Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
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
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
Appropriate Antibiotic Use in Daily Hospital Practice
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
General introduction
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
General introduction
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
General introduction
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
General introduction
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
Customized local version of SWAB-ID
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
Customized local version of SWAB-ID
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
Customized local version of SWAB-ID
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
Customized local version of SWAB-ID
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
Customized local version of SWAB-ID
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
Customized local version of SWAB-ID
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
Customized local version of SWAB-ID
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
Quality indicators for sepsis
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
Quality indicators for sepsis
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
Quality indicators for sepsis
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
Quality indicators for sepsis
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
Quality indicators for sepsis
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
Quality indicators for sepsis
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
Quality indicators for sepsis
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
Quality indicators for sepsis
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
Quality indicators for sepsis
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
Quality indicators for sepsis
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
Quality indicators for sepsis
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)
H.I. Bax, MD, PhD (infectious diseases 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)
C.W. Ang, MD, PhD (medical mi-crobiologist)
C.W. Ang, MD, PhD (medical mi-crobiologist)
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)
Prof. M.A. Boer-meester, MD (sur-geon)
Prof. M.A. Boer-meester, MD (sur-geon)
J.A. Schouten, MD, PhD (intensive care specialist)
J.A. Schouten, MD, PhD (intensive care specialist)
J.A. Schouten, MD, PhD (intensive care specialist)
Prof. P. Pickkers, MD (intensive care specialist)
Prof. P. Pickkers, MD (intensive care specialist)
Prof. P. Pickkers, MD (intensive care specialist)
J.J.W.M. Janssen, MD, PhD (hema-tologist)
J.J.W.M. Janssen, MD, PhD (hema-tologist)
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
Development of generic indicators
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
Development of generic indicators
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
Development of generic indicators
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
Development of generic indicators
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
Development of generic indicators
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
Development of generic indicators
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
Development of generic indicators
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
Development of generic indicators
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
Development of generic indicators
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
Development of generic indicators
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
Development of generic indicators
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
Development of generic indicators
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
Appropriate Antibiotic Use in Daily Hospital Practice
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
Applicability of generic indicators
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
Applicability of generic indicators
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
Applicability of generic indicators
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
Applicability of generic indicators
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
Applicability of generic indicators
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
Applicability of generic indicators
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
Applicability of generic indicators
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
Applicability of generic indicators
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
Applicability of generic indicators
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
Applicability of generic indicators
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
Length of hospital stay
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
Length of hospital stay
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
Length of hospital stay
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
Length of hospital stay
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
Total number of patients who start-ed with systemic antibiotic therapy
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
Total number of patients who start-ed with systemic antibiotic therapy
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
Length of hospital stay
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
Length of hospital stay
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
Length of hospital stay
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
Length of hospital stay
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.
4. 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.
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.
6. 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.
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
Length of hospital stay
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.
22. 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.
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
Length of hospital stay
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
Length of hospital stay
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
Length of hospital stay
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
Length of hospital stay
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.
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; 44: 159-77.
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.
5. 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.
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.
24. 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.
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
Nederlandse samenvatting
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
Nederlandse samenvatting
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
Nederlandse samenvatting
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
Nederlandse samenvatting
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
Nederlandse samenvatting
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
Nederlandse samenvatting
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.