al. Early switch to oral treatment in patients with moderate to severe community-acquired pneumonia: a meta-analysis. Drugs 2008; 68: 2469-81.

20. Fanning M, McKean M, Seymour K, et al. Adherence to guideline based antibiotic treatment for acute exacerbations of COPD in an Australian tertiary hospital. Intern Med J 2014.

21. Marschall J, Agniel D, Fraser VJ, et al. Gram-negative bacteraemia in non-ICU patients: factors associated with inadequate antibiotic therapy and impact on outcomes. J Antimicrob Chemother 2008; 61: 1376-83.


23. Hood HM, Allman RM, Burgess PA, et al. Effects of timely antibiotic administration and culture acquisition on the treatment of urinary tract infection. Am J Med Qual 1998; 13: 195-202.

24. Meehan TP, Fine MJ, Krumholz HM, et al. Quality of care, process, and outcomes in elderly patients with pneumonia. JAMA 1997; 278: 2080-4.

25. Berild D, Mohseni A, Diep LM, et al. Adjustment of antibiotic treatment according to the results of blood cultures leads to decreased antibiotic use and costs. J Antimicrob Chemother 2006; 57: 326-30.

26. Byl B, Clevenbergh P, Jacobs F, et al. Impact of infectious diseases specialists and microbiological data on the appropriateness of antimicrobial therapy for bacteremia. Clin Infect Dis 1999; 29: 60-6.

27. Roson B, Carratala J, Verdaguer R, et al. Prospective study of the usefulness of sputum Gram stain in the initial approach to community-acquired pneumonia requiring hospitalization. Clin Infect Dis 2000; 31: 869-74.

28. Briceland LL, Nightingale CH, Quintiliani R, et al. Antibiotic streamlining from combination therapy to monotherapy utilizing an interdisciplinary approach. Arch Intern Med 1988; 148: 2019-22.

29. Cremers AJ, Sprong T, Schouten JA, et al. Effect of antibiotic streamlining on patient outcome in pneumococcal bacteraemia. J Antimicrob Chemother 2014; 69: 2258-64.

30. van der Eerden MM, Vlaspolder F, de Graaff CS, et al. Comparison between pathogen directed antibiotic treatment and empirical broad spectrum antibiotic treatment in patients with community acquired pneumonia: a prospective randomised study. Thorax 2005; 60: 672-8.

31. Seaton RA, Nathwani D, Phillips G, et al. Clinical record keeping in patients receiving antibiotics in hospital. Health Bull (Edinb ) 1999; 57: 128-33.

32. Nathwani D, Rubinstein E, Barlow G, et al. Do guidelines for community-acquired pneumonia improve the cost-effectiveness of hospital care? Clin Infect Dis 2001; 32: 728-41.

33. Lopez-Cortes LE, Del Toro MD, Galvez-Acebal J, et al. Impact of an evidence-based bundle intervention in the quality-of-care management and outcome of Staphylococcus aureus bacteremia. Clin Infect Dis 2013; 57: 1225-33.

34. Marrie TJ, Lau CY, Wheeler SL, et al. A controlled trial of a critical pathway for treatment of community-acquired pneumonia. CAPITAL Study Investigators. Community-Acquired Pneumonia Intervention Trial Assessing Levofloxacin. JAMA 2000; 283: 749-55.

Chapter 6

136


137

6

APP

END

IX 1

.

TAB

LE 5

a. A

ssocia

tion

betw

een

QIs

and

in-h

ospi

tal m

orta

lity

Qua

lity

indi

cato

r

All

patie

nts (

n =

1890

)A

ppro

pria

te a

ntib

iotic

tr

eatm

ent

Inap

prop

riat

e an

tibio

t-ic

trea

tmen

tN

(%)

Mor

talit

y N

(%)

N (%

)M

orta

lity

N (%

)D

iffer

ence

(9

5% C

I)*

p-va

lue

Empi

rica

l the

rapy

in a

ccor

danc

e w

ith n

atio

nal g

uide

lines

(n

= 13

61)

563

(41)

18 (3

.2)

798

(59)

38 (4

.8)

-0.1

3

(-

0.6

– 0.

36)

0.59

Blo

od c

ultu

res t

aken

(n =

189

0)67

4 (3

6)27

(4)

1216

(64)

55 (4

.5)

-0.2

2

(-0.

6 –

0.21

)0.

32

Cul

ture

take

n of

susp

ecte

d si

te o

f inf

ectio

n (n

= 1

217)

59

5 (4

9)25

(4.2

)62

2 (5

1)28

(4.5

)-0

.02

(-

0.5

– 0.

47)

0.95

Doc

umen

tatio

n of

ant

ibio

tic p

lan

(n =

189

0)11

45 (6

1)54

(4.7

)74

5 (3

9)28

(3.8

)0.

135

(-

0.3

– 0.

54)

0.51

Earl

y sw

itch

if fu

lfilli

ng c

rite

ria

for s

afe

IV-o

ral s

witc

h (n

=

422)

134

(32)

4 (3

)28

8 (6

8)9

(3)

-0.0

1

(-0.

98 –

0.9

)0.

97

Path

ogen

-dir

ecte

d an

tibio

tic th

erap

y if

cultu

re b

ecom

es

posi

tive

(n =

453

)22

8 (5

0)4

(1.8

)22

5 (5

0)9

(4)

-0.2

4

(-1.

2 –

0.7)

0.61

Sum

Sco

re (n

=189

0)10

57 (5

6)45

(4.3

)83

3 (4

4)37

(4.4

)-0

.05

(-

0.4

– 0.

35)

0.82

*Adj

uste

d fo

r con

foun

ders

: age

, sev

erity

of il

lnes

s (M

EWS

scor

e), c

omor

bidi

ty (C

harls

on C

omor

bidi

ty In

dex)

TAB

LE 5

b. A

ssocia

tion

betw

een

QIs

and

in-h

ospi

tal m

orta

lity,

for c

omm

unity

-acq

uire

d in

fectio

ns

Qua

lity

indi

cato

r

Patie

nts w

ith co

mm

unit

y-ac

quir

ed in

fect

ion

(n =

139

1)A

ppro

pria

te a

ntib

iotic

tr

eatm

ent

Inap

prop

riat

e an

tibio

t-ic

trea

tmen

tN

(%)

Mor

talit

y N

(%)

N (%

)M

orta

lity

N (%

)D

iffer

ence

(9

5% C

I)*

p-va

lue

Empi

rica

l the

rapy

in a

ccor

danc

e w

ith n

atio

nal g

uide

lines

46

2 (4

5)11

(2.4

)57

1 (5

5)27

(4.7

)-0

.26

(-

0.8

– 0.

32)

0.38

Blo

od c

ultu

res t

aken

51

7 (3

7)22

(4.3

)87

4 (6

3)29

(3.3

)-0

.06

(-

0.6

– 0.

4)0.

81

Cul

ture

take

n of

susp

ecte

d si

te o

f inf

ectio

n

411

(46)

10 (2

.4)

487

(54)

19 (3

.9)

-0.1

(-0.

7 –

0.5)

0.75

Doc

umen

tatio

n of

ant

ibio

tic p

lan

876

(63)

34 (3

.9)

515

(37)

17 (3

.3)

0.1

(-0.

4 –

0.6)

0.70

Earl

y sw

itch

if fu

lfilli

ng c

rite

ria

for s

afe

IV-o

ral s

witc

h 11

5 (3

4)4

(3.5

)22

0 (6

6)7

(3.2

)0.

09

(-0.

95 –

1.1

)0.

87

Path

ogen

-dir

ecte

d an

tibio

tic th

erap

y if

cultu

re b

ecom

es

posi

tive

153

(49)

3 (2

)15

9 (5

1)6

(3.8

)-0

.16

(-

1.2

– 0.

96)

0.78

Sum

scor

e80

4 (5

8)28

(3.5

)58

7 (4

2)23

(3.9

)-0

.06

(-

0.5

– 0.

4)0.

8

*Adj

uste

d fo

r con

foun

ders

: age

, sev

erity

of il

lnes

s (M

EWS

scor

e)

Chapter 6

138


139

6

TAB

LE 5

c. A

ssocia

tion

betw

een

QIs

and

in-h

ospi

tal m

orta

lity f

or h

ospi

tal-a

cqui

red

infec

tions

Qua

lity

indi

cato

r

Patie

nts w

ith h

ospi

tal-a

cqui

red

infe

ctio

n (n

= 4

99)

App

ropr

iate

ant

ibio

tic

trea

tmen

tIn

appr

opri

ate

antib

iotic

tr

eatm

ent

N (%

)M

orta

lity

N (%

)N

(%)

Mor

talit

y N

(%)

Diff

eren

ce

(95%

CI)

*p-

valu

e

Empi

rica

l the

rapy

in a

ccor

danc

e w

ith n

atio

nal g

uide

lines

10

1 (3

1)7

(6.9

)22

7 (6

9)11

(4.8

)0.

39

(-0.

52 –

1.3

)0.

40

Blo

od c

ultu

res t

aken

157

(31)

5 (3

.2)

342

(69)

26 (7

.6)

-0.5

6

(-1.

4 –

0.3)

0.18

Cul

ture

take

n of

susp

ecte

d si

te o

f inf

ectio

n

184

(58)

15 (8

.2)

135

(42)

9 (6

.7)

0.19

(-

0.6

– 1)

0.66

Doc

umen

tatio

n of

ant

ibio

tic p

lan

269

(54)

20 (7

.4)

230

(46)

11 (4

.8)

0.33

(-0.

4 –

1)0.

36

Earl

y sw

itch

if fu

lfilli

ng c

rite

ria

for s

afe

IV-o

ral s

witc

h 19

(22)

068

(78)

2 (2

.9)

-0.4

8

(-3.

4 –

2.5)

0.75

Path

ogen

-dir

ecte

d an

tibio

tic th

erap

y if

cultu

re b

ecom

es

posi

tive

75 (5

3)1

(1.3

)66

(47)

3 (4

.5)

-0.4

2

(-2.

1 –

1.3)

0.64

Sum

scor

e25

3 (5

1)17

(6.7

)24

6 (4

9)14

(5.7

)0.

14

(-0.

6 –

0.8)

0.70

*Adj

uste

d fo

r con

foun

ders

: com

orbi

dity

(Cha

rlson

Com

orbi

dity

Inde

x)

TAB

LE 6

a. A

ssocia

tion

betw

een

QIs

and

ICU

adm

issio

n

Qua

lity

indi

cato

r

All

patie

nts (

n =

1890

)A

ppro

pria

te a

ntib

iotic

tr

eatm

ent

Inap

prop

riat

e an

tibio

tic

trea

tmen

tN

(%)

ICU

adm

is-

sion

N (%

)N

(%)

ICU

adm

is-

sion

N (%

)D

iffer

ence

(9

5% C

I)*

p-va

lue

Empi

rica

l the

rapy

in a

ccor

danc

e w

ith n

atio

nal g

uide

lines

(n

= 1

361)

563

(41)

21 (3

.7)

798

(59)

35 (4

.4)

-0.0

9

(-0.

5 –

0.4)

0.71

Blo

od c

ultu

res t

aken

(n =

189

0)67

4 (3

6)34

(5)

1216

(64)

43 (3

.5)

0.21

(-

0.2

– 0.

62)

0.32

Cul

ture

take

n of

susp

ecte

d si

te o

f inf

ectio

n (n

= 1

217)

59

5 (4

9)24

(4)

622

(51)

19 (3

.1)

0.14

(-

0.4

– 0.

66)

0.59

Doc

umen

tatio

n of

ant

ibio

tic p

lan

(n =

189

0)11

45 (6

1)47

(4.1

)74

5 (3

9)30

(4)

0.01

(-

0.4

– 0.

4)0.

96

Earl

y sw

itch

if fu

lfilli

ng c

rite

ria

for s

afe

IV-o

ral s

witc

h (n

=

422)

134

(32)

3 (2

.2)

288

(68)

8 (2

.8)

-0.0

8

(-1.

1 –

0.9)

0.87

Path

ogen

-dir

ecte

d an

tibio

tic th

erap

y if

cultu

re b

ecom

es

posi

tive

(n =

453

)22

8 (5

0)18

(7.9

)22

5 (5

0)13

(5.8

)0.

4

(-

0.3

– 1.

1)0.

27

Sum

scor

e10

57 (5

646

(4.4

)83

3 (4

4)31

(3.7

)0.

1

(-

0.32

– 0

.5)

0.67

*Adj

uste

d fo

r con

foun

ders

: non

e

Chapter 6

140


141

6

TAB

LE 6

b. A

ssocia

tion

betw

een

QIs

and

ICU

adm

issio

n, fo

r com

mun

ity-a

cqui

red

infec

tions

Qua

lity

indi

cato

r

Patie

nts w

ith co

mm

unit

y-ac

quir

ed in

fect

ion

(n =

139

1)A

ppro

pria

te a

ntib

iotic

tr

eatm

ent

Inap

prop

riat

e an

tibio

tic

trea

tmen

tN

(%)

ICU

adm

is-

sion

N (%

)N

(%)

ICU

adm

is-

sion

N (%

)D

iffer

ence

(9

5% C

I)*

p-va

lue

Empi

rica

l the

rapy

in a

ccor

danc

e w

ith n

atio

nal g

uide

lines

46

2 (4

5)14

(3)

571

(55)

24 (4

.2)

-0.1

7

(-0.

7 –

0.4)

0.56

Blo

od c

ultu

res t

aken

51

7 (3

7)23

(4.4

)87

4 (6

3)27

(3.1

)0.

19

(-0.

3 –

0.7)

0.44

Cul

ture

take

n of

susp

ecte

d si

te o

f inf

ectio

n

411

(46)

18 (4

.4)

487

(54)

14 (2

.9)

0.22

(-0.

4 –

0.8)

0.48

Doc

umen

tatio

n of

ant

ibio

tic p

lan

876

(63)

32 (3

.7)

515

(37)

18 (3

.5)

0.02

(-0.

5 –

0.5)

0.93

Earl

y sw

itch

if fu

lfilli

ng c

rite

ria

for s

afe

IV-o

ral s

witc

h 11

5 (3

4)3

(2.6

)22

0 (6

6)7

(3.2

)-0

.09

(-

1.1

– 1)

0.88

Path

ogen

-dir

ecte

d an

tibio

tic th

erap

y if

cultu

re b

ecom

es

posi

tive

153

(49)

12 (7

.8)

159

(51)

11 (6

.9)

0.4

(-0.

45 –

1.3

)0.

35

Sum

scor

e80

4 (5

8)32

(4)

587

(42)

18 (3

.1)

0.14

(-

0.36

– 0

.6)

0.59

*Adj

uste

d fo

r con

foun

ders

: non

e

TAB

LE 6

c. A

ssocia

tion

betw

een

QIs

and

ICU

adm

issio

n, fo

r hos

pita

l-acq

uire

d in

fectio

ns

Qua

lity

indi

cato

r

Patie

nts w

ith h

ospi

tal-a

cqui

red

infe

ctio

n (n

= 4

99)

App

ropr

iate

ant

ibio

tic

trea

tmen

tIn

appr

opri

ate

antib

iotic

tr

eatm

ent

N (%

)IC

U a

dmis

-si

on N

(%)

N (%

)IC

U a

dmis

-si

on N

(%)

Diff

eren

ce

(95%

CI)

*p-

valu

e

Empi

rica

l the

rapy

in a

ccor

danc

e w

ith n

atio

nal g

uide

lines

10

1 (3

1)7

(6.9

)22

7 (6

9)11

(4.8

)0.

26

(-

0.6

– 1.

2)0.

58

Blo

od c

ultu

res t

aken

157

(31)

11 (7

)34

2 (6

9)16

(4.7

)0.

48

(-

0.3

– 1.

3)0.

25

Cul

ture

take

n of

susp

ecte

d si

te o

f inf

ectio

n

184

(58)

6 (3

.3)

135

(42)

5 (3

.7)

-0.3

5

(-1.

6 –

0.9)

0.59

Doc

umen

tatio

n of

ant

ibio

tic p

lan

269

(54)

15 (5

.6)

230

(46)

12 (5

.2)

0.21

(-

0.6

– 1)

0.62

Earl

y sw

itch

if fu

lfilli

ng c

rite

ria

for s

afe

IV-o

ral s

witc

h 19

(22)

068

(78)

1 (1

.5)

-0.2

5

(-3.

3 –

2.7)

0.87

Path

ogen

-dir

ecte

d an

tibio

tic th

erap

y if

cultu

re b

ecom

es

posi

tive

75 (5

3)6

(8)

66 (4

7)2

(3)

0.7

(-0.

75 –

2.2

)0.

33

Sum

scor

e25

3(51

)14

(5.5

)24

6 (4

9)13

(5.3

)0.

02

(-0.

7 –

0.74

)0.

97

*Adj

uste

d fo

r con

foun

ders

: non

e

Chapter 6

142


143

6

TAB

LE 7

a. A

ssocia

tion

betw

een

QIs

and

read

miss

ion

Qua

lity

indi

cato

r

All

patie

nts (

n =

1890

)A

ppro

pria

te a

ntib

iotic

tr

eatm

ent

Inap

prop

riat

e an

tibio

tic

trea

tmen

tN

(%)

Re-

adm

is-

sion

N (%

)N

(%)

Re-

adm

is-

sion

N (%

)D

iffer

ence

(9

5% C

I)*

p-va

lue

Empi

rica

l the

rapy

in a

ccor

danc

e w

ith n

atio

nal g

uide

lines

(n

= 1

361)

563

(41)

42 (7

.5)

798

(59)

72 (9

)-0

.15

(-

0.5

– 0.

24)

0.46

Blo

od c

ultu

res t

aken

(n =

189

0)67

4 (3

6)65

(9.6

)12

16 (6

4)91

(7.5

)0.

16

(-0.

17 –

0.5

)0.

33

Cul

ture

take

n of

susp

ecte

d si

te o

f inf

ectio

n (n

= 1

217)

59

5 (4

9)43

(7.2

)62

2 (5

1)53

(8.5

)-0

.14

(-

0.55

– 0

.3)

0.50

Doc

umen

tatio

n of

ant

ibio

tic p

lan

(n =

189

0)11

45 (6

1)98

(8.5

)74

5 (3

9)58

(7.8

)0.

06

(-

0.3

– 0.

4)0.

71

Earl

y sw

itch

if fu

lfilli

ng c

rite

ria

for s

afe

IV-o

ral s

witc

h (n

=

422)

134

(32)

6 (4

.5)

288

(68)

19 (6

.6)

-0.3

(-1.

2 –

0.6)

0.48

Path

ogen

-dir

ecte

d an

tibio

tic th

erap

y if

cultu

re b

ecom

es

posi

tive

(n =

453

)22

8 (5

0)17

(7.5

)22

5 (5

0)20

(8.9

)-0

.13

(-

0.8

– 0.

6)0.

72

Sum

scor

e10

57 (5

6)90

(8.5

)83

3 (4

4)66

(7.9

)0.

02

(-0.

3 –

0.35

)0.

91

*Adj

uste

d fo

r con

foun

ders

: sev

erity

of il

lnes

s (M

EWS)

, prio

r use

of a

ntib

iotic

s

TAB

LE 7

b. A

ssocia

tion

betw

een

QIs

and

read

miss

ion,

for c

omm

unity

-acq

uire

d in

fectio

ns

Qua

lity

indi

cato

r

Patie

nts w

ith co

mm

unit

y-ac

quir

ed in

fect

ion

(n =

139

1)‘A

ppro

pria

te’ a

ntib

iotic

tr

eatm

ent

‘Inap

prop

riat

e’ a

ntib

iot-

ic tr

eatm

ent

N (%

)R

e-ad

mis

-si

on N

(%)

N (%

)R

e-ad

mis

-si

on N

(%)

Diff

eren

ce

(95%

CI)

*p-

valu

e

Empi

rica

l the

rapy

in a

ccor

danc

e w

ith n

atio

nal g

uide

lines

46

2 (4

5)37

(8)

571

(55)

53 (9

.3)

-0.1

(-0.

54 –

0.3

)0.

65

Blo

od c

ultu

res t

aken

51

7 (3

7)50

(9.7

)87

4 (6

3)73

(8.4

)0.

13

(-0.

2 –

0.5)

0.51

Cul

ture

take

n of

susp

ecte

d si

te o

f inf

ectio

n

411

(46)

31 (7

.5)

487

(54)

43 (8

.8)

-0.2

(-0.

67 –

0.3

)0.

4

Doc

umen

tatio

n of

ant

ibio

tic p

lan

876

(63)

81 (9

.2)

515

(37)

42 (8

.2)

0.11

(-

0.3

– 0.

5)0.

57

Earl

y sw

itch

if fu

lfilli

ng c

rite

ria

for s

afe

IV-o

ral s

witc

h 11

5 (3

4)5

(4.3

)22

0 (6

6)12

(5.5

)-0

.01

(-

1 –

0.9)

0.98

Path

ogen

-dir

ecte

d an

tibio

tic th

erap

y if

cultu

re b

ecom

es

posi

tive

153

(49)

12 (7

.8)

159

(51)

14 (8

.8)

-0.1

7

(-1.

01 –

0.7

)0.

71

Sum

scor

e80

4 (5

8)74

(9.2

)58

7 (4

2)49

(8.3

)0.

04

(-0.

3 –

0.42

)0.

82

*Adj

uste

d fo

r con

foun

ders

: age

, com

orbi

dity

(Cha

rlson

Com

orbi

dity

Inde

x), p

rior u

se of

ant

ibio

tics,

sex

Chapter 6

144

TAB

LE 7

c. A

ssocia

tion

betw

een

QIs

and

read

miss

ion,

for h

ospi

tal-a

cqui

red

infec

tions

Qua

lity

indi

cato

r

Patie

nts w

ith h

ospi

tal-a

cqui

red

infe

ctio

n (n

= 4

99)

App

ropr

iate

ant

ibio

tic

trea

tmen

tIn

appr

opri

ate

antib

iotic

tr

eatm

ent

N (%

)R

e-ad

mis

-si

on N

(%)

N (%

)R

e-ad

mis

-si

on N

(%)

Diff

eren

ce

(95%

CI)

*p-

valu

e

Empi

rica

l the

rapy

in a

ccor

danc

e w

ith n

atio

nal g

uide

lines

10

1 (3

1)5

(8.4

)22

7 (6

9)19

(8.4

)-0

.4

(-

1.3

– 0.

5)0.

39

Blo

od c

ultu

res t

aken

157

(31)

15 (9

.6)

342

(69)

18 (5

.3)

0.49

(-0.

19 –

1.2

)0.

16

Cul

ture

take

n of

susp

ecte

d si

te o

f inf

ectio

n

184

(58)

12 (6

.5)

135

(42)

10 (7

.4)

-0.1

(-0.

95 –

0.7

)0.

81

Doc

umen

tatio

n of

ant

ibio

tic p

lan

269

(54)

17 (6

.3)

230

(46)

16 (7

)-0

.1

(-

0.8

– 0.

6)0.

83

Earl

y sw

itch

if fu

lfilli

ng c

rite

ria

for s

afe

IV-o

ral s

witc

h 19

(22)

1 (5

.3)

68 (7

8)7

(10.

3)-0

.57

(-

2.6

– 1.

5)0.

58

Path

ogen

-dir

ecte

d an

tibio

tic th

erap

y if

cultu

re b

ecom

es

posi

tive

75 (5

3)5

(6.7

)66

(47)

6 (9

.1)

-0.0

32

(-1.

3 –

1.2)

0.96

Sum

scor

e25

3 (5

1)16

(6.3

)24

6 (4

9)17

(6.9

)-0

.06

(-

0.7

– 0.

6)0.

86

*Adj

uste

d fo

r con

foun

ders

: non

e

General Discussion

Chapter 7

Chapter 7

146

General Discussion

147

7

Many interventions and programs have been developed to improve appropriate antibiotic use in terms of choice of antibiotics, dosing, timing, de-escalation and duration. These programs are collectively known as antibiotic stewardship programs.1 Their objective is to optimize clinical outcome while minimizing unintended consequences of antibiotic use, including the emergence of resistance.2 An essential tool for an effective stewardship program in order to set priorities and focus improvement is the ability to reliably measure the appropriateness of hospital antibiotic use. This thesis focused on the development and validation of quality indicators for appropriate antibiotic use. MAIN FINDINGSComprehensiveness and compliance of local guidelinesA local, evidence-based hospital antimicrobial guide is an essential element of a successful stewardship program. In chapter 2 we studied the local antimicrobial guides in use in Dutch hospitals. These local guides are either based on the national antimicrobial guide (SWAB-ID), customized to local resistance patterns, or they are locally developed (non-SWAB-ID). In this chapter we compared the local SWAB-ID based with the non-SWAB-ID based antimicrobial guides for comprehensiveness and compliance with the national treatment guidelines. Fifty antimicrobial guides, covering nearly every hospital in the Netherlands, were scored on 199 pre-specified items to measure comprehensiveness (e.g. presence of recommendations for prophylaxis or the empirical treatment of cholangitis) and on 35 items to measure compliance with the national guidelines (e.g. the recommendation for empirical treatment of diverticulitis). The non-SWAB-ID based local antimicrobial guides (n=27) were significantly less comprehensive (p < 0.001) and less guideline-compliant (p < 0.001) than the SWAB-ID based local antimicrobial guides (n=23). In conclusion, the use of a local version of the national SWAB-ID antimicrobial guide significantly increased comprehensiveness and guideline-compliance of the local antimicrobial policy and the recommendations more often trace back to evidence-based guidelines. A comprehensive national guideline like SWAB-ID with customized local versions may therefore help to improve the quality of local antimicrobial policy.

Development of quality indicatorsIn 2010 SWAB published a guideline regarding antimicrobial use in hospitalized patients with sepsis.3 Sepsis is a common reason for clinicians to start with antibiotics, and severe sepsis or septic shock is a frequent cause of in-hospital mortality.4-6 In chapter 3 we utilized this guideline to develop a concise set of QIs to assess the appropriateness of antimicrobial use in hospitalized adult patients with sepsis, using the RAND-modified Delphi method. A Dutch multidisciplinary panel of 14 experts appraised and prioritized 40 key recommendations derived from the SWAB Sepsis guideline during two rounds of questionnaires with an in-between face-to-face consensus meeting. This systematic, stepwise method, which combined evidence with expert opinion, led to a concise and therefore feasible set of 5 QIs. The final set comprised the following QIs: obtain cultures; prescribe empirical antimicrobial therapy according to the national guideline; start intravenous drug therapy; start antimicrobial treatment within one hour; and streamline antimicrobial therapy.

The aim of the study described in chapter 4 was to develop generic quality indicators which can be used to measure the appropriateness of antibiotic use in the treatment of all bacterial infections in hospitalized non-ICU adult patients. The RAND-modified Delphi consensus method was once again used, after potential QIs were retrieved from the literature. An international multidisciplinary expert panel (17 experts) appraised and prioritized 24 potential QIs in two questionnaire mailings with an in-between face-to-face consensus meeting. Eleven QIs were selected. Nine QIs described recommended care at the patient level (i.e. process indicators): (1) take at least two sets of blood cultures before starting systemic antibiotic therapy, (2) take cultures from suspected sites of infection before starting systemic antibiotic therapy, (3) prescribe empirical antibiotic therapy according to the local guideline, (4) change empirical to pathogen-directed therapy as soon as culture results become available, (5) adapt dose and dosing interval of systemic antibiotics to renal function, (6) switch systemic antibiotic therapy from intravenous to oral antibiotic therapy after 48 -72 hours on the basis of the clinical condition and when oral treatment is adequate, (7) document antibiotic plan in the case notes at the start of systemic antibiotic treatment, (8) perform therapeutic drug monitoring, and (9) discontinue antibiotic therapy if infection is not confirmed. Two structure QIs describing recommended care at the hospital

Chapter 7

148

General Discussion

149

7

level were: (10) a local antibiotic guideline should be present in the hospital, and (11) these local guidelines should correspond to the national antibiotic guidelines. Both sets of QIs (sepsis and generic) can be used in antibiotic stewardship programs to determine for which aspects of antibiotic use there is room for improvement.

Quality indicators in daily hospital careAfter developing QIs, assessing their feasibility in daily practice is essential before using them to measure the quality of antibiotic use. This was done in chapter 5 by testing important clinimetric characteristics. An observational multicenter study was conducted and during a point prevalence measurement in 22 hospitals in the Netherlands on non-Intensive Care Unit departments, 1890 patients treated with antibiotics for a suspected bacterial infection were included. In this cohort we tested the measurability, applicability, reliability, room for improvement and case mix stability of the generic QIs. Low applicability (≤ 10%) was found for the QIs ‘perform therapeutic drug monitoring’, ‘adapt antibiotic dosage to renal function’ and ‘discontinue empirical therapy in case of lack of clinical and/or microbiological evidence of infection’. For the latter, we also found a low inter-observer agreement (kappa < 0.4). The QI ‘a local antibiotic guideline should be present’ showed a low improvement potential. In conclusion, seven of the eleven QIs (i.e. six process indicators and one structure indicator) had sound clinimetric properties. Case-mix correction was necessary for most process QIs. For all QIs, we found ample room for improvement and large variation between hospitals.

Quality indicators and outcomeChapter 6 addresses the impact of appropriate antibiotic use on length of hospital stay (LOS). Data from the previously described observational multicenter study in 1890 patients using antibiotics for a suspected bacterial infection were used. Appropriate antibiotic use in hospitalized patients was defined by the previous six process QIs with sound clinimetric properties. From the patients’ medical charts data were retrieved to determine QI performance scores and LOS. Performance scores were calculated for all QIs separately (appropriate or not) and a sum score described the performance on the total set of QIs. This sum score was defined as the overall performance on all QIs

(appropriate =1 and inappropriate =0) in a patient, divided by the number of QIs that applied to that specific patient. We divided the sum scores into two groups: low sum scores (0 – 49%) versus high scores (50 – 100%). LOS was log-transformed before analyses were done, to satisfy normality assumptions, and afterwards LOS was back-transformed for presentation as geometric mean (95% confidence interval). Multilevel mixed model analyses, correcting for confounders, were used to correlate QI performance (single and combined) with LOS. Sub-analyses were conducted for patients with community-acquired versus hospital-acquired infections. The geometric mean for LOS was 10.2 days form community acquired infections and 11.9 days for hospital-acquired infections. An association was demonstrated between an early i.v.-oral switch and a shorter LOS (geometric mean 6.5 vs. 11.2 days; P <0.001), also in both subgroups separately. Similarly, a high sum score was associated with a shorter LOS (geometric mean 10.1 vs. 11.2 days; P = 0.002), this also applied to the subgroup community-acquired infections (geometric mean 9.7 vs. 10.9 days; P = 0.007). The subgroup hospital-acquired infections (geometric mean 11.6 vs. 12.2 days; P = 0.08) was smaller and although patients with a high sum score had a shorter length of hospital stay, the difference was not significant. In conclusion, appropriate antibiotic use, defined by validated process QIs in hospitalized patients with a suspected bacterial infection, seems to reduce length of stay with one day and therefore positively contributes to patient outcome and healthcare costs.

IMPLICATIONS FOR PRACTICEUsing guidelines as a starting point for quality measurementsIn this thesis guidelines play an important role as starting point for defining and measuring appropriate antibiotic use. In chapter 2 we measured comprehensiveness and guideline-compliance of the Dutch local hospital antibiotic guides. The use of an online, locally customized version of the national SWAB-ID antimicrobial guide significantly increased comprehensiveness and compliance with the national guidelines of the local antimicrobial policy. Studies show that developing trustworthy evidence-based clinical practice guidelines requires considerable expertise, time and financial support, which are mostly not available at the local level.7, 8 Adaption of international clinical practice guidelines for the use at the national or the local hospital levels could save a lot of time and could also improve their implementation.8-10 It takes specific

Chapter 7

150

General Discussion

151

7

national or local contexts into account, for example the local resistance patterns of causative microorganisms of common bacterial infections, while limiting unnecessary duplication.8 The national antibiotic guide SWAB-ID, which can be locally customized, is a good example of this concept, and the fact that the local version is web-based makes it easy to update the guide when for instance resistance rates are changing. Therefore, we concluded that a comprehensive national guideline like SWAB-ID with customized local versions may be an effective way to improve the quality of local antimicrobial policy. Clinical practice guidelines are thought to be a cornerstone of evidence-based medicine and the international interest in clinical practice guidelines seems to increase, but the development of guidelines is not without problems.11, 12 Scientific evidence for recommendations is sometimes lacking, misleading or misinterpreted, and guidelines do not always meet the basic quality requirements.12-14 Therefore, rigorous methods for developing trustworthy evidence-based guidelines have been proposed, 11, 14-17 but less attention has been paid to the updating of a guideline or the development of guideline-based quality indicators.18, 19 A previous study by our group showed that two years after publication of a national guideline for complicated urinary tract infections, relatively high inadequate coverage rates of many guideline recommended empiric treatment options were found. This was probably because of continuously changing resistance rates and differences between the epidemiology of uropathogens assumed in the guideline and those in real-life.20

Even when recommendations are based on a low level of evidence, antibiotic guideline adherence has been shown to be associated with a decreased mortality, a lower admission rate to the intensive care unit (ICU) and a shorter length of hospital stay.21-27 In addition, we previously also showed that guideline adherence seems to increase coverage rates without prescribing unnecessary broad regimens.20 This is an important finding, because decreasing the use of broad-spectrum antimicrobial therapy is an important strategy to contain the emergence of resistant pathogens.28, 29

Delphi procedureGuidelines and international literature are used to systematically develop QIs.30-32 In a systematic review regarding the Delphi method, Boulkedid et al. concluded that a systematic and well-designed procedure increases the

reliability and rigorousness of the QIs, and they created a practical guidance with recommendations for planning, using and reporting the Delphi procedure.33 In chapter 3 and 4 we developed indicators using an international, multidisciplinary panel of experts, with discussion of the indicators in a face-to-face meeting, in accordance with the recommendations of Boulkedid et al.

When raising the question “what is appropriate care?” the answer will often be “it depends”. It depends on which clinicians are asked, where they live and work, what weight is given to different types of evidence and endpoints, and whether one considers the preferences of patients and families and level of resources in a given health system.34 Therefore, consensus methods are used to define appropriate care and these are important when assessing many aspects of performance in which evidence alone is insufficient to support recommendations.35 The RAND modified Delphi procedure is well-known and valuable for achieving consensus about issues or potential QIs when no QIs existed previously.33 However, the reproducibility of the Delphi procedure is not perfect; it varied from good to only moderate agreement, although the reliability of panels rating the same set of QIs was generally regarded as acceptable.34, 36,

37 Ayanian et al. showed that over a 1000 surveyed physicians agreed for most indications with the Delphi expert panel about the appropriateness of coronary angiography after myocardial infarction, concluding that well-designed expert panels can closely reflect the views of practicing physicians.38

Experts contribute to the content validity of indicators, because they interpret potential QIs and integrate results from studies with information from daily clinical practice.39 We carefully planned the Delphi procedures and the expert panel: an international multidisciplinary panel in which all the main specialism’s involved in antibiotic treatment were represented. This was important, because studies have shown that panel composition influences ratings, and heterogeneity in a panel contributes to different point of views about quality of care, which improves the result of the Delphi procedure.33, 40 Hutchings et al. evaluated the differences between a consensus procedure with or without a face-to-face meeting. They found that indeed opinions of experts are more likely to shift when groups meet.41 Discussion resulted in more complete or revised indicators that incorporated the views of multiple experts.42

Chapter 7

152

General Discussion

153

7

Generic Quality Indicators in Clinical PracticeIn chapter 4 and 5 we developed a valid set of generic QIs that can be used to measure, on the patient level, the various steps in the process of antibiotic use along the entire antibiotic pathway. An important strength of these indicators is that they are generic indicators, and therefore enable comparison of antibiotic use between different infectious diseases across hospitals. Instead of developing sets for each infectious disease, one set can be used to measure the appropriateness of antibiotic use in all infectious diseases. Moreover, it creates opportunities to measure the appropriateness of antibiotic use for more rare infectious diseases or when there are two possible diagnoses. Another potential benefit is that generic indicators are less susceptible to change, because they contain little detailed information.41

To test the QIs in daily clinical practice, we collected data by means of a point prevalence measurements and a retrospective chart review. Subsequently, to calculate the QI scores, data were inserted into constructed algorithms which captured the denominator and numerator of the QIs, to minimize variation when interpreting appropriateness of antibiotic use. This was a major strength of this study and can possibly also explain the high inter-observer reliability. Although these algorithms have been rigorously constructed, disadvantages of the method we used are that the assessment was retrospective, and potentially justified deviations from the guideline may have been disregarded, because not all data that influenced the physicians’ choices may have been documented properly, especially in complex treatment decisions. Collecting the data prospectively might have diminished this problem.

Generic Quality Indicators and antibiotic stewardship programsThese QIs can potentially represent “Antibiotic Stewardship QIs”, which can be used by inpatient treatment facilities and governmental agencies for the purposes of estimating appropriate antibiotic use in acute care hospitals. There is a great need for such reliable process indicators for acute care hospitals throughout the world, because stewardship programs have been implemented at many hospitals in Europe and the United States, but assessing the true effect of these programs remains challenging and their impact has been difficult to measure.43, 44 Measures to evaluate antimicrobial stewardship can be split up in four main categories: patient outcomes, unintended consequences, antimicrobial utilization and costs, and process measures.43

Antimicrobial utilization and costs are frequently used to assess stewardship programs, among others because this type of data is easy to collect and this measure is of great interest for administrators to justify these programs.43, 45-48 Generally there is a great interest in outcome measures, because they reflect multiple aspects of appropriate care and improving outcome is the ultimate goal of the intervention. Only, evaluating patient outcome data without some quantification of associated stewardship process strategies is not useful.43 So none of these measures is on its own adequate enough to assess the overall effect of stewardship programs. The combination of both process QIs and related patient-specific outcomes seem to be the most effective way to measure appropriate antibiotic care.43 Therefore in chapter 6 we addressed the impact of appropriate antibiotic use, defined by process QIs, on length of hospital stay (LOS). Controlling for multiple confounders, appropriate antibiotic use, defined by the sum score of these process QIs, was associated with a reduction in length of stay with one day.

It seems obvious that appropriate antibiotic use improves clinical outcome. However, measuring clinical outcome is challenging, since a large sample size is needed to find clinically meaningful and significant differences in outcome measures like mortality, ICU admission or readmission rate.49-51 Furthermore, outcomes like clinical response, LOS or mortality depend on various aspects of care and not only on the quality of antibiotic care. Controlling for confounding factors is essential and can be difficult.43

IMPLICATIONS FOR FUTURE RESEARCHMeasuring appropriate antibiotic useIn this thesis, the method of using generic quality indicators (QIs) for measuring appropriateness of antibiotic use has been validated. An important limitation of this measurement method was that the used method, retrospectively hand searching medical charts, was very resource- and labor intensive. To enhance future use of the QIs, we evaluated possibilities for data reduction. First, we examined the relationship between the various process QIs, with the ultimate aim of reducing the number of QIs, and therefore the data collection necessary to assess ‘appropriate antibiotic use’. Since no correlation was found, this was not possible. Second, we explored what the consequences would be when

Chapter 7

154

General Discussion

155

7

certain patient characteristics (allergies, pregnancy, and previous ESBL infection) were not taken into account when computing the algorithm of the QI ‘prescribe empirical antibiotic therapy according to the guideline’. We found that excluding allergies, pregnancy and previous ESBL infection changed the mean performance rate by only 1.1%. Omitting these parameters when assessing this QI could be considered, however still many patient parameters are needed when assessing appropriate antibiotic use. In conclusion, major data reduction to assess ‘appropriate antibiotic use’ using quality indicator performance was not possible and so this method is still very labour intensive. Therefore, it is important to consider other methods. Various methods can be used to evaluate the quality of antibiotic use in hospitals, ranging from continuously monitoring overall antibiotic use at an institutional- or organizational level, to performing point prevalence studies in which appropriate use is assessed in individual patients, like was done in this thesis. These methods have never been compared and at present, the (cost) effectiveness of these various options in measuring and providing feedback information on antibiotic use is unknown. Future research should evaluate whether monitoring of overall antibiotic use suffices, or whether labor-intensive and costly point prevalence studies are cost-effective when evaluating appropriateness of antibiotic use. At present, a multicenter study is being performed in 21 hospitals in the Netherlands, comparing the (cost) effectiveness of the various approaches (the “IMPACT”-study, funded by ZonMw).

Improving appropriateness of antibiotic useOnly a few studies, including ours, have been performed to examine the relationship between multiple care processes and the clinical outcome of hospitalized patients treated with antibiotics for a suspected bacterial infection.25, 27 For patients with lower respiratory tract infections (LRTI), complicated urinary tract infections (UTI) or sepsis, guideline-concordant empirical therapy has been associated with a lower mortality rate, a decreased admission rate to the intensive care unit (ICU) and a decreased length of hospital stay.21-27 However, the most important conclusion of chapter 6 was that not adherence to one single QI, but that adherence to a combination of concomitant processes of care resulted in a decreased duration of hospital stay.

These findings show resemblance with the concept of ‘care bundles’: a small set of practices (usually three to six elements), that, when implemented together, are expected to result in better outcome than when implemented individually. Care bundles should be delivered by one healthcare team at one point in time.52 An example is the bundle for the management of severe sepsis and septic shock. Extensive guidelines for these diseases have been summarized into the ‘Surviving Sepsis Campaign (SSC) care bundles’, including a bundle that should be completed within 3 hours and another bundle that must be done within 6 hours. Using the SSC bundle led to sustained quality improvement in sepsis care and was associated with reduced mortality.5 Also checklists have been shown to increase patient safety and patient outcome. The reasoning behind a checklist is the same as a bundle. For example, in general surgery, implementation of a comprehensive multidisciplinary SURgical Patient Safety System (SURPASS) was associated with an absolute reduction in surgical complications of 10.6 percentage points and a decrease of the mortality rate of 0.7 percentage points.53 Similarly, the introduction of a checklist to improve patient care among gynecologic oncology patients resulted into a decreased length of stay of one day.54

The next step, after developing and validating generic QIs, is the embedding of appropriate antibiotic use in daily clinical practice. The development of an antibiotic checklist can be a tool for this implementation. The construction of the antibiotic checklist can be based on the generic quality indicators. Translating the QIs into checklist items would mean separating the QIs into two bundles: an early checklist (day 0) with the QIs ‘perform blood cultures’, perform culture from suspected site of infection’, ‘prescribe empirical therapy in accordance with the national guideline’, ‘documentation of antibiotic plan’, followed by a checklist on day 3 with the QIs ‘switch from intravenous to oral treatment’ and ‘adapt empirical to pathogen-directed antibiotic therapy as soon as culture results become available’. At present, a cluster-randomized multicenter trial has started in the Netherlands to analyse the effect of the introduction of this antibiotic checklist on intensive care unit (ICU)- and hospital stay, adequate treatment, mortality rates, total antibiotic use, and costs (funded by ZonMw).55

Chapter 7

156

General Discussion

157

7

In conclusion, this thesis provides a reliable set of generic QIs which can be used to measure and improve the various steps in the process of antibiotic use in the hospital, by identifying for which step along the antibiotic pathway there is room for improvement. In particular adherence to a combination of generic process QIs, rather than just one QI, seems to decrease the duration of hospital stay. Prospective studies are needed to determine whether interventions that improve compliance with these validated generic QIs, could improve patient outcome and reduces length of stay.

Reference List

1. Bartlett JG. A call to arms: the imperative for antimicrobial stewardship. Clin Infect Dis 2011; 53 Suppl 1: S4-S7.


3. SWAB guidelines for Antibacterial therapy of adult patients with sepsis. In: 2010.4. Angus DC, Linde-Zwirble WT, Lidicker J, et al. Epidemiology of severe sepsis in

the United States: analysis of incidence, outcome, and associated costs of care. Crit Care Med 2001; 29: 1303-10.


6. Levy MM, Dellinger RP, Townsend SR, et al. The Surviving Sepsis Campaign: results of an international guideline-based performance improvement program targeting severe sepsis. Intensive Care Med 2010; 36: 222-31.

7. Eccles M, Clapp Z, Grimshaw J, et al. Developing valid guidelines: methodological and procedural issues from the North of England Evidence Based Guideline Development Project. Qual Health Care 1996; 5: 44-50.

8. Kristiansen A, Brandt L, Agoritsas T, et al. Adaptation of trustworthy guidelines developed using the GRADE methodology: A novel 5-step process. Chest 2014.

9. Fervers B, Burgers JS, Haugh MC, et al. Adaptation of clinical guidelines: literature review and proposition for a framework and procedure. Int J Qual Health Care 2006; 18: 167-76.

10. Harrison MB, Graham ID, van den Hoek J, et al. Guideline adaptation and implementation planning: a prospective observational study. Implement Sci 2013; 8: 49.

11. Development and validation of an international appraisal instrument for assessing the quality of clinical practice guidelines: the AGREE project. Qual Saf Health Care 2003; 12: 18-23.

12. Khan AR, Khan S, Zimmerman V, et al. Quality and strength of evidence of the Infectious Diseases Society of America clinical practice guidelines. Clin Infect Dis 2010; 51: 1147-56.

13. Lee DH, Vielemeyer O. Analysis of overall level of evidence behind Infectious Diseases Society of America practice guidelines. Arch Intern Med 2011; 171: 18-22.

14. Woolf SH, Grol R, Hutchinson A, et al. Clinical guidelines: potential benefits, limitations, and harms of clinical guidelines. BMJ 1999; 318: 527-30.

Chapter 7

158

General Discussion

159

7

15. Goldet G, Howick J. Understanding GRADE: an introduction. J Evid Based Med 2013; 6: 50-4.

16. Guyatt GH, Oxman AD, Kunz R, et al. Going from evidence to recommendations. BMJ 2008; 336: 1049-51.

17. Guyatt GH, Oxman AD, Vist GE, et al. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ 2008; 336: 924-6.

18. Martinez GL, Arevalo-Rodriguez I, Sola I, et al. Strategies for monitoring and updating clinical practice guidelines: a systematic review. Implement Sci 2012; 7: 109.

19. Shekelle P, Eccles MP, Grimshaw JM, et al. When should clinical guidelines be updated? BMJ 2001; 323: 155-7.

20. Spoorenberg V, Prins JM, Stobberingh EE, et al. Adequacy of an evidence-based treatment guideline for complicated urinary tract infections in the Netherlands and the effectiveness of guideline adherence. Eur J Clin Microbiol Infect Dis 2013; 32: 1545-56.

21. Asadi L, Eurich DT, Gamble JM, et al. Impact of guideline-concordant antibiotics and macrolide/beta-lactam combinations in 3203 patients hospitalized with pneumonia: prospective cohort study. Clin Microbiol Infect 2013; 19: 257-64.

22. Dambrava PG, Torres A, Valles X, et al. Adherence to guidelines’ empirical antibiotic recommendations and community-acquired pneumonia outcome. Eur Respir J 2008; 32: 892-901.

23. Lee SS, Kim Y, Chung DR. Impact of discordant empirical therapy on outcome of community-acquired bacteremic acute pyelonephritis. J Infect 2011; 62: 159-64.


25. Menendez R, Torres A, Reyes S, et al. Initial management of pneumonia and sepsis: factors associated with improved outcome. Eur Respir J 2012; 39: 156-62.

26. Shorr AF, Micek ST, Welch EC, et al. Inappropriate antibiotic therapy in Gram-negative sepsis increases hospital length of stay. Crit Care Med 2011; 39: 46-51.

27. Spoorenberg V, Hulscher ME, Akkermans RP, et al. Appropriate antibiotic use for patients with urinary tract infections reduces length of hospital stay. Clin Infect Dis 2014; 58: 164-9.

28. Safdar N, Maki DG. The commonality of risk factors for nosocomial colonization and infection with antimicrobial-resistant Staphylococcus aureus, enterococcus, gram-negative bacilli, Clostridium difficile, and Candida. Ann Intern Med 2002; 136: 834-44.

29. Singh N, Rogers P, Atwood CW, et al. Short-course empiric antibiotic therapy for patients with pulmonary infiltrates in the intensive care unit. A proposed solution for indiscriminate antibiotic prescription. Am J Respir Crit Care Med 2000; 162: 505-11.

30. Campbell SM, Braspenning J, Hutchinson A, et al. Research methods used in developing and applying quality indicators in primary care. BMJ 2003; 326: 816-9.

31. Coenen S, Ferech M, Haaijer-Ruskamp FM, et al. European Surveillance of Antimicrobial Consumption (ESAC): quality indicators for outpatient antibiotic use in Europe. Qual Saf Health Care 2007; 16: 440-5.

32. Schouten JA, Hulscher ME, Wollersheim H, et al. Quality of antibiotic use for lower respiratory tract infections at hospitals: (how) can we measure it? Clin Infect Dis 2005; 41: 450-60.

33. Boulkedid R, Abdoul H, Loustau M, et al. Using and reporting the Delphi method for selecting healthcare quality indicators: a systematic review. PLoS One 2011; 6: e20476.

34. Naylor CD. What is appropriate care? N Engl J Med 1998; 338: 1918-20.35. Kotter T, Blozik E, Scherer M. Methods for the guideline-based development of

quality indicators--a systematic review. Implement Sci 2012; 7: 21.36. Marshall MN, Shekelle PG, McGlynn EA, et al. Can health care quality indicators

be transferred between countries? Qual Saf Health Care 2003; 12: 8-12.37. Shekelle PG, Kahan JP, Bernstein SJ, et al. The reproducibility of a method to

identify the overuse and underuse of medical procedures. N Engl J Med 1998; 338: 1888-95.

38. Ayanian JZ, Landrum MB, Normand SL, et al. Rating the appropriateness of coronary angiography--do practicing physicians agree with an expert panel and with each other? N Engl J Med 1998; 338: 1896-904.

39. Braspenning J, Hermens R, Wollersheim H, et al. Meten van (veranderingen in) de zorg: de rol van indicatoren. In: Grol R, Wensing M, eds. Implementatie: effectieve verbetering van de patientenzorg. Amsterdam: Elsevier gezondheidszorg, 2010; 153-75.

40. Campbell SM, Hann M, Roland MO, et al. The effect of panel membership and feedback on ratings in a two-round Delphi survey: results of a randomized controlled trial. Med Care 1999; 37: 964-8.

41. Wennekes L. Development and Validation of Quality Indicators for Cancer Care. Thesis. 2013; 8-158.

42. Hutchings A, Raine R, Sanderson C, et al. A comparison of formal consensus methods used for developing clinical guidelines. J Health Serv Res Policy 2006; 11: 218-24.

43. Dodds Ashley ES, Kaye KS, DePestel DD, et al. Antimicrobial stewardship: philosophy versus practice. Clin Infect Dis 2014; 59 Suppl 3: S112-S121.

44. Kaki R, Elligsen M, Walker S, et al. Impact of antimicrobial stewardship in critical care: a systematic review. J Antimicrob Chemother 2011; 66: 1223-30.

45. Beardsley JR, Williamson JC, Johnson JW, et al. Show me the money: long-term financial impact of an antimicrobial stewardship program. Infect Control Hosp Epidemiol 2012; 33: 398-400.

Chapter 7

160

46. Ng CK, Wu TC, Chan WM, et al. Clinical and economic impact of an antibiotics stewardship programme in a regional hospital in Hong Kong. Qual Saf Health Care 2008; 17: 387-92.

47. Sick AC, Lehmann CU, Tamma PD, et al. Sustained savings from a longitudinal cost analysis of an internet-based preapproval antimicrobial stewardship program. Infect Control Hosp Epidemiol 2013; 34: 573-80.

48. Standiford HC, Chan S, Tripoli M, et al. Antimicrobial stewardship at a large tertiary care academic medical center: cost analysis before, during, and after a 7-year program. Infect Control Hosp Epidemiol 2012; 33: 338-45.

49. Khadem TM, Dodds AE, Wrobel MJ, et al. Antimicrobial stewardship: a matter of process or outcome? Pharmacotherapy 2012; 32: 688-706.

50. McGowan JE. Antimicrobial stewardship--the state of the art in 2011: focus on outcome and methods. Infect Control Hosp Epidemiol 2012; 33: 331-7.

51. Almirante B, Garnacho-Montero J, Pachon J, et al. Scientific evidence and research in antimicrobial stewardship. Enferm Infecc Microbiol Clin 2013; 31 Suppl 4: 56-61.

52. Marwick C, Davey P. Care bundles: the holy grail of infectious risk management in hospital? Curr Opin Infect Dis 2009; 22: 364-9.

53. de Vries EN, Prins HA, Crolla RM, et al. Effect of a comprehensive surgical safety system on patient outcomes. N Engl J Med 2010; 363: 1928-37.

54. Diaz-Montes TP, Cobb L, Ibeanu OA, et al. Introduction of checklists at daily progress notes improves patient care among the gynecological oncology service. J Patient Saf 2012; 8: 189-93.

55. van Daalen FV, Prins JM, Opmeer BC, et al. A cluster randomized trial for the implementation of an antibiotic checklist based on validated quality indicators: the AB-checklist. BMC Infect Dis 2015; 15: 134.

Chapter 8Appendices

Chapter 8

162

Nederlandse samenvatting

163

8


Een gedeelte van deze samenvatting is eerder gepubliceerd in het Nederlands Tijdschrift voor Anesthesiologie1. Deze samenvatting geeft een overzicht van de verschillende hoofdstukken (1 t/m 7) van dit proefschrift. De hoofdstukken worden hieronder afzonderlijk, maar ook in onderlinge samenhang besproken.

Hoofdstuk 1 Inleiding - antibiotica resistentieDe ontdekking van de antibiotica in de jaren dertig van de vorige eeuw veroorzaakte een dramatische afname van de mortaliteit van infectieziekten doordat het effectief behandelen van bacteriële infecties mogelijk werd. Antibiotica, vaak ook wel wonder drugs genoemd, hebben hierdoor de gezondheidszorg in de twintigste eeuw fundamenteel veranderd1,2. Vandaag de dag zijn antibiotica niet meer weg te denken uit de moderne geneeskunde. Het uitvoeren van grote operaties, orgaantransplantaties, plaatsen van protheses en chemotherapie zou niet mogelijk zijn zonder een effectieve behandeling voor bacteriële infecties1,2. Dit succesverhaal kent echter een keerzijde. Het veelvuldig gebruik van antibiotica is de belangrijkste oorzaak van het ontstaan van resistente micro-organismen. Dit brengt de effectiviteit van de wonder drugs in gevaar3. Naast de humane geneeskunde, draagt ook het gebruik van antibiotica in de veterinaire sector, en in de land- en tuinbouw fors bij aan het ontstaan van resistente micro-organismen2. Ondanks bekendheid met het resistentieprobleem neemt wereldwijd het antibioticagebruik en daarmee ook de antibioticaresistentie nog steeds toe. Dit gaat gelijktijdig gepaard met een forse afname van ontwikkeling van nieuwe antibiotica. Beide zijn een grote bedreiging voor de volksgezondheid4,5. Een voorbeeld is de opkomst van carbapenem-resistente Enterobacteriaceae die ernstige infecties met een hoge mortaliteit kunnen veroorzaken. Hiermee lijken we aan het begin van een postantibiotisch tijdperk te staan1,2,6.

De resistentie tegen antibiotica heeft echter niet alleen grote en directe gevolgen voor de patiënt (hogere morbiditeit en mortaliteit)7. Antibioticaresistentie leidt ook tot de noodzaak van duurdere tweedelijns antibiotica en maakt het toedienen van intraveneuze in plaats van orale antibiotica vaker noodzakelijk. Daarnaast liggen patiënten met infecties veroorzaakt door resistente pathogenen gemiddeld langer in het ziekenhuis. Alle genoemde factoren leiden tot een substantiële toename in de kosten1,8.

Om dit probleem onder controle te krijgen heeft de World Health Organization (WHO) verklaard dat het beter gebruik van antibiotica absoluut noodzakelijk is4. Gestandaardiseerde behandeling volgens richtlijnen moet worden aangemoedigd om het voorschrijven van incorrecte antibiotische therapie te verminderen4. Dit is namelijk geassocieerd met een verbetering in klinische uitkomst, een afname in resistentie-ontwikkeling en lagere kosten9-12. De European Academies Science Advisory Council (EASAC) heeft soortgelijke doelen geformuleerd13. Daarnaast is het belangrijk dat er onder artsen meer besef komt ten aanzien van de ernst van de situatie. Vaak zijn artsen vooral bezorgd over de effecten van de infectie en het starten van antibiotica bij een individuele patiënt en veel minder bezorgd over het risico bij te dragen aan de toename van resistentie tegen antibiotica binnen de samenleving1,14.

In verschillende studies komt naar voren dat 30 tot 40% van de patiënten niet die zorg krijgt die wordt aanbevolen volgens de richtlijnen15,16. Ongeveer 20 tot 30% van alle patiënten krijgen zelfs behandelingen die overbodig blijken te zijn16,17. Ditzelfde geldt voor het voorschrijven van antibiotica. Binnen het ziekenhuis wordt naar schatting 20 tot 50% van de antibiotica niet volgens een richtlijn voorgeschreven18,19. Voorbeelden zijn het afwijken zonder gemotiveerde reden van het eerste keus middel volgens de lokale richtlijn en een verkeerde therapieduur17,18. Studies binnen Nederlandse ziekenhuizen bij lagere luchtweginfecties en urineweginfecties bevestigen het vóórkomen van frequent inadequaat antibioticagebruik. Inadequaat gebruik werd in deze studies onder andere gedefinieerd als het achterwege laten van een sputum- of urinekweek, niet volgens de richtlijn voorschrijven van empirische therapie, een verkeerde therapieduur en te laat switchen van intraveneuze naar orale therapie1,20,21.

Implementatiestrategieën kunnen het voorschrijven van antimicrobiële therapie succesvol verbeteren en een toename in antibioticaresistentie verminderen22,23. Een studie onder huisartspraktijken in Engeland liet zien dat praktijken met de grootste daling in het voorschrijven van antibiotica over een periode van 7 jaar, ook significant de grootste daling in resistentie voor de ‘eerste keus antibiotica’ bij de behandeling van urineweginfecties hadden. Resistentie bleek dan ook omkeerbaar bij het verminderen van het voorschrijven van antibiotica bij de behandeling van urineweginfecties1,24.

Chapter 8

164


165

8

RichtlijnenInternationale, nationale en lokale richtlijnen ten aanzien van de behandeling rondom diverse infectieziekten zijn ontwikkeld. Correct en veilig omgaan met antibiotica staat beschreven in deze richtlijnen, waarin ook aanbevelingen voor de klinische praktijk worden gemaakt. In Nederland ontwikkelt de Stichting Werkgroep Antibiotica Beleid (SWAB) richtlijnen voor antibioticagebruik voor de meest relevante infecties (www.swab.nl). Het doel van richtlijnen is het ondersteunen van klinische beslissingen en het verminderen van ongewilde praktijkvariatie15. Met andere woorden, richtlijnen hebben als doelstelling het verbeteren van de kwaliteit van antibioticagebruik door de kennis binnen de wetenschap te verbinden met de klinische praktijk. Echter, het publiceren van een richtlijn zorgt niet automatisch voor het volgen van deze richtlijn en goed antibioticagebruik in de klinische praktijk1,25. In hoofdstuk 2 hebben wij gekeken naar de lokale antibioticaboekjes van praktisch alle ziekenhuizen binnen Nederland. Het lokale antibioticaboekje kan gekoppeld zijn aan het nationale antibioticaboekje (SWAB-ID), waardoor het lokale boekje eigenlijk gebaseerd is op het nationale antibioticaboekje en vervolgens aangepast is aan het lokale resistentiepatroon. Daarnaast kunnen ziekenhuizen ervoor kiezen om het antibioticaboekje zelf te schrijven (non-SWAB-ID). In hoofdstuk 2 hebben we deze 2 groepen met elkaar vergeleken en zowel gekeken naar de volledigheid als inhoudelijk naar de adviezen ten opzichte van de nationale richtlijn. In totaal zijn er 50 lokale antibioticaboekjes gescoord; 199 items ten aanzien van de volledigheid (bijvoorbeeld: is er een aanbeveling voor profylaxe of empirische therapie voor cholangitis) en 35 items over inhoudelijke adviezen (zoals: de aanbeveling van empirische therapie voor diverticulitis). De non-SWAB-ID lokale antibioticaboekjes (n=27) waren in vergelijking met de SWAB_ID groep (n=23), minder volledig (p < 0.001) en als gevolg hiervan lag de adherentie aan het nationale antibioticaboekje lager (p < 0.001). Het gebruik van een lokaal antibioticaboekje dat een afgeleide is van het nationale antibioticaboekje, zorgt ervoor dat de lokale antibioticaboekjes vollediger zijn en meer conform het nationale antibioticaboekje, waardoor aanbevelingen vaker herleid zijn van evidence-based richtlijnen. Dit verbetert de kwaliteit van de lokale antibioticaboekjes en verhoogt de uniformiteit.

KwaliteitsindicatorenHet ontwikkelen van richtlijnen en een antibioticaboekje is dan ook een belangrijke eerste stap voor het bevorderen van de kwaliteit van antibioticagebruik. De tweede stap is de beschikbaarheid van een instrument dat op een betrouwbare manier de kwaliteit van antibioticagebruik kan meten. Kwaliteitsindicatoren zijn meetbare elementen van handelen in de praktijk, waarvoor bewijs is of waarover consensus bestaat dat ze gebruikt kunnen worden om de kwaliteit van zorg en veranderingen daarin te evalueren26. Ze bestaan uit een teller en noemer; een voorbeeld is: neem, bij verdenking op een infectie, een kweek af van het verdachte focus voor de aanvang van de antibioticabehandeling. De noemer bestaat uit alle patiënten waar antibiotica wordt gestart vanwege een verdenking op een infectie en de teller uit het aantal van deze patiënten waarbij er daadwerkelijk een kweek is afgenomen1. Kwaliteitsindicatoren zijn grofweg onder te verdelen in structuur-, proces- en uitkomstindicatoren27. Een structuurindicator richt zich op organisatorische aspecten, zoals het aanwezig zijn van een lokaal antibioticaformularium. Procesindicatoren richten zich op de daadwerkelijk aan de patiënt geleverde zorg (onderzoek en/of behandeling), bijvoorbeeld het percentage patiënten dat na de uitkomst van de kweek versmalling van therapie krijgt. Een uitkomstindicator meet het gewenste of ongewenste uiteindelijke resultaat van de geleverde zorg, bijvoorbeeld de ziekenhuisduur bij patiënten met een bepaalde infectie1, 26. Kwaliteitsindicatoren kunnen ontwikkeld worden door middel van een wetenschappelijke consensus methode, zoals de RAND Delphi procedure, waar wetenschappelijk bewijs (onttrokken uit richtlijnen en de literatuur) en de opinie van experts wordt gebruikt om tot consensus te komen. Kwaliteitsindicatoren geven dus inzicht in de zorg die daadwerkelijk wordt gegeven in de dagelijkse praktijk. Ze kunnen de effecten meten van een verbeterstrategie of knelpunten binnen een ziekenhuis opsporen (interne kwaliteitsmeting). Daarnaast kunnen ze ook gebruikt worden voor externe verantwoording, door bijvoorbeeld ziekenhuizen onderling te vergelijken. Dit laatste wordt ook wel benchmarking genoemd1.

In 2010 heeft de SWAB een nationale richtlijn gepubliceerd over antimicrobiële therapie bij patiënten met een sepsis opgenomen in het ziekenhuis. In hoofdstuk

Chapter 8

166


167

8

3 hebben wij deze richtlijn gebruikt om een set kwaliteitsindicatoren te ontwikkelen welke de kwaliteit van antibioticagebruik bij patiënten opgenomen met een sepsis kan meten. Volgens de RAND Delphi procedure werden 14 Nederlands expert uitgenodigd om in totaal 40 aanbevelingen te scoren en prioriteren door middel van 2 schriftelijke vragenlijsten en een face-to-face consensus bijeenkomst. Dit heeft geleid tot een set van 5 kwaliteitsindicatoren, te weten: neem een set bloedkweken af, schrijf empirische antimicrobiële therapie voor volgens de nationale richtlijn, start de antibiotica intraveneus, start de antibiotica binnen 1 uur bij ernstige sepsis/septische shock, en stroomlijn de therapie indien de kweekuitslag bekend wordt.

RIANT-studieSpecifiek voor de Nederlandse situatie zijn indicatoren ontwikkeld die binnen ziekenhuizen de kwaliteit van antibioticatherapie van lagere luchtweg- en urineweginfecties bij patiënten meten20,21. In 2011 is de RIANT-studie van start gegaan: Development of Reliable generic quality Indicators for the optimalisation of ANTibiotic use in the hospital. Deze set van indicatoren heeft als doel om de belangrijkste stappen rondom het gehele proces van antibioticagebruik bij alle infecties binnen het ziekenhuis te meten en te monitoren op het niveau van de patiënt1. In hoofdstuk 4 beschrijven wij het eerste deel van de studie, namelijk het ontwikkelen van de generieke kwaliteitsindicatoren. Door middel van een literatuursearch in PubMed en Embase werden potentiële generieke indicatoren ontleend en gecombineerd met de al eerder ontwikkelde indicatoren voor sepsis, lagere luchtweg- en urineweginfecties20,21,27. De RAND Delphi methode werd toegepast en een internationaal expert panel van 17 experts uit Nederland, België, Spanje, Schotland, Kroatië en Zweden, scoorden alle 24 potentiële generieke indicatoren. Deze procedure bestond uit 2 vragenlijsten met tussendoor een eenmalige face-to-face bijeenkomst, waardoor uiteindelijk 11 generieke indicatoren overbleven, waarvan 9 proces- en 2 structuurindicatoren28. Deze set indicatoren is weergegeven in Tabel 1.

Het tweede deel van de studie wordt beschreven in hoofdstuk 5, waar de indicatoren zijn getest op meetbaarheid, betrouwbaarheid, verbeterbaarheid, case-mix stabiliteit en toepasbaarheid (bij hoeveel procent van de patiënten die antibiotica gebruikten was deze indicator toepasbaar). De data die voor

deze metingen nodig waren zijn afkomstig van 1890 volwassen patiënten uit 4 universitaire en 18 perifere (grotere en kleinere) ziekenhuizen door heel Nederland. Voor de inclusie van de patiënten is samengewerkt met PREZIES (PREventie van ZIEkenhuisinfecties door Surveillance). PREZIES is een Nederlands samenwerkingsverband tussen meer dan 90 participerende ziekenhuizen en het RIVM (Rijksinstituut voor Volksgezondheid en Milieu). Tijdens de reguliere halfjaarlijkse prevalentiemeting van PREZIES zijn alle patiënten die langer dan 24 uur antibiotica gebruikten vanwege een verdenking op een bacteriële infectie, ouder waren dan 18 jaar en niet op een intensive care/medium care waren opgenomen, geïncludeerd in de RIANT studie. Lage toepasbaarheid (≤10%) werd gevonden bij de kwaliteitsindicatoren: ‘dosering antibiotica aanpassen aan de nierfunctie’ en ‘therapeutic drug monitoring’. De indicator ‘het stoppen van empirische therapie bij gebrek aan bewijs voor infectie’ liet een lage betrouwbaarheid zien (kappa < 0.4). De mogelijkheid tot verbeteren was laag bij de indicator ‘een actuele lokale richtlijn moet aanwezig zijn’. Concluderend zijn zeven van de elf kwaliteitsindicatoren meetbaar, toepasbaar en betrouwbaar, en behoren daarmee tot de definitieve set generieke indicatoren, zie Tabel 1. Case mix correctie was nodig voor de meeste procesindicatoren. Voor alle indicatoren werd een grote variabiliteit tussen de ziekenhuizen gevonden en nog veel ruimte voor verbetering.

Kwaliteitsindicatoren en uitkomstenStudies laten zien dat patiënten met een lage luchtweginfectie of sepsis gebaat zijn bij goed antibioticagebruik (beschreven in richtlijnen). Goed antibioticagebruik zorgt namelijk voor betere patiëntuitkomsten en voor minder kosten. Een eerdere studie bij gecompliceerde urineweginfecties laat zien dat indien meer indicatoren per patiënt goed gescoord werden, de opnameduur in het ziekenhuis korter was29. In hoofdstuk 6 hebben wij daarom gekeken naar het effect van goed antibioticagebruik op de opnameduur in het ziekenhuis. Goed antibioticagebruik werd gedefinieerd als de zes procesindicatoren die meetbaar en betrouwbaar waren in hoofdstuk 5. In een observationele multicenter studie is er bij 1890 patiënten, opgenomen met antibiotica vanwege een infectie, gekeken naar de relatie tussen wel of niet voldoen aan deze indicatoren en opnameduur. Daarnaast is er een ‘sum score’ gemaakt die de prestatie weergeeft op de totale

Chapter 8

168


169

8

set indicatoren. De ‘sum score’ is weergegeven in twee groepen; de groep met de lage ‘sum score’ (in 0 – 49% wordt voldaan aan de indicatoren) versus de hoge ‘sum score’ (in 50 – 100% wordt voldaan aan de indicatoren). De resultaten laten een associatie zien tussen de indicator ‘tijdig switchen van intraveneus naar oraal’ en een kortere opnameduur (6.5 vs. 11.2 dagen; P <0.001). Daarnaast zien we dat als 50% of meer van de indicatoren goed gescoord worden, de patiënt een dag korter in het ziekenhuis ligt (10.1 vs. 11.2 dagen, P =0.002). Dit wordt ook gevonden bij de subgroep community-acquired infecties (9.7 vs. 10.9 dagen; P =0.007). , De subgroep hospital-acquired infecties lag ook korter in het ziekenhuis alleen was dit verschil niet significant (11.6 vs. 12.2 dagen; P =0.08). Concluderend, goed antibioticagebruik, gedefinieerd als de ‘sum score’ van alle gevalideerde indicatoren samen, bij patiënten opgenomen met een suspecte bacteriële infectie lijkt de opnameduur te reduceren met 1 dag en daarmee een bijdrage te leveren aan een goede patiëntenzorg en een reductie in de ziekenhuiskosten.

ConclusieIn het laatste hoofdstuk, hoofdstuk 7, wordt een overzicht gegeven van de resultaten zoals beschreven in dit proefschrift, waarna er aanbevelingen worden geformuleerd voor toekomstig onderzoek. De Inspectie voor de Gezondheidszorg heeft sinds 2014 verplicht gesteld dat elk ziekenhuis een antibiotica team (A-team) moet hebben, dat de kwaliteit van het antibioticabeleid in het ziekenhuis bewaakt. Een belangrijke taak van het A-team is het meten, en verbeteren van het antibioticagebruik binnen het ziekenhuis. Samenvattend, hebben de resultaten van de RIANT-studie geleid tot een gevalideerde set kwaliteitsindicatoren, wat een belangrijk instrument is om in de toekomst op een betrouwbare manier antibioticagebruik binnen ziekenhuizen te monitoren en te gebruiken bij projecten die gericht zijn op effectief, doelmatig en veilig voorschrijven van antibiotica.

De volgende stap is om deze set kwaliteitsindicatoren ook te gebruiken om antibioticagebruik te verbeteren. Het meten van de kwaliteitsindicatoren kan per ziekenhuis zichtbaar maken voor welke stap verbetering nodig is. Een belangrijke conclusie van hoofdstuk 6 was echter dat eerder de combinatie van meerdere procesindicatoren samen, zorgden voor een kortere opnameduur, dan één indicator alleen. Dit sluit aan bij de opvattingen over een ‘bundel’

of ‘checklist’. Daarom gaat deze set generieke indicatoren gebruikt worden om een checklist te ontwikkelen. Het effect van de implementatie van deze checklist wordt op dit moment op klinische eindpunten onderzocht (zie www.ABchecklist.nl). De filosofie achter deze checklist is hetzelfde als de checklist/briefing die al standaard wordt uitgevoerd bij patiënten die worden geopereerd. Door het implementeren van de SURPASS (surgical patient safety system) lieten Boermeester et al. een reductie van chirurgische complicaties (10.6%) en mortaliteit zien.30 De verwachting is door het hele proces rondom antibioticagebruik verder te standaardiseren, het antibioticabeleid in ziekenhuizen zal verbeteren, met als uiteindelijk resultaat een verbetering in de patiëntenzorg, een reductie in de ziekenhuiskosten en op de lange termijn een afname in de stijgende antibioticaresistentie.

Chapter 8

170


171

8

Tabe

l 1. L

ijst m

et ge

nerie

ke k

walit

eitsin

dica

tore

n vo

or h

et m

onito

ren

van

antib

iotic

ageb

ruik

bij

alle

bacte

riële

infec

ties b

inne

n he

t ziek

enhu

is.

Soor

t ind

i-ca

tor

ENof

de

indi

ca-

tor d

oor d

e pr

aktij

ktes

t is

gek

omen

Kw

alite

itsin

dica

tor

Alle

pat

iënt

en zi

jn: v

olw

asse

nen,

opg

enom

en

op in

het

ziek

enhu

is (b

ehou

dens

op

de IC

U

afde

ling)

met

een

ver

denk

ing

op e

en b

acte

riël

e in

fect

ie w

aarv

oor l

ange

r dan

>24

u sy

stem

isch

an

tibio

tica

wor

dt g

egev

en.

Bes

chri

jvin

g Te

ller:

A

lle p

atië

nten

zijn

: vol

was

sene

n,

opge

nom

en o

p in

het

ziek

enhu

is

(beh

oude

ns o

p de

ICU

afd

elin

g)

met

een

ver

denk

ing

op e

en b

acte

-ri

ële

infe

ctie

waa

rvoo

r lan

ger d

an

>24u

syst

emis

ch a

ntib

iotic

a w

ordt

ge

geve

n.

Bes

chri

jvin

g N

oem

er:

Alle

pat

iënt

en zi

jn: v

olw

asse

nen,

opg

e-no

men

op

in h

et zi

eken

huis

(beh

oude

ns

op d

e IC

U a

fdel

ing)

met

een

ver

denk

-in

g op

een

bac

teri

ële

infe

ctie

waa

rvoo

r la

nger

dan

>24

u sy

stem

isch

ant

ibio

tica

wor

dt g

egev

en.

Proc

es& JA

Empi

risch

e ant

ibio

tica m

oet w

orde

n vo

orge

s-ch

reve

n vo

lgen

s de l

okal

e ric

htlij

n.

*als

een

loka

le ric

htlij

n ni

et a

anwe

zig i

s sch

rijf d

an

voor

volge

ns d

e nat

iona

le ric

htlij

n of

and

ers v

olge

ns

de in

tern

atio

nale

richt

lijn.

Aan

tal p

atië

nten

waa

r em

piris

che

antib

iotic

a is g

esta

rt vo

lgen

s de l

okal

e ric

htlij

n

Tota

al aa

ntal

pat

iënt

en w

aar g

esta

rt is

met

em

piris

che a

ntib

iotic

a bij

verd

enki

ng o

p ee

n ba

cter

iële

infe

ctie

.

Proc

es& JA

Voor

het

star

ten

van

syst

emisc

he an

tibio

tica

moe

ten

min

sten

s 2 se

ts b

loed

kwek

en w

orde

n af

geno

men

.

Aan

tal p

atië

nten

waa

rbij

2 se

ts b

loed

k-w

eken

zijn

afge

nom

en vo

or st

art v

an

antib

iotic

a

Tota

al aa

ntal

pat

iënt

en w

aarb

ij ge

star

t is

met

syst

emisc

he an

tibio

tica b

ij ve

rden

king

op

een

bact

erië

le in

fect

ie

Proc

es& JA

Kwek

en va

n ve

rdac

hte f

oci m

oete

n zo

snel

mo-

gelij

k w

orde

n af

geno

men

, het

lief

st vo

or st

art v

an

syst

emisc

he an

tibio

tica.

*K

weke

n m

ogeli

jk to

t max

imaa

l 24u

na

start

ther

a-pi

e wor

den

afge

nom

en.

Aan

tal p

atië

nten

waa

rbij

kwek

en va

n ve

rdac

hte f

oci z

ijn af

geno

men

voor

st

art a

ntib

iotic

a dan

wel

<24

uur n

a st

art a

ntib

iotic

a

Tota

al aa

ntal

pat

iënt

en w

aarb

ij ge

star

t is

met

syst

emisc

he an

tibio

tica b

ij ve

rden

king

op

een

bact

erië

le in

fect

ie, e

n ee

n kw

eek

van

een

verd

acht

e foc

us m

oet w

orde

n af

ge-

nom

enPr

oces

& JA

Empi

risch

e ant

ibio

tica m

oet w

orde

n aa

ngep

ast

aan

de k

wee

kuits

lag

en w

aar m

ogel

ijk w

orde

n ve

rsm

ald,

zo sn

el m

ogel

ijk n

adat

de k

wee

k be

kend

is.

Aan

tal p

atië

nten

met

een

posit

ieve

kw

eeku

itsla

g w

aar e

mpi

risch

e the

rapi

e is

aang

epas

t aan

de k

wee

kuits

lag,

en

waa

r mog

elijk

is ve

rsm

ald.

Tota

al aa

ntal

pat

iënt

en m

et em

piris

che a

nti-

biot

ica e

n ee

n po

sitie

ve k

wee

kuits

lag,

waa

r aa

npas

sen

aan

de k

wee

kuits

lag

nood

zake

lijk

was

. *aa

npas

sen

was n

oodz

akeli

jk in

dien

het

m

icro-

orga

nism

e ook

sens

itief

was v

oor e

en

smal

ler a

ntib

iotic

um of

het

micr

o-or

gani

sme

resis

tent

was

voor

de e

mpi

risch

e the

rapi

e

Proc

es& NEE

Dos

erin

g en

inte

rval

van

syst

emisc

he an

tibio

tica

moe

t wor

den

aang

epas

t aan

de n

ierf

unct

ie

Aan

tal p

atië

nten

waa

r de a

ntib

iotic

a do

serin

g en

inte

rval

ook

daa

dwer

kelij

k w

erd

aang

epas

t aan

de v

erm

inde

rde

nier

func

tie

*ver

min

derd

e nier

func

tie w

erd

gede

finiee

rd a

ls ee

n es

timat

ed G

lom

eru-

lar F

iltra

tion

Rate

< 5

0 m

l/m

in/1

,73m

2

Tota

al aa

ntal

pat

iënt

en m

et an

tibio

tica d

ie

mog

elijk

moe

t wor

den

aang

epas

t aan

de

nier

func

tie, m

et ee

n ve

rmin

derd

e of o

nbek

-en

de n

ierf

unct

ie. *

verm

inde

rde n

ierfu

nctie

we

rd ge

defin

ieerd

als

een

estim

ated

Glo

mer

ular

Fi

ltrat

ion

Rate

< 5

0 m

l/m

in/1

,73m

2

Proc

es&

JA

Syst

emisc

he an

tibio

tica m

oet w

orde

n ge

switc

ht

van

intra

vene

us n

aar o

raal

bin

nen

48-7

2uur

op

basis

van

de k

linisc

he co

nditi

e van

pat

iënt

en

wan

neer

ora

le th

erap

ie ad

equa

at is

. *A

dequ

aat =

1. w

anne

er h

et a

ntib

iotic

um or

aal

besc

hikb

aar i

s. 2.

wan

neer

oral

e int

ake m

ogeli

jk is

en

abso

rptie

ade

quaa

t. 3.

ade

quaa

t t.a

.v. d

e dia

gnos

e (u

itzon

derin

gen

zijn

b.v.

endo

card

itis,

men

ingi

tis et

c.) A

anta

l pat

iënt

en m

et in

trave

neuz

e an

tibio

tica w

aar b

inne

n 48

-72

uur i

s ge

switc

ht n

aar o

rale

ther

apie

.

Tota

al aa

ntal

pat

iënt

en m

et in

trave

neuz

e an

tibio

tica v

oor 4

8-72

uur

, waa

r sw

itche

n na

ar o

raal

op

basis

van

klin

ische

cond

itie e

n an

dere

adeq

uate

cond

ities

mog

elijk

is.

Proc

es&

JA

Een

antib

iotic

a pla

n m

oet g

edoc

umen

teer

d w

orde

n in

de s

tatu

s bij

star

t van

syst

emisc

he an

ti-bi

otic

a. *a

ntib

iotic

a pl

an is

indi

catie

voor

ther

apie,

na

am, d

osis,

rout

e en

inte

rval

van

het a

ntib

iotic

um

Aan

tal p

atië

nten

waa

r een

antib

iotic

a pl

an w

as g

edoc

umen

teer

d in

de s

tatu

s To

taal

aant

al p

atië

nten

waa

r ges

tart

is m

et

syst

emisc

he an

tibio

tica b

ij ve

rden

king

op

een

bact

erië

le in

fect

ie

Proc

es& NEE

‘Ther

apeu

tic d

rug

mon

itorin

g’ m

oet w

orde

n ui

tgev

oerd

indi

en am

inog

lyco

side l

ange

r dan

3

dage

n en

vanc

omyc

ine l

ange

r dan

5 d

agen

wor

dt

gege

ven

Aan

tal p

atië

nten

waa

r min

imaa

l één

bl

oeds

pieg

elbe

palin

g va

n he

t des

betr-

effen

de an

tibio

ticum

is g

edaa

n

Tota

al aa

ntal

pat

iënt

en d

ie am

inog

lyco

side

lang

er d

an 3

dag

en o

f van

com

ycin

e lan

ger

dan

5 da

gen

krijg

en

Proc

es&

N

EE

Empi

risch

e ant

ibio

tica m

oet g

esto

pt w

orde

n bi

j ee

n ge

brek

aan

klin

ische

- dan

wel

mic

robi

olog

isch

bew

ijs vo

or in

fect

ie. D

e max

imal

e the

rapi

eduu

r is

7 da

gen.

Aan

tal p

atie

nten

waa

r de e

mpi

risch

e th

erap

ie is

ges

topt

bin

nen

7 da

gen

vanw

ege g

ebre

k aa

n kl

inisc

he- d

an w

el

mic

robi

olog

isch

bew

ijs vo

or in

fect

ie

Tota

al aa

ntal

pat

iënt

en w

aar g

esta

rt is

met

em

piris

che a

ntib

iotic

a maa

r waa

r klin

ische

- da

n w

el m

icro

biol

ogisc

h be

wijs

voor

infe

c-tie

nie

t aan

toon

baar

is

Stru

ctuu

r&

N

EE

Een

actu

ele l

okal

e ric

htlij

n t.a

.v. an

tibio

tica

moe

t aan

wez

ig zi

jn in

het

ziek

enhu

is, en

elke

3

jaar

moe

t geë

valu

eerd

wor

den

of ee

n up

date

no

odza

kelij

k is

Stru

ctuu

r& JA

De l

okal

e ric

htlij

n m

oet i

n pr

inci

pe o

vere

en-

kom

en m

et d

e nat

iona

le ri

chtli

jn, m

aar m

oet w

el

afw

ijken

op

basis

van

loka

le re

siste

ntie

Chapter 8

172


173

8

Literatuur

(1) van den Bosch CM and Geerlings SE. Generieke kwaliteitsindicatoren voor optimaliseren van antibioticagebruik in ziekenhuizen. Nederlands Tijdschrift voor Anesthesiologie. 2015 mei 28:42-46

(2) Laxminarayan R, et al. Antibiotic resistance-the need for global solutions. Lancet Infect Dis 2013; 13: 1057-98.

(3) Centers for Disease Control and Prevention. Report 2011 (revision): A public health action plan to combat antimicrobial resistance. Centers for Disease Control and Prevention 2011. Available from: URL: http://www.cdc.gov/drugresistance/pdf/public-health-action-plan-combat-antimicrobial-resistance.pdf

(4) WHO Global Strategy for Containment of Antimicrobial Resistance 2011. http://www.who.int/drugresistance/WHO_Global_Strategy_English.pdf

(5) Fischbach MA & Walsh CT. Antibiotics for emerging pathogens. Science 2009; 325: 1089-93.

(6) Vital signs: carbapenem-resistant Enterobacteriaceae. MMWR Morb Mortal Wkly Rep 2013; 62:165-70.

(7) de Kraker ME, et al. Burden of antimicrobial resistance in European hospitals: excess mortality and length of hospital stay associated with bloodstream infections due to Escherichia coli resistant to third-generation cephalosporins. J Antimicrob Chemother 2010; 66: 398-407.

(8) ECDC/EMEA JOINT TECHNICAL REPORT. The bacterial challenge: time to react. Stockholm: European Centre for Disease Prevention and Control 2011. http://www.ecdc.europa.eu/en/publications/Publications/0909_TER_The_Bacterial_Challenge_Time_to_React.pdf

(9) Arnold FW, et al. Improving outcomes in elderly patients with community-acquired pneumonia by adhering to national guidelines: Community-Acquired Pneumonia Organization International cohort study results. Arch Intern Med 2009; 169: 1515-24.

(10) Butler CC, et al. Containing antibiotic resistance: decreased antibiotic-resistant coliform urinary tract infections with reduction in antibiotic prescribing by general practices. Br J Gen Pract 2007; 57: 785-92.

(11) Larson EL, et al. Relationship of antimicrobial control policies and hospital and infection control characteristics to antimicrobial resistance rates. Am J Crit Care 2007; 16: 110-20.

(12) McCabe C, et al. Guideline-concordant therapy and reduced mortality and length of stay in adults with community-acquired pneumonia: playing by the rules. Arch Intern Med 2009; 169: 1525-31.

(13) Tackling antibacterial resistance in Europe. EASAC policy report. European Academies Science Advisory Counsil 2011.

(14) Allerberger F, et al. Optimization of antibiotic use in hospitals--antimicrobial stewardship and the EU project ABS international. Chemotherapy 2008; 54 : 260-7.

(15) Grol R. Successes and failures in the implementation of evidence-based guidelines for clinical practice. Med Care 2001; 39: II46-II54.

(16) Schuster MA, et al. How good is the quality of health care in the United States? Milbank Q 1998; 76: 517-63.

(17) Willemsen I, et al. Appropriateness of antimicrobial therapy: a multicentre prevalence survey in the Netherlands, 2008-2009. Euro Surveill 2010; 15: 19715.

(18) Behar P, et al. Assessing the antimicrobial prescription request process in a teaching hospital in Brazil: regulations and training. Braz J Infect Dis 2000;4: 76-85.

(19) Bugnon-Reber A, et al. Antibiotic misuse in medium-sized Swiss hospitals. Swiss Med Wkly 2004;134: 481-5.

(20) Hermanides HS, et al. Development of quality indicators for the antibiotic treatment of complicated urinary tract infections: a first step to measure and improve care. Clin Infect Dis 2008; 46: 703-11.

(21) Schouten JA, et al. Quality of antibiotic use for lower respiratory tract infections at hospitals: (how) can we measure it? Clin Infect Dis 2005;41: 450-60.

(22) Davey P, et al. Interventions to improve antibiotic prescribing practices for hospital inpatients. Cochrane Database Syst Rev 2005; 4: CD003543.

(23) Molstad S, et al. Sustained reduction of antibiotic use and low bacterial resistance: 10-year follow-up of the Swedish Strama programme. Lancet Infect Dis 2008; 8: 125-32.

(24) Butler CC, et al. Containing antibiotic resistance: decreased antibiotic-resistant coliform urinary tract infections with reduction in antibiotic prescribing by general practices. Br J Gen Pract 2007; 57: 785-92.

(25) Bero LA, et al. Closing the gap between research and practice: an overview of systematic reviews of interventions to promote the implementation of research findings. The Cochrane Effective Practice and Organization of Care Review Group. BMJ 1998; 317: 465-8.

(26) Braspenning J, et al. Meten van (veranderingen in) de zorg: de rol van indicatoren. In: Grol, Wensing (red.). Implementatie: effectieve verbetering van de patientenzorg. Amsterdam: Elsevier gezondheidszorg; 2010. pp 153-75.

(27) van den Bosch CM, et al. Development of quality indicators for antimicrobial treatment in adults with sepsis. BMC Infect Dis 2014; 14: 345.

(28) van den Bosch CM, et al. Quality indicators to measure appropriate antibiotic use in hospitalized adults. Clin Infect Dis 2015; 60: 281-91.

(29) Spoorenberg V, et al. Appropriate antibiotic use for patients with urinary tract infections reduces length of hospital stay. Clin Infect Dis 2014; 58: 164-9.

(30) de Vries EN, Prins HA, Crolla RM, et al. Effect of a comprehensive surgical safety system on patient outcomes. N Engl J Med 2010;363(20):1928-1937.

Chapter 8

174

Dankwoord

175

8

Dankwoord

Tot slot van dit proefschrift zou ik een ieder die zich heeft ingezet voor mijn promotie willen bedanken. Een aantal mensen in het bijzonder.

Allereerst wil ik mijn twee promotoren, prof.dr. Suzanne Geerlings en prof.dr. Jan Prins, en mijn copromotor prof.dr. Marlies Hulscher, bedanken. Al tijdens mijn wetenschappelijke stage kwam ik erachter dat begeleiding heel belangrijk is als je net start met onderzoek. Ik ben ontzettend blij dat ik jullie om mij heen heb gehad! Ik had ruimte genoeg om mijn eigen ding te doen, maar jullie waren altijd dichtbij zodat problemen ook weer snel opgelost werden. De bijeenkom-sten in Utrecht met z’n vieren staan in mijn geheugen gegrift. Wat werden daar vaak heftige discussies gevoerd, alle drie weer een totaal andere kijk op kwaliteit en antibiotica, echter de beste ideeën zijn daar ontstaan. Het is erg leerzaam geweest. Ik had het niet beter kunnen treffen!

Lieve Suzanne, wat ontzettend leuk dat ik jouw eerste promovenda ben!!! Ik ben heel blij dat jij mijn directe begeleider bent geweest; jij maakte altijd tijd, beantwoordde e-mails bij nacht en ontij en was altijd geïnteresseerd, in mijn werk maar ook privé. Jouw enthousiasme, efficiëntie en oplossingsgerichtheid zijn bijzonder en maakten dat we voor elk probleem snel een praktische oploss-ing vonden. Onze mooie fietstochtje samen met Floor over de Golden Gate Bridge zal ik niet vergeten!

Lieve Jan, dank je wel voor je betrokkenheid, vertrouwen en positiviteit. Jouw oog voor detail en je kritische blik op mijn data en het schrijven van de stukken zijn bijzonder, en hebben de inhoud van dit boekje verbeterd. Ik heb genoten van je droge humor, je opbeurende oneliners als je weer eventjes je hoofd om de hoek van de deur stak en natuurlijk van de vele pepermuntjes!

Lieve Marlies, ik ken niemand die met zoveel passie en overtuiging kan praten over kwaliteit van zorg en kwaliteitsindicatoren. Dit was precies wat ik nodig had toen ik begon met dit project, en zoveel artsen en infectiepreventie deskun-dige moest overtuigen van het belang van dit onderzoek. Jouw visie en kritische noot waren essentieel voor het opzetten en uitvoeren van dit project.

De overige leden van mijn commissie, prof.dr. L.P.H.J. Aarts, prof.dr. M.A. Boermeester, Dr. J.C.C. Braspenning, prof.dr. M.D. de Jong, prof.dr. B.J. Kull-berg en prof.dr. R.A.A. Mathôt. Hartelijk dank voor het kritisch beoordelen van mijn proefschrift en voor de bereidwilligheid zitting te nemen in mijn pro-motiecommissie.

Daarnaast ben ik heel veel dank verschuldigd aan alle 22 deelnemende zieken-huizen van de RIANT-studie, in het bijzonder de contactpersonen van de af-delingen Medische Microbiologie en Ziekenhuishygiene & Infectiepreventie! In elk ziekenhuis waren er een aantal mensen die enthousiast waren over dit project en mij geholpen hebben de rest van de afdeling en ziekenhuis te moti-veren om mee te doen. Dankzij jullie onvermoeibare inzet; het helpen zoeken van statussen, het creëren van werkplekken, en het regelen van pas en toegang, is het gelukt om bijna 1900 patiënten te includeren in de studie. Voor een over-zicht van de participerende ziekenhuizen verwijs ik naar de in dit proefschrift opgenomen ‘lijst van deelnemende centra’.

PREZIES (PREventie van ZIEkenshuisinfecties door Surveillance), onderdeel van het RIVM, heeft door zijn halfjaarlijkse prevalentie metingen een belan-grijke bijdrage geleverd aan het selecteren en includeren van patiënten voor de RIANT-studie. In het bijzonder wil ik Jan Wille nog bedanken voor zijn inzet.

Graag bedank ik ook mijn mede-auteurs voor hun bijdragen aan de studies uit mijn proefschrift. Beste Reinier (drs. R.P. Akkermans), het was heel fijn dat jij, als statisticus die ook begrijpt wat kwaliteitsindicatoren zijn, wilde meedenken over dit project. Heel erg bedankt voor de vele adviezen, uitleg en geduld als ik weer even tussendoor belde met een ‘kleine’ vraag! Lieve Emelie (drs E.C. Schuts), het begon als een wetenschappelijke stage maar al snel werd duidelijk dat jij veel meer in je mars hebt. Dank voor je inzet en immer positieve houding, ik ben blij dat het heeft geleid tot een mooie publicatie.

Gelukkig heb ik niet alle 1890 patiënten zelf geïncludeerd en werd ik geholpen door Evelien en Jose. Ik ben jullie heel erg dankbaar voor de vele stapel status-sen waar jullie je doorheen geploegd hebben! Kwaliteitsdenken is een vak apart. Het was daarom zeer leerzaam om deel te mogen nemen aan de Kennisgroep Transparantie bijeenkomsten bij IQ Health-care, onder leiding van Arna van Doorn en Dr. J.C.C. Braspenning.

Chapter 8

176

Dankwoord

177

8

Ik wil mijn collegae van de afdeling Anesthesiologie uit het LUMC en het Alri-jne Ziekenhuis bedanken voor de tijd die ik heb gekregen om mijn promotie af te ronden, in het bijzonder prof.dr. L.P.H.J. Aarts en dr. J. Vuijk uit het LUMC, en dr. G.M. Woerlee en dr. C.A.M. van Tulder uit het Alrijne ziekenhuis. Daar-naast wil ik mijn ‘knappe koppen’ uit het LUMC bedanken voor hun steun en interesse! Margot, Eva, Karlien, Linda, Marieke, Rineke, Astrid, Daniel en Vin-cent, super veel zin in onze trip naar NYC!!

Je kunt niet promoveren zonder gezellige en lieve collega’s om je heen. Marlous, Caatje-1, Niekie, Hanneke, Emelie, Frederike, Thijs, Marlot, Karien, Thomas en Olivier, wat ben ik blij dat jullie (de een langer als de ander) mij gezelschap hebben gehouden op F4! Het is zo fijn om frustraties en tegenslagen te del-en, maar ook mijlpalen en publicaties te vieren met mensen die door dezelf-de fase gaan als ik! Lieve Caatje-1, jouw energie is aanstekelijk! Jammer dat ik niet meer op de JP Heije woon en hierdoor onze zondagochtend koffie mis-loop.. Lieve Niek, je bent niet meer mijn collega maar gelukkig nog steeds mijn vriendinnetje. Ik ben zo blij dat jij er was toen ik begon, wij kunnen zo hard lachen samen en jij begrijpt mij zo goed. Dank je wel dat jij mijn paranimf wilt zijn!! Lieve Paulien, dank voor de gezellige pauzes met lekkere koffie op Plein G, Coffee Plaza of Coffee Company. Lieve Lonneke, dank voor de gave feestjes en gezellige lunches!

Dit dankwoord is niet compleet zonder dat ik al mijn lieve vrienden en (schoon)familie bedank voor hun steun en interesse de afgelopen jaren! Lieve An, Jas, Fleur, Martini, Marleen en Kim, ik ben heel blij dat jullie na 18 jaar nog steeds zulke goede vriendinnetjes van mij zijn en dat wij elkaar nog vaak zien!! Lieve Anna, wat hebben wij een aantal super mooie vakanties gehad samen! Ik denk er nog vaak aan terug. Lieve Tessa en Karin, wij hebben veel bijzondere dingen samen meegemaakt, ik zal onze avonturen in Tanzania nooit vergeten! Jammer dat wij nu een stuk verder bij elkaar uit de buurt wonen. Lieve Linnie en Han, jullie waren gelukkig mijn roomies gedurende m’n studententijd. Wat hebben wij het super gezellig gehad, en ik ben heel blij dat wij elkaar nog steeds zien. Dimitri, super bedankt voor het ontwerpen van de cover van m’n boekje. Lieve Jolle, ik hoop dat na deze promotie ik meer tijd heb om met jou in Haarlem te borrelen. Lieve Mink, super leuk dat wij nu dezelfde opleiding doen, ik hoop dat je over n paar jaar ook naar Haarlem komt! Lieve Susan, de laatste bbq bij

jou was zeer geslaagd. Lieve Daan, jij stond naast mij toen ik besloot te gaan promoveren, dank je wel dat je mij altijd hebt gesteund.

Lieve papa, ik ben trots op jou, als vader en als mens, en wil je bedanken voor alles wat je mij in mijn leven hebt meegeven. Lieve mama, bedankt voor je on-voorwaardelijke liefde en vertrouwen in mij. Ik ben zo dankbaar dat ik altijd bij jullie terecht kan! Dit boekje is voor jullie. Lieve Pieter en Pim, je familie kies je niet uit, maar jullie had ik zeker uitge-kozen! Wat een geluk om twee van die lieve broertjes te hebben, waar ik altijd zo mee kan lachen! Pim, ik ben heel blij dat jij m’n paranimf wilt zijn. Lieve Juul, Peet, Casper, Gerda en Ben, wat gezellig dat jullie nu ook bij onze familie horen. Lieve oma, wat zou het bijzonder zijn als u er 22 januari bij kunt zijn!

Lieve Thijs, ik eindig dit dankwoord met jou, want wat ben ik blij en dankbaar dat ik jou ben tegengekomen! Zeker de helft van dit boekje heb ik geschreven met mijn hoofd in de wolken.. dus als er typefouten in staan weet je alvast wie ik de schuld geef ;-) Ik heb zoveel zin in al onze toekomstplannen samen en heb je meer dan lief.

Caroline

PhD portfolio

179

8

Chapter 8

178

Lijst van deelnemende centra

Academisch Medisch Centrum, Meibergdreef 9, 1105 AZ Amsterdam Atrium Medisch Centrum Parkstad, Henri Dunantstraat 5, 6419 PC HeerlenDiakonessenhuis, Bosboomstraat 1, 3582 KE UtrechtFranciscus Ziekenhuis (nu het Bravis Ziekenhuis), Boerhaavelaan 25, 4708 AE

RoosendaalGelderse Vallei, Willy Brandtlaaan 10, 6716 RP EdeGemini Ziekenhuis, Huisduinerweg 3, 1782 GZ Den HelderGroene Hart Ziekenhuis, Bleulandweg 10, 2803 HH GoudaJeroen Bosch Ziekenhuis, Henri Dunantstraat 1, 5223 GZ Den BoschLangeLand Ziekenhuis, Toneellaan 1, 2725 NA ZoetermeerLeids Universitair Medisch Centrum, Albinusdreef 2, 2333 ZA LeidenLievensberg Ziekenhuis (nu het Bravis Ziekenhuis), Boerhaaveplein 1, 4624

VT Bergen op ZoomMaxima Medisch Centrum, De Run 4600, 5504 DB VeldhovenMedisch Centrum Haaglanden, Lijnbaan 32, 2512 VA Den HaagMedisch Spectrum Twente, Haaksbergerstraat 55, 7513 ER EnschedeOrbis Medisch Centrum, Dr. H. van der Hoffplein 1, 6162 BG Sittard-GeleenRadboud Universitair Medisch Centrum, Geert Grooteplein Zuid 10, 6525

GA NijmegenRijnstate Ziekenhuis, Wagnerlaan 55, 6815 AD ArnhemUniversitair Medisch Centrum Utrecht, Heidelberglaan 100, 3584 CX UtrechtVlietland Ziekenhuis, Vlietlandplein 2, 3118 JH SchiedamWestfriesgasthuis, Maelsonstraat 3, 1624 NP HoornZiekenhuis Rivierenland, President Kennedylaan 1, 4002 WP TielZiekenhuis Bernhoven, Nistelrodeseweg 10, 5406 PT Ude

AMC Graduate School for Medical SciencesPhD Portfolio

Summary of PhD training, teaching and publicationsName PhD student: Caroline van den Bosch

PhD period: December 2010 – October 2014Name PhD supervisor: Prof dr S.E. Geerlings

1. PhD training

General courses

- Evidence-based searching- Oral presentation in English- Scientific Writing in English for Publication- Career Development- BROK (‘Basiscursus Regelgeving Klinisch Onderzoek’)- Practical Biostatistics- Project managment- Reference Manager Basic- Clincial Data Management- Clinical Epidemiology- The AMC World of Science- Pubmed- Web of Science- Systematic Reviews

Year

20122012201220122012201120112011201120112011201120112011

Workload(ECTS)

0,30,81,50,80,91,10,60,30,30,60,70,30,30,3

Specific courses- Klinische epidemiologie op Schiermonnikoog Boerhaave Commissie- Computing in R- Crash Course: Basic Chemistry, Biochemistry and

Molecular Biology for MDs starting Scientific Research- Advanced Topics in Clinical Epidemiology- Advance Biostatistics- Infectious Diseases

2013

20132012

201220122010

2,0

0,40,2

0,62,11,3

Chapter 8

180

PhD portfolio

181

8

Seminars, workshops and symposia- Symposium Incidenten in de patientenzorg,

Amsterdam- Symposium Incidenten in de patientenzorg,

Amsterdam- Infectieziekten Symposium, Amsterdam- Kennisgroep Transparantie IQ Healthcare, Radboud Universiteit- Infectieziekten Symposium, Amsterdam- Infectieziekten Symposium, Amsterdam

2014

2013

20122012

20112010

0,25

0,25

0,251,0

0,250,25

Presentations- Poster: Quality indicators for monitoring appropriate

antibiotic use: an important tool for antibiotic stewardship. ICAAC (Interscience Conference on Antimicrobial Agents and Chemotherapy) conference, Washington

- Poster: Quality indicators for monitoring appropriate antibiotic use: an important tool for antibiotic stewardship. Symposium Incidenten in de patiëntenzorg, Amsterdam

- Poster: Development of generic quality indicators for Antibiotic treatment in hospitalized adults. IDSA (the Infectious Diseases Society of America) Annual congress, in San Francisco

- Poster: Development of quality indicators for antimicrobial treatment in adults with sepsis. Symposium Incidenten in de patiëntenzorg, Amsterdam

- Poster: Development of generic quality indicators for antibiotic treatment in hospitalized adults. Symposium Incidenten in de patiëntenzorg, Amsterdam

2014

2014

2013

2013

2013

0,5

0,5

0,5

0,5

0,5

- Oral: Development of quality indicators for antimicrobial treatment in adults with sepsis. ECCMID (European Congress of Clinical Microbiology and Infectious Diseases) 22nd annual congress, in London

- Poster: Patients with severe COPD requiring invasive mechanical ventilation: mortality rate and a 2-year follow-up. ESICM (European Society of Intensive Care medicine) 23rd annual congress, in Barcelona

2012

2010

0,5

0,5

(Inter)national conferences- Can Better – patientveiligheidscongres, Amersfoort- ICAAC (Interscience Conference on Antimicrobial

Agents and Chemotherapy) Annual conference, Washington

- IDSA (the Infectious Diseases Society of America) Annual congress, in San Francisco

- ECCMID (European Congress of Clinical Microbiology and Infectious Diseases) 22nd annual congress, in Londen

20152014

2013

2012

0,20,75

0,75

0,75

2. Teaching

Year Workload(ECTS)

Supervising- Daan Vermeulen, Bachelorthesis: Zorgt adherentie

aan de richtlijn voor antibioticagebruik tot een beter resultaat voor de patiënt?

- Emelie Schuts, wetenschappelijke stage: An online national antimicrobial guide with locally customized versions can promote appropriate antibiotic use in hospitals.

2011

2013

1

1

Total ECTS 23,6

Curriculum Vitae

183

8

Chapter 8

182

3. List of publications

Year- CMA van den Bosch and SE Geerlings. Generieke

kwaliteitsindicatoren voor optimaliseren van antibioticagebruik in ziekenhuizen. Nederlands Tijdschrift voor Anesthesiologie. 2015 mei 28:42-46

- CMA van den Bosch, MEJL Hulscher, S. Natsch, JM Prins and SE Geerlings.Qality indicators to measure appropriate antibiotic use in hospitalized adults. Clin Infect Dis. 2015 Jan ;60(2):281-91.

- CMA van den Bosch, MEJL Hulscher, S. Natsch, IC Gyssens, JM prins and SE Geerlings. Development of quality indicators for antimicrobial treatment in adults with sepsis. BMC Infect Dis. 2014 Jun 20;14:345.

- CMA van den Bosch, MEJL Hulscher, J Wille, S. Natsch, BHB van Benthem, JM Prins en SE Geerlings. Generieke kwaliteitsindicatoren antibioticagebruik in het ziekenhuis. Nederlands Tijdschrift voor Medische Microbiologie. 2011;19:nr 3:19-21

- CMA van den Bosch, MEJL Hulscher, J Wille, S. Natsch, BHB van Benthem, JM Prins en SE Geerlings. Generieke kwaliteitsindicatoren voor antibioticagebruik in het ziekenhuis. Nederlands Tijdschrift voor Geneeskunde. 2011;155:A3502

- CMA van den Bosch, JAE van Wijk, GMA Beckers, HJR van der Horst, MF Schreuder and A Bökenkamp. Urological and Nephrological Findings in Renal Ectopia. Journal of Urology. 2010 Apr;183(4):1574-8

2015

2015

2014

2011

2011

2010

Curriculum Vitae

Caroline van den Bosch was born in Hilversum, the Netherlands, on Decem-ber 1st 1982. In 2001 she finished her secondary education at R.S.G. Broklede in Breukelen. From 2001 to 2002 she obtained her propaedeutics for Health Sciences at the VU University Amsterdam. She started Medical School in 2002 at the VU University in Amsterdam. She went to Tanzania in 2008 for an elec-tive clinical internship Tropical Medicine in the St. Francis Designated District Hospital at Ifakara and did an elective clinical internship Anesthesiology in Kennemer Gasthuis Hospital, Haarlem. She obtained her medical degree in 2009.After graduation, she worked for 11 months as a critical care resident at the In-tensive Care Unit in Flevoziekenhuis, Almere. Hereafter she continued working at the same hospital for 7 months, only now as a resident Internal Medicine. At the beginning of 2011, she started her PhD research under supervision of Prof. J.M. Prins and Prof. S.E. Geerlings from the Academic Medical Center (AMC), and Prof. M.E.J.L. Hulscher from the Radboud University Medical Center. Since October 2014, she is a resident Anesthesiology at Leiden University Medical Center and Leiderdorp Hospital. During the past year, she combined her residency with her PhD research. After completion of this thesis, she will continue her residency for four years at the department of Anesthesiology.

Documents

UvA-DARE (Digital Academic Repository) Development of ...€¦ · inappropriate antibiotic use is imperative.6 Antibiotic stewardship programs are developed to optimize antibiotic