University of Groningen Impact of medical …...Paranimfen Corien Kuiper Anne Loohuis The work...

University of Groningen

Impact of medical microbiologyDik, Jan-Willem

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2016

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Dik, J-W. (2016). Impact of medical microbiology: A clinical and financial analysis. [Groningen]:Rijksuniversiteit Groningen.

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

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 22-08-2020

https://www.rug.nl/research/portal/en/publications/impact-of-medical-microbiology(229db4ca-ff7e-44a4-9310-8e4d3d1280f0).html

https://www.rug.nl/research/portal/en/persons/janwillem-dik(fe795c0e-b5f6-4ccf-8f72-82de09774fe8).html

https://www.rug.nl/research/portal/en/publications/impact-of-medical-microbiology(229db4ca-ff7e-44a4-9310-8e4d3d1280f0).html

Impact of Medical Microbiology

a clinical and financial analysis

ISBN: 978-90-367-9216-5 (printed)

ISBN: 978-90-367-9210-3 (e-book)

Cover design: Erik Buikema

Printed by Ipskamp Printing BV, Enschede

© Jan-Willem Hendrik Dik, 2016. No part of this publication may be reproduced or

transmitted in any form or by any means without permission in writing from the author. The

copyright of previously published chapters of this thesis remains with the publisher or the

journal.

Impact of Medical Microbiology

A clinical and financial analysis

Proefschrift

ter verkrijging van de graad van doctor aan de

Rijksuniversiteit Groningen

op gezag van de

rector magnificus prof. dr. E. Sterken

en volgens besluit van het College voor Promoties.

De openbare verdediging zal plaatsvinden op

maandag 7 november 2016 om 16.15 uur

door

Jan-Willem Hendrik Dik

geboren op 14 november 1986 te Groningen

Promotores Prof. dr. A.W. Friedrich Prof. dr. dr. B.N.M. Sinha Prof. dr. M.J. Postma Copromotor Dr. M.G.R. Hendrix Beoordelingscommissie Prof. dr. J.E. Degener Prof. dr. J.A.J.W. Kluytmans Prof. dr. A. Voss

Paranimfen

Corien Kuiper

Anne Loohuis

The work presented in this thesis was performed at the Department of Medical Microbiology

and Infection Prevention of the University of Groningen and the University Medical Center

Groningen. It was funded by the European Union, the federal German states of North Rhine-

Westphalia and Lower Saxony and the Dutch provinces Overijssel, Gelderland and Limburg

via the EurSafety Health-net project [Interreg IVa III-1-01=073]. Financial support for

printing this thesis was kindly provided by the University of Groningen and the Groningen

University Institute for Drug Exploration (GUIDE) of the Graduate School of Medical

Sciences (GSMS).

Contents

1. General introduction 2

2. Cross-border comparison of antibiotic prescriptions among children and adolescents between the north of the Netherlands and the north-west of Germany

16

3. An integrated stewardship model: Antimicrobial, Infection prevention, and Diagnostic (AID)

28

4. Measuring the impact of antimicrobial stewardship programs 42

5. Financial evaluations of antibiotic stewardship programs a systematic review 56

6. Automatic day-2 intervention by a multidisciplinary Antimicrobial Stewardship-Team leads to multiple positive effects

74

7. Cost-minimization model of a multidisciplinary Antibiotic Stewardship Team based on a successful implementation on a urology ward of an academic hospital

88

8. Cost-analysis of seven nosocomial outbreaks in an academic hospital 104

9. Positive impact of eight years infection prevention on nosocomial outbreak management at an academic hospital

114

10. Performing timely blood cultures in patients receiving IV antibiotics is correlated with a shorter length of stay

128

11. General conclusion, discussion and recommendations 144

Appendix I: Bibliography 156

Appendix II: English summary 184

Appendix III: Nederlandse samenvatting 190

Appendix IV: Publication lists 196

Appendix V: Acknowledgements / Dankwoord 202

General Introduction

1

3 | Chapter 1

Introduction

General introduction and research question

The Netherlands is cited to have “the best healthcare system in Europe”, but this comes at a

price; the country also has one of the highest per capita spending in healthcare (Björnberg,

2015). Since 2013 there is a strong political focus on lowering healthcare costs; providing top

quality healthcare in the most cost-effective way in order to keep the ever increasing

expenditures under control (Schippers and van Rijn, 2013). This led to a drop in the total

healthcare budget (as percentage of the gross domestic product) of the Netherlands in 2014

and 2015. Due to the healthcare system within the Netherlands, with an ‘open market’,

politicians have a low influence on direct daily practice. In general, insurance companies and

healthcare providers (e.g. the hospitals), together with medical specialists and patient groups,

are the main players and decision makers in the system. Insurance companies can use this

freedom, to negotiate lower prices with healthcare providers. This relatively open market will

in general lead to a strategic trade-off for healthcare providers between (cost-)efficiency and

quality, in order to be a competitive player. Ideally, the competitive healthcare system should

encourage providers to spend money in the most cost-efficient manner, while at the same time

improve the quality of healthcare on which providers are scored and judged. Concretely, this

means that departments within a hospital (but also healthcare providers outside hospitals, such

as commercial laboratories) will need to provide better, more transparent and more complete

accountability of their activities to their boards of directors for the near future. There is thus a

clear need for scientifically sound impact analyses, which evaluate clinical and financial effects

of current practices, to support the policy makers and further improve quality (Brook, 2011;

NZa, 2015; Ubbink, et al., 2014).

More and more, outcome measures such as mortality rates, waiting times, and patients’ ratings

are used to define quality of healthcare. Also infection-related outcome measures such as the

number of hospital-acquired infections (including surgical site infections) and antimicrobial

resistance levels are often taken into account. Medical Microbiology and Infection Prevention

is one of the departments that can influence those latter, infection-related outcome measures.

In Europe, the Netherlands scores well on antimicrobial use and resistance levels; keeping

both relatively low compared to other EU countries (European Centre for Disease Prevention

and Control, 2013). The way microbiology is organized within hospitals in the Netherlands,

with clinical microbiologists who are actively involved in patient care, is thought to have been a

large contributor to these scores (Bonten, 2008). However, departments of Medical

Microbiology within the Netherlands are under pressure of insurance companies and

subsequently, also hospital boards of directors, to provide cheaper diagnostics (VGZ, 2013). In

parallel, increasing levels of antimicrobial resistance are causing more infections and

complications that are difficult to treat, leading to increasing healthcare costs (Review on

General introduction | 4

Antimicrobial Resistance, 2016). Germany is often mentioned as an example of a country

where lower prices are being paid for the same microbiological diagnostic tests compared to

other EU countries and especially the Netherlands (VGZ, 2013). The question is however, if

such a comparison of diagnostic cost prices between Germany and the Netherlands is a fair

one. Not everyone thinks so (Kusters, et al., 2014; Tersmette, et al., 2012). Germany and the

Netherlands have different healthcare systems in place, making direct comparisons on cost

price difficult (Müller, et al., 2015). In Germany, it is common to have the Medical

Microbiology laboratory and their teams at distance of the patient and the majority of the

hospitals do not have a medical microbiologist at their premises (Kaan, 2015).

Therefore, to better visualize effects of Medical Microbiology, also compared to other

EU countries such as Germany, there have been debates how to better measure outcomes,

besides just using process indicators (Bonten, et al., 2014; Bonten, et al., 2015). Furthermore,

there has been a discussion on the overall position of Medical Microbiology and their focus.

The Dutch Society of Medical Microbiology (Nederlandse Vereniging voor Medisch

Microbiologie, NVMM) felt a need to act. They discussed their focus and the relevance and

necessity of the Medical Microbiology, also regarding the finances (NVMM, 2012).

Within the University Medical Center Groningen (UMCG), Medical Microbiology is

combined with an Infection Prevention unit into one hospital department. This was done to

stimulate collaboration and integration of the two former departments, with as goal

improvement of efficiency, patient safety and overall quality within the hospital. However,

because Infection Prevention activities are not reimbursable on their own within the Dutch

system, the UMCG has chosen to generate budget for Infection Prevention from the overhead

of microbiological diagnostics. This makes the financial situation complex and is a relevant

contributing factor when analyzing and evaluating such a department. Keeping in mind the

before-mentioned need for impact analyses of healthcare practices, in order to create an

objective and transparent overview of the work performed within the hospital, this research

project and PhD thesis was therefore set out to answer the following question:

What is the clinical and financial impact of the combined activities of Medical Microbiology and Infection

Prevention on relevant outcome measures, at an academic hospital such as the UMCG?

Theoretical framework

The department of Medical Microbiology and Infection Prevention in the UMCG comprises

the broad spectrum of microbiology on clinical, teaching and research levels. For this research

project, the focus will be mainly on (infection prevention-related) bacteriology (the virology

aspect is covered by a second impact analysis and falls outside the scope of this thesis). To

5 | Chapter 1

adequately analyze and evaluate the department of Medical Microbiology, a model was

developed which puts all activities into context. The development of this model is also part of

this thesis (see chapter 3). In short, the model integrates all activities into three complementary

so-called “stewardship programs”. An Antimicrobial Stewardship Program (ASP), Infection

prevention Stewardship Program (ISP) and Diagnostic Stewardship Program (DSP); i.e. the

AID Stewardship Program. It was developed keeping in mind that infection management

should comprise of integrative actions focused on antimicrobial therapy, infection

prevention/control, and diagnostics, as well as being more patient-centered. This thesis is also

structured according this AID model.

General background

Medical Microbiology has a long history, starting with the Dutch scientist Antonie van

Leeuwenhoek who in 1676 discovered “small animals”, that later became known as bacteria

(van Leeuwenhoeck, 1677). Besides Van Leeuwenhoek, Louis Pasteur, who worked on the

principle of vaccination and the germ theory (Pasteur, 1881) and Robert Koch, who performed

ground breaking work on multiple bacteria, including Bacillus anthracis and Mycobacterium

tuberculosis (Koch, 1876) matured the field of Medical Microbiology. These three researchers are

therefore considered to be the “fathers of the microbiology”. However, by the turn of the

twentieth century, infections were still not treatable leading to high mortality rates. Patients

infected, for example with tuberculosis, were isolated when possible to prevent further spread

to others. All of this changed during the Second World War. In the years before, the Scottish

biologist Alexander Fleming discovered penicillin (Fleming, 1929). During the war, his

discovery was put into use for the wounded soldiers and became the first mass-produced

antibiotic. Both Robert Koch and Alexander Fleming received a Nobel Prize for their

respective work. During his acceptance lecture at the Nobel Prize ceremony, Fleming already

warned for the negative side effects of antibiotics. He observed that bacteria can develop

resistance quite easily in vitro when they are continuously exposed to low levels of penicillin. In

his speech he foresaw a future where antibiotics are freely available for everyone and resistance

could become a serious problem (Fleming, 1964). He could not have been more right…

Nowadays antimicrobial resistance has become a reality in modern medicine and is considered

a worldwide health threat by the World Health Organization (WHO) (World Health

Organization, 2012). The problems of resistance are mainly due to inappropriate use of

antibiotics. As Fleming observed already in the 1940’s, bacteria will adapt to survive.

Antibiotics are a unique treatment option, for the reason that they can eradicate the cause of

the disease, e.g. the pathogenic bacteria. Bacteria in turn want to survive and have multiple

ways of protecting themselves against antibiotics. By means of special resistant genes, either in

the core genome of the bacterium itself or in exchangeable plasmids, they can for example


produce proteins to breakdown antibiotics. In a normal situation, when there are no antibiotics

present, there is also no advantage in having these genes; it might even be a disadvantage (e.g.

due to reduced fitness). However, if antibiotics are introduced, the bacteria that do have

resistance genes now have an advantage and can thrive over the rest without them; a perfect

example of Darwin’s survival of the fittest under given circumstances. Incorrect use of

antibiotics can stimulate this even more. Subsequent suboptimal treatment, when antibiotics

are given but in such a low doses that they cannot kill the bacteria, can lead to selection of

resistance. Furthermore, therapy that is not finished completely can kill the bacteria without

resistance, leaving room for more resistant bacteria that were not (yet) killed because therapy

was cut short, to grow. The use of broad-spectrum antibiotics can have the same effect of

killing non-resistant bacteria, creating room for the more resistant ones to grow and thrive. A

next infection is then more likely to be caused by the more prevalent resistant bacteria for

which antibiotic treatment is more likely to fail. This resistance development happened already

to penicillin in the 1950’s.

The discovery of penicillin however, also led to a successful search for more

antibiotics. These findings and ultimately general use of each new antibiotic however, also led

to resistance development for these respective new antibiotics. This continued until the 1980’s.

Around that time new discoveries of antibiotics became less frequent and some people began

to see the destructive effects of 40 years of uncontrolled use. Last-resort antibiotics were

defined and reserved specifically for highly resistant infections. But also against these more

scarcely used antibiotics, resistance is developing and spreading. At the moment, no new

classes of antibiotics are expected to become available and new types of antibiotics in the

current classes are scarce, although there are some in the final clinical trials (Bettiol and

Harbarth, 2015; Draenert, et al., 2015). Furthermore, there are some potential alternatives to

antibiotics (e.g. antibodies, probiotics, lysins, bacteriophages and vaccination). However, it was

estimated that at least another 10 years and 1.5 billion dollar is needed to finalize research on

the top 10 most potential antimicrobials (Czaplewski, et al., 2016). The lack of research into

new antibiotics by the pharmaceutical industry is partly financially driven, because the

business-case is not interesting enough (EFPIA MID3 Workgroup, et al., 2016). Newly

discovered antibiotics will inevitably be reserved as last-resort drugs, meaning that sales will be

low and investments difficult to earn back. Several governmental programs by the FDA and

the EU are therefore in place to stimulate research and development of antibiotics (see for

example Harbarth, et al., 2015). However, a future where infections are untreatable once more,

a so-called post-antibiotic era is a real risk (Fowler, et al., 2014; World Health Organization,

2014).

Vancomycin resisitant Entrococcus spp. and carbapenem resistant gram-negative rods

(e.g. Klebsiellae pneumonia producing the NDM-1 enzyme) are just two examples of resistant

microorganisms, of which the latter is considered the biggest threat right now. In November

2015, Chinese researchers discovered Escherichia coli carrying a plasmid coding for polymyxin

resistance (Lui, et al., 2015). This was the first finding of such a plasmid, making them resistant

7 | Chapter 1

against the last group of available antibiotics (i.e. colistin) and more importantly, thanks to the

plasmid, giving the bacteria the possibility to more easily exchange this resistance. In the

following months, plasmids containing the gene for colisitin resistance (mcr-1 gene) were

found in human fecal samples on multiple continents suggesting worldwide spread (Arcilla, et

al., 2016; Malhotra-Kumar, et al., 2016). Some saw these findings, as a final confirmation that a

world with patients infected with effectively untreatable bacteria due to resistance against every

single clinically available antibiotic, is imminent (Gallagher, 2015).

Implications of such a situation are severe. Clinically, patients would be at risk of dying again

from manageable diseases such as tuberculosis and pneumonia. Financially, costs are estimated

to be hundreds of billions of dollars worldwide (Review on Antimicrobial Resistance, 2016;

Smith and Coast, 2013). These costs might seem as an exaggeration to some, but it is

important to understand the huge (societal) consequences of a post-antibiotic world. There will

for example, be a large impact on the workforce due to infections that become untreatable,

thereby incapacitating those patients from functioning. Tuberculosis is one example that is

often hailed as likely threat. Already, an alarming number of extensively drug-resistant

tuberculosis cases (XDR-TB), where treatment options are highly limited, more toxic, and

much more expensive, are reported around the world (Matteelli, et al., 2014). If this worrisome

development continues, tuberculosis might once more become a highly dangerous and fatal

disease that requires strict isolation in sanatoria, impacting our society (and our economy)

tremendously (Review on Antimicrobial Resistance, 2016; Wilson and Tsukayama, 2016).

Hospitals need more isolation rooms, patients are unable to work and at risk of dying,

healthcare workers treating those patients are at risk, sanatoria need to be built again and the

medication that is available is costly. It was estimated, that by 2050 an additional 10 million

people worldwide would die every year from resistant infections alone (Review on

Antimicrobial Resistance, 2016). This number not even takes into account the possible impact

if prophylactic use of antibiotics in surgical procedures becomes ineffective. Post-operative

wound infections will increase and certain operations (e.g. complex organ transplantations,

intensive care, prosthetic implants, etc.) are most likely not safe to perform anymore, once

again impacting our healthcare system and our economy hugely.

It is therefore not just the WHO, but also world leaders who advocate taking action.

The G7 recognized the threat of antibiotic resistance and agreed on the need for a worldwide

“one health” approach (G7 Germany, 2015). The Obama Administration made antibiotic

resistance one of the focus points (Task Force for Combating Antibiotic-Resistance Bacteria,

2015) and the General Assembly of the United Nations discussed antibiotic resistance in

September 2016. The Netherlands is one of the countries that is taking the lead in Europe. The

Dutch Minister of Healthcare Mrs. Schippers is actively involved in promoting awareness and

supporting research. As chair country of the European Union in the first half of 2016,

antibiotic resistance was one of the topics put on the agenda by the Netherlands (Koenders,


2015), resulting in, among others,. a European congress on antimicrobial resistance in

Amsterdam in March 2016. This meeting aimed at focusing on the main problems leading

towards the development of antimicrobial resistance.

Figure 1.1: Percentages of isolates that are classified as resistant as part of the total number of

isolates of four different microorganisms in Europe. Data from the 2015 EARS-net report on

antimicrobial resistance in Europe. Depicted are the percentages of methicillin-resistant Staphylococcus

aureus (MRSA), Escherichia coli extended-spectrum beta-lactamase (ESBL), vancomycin resistant Enterococcus

faecium (VRE), and Klebsiella pneumonia producing carbapenemase (KPC) in 2014 (except for Poland which

is 2013 data) on a logarithmic scale. The Netherlands and Germany are highlighted in green.

9 | Chapter 1

Within the Netherlands the resistance threat was recognized quite early and already in 1980 the

first interdisciplinary committee, the Working party Infection Prevention (Werkgroep

Infectiepreventie, WIP) was established. This was followed by a specific Working Party on

Antibiotic Policy (Stichting Werkgroep Antibioticabeleid, SWAB) in 1996. Within these

committees, specialists from Infectious Diseases, Internal Medicine, Pharmacy, Medical

Microbiology and Infection Prevention are working together. Both committees are responsible

for writing and maintaining binding national guidelines on infection prevention and antibiotic

use.

Within healthcare centers, the first major problem of highly resistant bacteria was the

emergence of outbreaks with methicillin-resistant Staphylococcus aureus (MRSA) in the USA in

the late 60’s. Slowly MRSA started to become endemic to healthcare institutions worldwide.

The Netherlands responded by implementing a strict search-and-destroy policy. This entails

proactive surveillance to detect (“search”) and then isolate and if possible decolonize patients

(“destroy”) that were found positive for MRSA. The search-and-destroy policy was a big

success and nowadays the Netherlands is one of the few countries that continues to keep a low

prevalence (besides the Scandinavian countries, which have similar policies) (see Figure 1.1).

Within a large EUREGIO project (MRSA-Net, see also the info box), German border regions

implemented a search-and-follow strategy for MRSA, adopting the Dutch principles. This has

led to substantial and significant drop in MRSA incidence, once more showing the

effectiveness of such a program (Jurke, et al., 2013). Next to a proactive screening policy, there

is also a strong focus on appropriate antibiotic use. Also in this respect, the Netherlands

performs exceptionally well compared to many neighboring countries (European Centre for

Disease Prevention and Control, 2013).

A department of Medical Microbiology should stimulate compliance with the guidelines on

correct antimicrobial use and infection prevention, as in both cases the results from

microbiological cultures provide essential diagnostic information. Nowadays the department is

often combined with an infection prevention and/or hospital hygiene unit. In that case, the

department will consist of clinical microbiologists who actively work together with infection

prevention specialists (Deskundige InfectiePreventie, DIP, in the Dutch system). They are

supplemented with additional specialists depending on the hospital. Academic microbiology

departments will most often also include molecular microbiologists and sometimes also

infectious disease specialists. Together they are responsible for the microbiological diagnostics

within the hospital. This includes among others bacterial cultures, virological assays and next-

generation diagnostics such as whole genome sequencing tools.


With these diagnostics it is possible to

identify as quickly as possible the micro-

organism(s) that might be responsible for the

patient’s problem. Also part of the identification is

a resistance pattern in order to advise the treating

physicians on the correct antibiotic(s) for which the

pathogen is susceptible. Something that is

becoming more important with growing resistance

levels.

Patients in academic hospitals, such as the

University Medical Center Groningen, or other

referral hospitals, are coming more frequently from

further away (even from abroad), instead of from a

small region surrounding the hospital. Also,

patients are more often transferred from another

general hospital to an academic center (Donker, et

al., 2010). This means that the bacteria that patients

are carrying are also coming from a much larger

region compared to small hospitals, which directly

influences the local resistance rates in the hospital.

It is thus important to take this into account when

thinking about antibiotic policies and infection

prevention (Ciccolini, et al., 2013).

Especially with patients coming from a

country with higher resistance rates as compared to

the Netherlands (e.g. Germany). The risk of

importing resistant bacteria becomes much higher and must be taken into account. Regional

collaboration is therefore crucial (Donker, et al., 2015). The Medical Microbiology department

of the UMCG works together in a regional collaboration of healthcare centers and

microbiology laboratories (Regionaal Microbiologisch Infectiologisch Symposium, REMIS+).

Furthermore it is part of a Euregional project, which is a large collaboration of Dutch and

German healthcare centers and professionals within the border region. These Interreg-V

projects, called health-i-care and EurHealth-1Health started in 2016. Previous Interreg projects

include EurSafety Health-Net (see box) and MRSA-Net. All programs focus on improving

infection prevention interdisciplinary, intersectional and cross-border.

The Euregional Interreg-IV project

EurSafety Health-Net started in 2009

after the successful Interreg-V project

MRSA-Net ended. The goal of the

EurSafety project was to improve

patient care and safety and a better

protection for patients against

hospital associated infections. The

project had numerous partners and

project leaders, covering the whole

border of the Netherlands (and

Belgium) with Germany. Among

others, it introduced a quality seal for

healthcare institutions and numerous

publications describing infection

prevention related issues between the

Netherlands and Germany, in order

to learn from each other and improve

patient care. In 2014 the European

Commission elected the project as a

3-star project. The related Interreg-V

projects health-i-care and EurHealth-

1Health commenced in May 2016.

See for more information:

http://www.eursafety.eu.

11 | Chapter 1

Scope of this thesis

Regional connectivity as mentioned above, has consequences for an antimicrobial stewardship

and infection prevention measures within a hospital. For example, for a hospital like the

UMCG, it means that they receive more patients from Germany than hospitals in the west of

the Netherlands. Antimicrobial use and resistance data from Germany are thus important

information to take into account and on some levels (such as MRSA) we know that there are

large differences between the two countries (van Cleef, et al., 2012). Chapter 2 of the thesis is

an example of a cross-border evaluation study of specific antimicrobial use between the two

bordering regions. Such studies are important in order to detect and analyze antimicrobial

resistance levels of patients. If needed, this information can be used to modify or update

guidelines regarding antimicrobial stewardship and infection prevention, in order to provide

optimal care within a hospital.

When a patient enters the hospital and is suspected of having an infection, many different

medical disciplines will contribute to the care of this patient at one time or another. Most

notably of course the department of Medical Microbiology, but also Infectious Diseases, the

Pharmacy and a Surgical or Internal Medicine discipline. This integrative approach is laid out in

more detail in Chapter 3. Here the process of different interventions is explained and

described as a novel model that incorporates three complementary stewardship programs:

Antimicrobial, Infection Prevention, and Diagnostic Stewardship (AID). The stewardship

programs are a set of different interventions and actions initiated by the Department of

Medical Microbiology. This model provides the theoretical framework and structure for the

rest of this thesis.

Firstly the Antimicrobial Stewardship Program (ASP) is discussed. The Society for Healthcare

Epidemiology of America (SHEA) and the Infectious Diseases Society of America (IDSA)

gave a clear definition of an ASP in their 2012 position paper:

“… [it] refers to coordinated interventions designed to improve and measure the appropriate use of antimicrobial

agents by promoting the selection of the optimal antimicrobial drug regimen including dosing, duration of

therapy, and route of administration. The major objectives of antimicrobial stewardship are to achieve best

clinical outcomes related to antimicrobial use while minimizing toxicity and other adverse events, thereby limiting

the selective pressure on bacterial populations that drives the emergence of antimicrobial-resistant strains.”

(Society for Healthcare Epidemiology of America, et al., 2012).

An ASP is effectively a catch-all for all kinds of interventions done to improve

antimicrobial therapy. Although there are certain interventions that are often associated with

an ASP (e.g. a switch program intravenous to oral, or an audit-and-feedback program), there is


no clear definition of the contents of an ASP. This is also strongly dependent on the setting in

which an ASP will be implemented. Thus, before starting it is therefore important to make an

assessment of the baseline situation (van den Bosch, et al., 2015). As stated, an ASP should

focus on improving antimicrobial therapy by promoting the correct therapeutic in the correct

dosage and route of administration for the correct duration. Different interventions can be

necessary to achieve all these different goals. However, optimal therapy is not the only goal,

but also a means to improve sustainability of patient care by reducing selective pressure that

drives bacteria to become resistant.

As mentioned before, it is essential to evaluate practices within the hospital. Especially with an

ASP, consisting of several interventions, it is important to know which interventions are

effective and which not. This should be done not just by looking at clinical outcomes, but also

by looking at the financial aspects of each intervention. The ASP section therefore starts with

an overview of different methods to evaluate an ASP. Chapter 4 is an expert review that

discusses ASP interventions and focuses mainly on the different methods to evaluate those

interventions, clinically and financially. These financial evaluations are sometimes overlooked

or underestimated evaluative analyses. This has consequences for the quality of these

evaluations, because it is highly undesirable to draw conclusions on the effectiveness of an

intervention if the costs and benefits are unknown or unclear (Drummond, et al., 2005). We

systematically reviewed publications that included financial evaluations in Chapter 5.

At the UMCG, as part of the local ASP, an Antimicrobial Stewardship-Team or A-

Team was implemented in 2012 (Lo-Ten-Foe, et al., 2014). The A-Team works following a

consensus-based “day-2 bundle”. The evaluation is part of this thesis. Using a specially created

email alert, antimicrobial therapy is evaluated by an A-Team member on the second day of

therapy. The goal is to streamline therapy as quickly as possible using the first results of the

microbiological diagnostics, together with other available data and the clinical course. By

focusing on simple improvements, such as an early stop, or switch to oral administration,

antimicrobial therapy should be optimized relatively uncomplicated (Goff, et al., 2012; Pulcini,

et al., 2008). The effects of the implemented A-Team was first evaluated looking at clinical

outcome measures (e.g. length of stay and antimicrobial use) (Chapter 6) and subsequently

also financially (Chapter 7).

Regarding Infection Prevention Stewardship, two studies were performed to assess the impact

on a subset of outcome measures (e.g. costs, number of patients colonized with resistant

microorganisms per year, and number of outbreak patients per year). An important task of the

Infection Prevention Division is recognizing, controlling and finally clearing outbreaks caused

by resistant microorganisms. Actions performed to do so cost time, consumables, manpower,

and revenue due to closed beds. Until now however, it was unclear what exactly is done when

13 | Chapter 1

the UMCG has an outbreak and what the costs are to control it. Seven different outbreaks that

occurred between 2012 and 2014 were therefore retrospectively evaluated in Chapter 8. With

rising resistance levels worldwide, but also in the Netherlands, there are more and more

patients entering the hospital each year that carry (resistant) bacteria with them that can cause

outbreaks. Prevention is therefore becoming more and more important as well. The hospital

and department of Medical Microbiology invest each year extra money to keep up with the

growing number of risk patients. Is it cost-beneficial to do these investments? Chapter 9

describes eight years of infection prevention in the UMCG looking at costs, control measures,

utensil use, number of expected outbreak patients and the actual found number of patients, to

make an estimation on the financial costs and benefits.

Finally, regarding Diagnostic Stewardship, the effect of taking blood for cultures was studied.

Blood cultures take up a major part of (diagnostic) cultures that are performed. The UMCG

performs over 10.000 every year, making it the second most frequent material for cultures after

urine. Using the data of 5 years of admissions, a large dataset of (infectious) patients receiving

broad-spectrum antibiotics from the start of their admission was constructed. Performing

blood cultures is included in almost all guidelines for severe infections. The use of antibiotics

can influence the result of a culture; so appropriate diagnostics should be performed before the

start of therapy. Until now however, effects of performing these blood cultures remain unclear.

Therefore, this evaluation in Chapter 10 focuses on several clinical outcome measures for

patients receiving antimicrobial therapy with and without blood cultures.

Using all evaluations on the three different focus points of the department (on antimicrobial

therapy, infection prevention and control, and diagnostics), a more wide-ranging and

comprehensive impact analysis on relevant outcome measures can be performed leading to a

general conclusion of this research as well as to some recommendations to improve the

healthcare system and the overall quality of healthcare (Chapter 11).

Cross-border comparison of antibiotic prescriptions among

children and adolescents between the north of the

Netherlands and the north-west of Germany

Jan-Willem H. Dik, Bhanu Sinha, Alex W. Friedrich, Jerome R. Lo-Ten-Foe,

Ron Hendrix, Robin Köck, Bert Bijker, Maarten J. Postma, Michael H. Freitag,

Gerd Glaeske, Falk Hoffmann

Antimicrobial Resistance and Infection Control. 2016; 5:14

2

17 | Chapter 2

Abstract

Antibiotic resistance is a worldwide problem and inappropriate prescriptions are

a cause. Especially among children, prescriptions tend to be high. It is unclear

how they differ in bordering regions. This study therefore examined the

antibiotic prescription prevalence among children in primary care between

northern Netherlands and north-west of Germany.

Two datasets were used: The Dutch (IADB) comprises representative data of

pharmacists in North Netherland and the German (BARMER GEK) includes

nationwide health insurance data. Both were filtered using postal codes to

define two comparable bordering regions with patients under 18 years for 2010.

The proportion of primary care patients receiving at least one antibiotic was

lower in northern Netherlands (29.8%; 95% confidence interval [95% CI]: 29.3-

30.3), compared to north-west Germany (38.9%; 95% CI: 38.2-39.6). Within the

respective countries, there were variations ranging from 27.0 to 44.1% between

different areas. Most profound was the difference in second-generation

cephalosporins: for German children 25% of the total prescriptions, while for

Dutch children it was less than 0.1%.

This study is the first to compare outpatient antibiotic prescriptions among

children in primary care practices in bordering regions of two countries. Large

differences were seen within and between the countries, with overall higher

prescription prevalence in Germany. Considering increasing cross-border

healthcare, these comparisons are highly valuable and help act upon antibiotic

resistance in the first line of care in an international approach.

Cross-border comparison of antibiotic prescriptions among children and adolescents between the north of the | 18


Introduction

Inappropriate use of antibiotics leads to significant clinical and economic problems due to

resistant bacteria and increasingly limited anti-infective treatment options (Cosgrove, 2006;

Goossens, 2009). The majority of antibiotics are prescribed in primary care, and children

receive a large portion of these prescriptions (Brauer, et al., 2015; European Centre for Disease

Prevention and Control, 2014). These primary care prescriptions of antibiotics are contributing

to the world-wide resistance problem (Costelloe, et al., 2010) and promotion of appropriate

use is still of great importance (Earnshaw, et al., 2014). Within the European Union there are

large variations in outpatient antibiotic prescriptions. In the Netherlands, the level of

prescriptions has traditionally been one of the lowest in Europe. In Germany, generally the

prescription level is also relatively low (European Centre for Disease Prevention and Control,

2014). However, when it comes to antibiotic prescriptions for children and adolescents,

prescriptions in Germany seem to be higher than in other countries (Holstiege and Garbe,

2013; Holstiege, et al., 2014). Regarding antibiotic resistance Germany has also slightly higher

levels than the Netherlands (European Centre for Disease Prevention and Control, 2013).

Due to a recent EU directive, patients can more easily obtain health services in other

European member states. Directive 2011/24/EU creates the possibility for EU citizens to

cross borders and seek healthcare in another country. This possibility for cross-border care is

especially relevant for bordering regions such as the north of the Netherlands and north-west

Germany, where Dutch healthcare centers can treat German patients and vice versa.

Considering the potential effects of inappropriate antibiotic prescriptions in primary care, in

particular antibiotic resistance development, it is relevant to look at differences in prescriptions

between countries. It is therefore interesting to approach the topic of antimicrobial therapies

and their possible unwanted effects by international cross-border collaboration, especially

between bordering regions (Friedrich, et al., 2008b). Notably, children and adolescents

represent large group of recipients of antibiotics, in particular in outpatient care (Brauer, et al.,

2015; European Centre for Disease Prevention and Control, 2014). The goal of this study was

to examine the prevalence and most frequently prescribed antibiotics among children and

adolescents in outpatient care in adjacent Dutch and German regions, and to compare this

data. We also aimed to answer if different healthcare systems influence prescribing or if

prescribing in regions gets more comparable the nearer they are to the border.

Methods

Study design and setting

To address the question, we chose a retrospective cross-sectional study in a predominantly

rural region including analysis of outpatients’ data of a health insurance company and a

pharmacy research database for north western Germany (postal codes 26xxx; part of Lower

19 | Chapter 2

Saxony) and the northern Netherlands (postal codes 9xxx, Groningen/Drenthe). Both regions

have no major geographic and infection risk differences, have about 1 million inhabitants and

share a common green border (Figure 2.1). Data comprises the year 2010 and we focused on

persons aged 0 to 18 years. Orally administered antibiotics were selected based on the

Anatomical Therapeutic Chemical (ATC) code J01 in the outpatient setting.

Data sources

The German data including

pharmacy dispensing data

come from one of the largest

health insurance companies,

the BARMER GEK,

representing approximately

8.4 million persons

nationwide (of them about

104,000 in postal codes

26xxx). In total, there were

160 different statutory health

insurance companies

(including BARMER GEK)

covering a total of 70 million

persons (87% of the German

population) at the end of

2010, while the remaining

inhabitants are privately

insured. The study cohort

consisted of persons who

were insured at least one

day in each of the four

quarters of 2010. Due to this criterion, the vast majority of our study cohort has been

continuously insured throughout the whole year but infants born in the first quarter could also

be analysed (Hoffmann, et al., 2012).

The Dutch data were derived from pharmacy dispensing data of the pharmacy

research database (http://www.IADB.nl). The IADB comprises prescriptions derived from 55

community pharmacies in the northern part of the Netherlands and has in total approximately

600,000 persons in the database (of them about 263,000 in postal codes 9xxx). Since Dutch

patients are in general registered at one single local pharmacy in their hometown, the chance of

multiple prescriptions of one person being counted in different areas is relatively small.

Registration is furthermore irrespective of healthcare insurance and age, gender and

Figure 2.2: Study regions. North of the Netherlands (postal codes

9xxx; n=36,747) and north-west Germany (postal codes 26xxx;

n=18,374).



prescription rates among the database population have been representative for the whole of

the Netherlands (Visser, et al., 2013). Prescription records are virtually complete due to the

high patient-pharmacy commitment in the Netherlands, except for medication dispensed

during hospitalization (Visser, et al., 2013).

Statistical analysis

Our main outcome was outpatient prescription prevalence, e.g. the proportion of children and

adolescents receiving at least one prescription for a systemic antibiotic in 2010. Prevalence was

stratified by sex, age groups (0-2, 3-6, 7-10, 11-13 and 14-17 years) and region of residence (six

areas for the Netherlands and nine for Germany) alongside with 95% confidence intervals

(95% CI). Furthermore, the most frequently prescribed antibiotic substances (by different

ATC-codes) were studied. Statistical analyses were performed with SAS, Version 9.2 (SAS

Institute Inc., Cary, NC). Maps were created with ESRI ArcGIS®, Version 10.2.2.

Results

The study cohort consisted of 36,747 children and adolescents under the age of 18 years, living

in the northern Netherlands (ranging between 649 - 20,739 individuals per region) and 18,374

from north-west Germany (ranging between 988 – 4,028 individuals per region). For the

Netherlands, the distribution male vs. female was 50% vs. 50% and for Germany 51% vs.

49%.

Overall, the proportion

of children and

adolescents receiving at

least one antibiotic

course in 2010 was

lower in the northern

Netherlands (29.8%;

95% CI: 29.3-30.3)

compared to north-

west Germany (38.9%;

95% CI: 38.2-39.6).

There were small area

variations ranging from

27.0 to 36.4% in the

northern Netherlands and from 35.1 to 44.1% in north-west Germany (Figure 2.2). Prevalence

stratified by region, sex and age groups is shown in Tables 2.1 and 2.2. The age groups with the

highest proportion of prescriptions were children between 0-2 years (northern Netherlands vs.

Figure 2.3: Regional variations (by postal codes) in the proportion of

children and adolescents with prescriptions of antibiotics in the

northern Netherlands and north-west Germany.

21 | Chapter 2

north-west Germany: 43.1% vs. 49.9%) and those between 3-6 years (37.4% vs. 54.8%). The

proportion was considerably higher in Germany than in the Netherlands in all age groups, for

males and for females. Males had a lower prevalence of antibiotic prescriptions than females

for both Netherlands and Germany in all age groups, except for children aged 0-2 years old

(Table 2.2).

Distributions of the antibiotic substances varied between the two bordering regions.

Amoxicillin was the most frequently prescribed substance in both regions, 49.6% of all

prescriptions in the Netherlands versus 21.1% in Germany. Another profound difference was

found for second-generation cephalosporins, which in the Netherlands is reserved as a second

line antibiotic. These antibiotics comprised 25% of the prescriptions in Germany and less than

0.1% in the Netherlands.

The five most frequent prescribed antibiotics for all paediatric age groups covered 81.0 % of

the total number of prescriptions in the Netherlands vs. 64.3 % in Germany (Table 2.3). In

both countries, the percentage of top 5 prescriptions decreased with age: in the Netherlands

from 94.2% in 0-2 year olds to 69.0% in 14-17 year olds and in Germany from 76.4% in 0-2

year olds to 53.6% in 14-17 year olds.

Table 2.1. Small area variations (by postal codes) in the proportion of children and adolescents with prescriptions of antibiotics in the northern Netherlands and north-west Germany (with 95% CI), by sex. Region (postal code) Males Females Total

Netherlands 95, 96 (n=10,439) 31.6% (30.3-32.9) 34.7% (33.4-36.0) 33.2% (32.3-34.1)

90, 91, 99 (n=649) 27.7% (23.0-32.8) 26.1% (21.3-31.4) 27.0% (23.6-30.6) 92, 98 (n=1,898) 36.2% (33.2-39.4) 36.6% (33.5-39.8) 36.4% (34.2-38.6) 93 (n=868) 28.0% (23.9-32.5) 28.0% (23.8-32.5) 28.0% (25.0-31.1) 94 (n=2,154) 28.0% (25.3-30.8) 29.2% (26.5-31.9) 28.6% (26.7-30.6) 97 (n=20,739) 26.7% (25.8-27.6) 28.9% (28.0-29.8) 27.8% (27.2-28.4) Germany 261 (n=4,028) 35.3% (33.2-37.4) 34.9% (32.8-37.0) 35.1% (33.6-36.6) 262 (n=988) 35.0% (30.8-39.3) 42.1% (37.6-46.6) 38.5% (35.4-41.6) 263 (n=1,744) 31.9% (28.8-35.1) 38.4% (35.2-41.7) 35.1% (32.9-37.4) 264 (n=1,959) 36.5% (33.5-39.6) 34.9% (31.8-37.9) 35.7% (33.6-37.8) 265 (n=1,102) 40.8% (36.7-45.1) 41.4% (37.2-45.6) 41.1% (38.2-44.1) 266 (n=2,476) 38.5% (35.8-41.2) 40.0% (37.2-42.8) 39.2% (37.3-41.2) 267 (n=1,263) 40.7% (36.7-44.7) 43.6% (39.8-47.5) 42.2% (39.5-45.0) 268 (n=3,642) 42.0% (39.8-44.3) 46.2% (43.9-48.6) 44.1% (42.5-45.7) 269 (n=1,172) 41.1% (37.2-45.1) 40.8% (36.7-45.0) 41.0% (38.1-43.8)



Discussion

Study findings and implications

This is one of the first studies to

look at outpatient antibiotic

prescriptions among children in the

bordering regions of two countries.

Considering the importance of

appropriate antibiotic use, the

ability for patients to seek

healthcare services abroad, and the

large differences between

healthcare systems and guidelines, a

cross-border comparison is of great

interest. Furthermore, these

comparisons may be beneficial to

inform healthcare providers and to

discuss best clinical practice. We

observed considerable differences

between the two countries. This

may indicate that improvements

can be achieved. Mainly on the

distribution of substances,

differences were profound: second

generation cephalosporins (i.e.

mostly oral cefuroxime) were

prescribed in 25% of the cases for

the German patients, while almost

none of the Dutch children

received this type (< 0.1%). Given

the low rate of oral bioavailability

and the high selective pressure due

to these substances, they are

avoided wherever possible in

ambulatory paediatric care in the

Netherlands. It would be highly

interesting and relevant to perform

further research into the reasons

for prescribing second-generation

cephalosporins in Germany and its

long term effects. Tab

le 2

.2:

Pro

po

rtio

n o

f ch

ild

ren

an

d a

do

lesc

en

ts w

ith

pre

scri

pti

on

s o

f an

tib

ioti

cs

in t

he n

ort

hern

Neth

erl

an

ds

an

d n

ort

h-w

est

Germ

an

y (

wit

h 9

5%

CI)

, b

y s

ex

an

d a

ge g

rou

p.

To

tal

Germ

an

y

(n=

18,3

74)

49.9

%

(47.6

-52.3

)

54.8

%

(53.2

-56.5

)

36.4

%

(35.0

-37.9

)

28.3

%

(26.9

-29.8

)

34.0

%

(32.7

-35.3

)

38.9

%

(38.2

-39.6

)

Neth

erl

an

ds

(n=

36,7

47)

43.1

%

(42.0

-44.2

)

37.4

%

(36.2

-38.5

)

23.4

%

(22.4

-24.4

)

17.9

%

(16.8

-18.9

)

23.4

%

(22.6

-24.3

)

29.8

%

(29.3

-30.3

)

Fem

ale

s

Germ

an

y

(n=

9,0

91)

46.6

%

(43.2

-50.0

)

53.3

%

(50.9

-55.7

)

37.4

%

35.4

-39.5

)

28.9

%

26.9

-31.0

)

39.1

%

(37.2

-41.0

)

39.9

%

(38.9

-40.9

)

Neth

erl

an

ds

(n=

18,5

18)

41.5

%

(39.9

-43.1

)

39.9

%

(38.2

-41.5

)

27.7

%

(26.1

-29.2

)

19.6

%

(18.0

-21.2

)

25.2

%

(24.0

-26.3

)

30.9

%

(30.3

-31.6

)

Male

s

Germ

an

y

(n=

9,2

83)

53.1

%

(49.8

-56.4

)

56.2

%

(53.9

-58.5

)

35.4

%

(33.3

-37.4

)

27.7

%

(25.6

-29.8

)

29.1

%

(27.4

-30.9

)

37.9

%

(36.9

-38.9

)

Neth

erl

an

ds

(n=

18,2

29)

44.5

%

(43.0

-46.0

)

35.2

%

(33.8

-36.8

)

19.6

%

(18.3

-20.9

)

16.3

%

(14.9

-17.8

)

20.7

%

(19.4

-22.1

)

28.7

%

(28.0

-29.4

)

Ag

e

0-2

yrs

.

(n=

8,0

75 r

esp

. n

=1,7

70)

3-6

yrs

.

(n=

7,4

64 r

esp

. n

=3,4

88)

7-1

0 y

rs.

(n=

6,9

16 r

esp

. n

=4,2

92)

11-1

3 y

rs.

(n=

5,0

78 r

esp

. n

=3,7

33)

14-1

7 y

rs.

(n=

9,2

14 r

esp

. n

=5,0

91)

To

tal

(n=

36,7

47 r

esp

. n

=18,3

74)

23 | Chapter 2

These differences in prescriptions of antibiotics for outpatients between European

countries have been reported earlier, although not for bordering regions. There are however

comparisons showing differences between Germany and the Netherlands in general (European

Centre for Disease Prevention and Control, 2014), as well as for children (European Centre for

Disease Prevention and Control, 2013). In the Netherlands, guideline adherence is higher

compared to Germany (Philips, et al., 2014), which might also be a reason for different

distributions of substances between both countries. In 7-17 years old children, nitrofurantoin

was among the top 5 antibiotics in the Netherlands but not in Germany. In Germany, there

may still be hesitance to prescribe nitrofurantoin due to a history of warnings in the past

(pulmonary fibrosis, neuropathy and liver damage) (AT, 1993).

In addition, the occurrence of resistant bacteria such as MRSA differs also

significantly between the bordering regions of the Netherlands and Germany, with up to 32-

fold higher MRSA incidence in the German border region compared to the adjacent Dutch

border region (van Cleef, et al., 2012). In addition, higher resistance rates were also observed

for classical community-acquired pathogens such as pneumococci where penicillin-resistance

involved 1.9% of invasive isolates in Germany vs. 0.2% in the Netherlands in 2013 (European

Centre for Disease Prevention and Control, 2015). The Dutch prevalence data observed in this

study are comparable to earlier studies, indicating a quite stable use and also corroborating the

fact that the IADB database can be considered as representative for the whole county (de Jong,

et al., 2008; European Centre for Disease Prevention and Control, 2013). Comparing various

German studies there is a bit more variation, but it is known that there is a quite large regional

variation of outpatient antibiotic prescriptions within the country. The north-western part of

Germany, which is included in this study, is one of the higher prescribing regions (Kern, et al.,

2006; Koller, et al., 2013).

For both datasets, clinical indications for the prescriptions analyzed were not known.

For the Netherlands, a large survey showed that the primary diagnosis for children coming to

general practitioners is lower respiratory tract infections (van der Linden, et al., 2005).

Antibiotic prescriptions by general practitioners for children in the Netherlands are mainly for

acute otitis media and bronchitis, and especially broad-spectrum antibiotics are still prescribed

inappropriately (Otters, et al., 2004). Dutch guidelines regarding antibiotic treatment are very

strict. Otitis media guidelines recommend antibiotics only when there are other risk factors for

complications or with severe general symptoms (NHG, 2014). For respiratory tract infections,

antibiotics are only recommended in the case of pneumonia (NHG, 2013). In Germany most

diagnoses for children (0-15 years) in an outpatient setting were upper respiratory tract

infections without a focus, fever without a focus and acute bronchitis (Grobe, et al., 2012).

Antibiotic prescriptions for this group are mainly given for acute tonsillitis, bronchitis and

otitis media, for all of which appropriateness of antibiotics is debatable (Holstiege, et al., 2014).

The German guideline for otitis media recommends antibiotic treatment only to be started

after two days, thereby being less conservative than the Dutch counterpart (DEGAM, 2014b).

The general guideline for bronchitis states that when uncomplicated, antibiotics are not

recommended and should be avoided (DEGAM, 2014a).



Table 2.3. The top 5 most prescribed antibiotic substances in the northern Netherlands and north-west Germany, by age group and the total. Age Netherlands % Germany %

0-2 yrs. 1st Amoxicillin 70.7% Cefaclor 29.5% 2nd Amoxicillin-clavulanate 10.5% Amoxicillin 22.0% 3rd Clarithromycin 6.6% Erythromycin 14.2% 4th Sulfamethoxazole and

trimethoprim 3.7% Cefuroxime 5.9%

5th Pheneticillin 2.7% Phenoxymethylpenicillin 4.8% Total of the 5 most prescribed 94.2% 76.4%

3-6 yrs. 1st Amoxicillin 56.1% Amoxicillin 22.2% 2nd Amoxicillin-clavulanate 13.6% Cefaclor 21.9% 3rd Clarithromycin 8.9% Phenoxymethylpenicillin 10.5% 4th Pheneticillin 4.5% Erythromycin 10.0% 5th Sulfamethoxazole and

trimethoprim 4.2% Sulfamethoxazole and

trimethoprim 7.9%

Total of the 5 most prescribed 87.4% 72.6%

7-10 yrs. 1st Amoxicillin 43.5% Amoxicillin 19.7% 2nd Amoxicillin-clavulanate 13.5% Cefaclor 18.1% 3rd Clarithromycin 9.6% Erythromycin 11.2% 4th Nitrofurantoin 7.8% Phenoxymethylpenicillin 11.1% 5th Flucloxacilline 6.6% Sulfamethoxazole and

trimethoprim 8.1%


11-13 yrs. 1st Amoxicillin 36.5% Amoxicillin 22.4% 2nd Amoxicillin-clavulanate 13.2% Cefaclor 12.6% 3rd Clarithromycin 10.0% Cefuroxime 10.7% 4th

Nitrofurantoin 8.9% Sulfamethoxazole and trimethoprim

8.7%

5th Sulfamethoxazole and trimethoprim

5.8% Phenoxymethylpenicillin 8.1%


14-17 yrs. 1st Nitrofurantoin 19.7% Amoxicillin 19.7% 2nd Amoxicillin 16.0% Cefuroxime 9.4% 3rd Doxycycline 14.8% Azithromycin 8.4% 4th Amoxicillin-clavulanate 10.5% Phenoxymethylpenicillin 8.1% 5th Pheneticillin 8.1% Sulfamethoxazole and

trimethoprim 8.0%


0-17 yrs. 1st Amoxicillin 49.6% Amoxicillin 21.1% 2nd Amoxicillin-clavulanate 11.9% Cefaclor 17.3% 3rd Clarithromycin 7.7% Phenoxymethylpenicillin 9.0% 4th Nitrofurantoin 7.0% Erythromycin 9.0% 5th Pheneticillin 4.8% Cefuroxime 7.9% Total of the 5 most prescribed 81.0% 64.3%

25 | Chapter 2

Influence of parents on the prescribed antibiotic treatment seems to be relatively small. A

European survey, although not performed in the Netherlands or Germany, showed that

patients tend to adhere to the decision of the general practitioner even when they disagree

(Brookes-Howell, et al., 2014). When they disagree, they have a tendency to be more

conservative than the physician (Hawkings, et al., 2008). A survey in Germany confirms this

and shows that the large majority of patients understand the limitations of antibiotic treatment

for indications like the common cold (Faber, et al., 2010). The influence of the family

practitioner or paediatrician thus seems to be often underestimated, whereas they show indeed

a high inter-individual variation in their prescription pattern (Cars and Håkansson, 1997).

Perceptions of antibiotic resistance among general practitioners also differs between countries

(Wood, et al., 2013), probably also leading to different prescription behaviour. A combination

of these aspects is most likely leading to the differences between the Netherlands and

Germany.

Strengths and limitations

The major strength of this study is the unique dataset of two bordering regions coming from

countries with different healthcare systems and antibiotic prescribing policies. Other studies

compared nationwide data (either from a subset of databases or up to 100% coverage such as

most of ESAC). However, as shown here for the first time, there are also large small area

variations among and between bordering regions from two different countries. Studies

comparing national consumption data, aggregate these data and variations within a country are

then lost, making it impossible to effectively compare bordering regions. We were able to

include about 37,000 children and adolescents living in the northern Netherlands as well as

18,000 living in north-west Germany. However, especially in the Netherlands our cohort size

differs relevantly per area (depending on the distribution of pharmacies included in the

IADB.nl database). Therefore, in some areas several postal codes were combined. Age

distribution and drug utilization of the patients included in the database is, however,

representative for the total Dutch population (Visser, et al., 2013). For Germany, the sample

size was somewhat smaller than for the Netherlands. German data were derived from a large

health insurance fund and we know that differences exist regarding socioeconomic status and

morbidity between these insurance funds (Hoffmann and Icks, 2012). Such differences were

found in children and adolescents, too, but the utilization of medications within the specific

fund we used was quite comparable to the complete German population (Hoffmann and

Bachmann, 2014). Unfortunately, we had no access to diagnoses and indications for which

antibiotics were prescribed. These data would be relevant in determining the appropriateness

of the (antibiotic) treatments, and also could shed light on the question if some patients might

even be undertreated. It seems that coming closer to the border increases antibiotic

consumption in both countries. One explanation might be, that these parts of the country are

furthest away from an academic centre. These (rural) parts of the countries are also socio-

economically weaker compared to the more densely populated parts. Such a lower socio-



economic status appears to influence antibiotic prescribing, although it is unclear to which

extent (Covvey, et al., 2014; Ternhag, et al., 2014). However, more precise information is not

available within the datasets used. This study should form a starting point for (regional)

antimicrobial stewardship programs focusing on general practitioners and outpatients. This

group is still somewhat neglected in stewardship programs, but these data show that there is a

lot to gain.

It is important to keep in mind that the structures and organization of the two healthcare

systems differ substantially between the Netherlands and Germany. In the former, there are

only family practitioners in private practice, whereas nearly all specialists are working in

outpatient clinics in (larger) hospitals. Hence, the data analysed in this study primarily contain

prescription data for these family physicians. In Germany, the majority of medical specialists,

including paediatricians, is working in private practice (or consortiums). The German data set

comprised also the data from these specialists. Previous analysis for the whole of Germany

showed that 49% of the prescriptions came from paediatricians and 35% from general

practitioners (Koller, et al., 2013). This may also influence individual prescribing patterns due

to a variety of reasons and cause a bias due to more severe cases treated by specialists on the

German side of the border, although we hypothesize that severe cases are most likely send to a

hospital and are thus not included in this dataset. One may speculate that the more

individualized healthcare system for primary care in Germany might lead to a more

heterogeneous healthcare behaviour than the more peer group-dependent gatekeeper system in

the Netherlands. Prescribing patterns are influenced by many different factors. Other

differences between the Netherlands and Germany (e.g. medical education, performing

microbiological diagnostics or healthcare insurance system) are most likely also of influence,

however, to which degree is uncertain and should be subject to further investigation.

Conclusions

Concluding, this study shows clear differences between primary care antibiotic prescriptions

for children and adolescents in the bordering regions of the Netherlands and Germany.

Especially in the age group of 3-6 year-old children, prescriptions are more frequent in

Germany. Overall there also seems to be a tendency to prescribe broader substances in

Germany compared to the Netherlands. An evaluation like this is especially interesting,

considering that most antibiotics in a primary care setting are prescribed for children. Keeping

in mind the effects that sub-optimal antibiotic treatments can have, these comparisons of

bordering regions between two countries provide an opportunity to learn from each other and

collaborate internationally in order to counteract the problems of rising antibiotic resistance

from a first line of care perspective.

28

An integrated stewardship model:

Antimicrobial, Infection prevention, and Diagnostic (AID)

Jan-Willem H. Dik, Randy Poelman, Alex W. Friedrich, Prashant Nannan Panday, Jerome R.

Lo-Ten-Foe, Sander van Assen, Julia E.W.C. van Gemert-Pijnen, Hubert G.M. Niesters,

Ron Hendrix, Bhanu Sinha

Future Microbiology 2016; 11(1):93-102

3

29 | Chapter 3

Abstract

Considering the threat of antimicrobial resistance and the difficulties this entails

for treating infections, it is necessary to cross borders and approach infection

management in an integrated multi-disciplinary manner.

We propose the AID stewardship model consisting of three intertwined

programs: Antimicrobial, Infection prevention and Diagnostic Stewardship,

involving all stakeholders. The focus is a so-called “theragnostics” approach.

This leads to a personalized infection management plan, improving patient care

and minimizing resistance development. Furthermore, it is important that

healthcare regions nationally and internationally are working together; ensuring

that patient (and microorganism) transfers will not be causing problems in a

neighboring institution.

This AID stewardship model can serve as a blue print to implement innovative,

integrative infection management.

An integrated stewardship model: Antimicrobial, Infection prevention, and Diagnostic (AID) | 30

Introduction

Antimicrobial resistance is a growing public health threat, signaling the beginning of a post-

antimicrobial era (World Health Organization, 2012). Infections caused by Multi Drug

Resistant Organisms (MDRO) are associated with a significant deterioration of clinical

outcomes. This includes an increased risk of mortality and morbidity and it is associated with

an increase in costs (Smith and Coast, 2013). Recently, consensus has been achieved that the

global community should act to limit these emerging problems as much as possible

(Laxminarayan, et al., 2013; Smith and Coast, 2013; World Health Organization, 2012).

Emergence of antimicrobial resistance is strongly correlated with incorrect antimicrobial

prescribing patterns and the lack of consistent diagnostic procedures to identify the pathogens

involved, whether viral, bacterial or fungal. While antimicrobial medication has undoubtedly

reduced mortality due to infections, resistance to these drugs renders them ineffective. To

address these issues, healthcare institutions are implementing Antimicrobial Stewardship

Programs (ASPs) to manage antimicrobial usage with the goal of improving patient outcomes,

minimizing collateral damage, and reducing the incidence of MDRO infections. They aim at

treating patients according to the pathogens involved, based on a diagnostic strategy, that

ultimately keeps control on increasing expenditures in healthcare (Davey, et al., 2013; Dik, et

al., 2015c; van Limburg, et al., 2014). Participating in ASPs is mainly seen as a task of clinical

microbiologists and/or infectious diseases specialists, together with (hospital) pharmacists.

However, as patients move within the hospital, many more stakeholders are involved: bedside

doctors, nurses, and boards of directors and diagnostic laboratory directors who are

responsible for providing financial resources. Furthermore, patients are not only moving

within one particular hospital but also between different hospitals and other healthcare

institutions as part of a comprehensive healthcare network. Therefore, infection management

in a specific (part of an) institution will affect patients throughout the entire network (Ciccolini,

et al., 2013; Donker, et al., 2010). Infection management is thus a responsibility of all

stakeholders involved in this network. Additionally, recent developments in cross-border

patient care have even extended these current healthcare networks to different countries,

following the new EU directive on patients’ rights in cross-border healthcare (directive

2011/24/EU) (Deurenberg, et al., 2009; Friedrich, et al., 2008a; Köck, et al., 2009). Because

microbes will not adhere to any borders made by humans (i.e. departments, institutions or

countries) we are forced to think and act in the same manner and closely collaborate in

developing and implementing strategies that prevents patients from being colonized or

infected with MDRO.

Cohen et al., recently advocated the use of a “multifaceted ‘omics’ approach”, to decrease the

turn-around-time of microbiological diagnostics (Cohen, et al., 2015). We agree with this

vision, but think this solution is part of a bigger picture. New diagnostic tools are important,

but so is the interpretation and translation of the results into the day-to-day infection

management. This requires cooperation of everyone involved in one large network. We

31 | Chapter 3

therefore strongly advocate collaboration of several disciplines within and between involved

(healthcare) institutions, regionally and internationally, thereby crossing multiple borders. The

main challenge is to develop an integrated system of not only Antimicrobial Stewardship

Programs (ASPs), but also Infection Prevention Stewardship Programs (ISP) as well as

Diagnostic Stewardship Programs (DSP), combined in an “AID stewardship model”. These

combined stewardship programs should be consultancy-based systems, involving and

supporting stakeholders within the healthcare network (Figure 3.1), aiming at optimizing

(laboratory) diagnostics, interpreting results, and initiating correct and appropriate

antimicrobial therapy (Figure 3.2).

Figure 3.1: Multi Stakeholder Platform of the AID stewardship model. Pyramid platform showing

the interdisciplinary stakeholder connections between the Antimicrobial Stewardship Program (ASP),

Infection prevention Stewardship Program (ISP) and Diagnostic Stewardship Program (DSP) within the

AID stewardship model. This model represents the complexity of the patients (green for low complex,

orange for intermediate complex and red for high complex) that corresponds with number of patients

and treating staff (width of the pyramid) and the experience level of the treating staff (height of the

pyramid). The more complex a patient, the more he/she requires experienced specialists from multiple

disciplines supplemented with correct, on-time performed diagnostics and eHealth tools. This together is

needed to adequately deal with the specific infectious problems, whereby the complexity of the patient is

not a fixed label, but a continuous changing state which varies over time.


Furthermore, they act at the network level, aiding in taking the right infection control measures

in order to provide a safe environment for patients and healthcare workers. Ultimately, this

should also lead to more cost-effective healthcare in the mid to long term. In this view much

can be learned from the theragnostic approach in cancer therapy, a holistic approach

combining genetics, nuclear medicine and laboratory diagnostics guiding to adjust therapy

(Pene, et al., 2009).

Modern and rapid diagnostics should focus on individual patient care. Thus, from the

viewpoint of individual patients, fast identification of the causative agent, initiating optimal

therapy and preventing the spread of highly infectious and pathogenic microorganisms are

crucial. A more classical approach divides the world into microbes, host and population, and

leads to a division between diagnostics, patient care and public health. This is particularly clear

in countries were microbiology laboratories are located far away from patient care and even

from clinical expertise. A stringent patient-oriented and personalized approach is needed where

clinical consequences are directly dependent on timely generated microbiological (molecular)

diagnostics. Consequently, in the field of antimicrobial therapy, appropriate and timely

diagnostics needs to be initiated before starting therapy. This will improve infection

management on patient level by avoiding inadequate therapy and thus preventing collateral

damage such as toxicity, driving antimicrobial resistance, and spread of (MDR) organisms

within the healthcare network.

To attain this goal of a theragnostic approach to infection management, we have developed a

cross-border AID stewardship program, which is already partly implemented. Many are

contributing to this integrated infection management system, each with a specific background

and training. To facilitate this innovative and integral approach of infection management, the

implementation of novel eHealth systems is essential as well as input of social sciences to take

into account the influence of behavior (Janes, et al., 2012; van Gemert-Pijnen, et al., 2013; van

Gemert-Pijnen, et al., 2011). Our developed program is an example of how the AID

stewardship model could be implemented and can provide a blueprint or inspiration for others

in the field of infection management and beyond.

33 | Chapter 3

Figure 3.2: Master scheme AID stewardship model. Flow chart that depicts the path of care of a

patient from top to bottom, surrounded by (several of) the building blocks of the three different, but

supplemental, stewardship programs and how they are intertwined. Notice the overlap, thereby showing

the necessity of all three stewardship programs and integrative nature of the model.


Antimicrobial stewardship

Within the AID stewardship model, the antimicrobial stewardship program (ASP) is based on

the American (SHEA/IDSA) guidelines and the Dutch and German guidelines for ASPs

(Dellit, et al., 2007; SWAB, 2012; With de, et al., 2013). It represents a program consisting of

various building blocks as shown in the master scheme (Figure 3.2). Key components of this

scheme are: appropriate and timely microbiological diagnostics, calculated empirical therapy

based on up-to-date regional/local epidemiology, close cooperation with the pharmacy

department (Cooke and Holmes, 2007; Lo-Ten-Foe, et al., 2014; Pulcini, et al., 2008a), and

continuous clinical and proper financial outcome evaluations (Dik, et al., 2015c).

Our ASP program within the AID stewardship model especially focuses on a day-2 bundle

with intervention taking place on the second day of antimicrobial therapy through face-to-face

case audits by the Antimicrobial Stewardship-Team (A-Team) (Dik, et al., 2015a; Lo-Ten-Foe,

et al., 2014). Triggered by an email-alert an A-Team member will go to the ward and discusses

the antimicrobial therapy with the bed-side physician with as goal to improve and streamline

the therapy using the by-then available diagnostics. This A-Team consists of clinical

microbiologists, infectious diseases specialists and hospital pharmacists. They are supported by

designated, trained doctors (link doctors) and nurses (link nurses) on every hospital ward. This

bundle intervention aims at pro-active discussions with prescribing doctors, leading to a

consensus-driven intervention optimizing antimicrobial therapy, depending on patient status

and the available diagnostic results. First results are highly positive, both clinically and

financially (Dik, et al., 2015a; Dik, et al., 2015b). This highly interactive, face-to-face approach

is dependent on the complexity of the patient’s condition, whereby the more complex patients

demand more sophisticated diagnostics and the expertise of (senior) specialists from different

disciplines (Figure 3.1). This implies crossing traditional borders between disciplines and

combining the knowledge and expertise of the treating physicians and the A-Team members

optimally for the benefit of the individual patient and the healthcare institution.

An important and in practice often somewhat neglected area is the optimization and

personalization of therapy. This should take into account basic patient factors such as

preliminary diagnosis, compartment of the infection, body weight, pharmacokinetic aspects

(including organ dysfunction, current volume of distribution, etc.), pharmacodynamic

characteristics of the drugs (e.g. killing activity, etc.), and characteristics of the microorganisms

(most relevantly the MIC – minimal inhibitory concentration). An integration of these factors

can be performed using PK/PD (pharmacokinetic and pharmacodynamic) data and models.

This approach can be used to optimize empiric therapy (based on population data for both

patients and microorganisms), as well as personalizing therapy for individual patients over

time. Especially for the latter aspect, appropriate TDM (therapeutic drug monitoring) needs to

35 | Chapter 3

be implemented. When integrating all available data, both optimizing the effect of

antimicrobial therapy as well as limiting collateral damage (toxicity and emergence of

resistance) can be addressed.

Before starting the development and implementation of an ASP, it is important to assess the

already implemented activities, as well as the specific requirements of the healthcare institution.

Therefore, we have developed an ASP maturity model approach, which can be used for such

an assessment. This will help tailor an ASP program to the specific needs of the healthcare

institution making it easier to integrate within an overarching model such as the AID model

(van Limburg, et al., 2014). Furthermore, preparing for a future with more cross-border patient

movements, it is vital to also facilitate cross-border capacity building for the development of

A-Teams in the clinical practice. This fosters collaboration and knowledge transfers between

science, health and small/medium enterprises ultimately leading to new innovative eHealth

technologies from an end-user involvement perspective (Janes, et al., 2012; van Gemert-Pijnen,

et al., 2013; van Gemert-Pijnen, et al., 2011). For the support of activities of the A-Teams, we

already have developed several of these eHealth tools, notably mobile applications and

automatic e-mail alerts, in order to facilitate easy access to diagnostic and therapeutic data,

which significantly improved the decision making processes (Beerlage-de Jong, et al., 2014;

Wentzel, et al., 2014). However, more eHealth tools are available, most notably different

clinical discussion support systems (CDSSs), which can aid and support the prescriber and/or

the A-Team in optimizing antimicrobial therapy (Beaulieu, et al., 2013). It is important to

create an ICT-structure that can be used to transfer patient data safe and confidently. This is

not yet implemented for microbiological diagnostics. However, within our healthcare region

this is for example already achieved for radiographic data.

Infection prevention stewardship

It is vital that infection control and prevention measures are integrated into this unified

program to improve overall infection management. Without the proper infection prevention

measures, other interventions such as ASPs and DSPs will not yield the optimal effect. Within

the AID stewardship model, infection prevention stewardship entails early detection and close

surveillance of MDROs, as well as an adequate rapid reaction to every possible transmission

(Figure 3.2). This task is dependent on easy access to rapid microbiological diagnostics for

both direct patient care and surveillance purposes. Therefore, our units of infection control

and prevention and medical microbiology (bacteriology, virology, mycology, and parasitology)

are organized within one single department for maximal collaboration and cooperation and

work closely together with the internal medicine department and the hospital pharmacy.

Together, they are responsible for e.g. hand hygiene guidelines and audits, isolation measures

and patient-specific consultations, and hospital-wide infrastructure (e.g. contribution to in-


hospital guidelines for infections and the antimicrobial formularium). The success story of the

containment of MRSA within the Netherlands by a search-and-destroy strategy is an example

of the substantial positive effect of close cooperation between clinical microbiology and

infection prevention specialists (Köck, et al., 2014). This applies to bacteria as well as viruses

where similar screening programs are also beneficial (Poelman, et al., 2015; Rahamat-

Langendoen, et al., 2013). Continuous communication with other ISP consultants within the

healthcare network and developing and updating unified guidelines at the regional level is

important (Ciccolini, et al., 2013; Müller, et al., 2015). In ISP, the use of eHealth technology,

developed in a multi-disciplinary environment by medical experts together with eHealth

experts, has considerable potential to facilitate these work processes (van Gemert-Pijnen, et al.,

2013; van Gemert-Pijnen, et al., 2011). As an example for this: it has been shown that web-

based guidelines are sub optimally used by nurses (Verhoeven, et al., 2008). With the aid of

user-centered and persuasive eHealth systems the adherence to guidelines improved and errors

in using them reduced significantly (Verhoeven Fenne F, et al., 2010-1). These systems were

developed as part of MRSA-net (http://www.mrsa-net.nl) and EurSafety Health-net

(www.eursafety.eu) in the Dutch-German Euregio, exemplifying the potential of eHealth for

ISPs (Jurke, et al., 2013; Wentzel, et al., 2014).

Diagnostic stewardship

To provide the optimal therapy for individual patients but also for infection control and

prevention purposes, it is essential that state-of-the-art diagnostics are performed timely,

before initiating antimicrobial therapy. Diagnostics must be appropriate for the individual

patient, target all pathogens causing acute infections, and detect colonization and/or infection.

Helping individual physicians in selecting and interpreting diagnostic tests on the appropriate

clinical specimens is the major goal of diagnostic stewardship. To be most effective, these

diagnostic tests should provide relevant clinical data as soon as possible, but for sure within the

first hours of admission (viral) or the first 24-48 hours of admission (bacterial and fungal).

Molecular diagnostics can largely meet these requirements. With new point-of-care (or point-

of-impact) assays (e.g. testing different biomarkers) becoming available, the turnaround time

for an increasing number of viral and bacterial/fungal pathogens can be reduced to less than 2

hours, supporting clinical decision-making. An important issue is furthermore supporting a

non-infectious differential diagnosis for certain conditions if appropriate diagnostics rapidly

yield negative results. Similar as the theragnostics approach in oncology, this should ultimately

lead to a more personalized infection management plan for the patient, whereby diagnostics,

therapy and infection prevention are integrated.

The use of innovative methods (e.g. next-generation sequencing) is an exciting evolving field

within clinical microbiology and infection control and is advocated by more as one of the

37 | Chapter 3

solutions for antimicrobial resistance (Cohen, et al., 2015). It fits within the theragnostic

approach to treatment and therefore, innovative point-of-impact technologies and real-time

sequencing tools are already put into use within the AID stewardship model and will be fully

implemented in standard daily care in the near future (Rahamat-Langendoen, et al., 2013;

Zhou, et al., 2015). We expected that also novel comprehensive diagnostic tools will provide

results much faster and more accurately, as already shown to be the case in combined

virological and bacteriological diagnostics (e.g. multiplex molecular tests) (Popowitch, et al.,

2013; Rogers, et al., 2015). This can provide the basis for better and more personalized therapy

for patients, contributing to an optimized use of antimicrobials, surveillance and infection

control, all leading to a more theragnostic approach of infection management. Of course, that

does not disqualify the use of conventional, culture-based diagnostics like conventional blood,

urine sputum and other cultures combined with susceptibility testing. Bacterial cultures for

example are still an effective and relatively cheap diagnostic tool for the diagnosis of

infections/colonization. In addition, they are the only means to assist in storage and typing of

specific isolates. However, the turnaround times of these tests often are too long to be useful

in the early stages of treatment. Especially in regions were diagnostic facilities are situated far

away from actual point of patient care, logistics can negatively impact the turnaround time.

Making these tools available in the near proximity of patient care will significantly reduce

turnaround times.

With the implementation of rapid point-of-impact technologies, subsequent rapid decision-

making will be beneficial for the optimal use of resources (for instance bed management and

isolation room capacity). These diagnostics assays and next-generation diagnostics are mostly

based on molecular technologies and are therefore more expensive compared to classical

culture-based methodology. They are however faster, delivering results within hours, thereby

enabling a theragnostic approach for infection management. A proper cost-effectiveness study

can provide the validation for the implementation of these tests. However, from a managerial

point of view and to support health economical decision making, the so-called “€hr concept”

can easily make turnaround times visible in relation to the overall costs (such as costs for

unnecessary isolation). By multiplying costs and turnaround time it provides a quick,

understandable figure, assuming that quality remains high and therefore equals one. This

provides the basis of the diagnostics needed for the AID stewardship (Figure 3.3).


Integrated stewardship: crossing multiple borders

In our view, infection management cannot be adequately performed by one single doctor, by

one medical specialty, by one hospital, or even by one country. Infection management in

healthcare networks is based on an interdisciplinary and interregional approach (Figure 3.1).

Movement of patients within and between healthcare institutions entails movement of

microorganisms within and to other institutes as well. For this reason, we have developed an

integrated AID stewardship model in close cooperation with healthcare institutions and

diagnostic laboratories within the region.

Due to the regional nature, this program is part of a larger Euregional collaboration between

the Netherlands and Germany, initially started on a relatively small scale with the MRSA-net

project. It has meanwhile been extended to the EurSafety Health-net project

(www.eursafety.eu) covering the complete Dutch-German border region (> 8 million

inhabitants). Over the years, this resulted in an intense collaboration of more than 115 acute

care hospitals, more than 200 long-term care facilities and a large array of healthcare

institutions, working together on infection management. Optimizing patient care and infection

control by building networks of stakeholders, knowledge transfer and consolidating the best of

both worlds has always been the primary focus. Within this cross-border network, the focus

lies on ASP, ISP and DSP. In this approach, eHealth contributes significantly to the different

goals in these projects. Among many other results, this has led to tools for clear and easy

presentation of MRSA protocols (Jurke, et al., 2013; Verhoeven, et al., 2008), but also

middleware solutions to improve molecular diagnostics like FlowG (http://www.flowg.nl). In

order to facilitate the implementation of the AID stewardship model, the EurSafety Health-net

project has led to various other applications that were developed and can be used on-site by

infection experts, healthcare providers and patients, implementing eHealth in daily cross-

border practice (Jurke, et al., 2013; van Gemert-Pijnen, et al., 2011; Wentzel, et al., 2014).

Figure 3.3: €hr concept within

the AID stewardship model. A

diagram showing a top-view of

Figure 1. Diagnostics should be

present to provide an integrative

and effective stewardship

program such as AID. From a

managerial point-of-view, the

effectiveness of these diagnostics

can be described with the €hr

concept, visualizing the most

important aspects: turnaround

time and costs, assuming that

quality remains high and therefore

equals one.

39 | Chapter 3

Conclusions and perspectives

In our view, managing antimicrobial infections can only be achieved by a holistic approach

based on the theragnostic principles as exemplified by the AID stewardship model, with the

implementation of timely (innovate) diagnostics, woven into the treatment management plan

and infection prevention. When combining Antibiotic Stewardship, Infection Prevention

Stewardship and Diagnostic Stewardship, one is able to target the complete healthcare

network, which can be supported by health economics, social sciences, technological sciences,

and by using eHealth technologies.

This technology is crucial to support multidisciplinary and cross-border infection management,

to share knowledge and to support healthcare workers and infection specialists with easy

accessible and user-friendly instructions for stewardship programs. First, to share knowledge

and thoughts about integrated AID stewardship in healthcare; second to provide systems to

support staff in implementing stewardship programs in daily work; third to guide the

implementation process related to local situations and global guidelines for safe work and

reduction of antimicrobial resistance rates. The University of Twente develops the ecosystem

for novel technologies that address these questions. In the EurSafety Health-net project several

technologies have already been developed to support the administration of antimicrobials

(Wentzel, et al., 2014), to improve surveillance of infections (Beerlage-de Jong, et al., 2014),

and to implement stewardship programs (van Limburg, et al., 2014). These technologies are

created in co-design with the end-users (e.g. nurses, physicians, and infection prevention

specialists). Key stakeholders in the cross-border network need to be involved and contribute

appropriately to the system to guarantee the model will be effectively implementable (van

Limburg, et al., 2011; Wentzel, et al., 2012).

Within this model, it is vital to work together and to utilize expertise of various stakeholders

and organizations to intervene at various time points in infection management. The three

different stewardship aspects are highly intertwined and must be implemented adequately.

Especially in regions with higher resistance rates, this integration is even more important in

order to control the situation and work towards an improvement. The AID stewardship model

is in that respect generally applicable because the main focus point is the integration of the

three stewardship programs. The specific actions performed within these programs are

depending on the setting of the healthcare institution. Implementing a monovalent

antimicrobial stewardship program, as advocated in several countries, is of limited use without

the implementation of appropriate and more sophisticated diagnostics close patient care and

vice versa. Since hospitals are part of comprehensive healthcare networks, it is not helpful if

one hospital has implemented a complete stewardship program, but the hospital next-door has

not made proper, similar arrangements.


Therefore, we encourage stakeholders to cross borders and organize patient-centered infection

management in a theragnostic, multifaceted, multidisciplinary, interregional approach to

counteract antimicrobial resistance problems. In this manner smaller institutions without

extensive resources can benefit from academic centers and these centers in turn benefit with

the referral of patients. The AID stewardship model that we are enrolling is an example of

such an approach, which could serve as a blueprint in the field of infection management

worldwide.

Future perspective

An increasing amount of microbiology data is becoming available for clinicians to be used

when treating patients, including genomic data and information on the potential pathogen.

Furthermore, it will get easier to link already existing data from different databases to provide a

real-time up-to-date overview of the patient and his treatment (such as MIC values and

pharmacodynamics and pharmakokinetics [PK/PD] information), that clinician can use.

Clinical decision support systems are already becoming available that can help the clinician in

interpreting this increasing amount of data. These programs are expected to become smarter

and more comprehensive in the near future. All this should hopefully lead to a situation where

personalized treatment can be started right away or at least within the first hours after a patient

enters the hospital with a (suspected) infectious problem. This will be facilitated by the use of

novel diagnostic technologies (e.g. third / fourth generation sequencing and spectroscopy) and

smart decision support systems,. Antimicrobial Stewardship Programs can utilize this data to

ensure optimal treatment at all times, acting upon –PK/PD data in real-time. This should limit

the amount of inadequately prescribed antimicrobials, thereby minimizing resistance pressure

and collateral damage to patients. It would optimize the complete infection management

process, thus providing better and more (cost-)efficient patient care. Furthermore, this data can

facilitate optimized infection prevention within the healthcare region, by providing real-time

patient and pathogen characteristics to all involved healthcare institutions. It is vital that

stakeholders cross borders and start collaborating as soon as possible, in order to provide an

integrative infection management such as the proposed integrative stewardship model. By an

integrative stewardship approach, such our proposed AID model, healthcare institutions will

immediately be streamlining the infection management process. This will form a foundation of

stakeholder collaborations that are necessary to utilize the upcoming flow of data and integrate

these data in day-to-day patient care.

42

Measuring the impact of

antimicrobial stewardship programs

Jan-Willem H. Dik, Ron Hendrix, Randy Poelman, Hubert G Niesters, Maarten J. Postma

Bhanu Sinha, Alex W Friedrich

Expert Reviews of Anti-Infective Therapy 2016; 14(6):569-575

4

43 | Chapter 4

Abstract

Antimicrobial Stewardship Programs (ASPs) are being implemented worldwide

to optimize antimicrobial therapy, and thereby improve patient safety and

quality of care. Additionally, this should counteract resistance development. It

is, however, vital that correct and timely diagnostics are performed in parallel,

and that an institution runs a well-organized infection prevention program.

Currently, there is no clear consensus on which interventions an ASP should

comprise. Indeed this depends on the institution, the region, and the patient

population that is served. Different interventions will lead to different effects.

Therefore, adequate evaluations, both clinically and financially, are crucial.

Here, we provide a general overview of, and perspective on different

intervention strategies and methods to evaluate these ASP programs, covering

before mentioned topics.

This should lead to a more consistent approach in evaluating these programs,

making it easier to compare different interventions and studies with each other

and ultimately improve infection and patient management.

Measuring the impact of antimicrobial stewardship programs | 44

Introduction

Antimicrobial resistance is a growing problem. One of the major drivers of this disconcerting

development is inadequate use of antimicrobials, both in healthcare centers and out-patient

settings (Goossens, 2009), as well as in livestock (Marshall and Levy, 2011). The so-called One

Health approach targets resistance development on all the before-mentioned levels. Such a

broad approach is considered crucial, in order to effectively minimize the worldwide healthcare

antimicrobial resistance threat (Renwick, et al., 2016). Part of this approach is improving the

usage of antimicrobials in healthcare centers and out-patient settings, which in turn helps

reducing resistance development (World Health Organization, 2012). Antimicrobial

Stewardship Programs (ASPs) are being hailed as a solution to improve antimicrobial therapies

and thus results in a better patient outcome and safety. Different national and international

guidelines are available for hospitals, long-term care facilities and general practitioners (de

With, et al., 2016; Dellit, et al., 2007; SWAB, 2012). There is, however, no clear consensus on

the impact of different interventions (Davey, et al., 2013; Schuts, et al., 2016). Effects (clinical

and financial) in specific settings or patient populations are difficult to compare and/or

evaluate. Some interventions might even be redundant or counterproductive, although in

general, published results are often favorable (Davey, et al., 2013; Schuts, et al., 2016). This

inconclusiveness, necessitates performing scientifically sound (cost-)effectiveness studies on

ASP interventions (McGowan, 2012). There are multitudes of methods to evaluate several

interventions, but in general, they lack uniformity (Evans, et al., 2015; McGowan, 2012;

Morris, 2014). In this review, we will discuss stewardship in general, its (pre)requisites and the

main interventions and their approaches for evaluation, thereby giving a general and up-to-date

overview.

Importance of a broad stewardship program

The term “antimicrobial stewardship” has been coined roughly 20 years ago (McGowan and

Gerding, 1996). Stewardship programs are now being implemented worldwide and hundreds

of articles are published yearly (Howard, et al., 2015). As it became clear that inadequate

antimicrobial use (prophylactic and therapeutic) contributes to resistance development,

improving antimicrobial usage became a focus for many healthcare institutions, using a subset

of different interventions (Davey, et al., 2013; Dellit, et al., 2007). However, it is often be

overlooked that reducing resistance rates should not be the primary goal. The ultimate goal is

to with improve clinical outcome and patient safety by providing optimal patient care. Patients

should be the main focus and they have important questions related to infectious problems: i)

How can I be protected from a (resistant) infection? ii) Do I have an infection, and if yes, what

is causing it? iii) What is the optimal treatment to cure it? Answering these questions requires a

broader approach than just an ASP and also entails that certain requirements are met. Besides

implementing an ASP, an Infection prevention Stewardship Program (ISP) should be present

to ensure that other patients do not get infected by pertinent (resistant) pathogens, which are

45 | Chapter 4

often easy transmissible. Furthermore, optimal and timely diagnostics that can adequately and

rapidly diagnose the patient’s problems are vital (Diagnostic Stewardship Program [DSP]).

Only if all three aspects are covered - optimal treatment, prevention and diagnostics (an

integrated, Antimicrobial, Infection prevention & Diagnostic [AID] stewardship program) -

and all involved stakeholders have the necessary meta-competence (meaning a broad

understanding of all relevant above mentioned aspects), healthcare centers can optimally treat

infectious patients and tackle the development of antimicrobial resistance (Dik, et al., 2016b;

Lammie and Hughes, 2016). Because patient transfers between institutions are also pathogen

transfers (Donker, et al., 2010), these three aspects should not only be covered within one local

center, but in a (regional) healthcare network. This entails close collaboration of all healthcare

facilities (i.e. hospitals, but also general practices and long-term care facilities) within a clearly

defined region (Ciccolini, et al., 2013). Harmonization of guidelines and practices can be a first

start regarding this aspect (Müller, et al., 2015). In the near future, such an integrated AID-

approach should lead to a more personalized treatment plan, which is optimally adapted to the

specifics of each single patient.

Importance of diagnostics

It is thus important that adequate diagnostics are performed on time, and provide rapid results

to have impact on patient care (Caliendo, et al., 2013). Ideally, results, including resistance

patterns, should be available before the patient is started on antimicrobial treatment. Three

parameters influence the value of diagnostic tests: quality, cost and time. Overall, the sensitivity

and specificity of new commercial and often multiplex-based molecular, Point-Of-Care (POC)

assays approach the quality of Laboratory Developed Tests (LDTs). In this situation, lower

costs and/or shorter turnaround times become the main drivers for increased value. From a

managerial point of view, we introduced the euro-hour (€hr) concept, comparable with

kilowatt hour to easily visualize the impact of both parameters (Dik, et al., 2016b). In this

concept, the costs of a test are multiplied by the turnaround time and therefore represent the

impact of implementing a POC test. POC tests can only have an impact on antimicrobial

therapy and patient management if results are timely available, interpreted and followed-up by

a medical specialist (Buehler, et al., 2016; Rogers, et al., 2015). When implemented, it increases

the probability of a correct (preliminary) diagnosis – including the reduced need for further

diagnostics, streamline antimicrobial treatment sooner if needed (thereby also minimizing the

risk for toxicity), and improve infection prevention measurements. In the field of oncology this

so-called theragnostics approach is under continuous development during recent years and it

would be a powerful tool in personalized infection management as well (Dik, et al., 2016b;

Lammie and Hughes, 2016). Examples of POC tests or rapid diagnostics (e.g. multiplex PCR

and MALDI-TOF MS) already implemented for ASPs are: MRSA screening and testing

(Mather, et al., 2016; Tacconelli, et al., 2009b; Wassenberg, et al., 2012); resistance screening

(Evans, et al., 2016); the use in septic/bacteremic patients (Banerjee, et al., 2015; Bauer, et al.,

2010a; Clerc, et al., 2014); use of biomarkers (of which procalcitonin probably shows the most


promising results) (Bouadma, et al., 2010; de Jong, et al., 2016; Nobre, et al., 2008; Schuetz, et

al., 2012); and with viral infections, such as for respiratory illness (Müller, et al., 2015;

Popowitch, et al., 2013).

Basics of ASPs

Often ASP interventions are subdivided into three groups: the front-end approach, the back-

end approach and supplemental interventions. Front-end interventions focus on the start of

empirical therapy such as pre-analytic consultations and guidelines for (empiric) therapy. Back-

end interventions focus on optimization of therapy after e.g. two or three days . For example,

an intravenous to oral switch promotion; de-escalation; or timely stop of therapy, as

appropriate. Finally, there are interventions that supplement an ASP such as using resistance

data to keep local guidelines up to date and the availability of educational programs (Dellit, et

al., 2007). The evaluation of the program is also an important aspect of the latter group. Such a

different array of interventions implies that the timeline of impact of an ASP varies broad, with

effects on the short, middle and long term (see Figure 4.1 for a schematic overview).

An ASP should focus on improving patient care and safety, by increasing

appropriateness of all antimicrobial use (i.e. prophylactics, empiric therapy and directed

therapy). When looking at prophylactic antimicrobials, there are special cases like Selective

Digestive (or Intestinal) Decontamination (SDD) and Selective Oropharyngeal

Decontamination (SOD). SDD and SOD are generally implemented at ICUs, and thus are

often not part of a standard ASP. They are, however, important forms of antimicrobial use that

can influence resistance development, and will impact overall policy of antimicrobial usage. It

is thus imperative to take these interventions into account when implementing (and evaluating)

antimicrobial use and/or interventions strategies (Daneman, et al., 2013; Plantinga and Bonten,

2015).

Unresolved issues with evaluating ASPs

The recent systematic Cochrane review on interventions to improve antimicrobial therapy

systematically looked at all studies describing one or more interventions, and evaluated their

quality and strength of evidence (Davey, et al., 2013). This review mainly focuses on the clinical

effects. One of the main conclusions that can be drawn from the review is the lack of quality

of evaluations reported. This is exemplified by the fact that the majority of the studies could

not be included (Davey, et al., 2013). A recent highly comprehensive systematic review and

meta-analysis of 14 different antimicrobial stewardship objectives found similar results, albeit

generally with a low quality of evidence (Schuts, et al., 2016).

47 | Chapter 4

Figure 4.1: Schematic overview of expected timeline of impact of a subset of different ASP

aspects.

Financial effects are equally important. In 2015 two reviews on financial evaluations of ASP

studies were published. Both conclude that ASPs are evaluated inconsistently and often even

poorly, making it almost impossible to compare studies with another (Coulter, et al., 2015; Dik,

et al., 2015c). Keeping these results in mind, we will provide a general overview of the different

methods to evaluate ASP interventions both clinically and financially (leaving out structural

and process-focused aspects). Furthermore, we will mention some pros and cons for each

method.

Different methods of evaluating ASPs

Randomized controlled trials (RCTs) are considered as the gold standard and most preferable

type of study. However, they are often less suitable for antimicrobial intervention studies, due

to logistics, ethics and costs. Nevertheless, in recent years there were a couple examples

looking at ASP interventions in a randomized controlled manner (Banerjee, et al., 2015; Fleet,

et al., 2014; Lesprit, et al., 2013). The large majority of published evaluations are however

observational studies (e.g. case-control studies, interrupted time series analyses [ITS], etc.)

(Davey, et al., 2013; Howard, et al., 2015; Schuts, et al., 2016). For these studies, comparable

cohorts of patients are a major source for bias. This can be even more influenced by changes

over time, because the control period is usually several years prior to the intervention period.

With regard to economic evaluations, the preferred method would be to do a cost-

effectiveness/-utility study from a societal perspective (Drummond, et al., 2005). Generally

speaking, the level of expertise and the time required to do such an analysis are often too

scarce to be practically accessible. In practice, this has led to the fact that economic evaluations

performed on ASPs are often cost-minimization analyses (Dik, et al., 2015c). Finally, of

relevance is the fact that ASPs are implemented on specific wards and/or for specific patient


groups (e.g. ICUs, long-term care facilities, septic patients, pediatric patients, and MRSA

infections). This makes comparability difficult and it is therefore essential to mention in detail

the patient characteristics, as well as the setting of implementation.

Clinical outcome measures

The most important goal for an ASP should be to improve quality of patient care. A number

of different measures are used that describe some aspects of clinical outcomes. Most important

are mortality rates. In general, most studies that evaluate mortality conclude it is not

compromised, and that an ASP is thus a safe intervention (a non-inferiority analysis).

Especially in ASPs targeted at the more severe patient groups (e.g. septic patients), mortality

can be an important outcome. For less severe infections (e.g. urinary tract infections), the use

of mortality as an outcome measure might be less informative. Length of stay (LOS),

(secondary) infection rates, and re-admission rates are also often measured and evaluated

(Davey, et al., 2013; Schuts, et al., 2016). Other less frequently studied outcomes are toxicity

and possible side-effects (e.g. IV catheter-related problems or phlebitis) (see Table 4.1). LOS is

one of the more accessible variables to obtain. It is, however, important to take possible

secular trends towards earlier discharge into account when evaluating an ASP in a case-cohort

setting, especially if the time-period spans multiple years (e.g. did the institution in general saw

a drop in LOS over time). Besides the overall hospital LOS, ward specific LOS (such as ICU

stay) is an option. The latter is of course most interesting if the program is also ward

specifically implemented. If treatment improves and infections are cured more effectively, the

re-lapse rates may decrease and re-admission rates consequently will go down. However, this

outcome measure is biased if there are other hospitals in the vicinity where patients might be

readmitted, for example. in clusters of academic centers and surrounding general hospitals. As

a more indirect effect, the infection rate for Clostridium difficile can be taken as an outcome

measure (see e.g. Nathwani et al, for a successful program (Nathwani, et al., 2012)). In some

studies, a direct correlation of C. difficile infection with antimicrobial use is suspected, especially

regarding cephalosporins and clindamycin (Slimings and Riley, 2014). This rate might

consequently be used as an indirect indicator for antimicrobial use and therefore as an ASP

outcome measure.

Microbiological outcome measures

Besides the important clinical outcomes (that directly impact patient care and safety), a

secondary important goal is the reduction of resistance levels (see Table 4.1). There are

multiple ways to evaluate this goal. Resistance levels can be measured as percentage of patients

(or cultures) with microorganisms “resistant” for a certain antibiotic compared to the number

with “susceptible” microorganisms. This parameter can be measured in infected patients or

colonized patients. Furthermore, the rate of infections with a resistant microorganism can be

49 | Chapter 4

taken as a measure (preferably as a percentage of patients infected with the susceptible variant).

Difficulties arise due to the longer time that is required before a change in resistance levels can

be observed. In addition, reliability of certain trends in the data is difficult to estimate when

looking at small numbers. The long time-frame also implies that the influence of possible

confounders becomes greater. These slow, subtle changes make that antibiograms have been

shown to be inconclusive as separate outcome measures and the application of an interrupted

time series analysis is therefore a better and preferred method for resistance measures (Schulz,

et al., 2012). Furthermore, the baseline level of resistance is also of importance: countries with

high resistance levels will most likely see larger effects in a shorter time, provided there is no

major influx of resistant microorganisms. A final complicating factor is that “resistant” bacteria

reside in the community and in neighboring healthcare centers. If an ASP is not implemented

regionally, positive results might not be achievable, namely at referral centers. In this setting a

majority of the patients carrying “resistant” bacteria, will come from the surrounding

healthcare region (Donker, et al., 2010; Donker, et al., 2015). The quality of evidence of ASP

effects on resistance is still low (Davey, et al., 2013; Tacconelli, et al., 2016). To improve

studies specifically looking at correlating antimicrobial use and resistance development, a

Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) tool was

developed (STROBE-AMS) (Tacconelli, et al., 2016). Concerning SDD/SOD, there is still

discussion regarding possible resistance development and it is therefore highly advisable to

monitor possible resistance development when implementing such an intervention (Health

Council of the Netherlands, 2015).

Antimicrobial consumption outcome measures

ASPs mainly focus on accomplishing changes in broad-spectrum IV prescriptions, because

broad spectrum antimicrobials are more likely to promote resistance development and IV

treatment is more likely to cause secondary infections/complications. The programs focus on

either support of the prescribing physician at the start during decision making regarding

therapy, or after a few days to support the evaluation of diagnostics and subsequent possible

adjustments of therapy. An IV to oral switch program is one of the most frequently

implemented interventions (Nathwani, et al., 2015). To measure the effect on therapy different

outcome measures can be used (see Table 4.1). Often it is chosen to quantify the antimicrobial

therapy as Defined Daily Doses (DDDs, as defined by the WHO), with a denominator

correcting for clinical activity such as bed days or admissions

(http://www.whocc.no/atc_ddd_index). This can be done either by looking at dispensing data

or at purchasing data, which are strongly correlated with each other (Tan, et al., 2016). From a

national point of view, a broader denominator such as inhabitants of the healthcare region is

also valuable. Because, as mentioned before, people are transferred through different

healthcare centers within a region, transferring bacteria with them as well. IDSA/SHEA

advocates DDDs per 1000 patient days as universal outcome measure for ASP programs

(Dellit, et al., 2007). This is, however, not always suitable as it is a highly generalizing method


that does not take into account patient specifics (such as complicated infections which require

high dose therapy as for e.g. endocarditis), and is known to overestimate the use (de With, et

al., 2009). With respect to pediatric populations, these outcomes are far from optimal, because

they are based on adult dosages. This should be taken into account, although up to now it is

unclear what measure should preferably be used instead of DDDs (Fortin, et al., 2014). It

should thus be noted that a change in DDDs is not entirely suitable for drawing conclusions

on the success of an ASP. Optimal antibiotic therapy can also mean that undertreated patients

should receive more antibiotics (for example deep-seated infections or overweight patients).

Table 4.1: Overview of different outcome measures and some general remarks. Outcome measures Remarks

Clinical Mortality Important, but less suitable for mild infection (e.g.

uncomplicated UTI) LOS General or ward-specific (e.g. ICU stay); easy to obtain, but

highly sensitive to biases Complications Complications such as IV catheter-related problems and

phlebitis Clostridium difficile Also indirect measure for antimicrobial use Re-admission rates Due to relapse, but important to consider effect of

neighbouring institutions Other complications For example, IV catheter-related thrombosis and non-

infective phlebitis

Microbiological Resistance levels Difficult to measure due to generally long time frame

(months to years)

Antimicrobial consumption Total use Often measured in DDDs however, discrepancies are

known due to generalization IV/PO ratio Of interest with an active IV-to-PO switch program Broad/narrow ratio Potentially relevant with regard to resistance development Appropriateness of therapy Labour intensive and possibly subjective, but of importance

Financial Cost-benefit ratio Preferably done as cost-effectiveness study, including all

costs and benefits (at least at hospital level, but preferably at societal level)

UTI: urinary tract infection; LOS: length of stay; ICU: intensive care unit; IV; intravenous; DDD: daily defined dosage; PO; per os

51 | Chapter 4

Personalized therapy measures such as Prescribed Daily Doses (PDDs) or Recommended

Daily Doses (RDD) are a more patient specific approach to quantify antimicrobial treatment

and might give more suitable results (de With, et al., 2009; Gagliotti, et al., 2014). Furthermore,

the length of the therapy (in days) can be evaluated (duration of therapy, DOT). A discrepancy

when compared to DDDs is known, due to the difference between administered dose and the

WHO DDD values (Polk, et al., 2007). Because an ASP focuses on optimizing therapy, often

by promoting narrow spectrum oral medication, it can be worthwhile evaluating effects of

these interventions specifically. This can be done by looking at the percentage of IV

medication versus oral (in DDDs/PDDs/RDDs or DOTs) and/or the percentage broad

versus narrow spectrum antibiotics (in DDDs/PDDs/RDDs or DOTs). If it is expected that

improvements on a ward will be more systemic of nature, it is also worthwhile to evaluate the

percentage of patients receiving antimicrobial therapy. Finally, appropriateness of therapy can

be evaluated. This is a more labor intensive method, while it often requires reviewing single

patient’s files, making it also less objective than ‘hard numbers’ such as DDDs (DePestel, et al.,

2014). However, because the goal of an ASP is optimal therapy (according to protocol),

appropriateness of therapy is an outcome measure that directly evaluates the main goal of the

intervention. An example of this, is the analysis with regard to the appropriateness therapy of

urinary tract infections (Stewardson, et al., 2014).

Financial outcome measures

With regard to financial evaluations of an ASP, there is much room for improvement (Coulter,

et al., 2015; Davey, et al., 2013; Dik, et al., 2015c). Most notable issue is that not all costs are

taken into account, but just a subset of costs and benefits chosen based on data availability,

potentially leading to other cost-effectiveness results. Obviously, for correct interpretation of

costs and benefits of a stewardship program, it is important to take into account all costs (and

benefits) besides the obvious ones (e.g. antimicrobial costs) (Drummond, et al., 2005).

Preferable, this collection of costs should be done prospectively with up front agreed-upon

variables and parameters. This minimizes the chances that certain cost types are neglected or

cannot be informed. Often various types appear not to be collected when an evaluation is

performed retrospectively. Although highly desirable, it is not always necessary or feasible to

be cost saving. It is however important, to know if the intervention is the most cost-effective

way to reach the preferred outcome(s), compared to other potential interventions or the

baseline situation. If indeed it is not cost saving, it is worthwhile to take into account a certain

threshold of maximum cost per outcome (i.e. cost or willingness to pay per quality-adjusted life

year [QALY], life-year saved, or other chosen outcome) to enhance optimal allocations of

budgets.

An obvious start for integrative costing is to consider all costs that had to be made to

implement the program or intervention. This definitely includes time spent by the staff

involved, both specifically hired and those already working in the institution. In the latter case


this formally concerns so-called opportunity costs. Furthermore, the required infrastructure

(for example costs for the introduction of an IT program) and consumables (for example extra

or new diagnostics) should be considered. In short, all resources and costs of running the

program should be included, consistently measured by opportunity costing which reflects the

alternative next-best application of these resources and costs (applying to the people involved

in the intervention, but also maintenance contracts for IT programs or depreciation costs of

laboratory equipment).

If implementation and daily execution costs are known, one can relate these possible

savings or benefits, and draw conclusions on the cost-efficiency or cost-effectiveness.

Preferably, all outcome measures that were evaluated clinically are quantified and transformed

into monetary and/or utility values. In general, this will include: LOS, antimicrobial use, other

procedures done to treat patients (including nursing time), changes in re-admissions, infections

and other complications. Quantification into monetary values can be open to interpretation

and, for example for LOS high variations in willingness to pay were shown (Stewardson, et al.,

2014), meaning that proper justification is important, inclusive inspection of guidelines for

pharmacoeconomic research (http://www.ispor.org).

For ASPs, costs are usually calculated from a hospital perspective regarding monetary

outcomes and related to survival and quality-of-life as humanities’ outcomes. When taking a

societal perspective other outcomes should also be included, for example costs due to reduced

labor productivity (Dik, et al., 2015c). Unfortunately, ASP evaluations that include financial

outcomes often only include direct antimicrobial costs within a very limited perspective,

making it nearly impossible to draw conclusions from current literature on full cost-

effectiveness of ASPs and hampering comparison with cost-effectiveness of other

interventions in healthcare that are often done from this societal perspective (for example,

drugs) (Coulter, et al., 2015; Davey, et al., 2013; Dik, et al., 2015c).

Conclusions and recommendations

Antimicrobial stewardship programs are an important topic with hundreds of publications

appearing yearly. In the last few years, asides numerous studies focusing on antibiotics, many

studies were published on antifungals with a comparable set-up as antibiotic stewardship

studies and thus also comparable quality issues (Brüggemann and Aarnoutse, 2015; Muñoz, et

al., 2015; Valerio, et al., 2015). Because ASPs are consisting of multiple interventions and not

every healthcare center is implementing the same interventions, outcome evaluations are also

highly diverse. In this respect, a maturity model can for example help to establish the current

status of an ASP (van Limburg, et al., 2014). This complexity is further increased by the

method of evaluation (e.g. RCT, ITS or case-control study), and different outcome measures

used. Finally there are multiple challenges to obtain high quality data on effectiveness of an

ASP, for example a lack of data of presumptive diagnosis at time of prescribing (or not

53 | Chapter 4

prescribing) antimicrobials (and subsequent evaluation of appropriateness), as well as exact

timing of diagnostics versus start of therapy. Comparing and interpreting different ASP studies

is therefore extremely challenging. Cleary, there is a need for appropriate and well-standardized

definitions of interventions of an ASP, of the preferred method of evaluation, and of the

preferred outcome measures, inclusive those from the financial-economic perspective. Until

then, authors should clearly explain and discuss their methods of evaluation in order to make

the field of ASPs more transparent.

Five year view

Within the foreseeable future, more tools will be available within daily practice to guide

antimicrobial therapy in the best possible way. Faster diagnostics, genomic data, and smarter

clinical decision support systems are some of these examples, as well as the growing

importance of regional healthcare networks and integrative, interdisciplinary collaboration

between specialists. This entails that ASPs will continue to develop and that interventions are

expected to become easier to implement. Such developments also impact the evaluations of

ASPs. It is therefore even more important that ASP evaluations will be performed in a

transparent and comparable manner to help streamline the development process.

56

Financial evaluations of antibiotic stewardship programs –

a systematic review

Jan-Willem H. Dik, Pepijn Vemer, Alex W. Friedrich, Ron Hendrix, Jerome R. Lo-Ten-Foe,

Bhanu Sinha, Maarten J. Postma

Frontiers Microbiology 2015; 6:317

5

57 | Chapter 5

Abstract

There is an increasing awareness to counteract problems due to incorrect

antimicrobial use. Interventions that are implemented are often part of an

Antimicrobial Stewardship Program (ASPs). Studies publishing results from

these interventions are increasing, including reports on the economical effects

of ASPs.

This review will look at the economical sections of these studies and the

methods that were used. A systematic review was performed of articles found in

the PubMed and EMBASE databases published from 2000 until November

2014. Included studies found were scored for various aspects and the quality of

the papers was assessed following an appropriate check list (CHEC criteria

list).1233 studies were found, of which 149 were read completely. 99 were

included in the final review.

Of these studies, 57 only mentioned the costs associated with the antimicrobial

medication. Others also included operational costs (n=23), costs for hospital

stay (n=18) and/or other costs (n=19). 9 studies were further assessed for their

quality. These studies scored between 2 and 14 out of a potential total score of

19.This review gives an extensive overview of the current financial evaluation of

ASPs and the quality of these economical studies.

We show that there is still major potential to improve financial evaluations of

ASPs. Studies do not use similar nor consistent methods or outcome measures,

making it impossible draw sound conclusions and compare different studies.

Finally, we make some recommendations for the future.

Financial evaluations of antibiotic stewardship programs – a systematic review | 58

Introduction

The therapeutic use of antimicrobials in clinical medicine is continuing to be suboptimal. Both

overtreatment, with regard to spectrum and duration, and suboptimal treatment, with regard to

dosage and most effective therapy, are areas of concern. Either one can lead to an increase in

the resistance of bacteria (Goossens, 2009), and unnecessary side effects, including a

potentially large economic burden (Gandra, et al., 2014). It is thus imperative that action is

taken (Carlet, et al., 2011; World Health Organization, 2012). Fortunately, many hospitals are

aware of this, and act accordingly by implementing Antimicrobial Stewardship Programs

(ASPs) in their institutions. These programs differ in approach, but the consensus is that when

implemented correctly, considerable positive (clinical) effects can be attained by optimizing

patients’ antimicrobial therapy (Davey, et al., 2013). Notably, problems such as increased

antimicrobial resistance, consequent treatment failure, and the spread of nosocomial infections

can be prevented with ASPs, inclusive its financial consequences (Roberts, et al., 2009).

Implementing an ASP can thus also have a considerable positive financial impact within a

hospital, which is crucial in times where healthcare costs are rising.

There are various guidelines published describing to design an ASP (Dellit, et al.,

2007; SWAB, 2012; With de, et al., 2013), consisting of a set of interventions and services, all

with the goal to stimulate the correct use of antimicrobials, but intervening at different

moments in the chain of care. One or more ASPs can be implemented, depending on the type

of hospital, the types of patients and the local challenges that physicians face. In general, all

tasks within an ASP can be categorized within three blocks:

• The first block consists of tasks that can be performed during the start of empirical

antimicrobial therapy (the so-called “front-end” approach). Examples are pre-analytic

consultations and providing therapy guidelines and education for prescribing doctors.

• A second block consists of tasks to assist by the optimization of the therapy around

day 2 to 3, such as interventions to promote IV to oral switch, de-escalation and a timely stop

when appropriate (the “back-end” approach).

• Finally, a last block of supplemental tasks should assure evaluation of hospital data

and tasks that act upon those data accordingly, like updating guidelines using local resistance

rates and processes, as well as promoting surveillance studies (Dellit, et al., 2007; SWAB, 2012;

With de, et al., 2013).

Whether the intervention is restrictive or persuasive does not seem to differ on the long term –

i.e. after 6 months of implementation (Davey, et al., 2013).

During the last years, there is a steady rise in papers published on above mentioned

interventions. Coinciding with that rise, the number of economic evaluations of an ASP is also

increasing. This reflects the growing importance of health economic evaluations in general,

which is seen across the world, as well as that of ASPs in particular (Hjelmgren, et al., 2001).

59 | Chapter 5

However, there seems to be a large variation in the way ASPs are financially evaluated. Often

only the direct costs for antimicrobials is taken into account, whereas other (in)direct costs may

have a much larger impact (Davey, et al., 2013; McGowan, 2012). Many papers mention some

costs and/or benefits, but only a few give usable, in-depth data, analyses and conclusions.

When comparing financial ASP results, for example, within the latest Cochrane review, the

conclusion was that economic evaluations are done in a ‘disappointingly’ low number of

studies and often lack reliable data (Davey, et al., 2013).

This study provides a systematic methodological review of published economic

evaluations of ASP studies (intervention studies with an economic evaluation paragraph or

complete economic evaluations). Ideally, it will shed light on the divergence that is present in

these studies and where improvement is necessary. Keeping in mind that decision makers use

these economic evaluations in their daily practice nowadays, and costs are considered a barrier

to implement an ASP, correct and valuable studies are becoming more important (Johannsson,

et al., 2011). This review will therefore in particular look at the methods’ usability for others

that might want to implement an ASP in their hospital.

Material and Methods

A search was performed within the PubMed and EMBASE databases in November 2014,

using the following search strings: “antimicrobial stewardship”, “antimicrobial management”,

“antimicrobial prescribing intervention” and “antimicrobial program intervention”. All strings

were in combination with the words “cost(s)”, “financial”, “economic”, “dollar” or “euro”, or

the respective symbols for the latter two. All abstracts found were read and original studies

written in English, Dutch or German that discussed an intervention the related economic

analysis of that intervention within a hospital were included. Considering the fast

developments within the field of antimicrobial stewardship as well as within health economics,

studies before 2000 were excluded. Outpatient settings were excluded (see Figure 5.1 for the

complete flow chart).

The final set of included papers was read completely, and for each study the type of economic

evaluation done was scored. Keeping in mind that many studies are clinical effect studies and

not economic analyses studies, the papers that only mentioned direct antimicrobial costs

without mentioning any of the other relevant costs or savings, were categorized as an

Antimicrobial Cost Analysis (ACA). The remainder of the studies contained enough essential

financial parameters to be classified as different economic analyses (e.g. as described in

(Drummond, et al., 2005)). For studies looking at the effects of two methods and that

converted those effects into monetary values the classification Cost-Benefit Analyses (CBA)

was used. Those that evaluated the relative costs and effects of two different methods were


classified as Cost-Effectiveness Analyses (CEA). Studies observing the different costs and

effects of two methods were classified as Cost-Consequence Analyses (CCA). Studies that

looked at the costs but not at the effects were scored as Cost-Analyses (CA) and studies that

looked at the differences in costs assuming similar effects between the two evaluated methods

were scored Cost-Minimization Analyses (CMA). For the papers the following parameters

were scored: year, journal, country of research, study design, setting, number of participants,

outcome measures, price adjustment and/or discounting measures, and conclusions. The

types of intervention per study were scored and categorized.

Within an economic analysis, different costs can be taken into consideration. Drummond et al

recommend calculating operational costs and capital costs that are related to the intervention

(Drummond, et al., 2005). To further specify this for an ASP intervention study, the outcome

measures from the Cochrane review were taken as specific parameters (Davey, et al., 2013).

Depending on the perspective chosen, the following parameters were scored: implementation

costs, operational costs (personnel and/or equipment costs of the intervention), antimicrobial

costs, hospital day costs, morbidity and/or mortality costs (costs associated with hospital

procedures, treatment etc.), societal costs (costs occurring outside the hospital from a societal

perspective, e.g. loss of productivity) and other costs (costs mentioned by a study that are

different than already mentioned here).

In order to assess the level of quality of the included papers, an appropriate quality

criteria list (Consensus on Health Economics Criteria [CHEC] list) was used (Evers, et al.,

2005). A criteria list specifically intended to give insight in the quality of economic evaluations.

Considering the fact that many studies lacked a proper economic analysis it was not deemed of

great interest or informative doing an in depth quality assessment on all papers of which the

majority would not meet minimum criteria. We therefore decided to only formally assess the

quality in detail if the paper met minimum standards. Two parameters were assumed as most

basic and essential for an economic evaluation studies and used as an inclusion criterion:

implementation costs and/or operational costs and appropriate valuation of all costs. Articles

that included these parameters were consequently scored following the CHEC list by two

authors independently.

This review study followed the PRISMA criteria where possible (Moher, et al., 2009).

The complete checklist can be found as supplemental data (Supplemental Table S5.1).

Results

For this review, a total of 1233 papers were found using our search strings. Of these papers,

1083 were excluded based upon the fact that they did not meet our inclusion criteria (for

example, because they were review papers, there were no cost outcome measures mentioned or

61 | Chapter 5

they were published in a non-included language. 149 papers were included and read

completely. Of these papers, a further 50 were excluded for various reasons. A set of 99 papers

was included in the final analysis (Figure 5.1).

Figure 5.1: Flow chart of the search method. The followed search method as

performed on 4 November 2014.


Baseline characteristics

From the total of 99 papers, the majority came from the United States followed by Europe and

most included studies were published within the last 3 years. Studies were performed in

hospitals ranging from as little as 39 beds to as much as 1800 beds and included between 50

and 40,000 patients. For a complete overview see Table 5.1.

Types of interventions

Categorizing the stewardship interventions in block 1 (front-end approach), block 2 (back-end

approach) and block 3 (supplemental measures), most studies implemented one or more

interventions from block 2 (Table 5.2). Particularly, the implementation of an audit of and/or

feedback on the therapy provided at certain time-point(s) was performed frequently (62

studies; 52% of all interventions). Second most frequent was the creation and implementation

of antimicrobial therapy guidelines (16 studies, 13% of all interventions). 18 studies (18% of all

studies) implemented more than one intervention at the same time, providing a “bundle” of

services.

Types of analyses

57 (58%) papers only included antimicrobial costs as an economic outcome measure without

any of the other relevant costs. Although these studies looked at costs and effects, they could

not be classified as a proper economic analysis, because an appropriate economic evaluation

was not done. They were therefore classified as an ACA. Of the rest, the majority, 32 were

scored as a CBA. 3 studies evaluated the relative costs and effects classifying them as a CEA.

There were 3 CCAs, 2 studies only looked at the costs making it a CA and 2 studies were

CMAs (Table 5.3).

Cost outcome measures

For every study, the cost outcome measures they used were scored. None of the papers

included all. Every study took a hospital perspective when looking at the costs and benefits,

although only a few explicitly mention this. None of the studies performed their analysis from

a societal perspective, although comments were made on the importance of including these

costs. Disregarding the societal costs, there were 3 (3%) studies that included all of the other

parameters (Frighetto, et al., 2000; Gross, et al., 2001a; Hamblin, et al., 2012), 3 (4%) more

63 | Chapter 5

studies included all but implementation costs (and, as stated, societal costs) (Bauer, et al.,

2010b; Niwa, et al., 2012; Perez, et al., 2013). 57 studies (58%) only included antimicrobial

costs (Table 5.4).

Table 5.1: General characteristics of the reviewed studies (n=99).

Characteristic Number Percentage

Geography North America 51 52% South America 3 3% Europe 28 28% Asia 14 14% Africa 2 2% Australia 1 1% Publication year 2000 – 2003 8 8% 2003 – 2006 14 14% 2006 – 2009 17 17% 2009 – 2012 13 13% 2012 – 2014 47 47% Study design ITS 8 8% Quasi-experimental study 65 66% Retrospective evaluation 12 12% (R)CT 8 8% Cost-analysis 2 2% Observational study 3 3% Unclear 1 1% Number of beds in hospital < 150 10 10% 150 – 500 24 24% 500 – 1000 27 27% > 1000 12 12% Unclear 23 23% Number of patients included < 100 10 10% 100 – 250 22 22% 250 – 500 14 14% 500 – 1000 6 6% 1000 – 1500 6 6% > 1500 10 10% Unclear 31 31%


Table 5.2: Types of interventions of the reviewed papers. Performed interventions per category, the number of studies that evaluated this intervention, the percentage of the total and the total number of patients (if this was mentioned in the papers).

Block Interventions Number Percentage Patients

1 Altered therapy guidelines 16 16% 32,103 Antibiotic restriction lists or pre-authorization 12 12% 70,446 Giving education 10 10% 21,913 Antibiotic cycling 1 1% - Pre-analytic consultations 1 1% 100 New therapy 1 1% 2,888 2 Therapy evaluation, review and/or feedback 62 63% 51,506 Rapid diagnostic tools 9 9% 701 New biomarkers 2 2% - 3 Producing local use and resistance data 5 5% 19,390

Table 5.3: Types of economic evaluation of the reviewed papers.

Table 5.4: Scored outcome parameters of the reviewed papers.

Type of analysis

Number Percentage Cost outcome

measures Number Percentage

CA 2 2% Implementation 11 11% CMA 2 2% Antimicrobial 97 98% CBA 32 32% Operational costs 23 23% CCA 3 3% LOS 18 18%

CEA 3 3% Morbidity/

mortality 14 14%

CUA 0 0% Other 19 19% ACA 57 57% Societal 0 0%

CA: Cost-Analysis; CMA: Cost-Minimization Analysis; CBA: Cost-Benefit Analysis; CCA: Cost-Consequence Analysis; CEA: Cost-Effectiveness Analysis; CUA: Cost-Utility Analysis; ACA: Antimicrobial Cost Analysis

LOS: length of stay

65 | Chapter 5

Table 5.5: Results of CHEC list score.

Al-

Eid

an

et

al.

Oo

ster

-

hee

rt e

t al

.

Rütt

i-

man

n e

t al

.

Fri

ghet

to

et

al.

Ham

blin

et

al.

Gro

ss

et

al.

Zah

ar

et

al.

An

sari

et

al.

Sic

k

et

al.

Item

Is the study population clearly described?

Yes No No Yes No Yes Yes Yes No

Are competing alternatives clearly described?

No No No Yes No Yes No Yes Yes

Is a well-defined research question posed in answerable form?

Yes No No No Yes Yes Yes Yes Yes

Is the economic study design appropriate to the stated objective?

Yes No Yes Yes Yes Yes Yes Yes Yes

Is the chosen time horizon appropriate to include relevant costs and consequences?

Yes No No No Yes No Yes Yes Yes

Is the actual perspective chosen appropriate?

Yes Yes Yes Yes Yes Yes Yes Yes Yes

Are all important and relevant costs for each alternative identified?

No No Yes Yes No Yes Yes No Yes

Are all costs measured appropriately in physical units?

Yes No Yes No Yes No Yes Yes No

Are costs valued appropriately? Yes No Yes No No No Yes Yes Yes Are all important and relevant outcomes for each alternative identified?

Yes No Yes Yes No Yes No Yes Yes

Are all outcomes measured appropriately?

Yes No Yes Yes No No No Yes Yes

Are outcomes valued appropriately?

Yes No Yes Yes No Yes No Yes Yes

Is an incremental analysis of costs and outcomes of alternatives performed?

No No No No No No No No Yes

Are all future costs and outcomes discounted appropriately?

Yes No Yes Yes Yes Yes Yes Yes Yes

Are all important variables, whose values are uncertain, appropriately subjected to sensitivity analysis?

No No No No No No No No Yes

Do the conclusions follow from the data reported?

No No Yes Yes Yes No No Yes Yes

Does the study discuss the gene-ralizability of results to other set-tings and patient/client groups?

Yes No Yes No No No No No No

Does the article indicate that there is no potential conflict of interest of study researcher(s) and funder(s)?

Yes Yes No Yes Yes No Yes No No

Are ethical and distributional issues discussed appropriately?

No No No No No No No No No

Total score 13 2 11 11 8 9 10 13 14


Quality assessment

Following the criterion as stated in the material and methods section, 9 studies included

operational costs of the intervention and appropriately valued their costs by taking into

account inflation and/or price changes (Al Eidan, et al., 2000; Ansari, et al., 2003; Frighetto, et

al., 2000; Gross, et al., 2001a; Oosterheert, et al., 2005; Rüttimann, et al., 2004; Sick, et al.,

2013). For these studies, quality was assessed following the CHEC list and results are

mentioned in Table 5.5. Results ranged between 2 and 14 positives on the criteria of a

maximum of 19, with 3 studies scoring less than 10. An incremental analysis of the costs and

health outcomes, and sensitivity analyses were among the items most frequently missed.

Discussion

Concluding it can be said that economic evaluations of antimicrobial stewardship programs

have room for improvement. Often the methods chosen are insufficient to be conclusive,

because essential parameters are missing and multiple different approaches to evaluate an ASP

are being used. It is therefore in most cases difficult if not impossible to translate results to

other settings. Even relatively simple parameters, such as number of patients, study design and

setting, perspective chosen, performed statistics and inflation corrections are frequently

missing. Often the evaluation was done on a set of interventions, making their respective

benefits undistinguishable. The lack of quality is further reflected in the fact that only 9 studies

could be included in the quality assessment and none scored above 14 (of a maximum of 19),

underlining the need for a more systematic approach for these types of studies. A meta-analysis

of published results, which had the preference, could therefore not be performed. Thus, the

focus of this review was switched to highlighting the possible areas for improvement.

The primary goal for an ASP should be to improve patient outcomes and quality of care. It is

shown that running an effective ASP can optimize antimicrobial therapies thereby improving

patients’ treatment (Davey, et al., 2013). This can in turn positively affect local resistance rates

and local nosocomial infections rates. Some studies also showed the positive effect an ASP

could have on the length of stay (LOS) of a patient (e.g. (Al Eidan, et al., 2000; Frighetto, et al.,

2000; Niwa, et al., 2012; Perez, et al., 2013; Rimawi, et al., 2013; Sick, et al., 2013; Yen, et al.,

2012). Such a reduction can improve patient safety and quality of care as well. Keeping this in

mind, it is worthwhile to also look at the effectiveness of stewardship programs, especially

from an economical point of view. Obviously, it is highly preferably if this is performed

according to a set of guidelines (Drummond, et al., 2005). We have shown that major room for

improvement in this area still exists.

67 | Chapter 5

Ideally, an economic evaluation begins with defining the type of intervention that is performed

and the perspective that is chosen to do the evaluation. Most papers reviewed seemingly chose

a hospital perspective, although few mention this explicitly. A societal perspective might be

preferred in many cases, since patients and society in general will benefit from the intervention,

especially when a reduction in the antimicrobial resistance rates is achieved, or when LOS is

reduced. By merely looking at it from a hospital perspective, such benefits are lost or

marginalized (Chen, 2004). Effects of interventions are dependent on the local epidemiological

data on resistance, and consequently this will influence the total benefits and should thus not

be overlooked.

If the economic evaluation is done from a hospital perspective, the costs of the

program will consequently be the ones made by the hospital. These can be categorized as fixed

and variable. Fixed costs are those that do not vary with the quantity of output (i.e. number of

patients) in the short run and include for example rent, equipment, maintenance etc.

(Drummond, et al., 2005). For hospitals, studies showed that fixed costs can range from 65%

up to 84% of the total budget. The latter percentage also included the personnel costs in their

calculations (Roberts, et al., 1999; Taheri, et al., 2000). Staff salaries can be considered fixed or

semi-fixed, especially on the short term. This means that for an ASP, which is often evaluated

on the short term, direct cost reductions in practice can primarily be expected in the variable

costs. There is a need for a longer term perspective in economic evaluations of ASPs.

Almost all interventions require time, resources and sometimes equipment to

implement. These costs should not be overlooked. Not all community hospitals have the same

resources available as large academic centers, and implementation costs can therefore be a

major hurdle (Johannsson, et al., 2011). Of the reviewed papers however, only 11 studies

(11%) mentioned the costs spent on implementation.

For a majority of the reviewed studies (58%), the only included cost outcome

parameter was the cost price of antimicrobials. These costs are obviously easier to obtain and

measure than others. It is also one of the only ways for an ASP to significantly influence the

variable budget of the hospital. When evaluating the costs of antimicrobials however, it is still

important to adjust prices to a single year/level in order to remove the bias of changing

prices/inflation. Of the included studies only 14 (14%) did so. Prices of antimicrobials change

over time, due to inflation, but also because patents can expire and (cheaper) generic

substitutes can become available. This becomes especially important when a study evaluates

several years of antimicrobial acquisition costs. The longer the study period is, the higher the

chance that more generic substitutes became available due to expired patents. With more than

5 generic products entering the market, prices of the original brand antibiotic can drop by

more than 75% (FDA, 2010). Furthermore, prices yearly change in reality due to changing

agreements between hospitals and external parties. It is therefore essential that during a

financial evaluation of an ASP the prices are fixed for the whole study.


Performing the interventions, costs time and time spent invokes opportunity costs.

This implies that the time doctors or pharmacists spend on the ASP cannot be spent on

something else. This time should thus also be included in an economic evaluation. There are

multiple ways to measure the time that was put into the program and each method has his pros

and cons, as indicated by Page et al (2013) in an overview of various methods (Page, et al.,

2013). When looking at the total costs of the time spent, it is important to include all personnel

costs (salary and all related attributable on-costs) to give the most realistic monetary output.

One of the results of an ASP can be the reduction of LOS. From a financial point of view such

a reduction is highly interesting because of the high costs that are associated with it. Costs for a

hospital day are however almost completely fixed and will therefore not change over time due

to fewer patient hospital days, unless wards or beds are closed, but even then depreciation

costs are made (Rauh, et al., 2010). This is something to take into consideration. An important

question is thus how to value a possible reduction of the LOS. For hospital days there is no

market to establish a price (Scott, et al., 2001). This makes it difficult to aptly include the effect

on LOS in an economic evaluation. Often, studies look at the costs (fixed and variable) made

at a department or hospital and divide this by the patient days. An indication for these costs

can be calculated by accounting all costs; by looking at the incremental costs for the last

(cheaper) day(s) (Taheri, et al., 2000); or by willingness to pay (Stewardson, et al., 2014). The

latter study gave a fascinatingly low figure compared to the other methods, showing that high

prices for a hospital day are not always realistic depending on the perspective. Of the reviewed

papers that mentioned their cost for a hospital day, the average in 2013 Euro level, was

€649.04 per hospital day (Al Eidan, et al., 2000; Barenfanger, et al., 2000; Forrest, et al., 2006;

Frighetto, et al., 2000; Maddox, et al., 2014; Niwa, et al., 2012; Oosterheert, et al., 2005; Yen, et

al., 2012). However, because it was not always clear which costs were included in this figure, it

is almost impossible to draw conclusions based on these numbers. A reduction in LOS can

thus have a difficult to estimate effect on the balance. Freeing up beds does have however an

important positive benefit in the form of an increased possibility to boost the hospital’s

turnover. Reduction of hospital days is thus especially interesting if the backfill of a hospital is

large enough to fill up the freed beds. If this is not the case, fixed costs are divided over less

patient days, and depending on the cost structure, this can even mean hospital costs rise per

patient. Additionally, from a patient perspective there are huge benefits in leaving a hospital

earlier. These are covered by taken a broader perspective in the analysis.

A subsequent broader perspective for economical evaluations is looking at an

intervention from a health payer perspective. This entails the inclusion of costs made by the

patient, such as home medication and costs for a general practitioner. For a societal

perspective, additional indirect and non-medical costs that occur outside the hospital are

included, such as expenses made by patients like transportation and spent time, but also loss of

productivity for the society (Drummond, et al., 2005). Furthermore, a patient can live longer or

with better quality of life due to an intervention. One method to take these effects into account

69 | Chapter 5

is to measure the quality adjusted live years (QALYs) in a long-term analytic approach.

Together with the costs of the intervention, a cost-effectiveness ratio can then be calculated.

Depending on the threshold for a QALY gained, an intervention can be judged financially

worthwhile to implement or not. The inherent difficulties of this method (e.g. time and

knowledge needed to perform this extensive evaluation), especially for a relatively small

intervention program as an ASP, are illustrated by the fact that no study performed such an

analysis. In an outpatient setting however, (Oppong, et al., 2013), is a nice example of such a

study that did this analysis.

Results of an economic evaluation of an ASP can be used to convince a board of directors that

pro-actively spending money on improving antimicrobial therapies can give a positive return of

investment (Johannsson, et al., 2011). However, when this is the main goal of the evaluation, a

simple cost (benefit) analysis is not enough, because it will miss effects outside the hospital.

Cost effectiveness/utility is more precise, but complex. Preferably, the department or the

hospital therefore makes a business case model for an ASP in order to give an, as complete as

possible financial overview. Both (Stevenson, et al., 2012) and (Perencevich, et al., 2007)

proposed a similar set-up for this.

To assess the quality of the included papers within this review, the CHEC list was used (Evers,

et al., 2005). The Consolidated Health Economic Evaluation Reporting Standards (CHEERS)

Statement has been used to evaluate the quality of CEAs in reviews (Husereau, et al., 2013a).

See for example (Abu Dabrh, et al., 2014), (Hop, et al., 2014), and (Kawai, et al., 2014).

However, in our opinion, the CHEERS checklist is not suitable for this specific review. It only

considers whether something is reported, not whether the choices were appropriate or

justified. As such, a checked box on the CHEERS checklist is not an assessment of quality,

merely one of completeness. Another recent publication, the checklist by Caro et al., was

specifically designed for a reliability and credibility assessment. However, this checklist is only

applicable for decision models, and as such was not appropriate for our review (Caro, et al.,

2012). Therefore, the CHEC list was chosen, while this, as said, specifically gives insight in the

quality of economic evaluations (Evers, et al., 2005).

Concluding, it can be said that there is still much room to improve economic evaluations of

ASPs. Of the papers reviewed, none can really be used to draw strong economical conclusions.

Using more standardized methods to financially evaluate an ASP will contribute to the

advancement needed in this field and notably; further research should focus on the

harmonization of this field. Finally, inclusion of the societal perspective, real-world pricing and

a potentially longer time horizon for analysis can be recommended. Ultimately, these


improvements should provide a more solid basis for decision making, potentially leading to

better patient care.

71 | Chapter 5

Risk of bias across 15 Specify any assessment of risk of bias that may affect the N/A

Supplemental Table S5.1: PRISMA checklist.

Section/topic # Checklist item Reported on page #

TITLE

Title 1 Identify the report as a systematic review, meta-analysis, or both.

1 and 5

ABSTRACT

Structured summary

2

Provide a structured summary including, as applicable: background; objectives; data sources; study eligibility criteria, participants, and interventions; study appraisal and synthesis methods; results; limitations; conclusions and implications of key findings; systematic review registration number.

2-3

INTRODUCTION

Rationale 3 Describe the rationale for the review in the context of what is already known.

4-5

Objectives 4 Provide an explicit statement of questions being addressed with reference to participants, interventions, comparisons, outcomes, and study design (PICOS).

5

METHODS

Protocol and registration

5 Indicate if a review protocol exists, if and where it can be accessed (e.g., Web address), and, if available, provide registration information including registration number.

N/A

Eligibility criteria 6

Specify study characteristics (e.g., PICOS, length of follow-up) and report characteristics (e.g., years considered, language, publication status) used as criteria for eligibility, giving rationale.

6

Information sources

7 Describe all information sources (e.g., databases with dates of coverage, contact with study authors to identify additional studies) in the search and date last searched.

6 and Figure S1

Search 8 Present full electronic search strategy for at least one database, including any limits used, such that it could be repeated.

6

Study selection 9 State the process for selecting studies (i.e., screening, eligibility, included in systematic review, and, if applicable, included in the meta-analysis).

6-7

Data collection process

10 Describe method of data extraction from reports (e.g., piloted forms, independently, in duplicate) and any processes for obtaining and confirming data from investigators.

6-7

Data items 11 List and define all variables for which data were sought (e.g., PICOS, funding sources) and any assumptions and simplifications made.

6-7

Risk of bias in individual studies

12

Describe methods used for assessing risk of bias of individual studies (including specification of whether this was done at the study or outcome level), and how this information is to be used in any data synthesis.

N/A

Summary measures

13 State the principal summary measures (e.g., risk ratio, difference in means).

N/A

Synthesis of results

14 Describe the methods of handling data and combining results of studies, if done, including measures of consistency (e.g., I2) for each meta-analysis.

N/A


studies cumulative evidence (e.g., publication bias, selective reporting within studies).

Additional analyses 16 Describe methods of additional analyses (e.g., sensitivity or subgroup analyses, meta-regression), if done, indicating which were pre-specified.

7

RESULTS

Study selection 17 Give numbers of studies screened, assessed for eligibility, and included in the review, with reasons for exclusions at each stage, ideally with a flow diagram.

8 and Figure S1

Study characteristics

18 For each study, present characteristics for which data were extracted (e.g., study size, PICOS, follow-up period) and provide the citations.

N/A

Risk of bias within studies

19 Present data on risk of bias of each study and, if available, any outcome level assessment (see item 12).

N/A

Results of individual studies

20

For all outcomes considered (benefits or harms), present, for each study: (a) simple summary data for each intervention group (b) effect estimates and confidence intervals, ideally with a forest plot.

N/A

Synthesis of results 21 Present results of each meta-analysis done, including confidence intervals and measures of consistency.

N/A

Risk of bias across studies

22 Present results of any assessment of risk of bias across studies (see Item 15).

N/A

Additional analysis 23 Give results of additional analyses, if done (e.g., sensitivity or subgroup analyses, meta-regression [see Item 16]).

10

DISCUSSION

Summary of evidence

24

Summarize the main findings including the strength of evidence for each main outcome; consider their relevance to key groups (e.g., healthcare providers, users, and policy makers).

10-11

Limitations 25 Discuss limitations at study and outcome level (e.g., risk of bias), and at review-level (e.g., incomplete retrieval of identified research, reporting bias).

14-15

Conclusions 26 Provide a general interpretation of the results in the context of other evidence, and implications for future research.

15

FUNDING

Funding 27 Describe sources of funding for the systematic review and other support (e.g., supply of data); role of funders for the systematic review.

16

74

Automatic day-2 intervention by a

multidisciplinary Antimicrobial Stewardship-Team

leads to multiple positive effects

Jan-Willem H. Dik, Ron Hendrix, Jerome R. Lo-Ten-Foe, Kasper Wilting, Prashant Nannan

Panday, Lisette E. van Gemert, Annemarie M. Leliveld, Job van der Palen, Alex W. Friedrich,

Bhanu Sinha

Frontiers Microbiology 2015; 6:546

6

75 | Chapter 6

Abstract

Antimicrobial resistance rates are increasing. This is, among others, caused by

incorrect or inappropriate use of antimicrobials. To target this, a

multidisciplinary Antimicrobial Stewardship-Team (A-Team) was implemented

at the University Medical Center Groningen on a urology ward. Goal of this

study is to evaluate the clinical effects of the case-audits done by this team,

looking at length of stay (LOS) and antimicrobial use.

Automatic e-mail alerts were sent after 48 hours of consecutive antimicrobial

use triggering the case-audits, consisting of an A-Team member visiting the

ward, discussing the patient’s therapy with the bed-side physician and together

deciding on further treatment based on available diagnostics and guidelines.

Clinical effects of the audits were evaluated through an Interrupted Time Series

analysis and a retrospective historic cohort.

A significant systemic reduction of antimicrobial consumption for all patients

on the ward, both with and without case-audits was observed. Furthermore,

LOS for patients with case-audits who were admitted primarily due to

infections decreased to 6.20 days (95% CI: 5.59-6.81) compared to the historic

cohort (7.57 days; 95% CI: 6.92-8.21) (p=0.012). Antimicrobial consumption

decreased for these patients from 8.17 DDD/patient (95% CI: 7.10-9.24) to

5.93 DDD/patient (95% CI: 5.02-6.83) (p=0.008). For patients with severe

underlying diseases (e.g. cancer) these outcome measures remained unchanged.

The evaluation showed a considerable positive impact. Antibiotic use of the

whole ward was reduced, transcending the intervened patients. Furthermore,

LOS and mean antimicrobial consumption for a subgroup was reduced, thereby

improving patient care and potentially lowering resistance rates.

Automatic day-2 intervention by a multidisciplinary Antimicrobial Stewardship-Team leads | 76

to multiple positive effects

Introduction

Antimicrobial resistance is a world-wide problem. Suboptimal prescription and subsequent

inappropriate use of antimicrobials contributes to increasing development of resistance

(Goossens, 2009; Tacconelli, et al., 2009a). The optimization of antimicrobial therapy in

hospitalized patients is therefore an urgent global challenge (Bartlett, et al., 2013; World Health

Organization, 2012). Urology departments are even more vulnerable because of a high number

of (high risk) gram-negative bacteria species encountered (Wagenlehner, et al., 2013). Several

aspects regarding antimicrobial therapy such as choice of drug (and spectrum), duration, mode

of administration and dosage should be subject for improvement (Braykov, et al., 2015). This is

also true for countries with a relatively low antimicrobial prescription rate and low resistance

rates, such as the Netherlands (de Kraker, et al., 2013; European Centre for Disease

Prevention and Control, 2013). Therefore, Dutch government has made an Antimicrobial

Stewardship Program (ASP) with Antimicrobial Stewardship-Teams (A-Teams) mandatory for

every hospital from January 2014 (SWAB, 2012).

Antimicrobial stewardship addresses many aspects of infection management, of which one is

audit and feedback of the therapy (Davey, et al., 2013). In recent years this has proven to

improve on the appropriateness of antimicrobial therapy (Cisneros, et al., 2014; Liew, et al.,

2015; Pulcini, et al., 2008b; Senn, et al., 2004). At the University Medical Center Groningen

(UMCG) in the Netherlands, this case-audit is combined with face-to-face consultation on day

2. The goal is to optimize and streamline the antimicrobial therapy as early as possible using

microbiological diagnostics, thereby improving patient care. Face-to-face consultations are

used explicitly to create an effective learning moment for prescribing physicians (Lo-Ten-Foe,

et al., 2014).

The aim of this study is evaluating this implemented A-Team on a urology ward in an

academic hospital setting, focusing on two clinical outcome measures: antimicrobial use and

length of stay (LOS). Because there are considerable risks of acquiring a nosocomial infection

with each extra day of hospitalization (Rhame and Sudderth, 1981) and acquiring a catheter-

related infection with each extra day of an intravenous line (Chopra, et al., 2013; Sevinç, et al.,

1999), changes in LOS and/or antimicrobial administration can have a large impact on the

quality of care and patient safety. The direct return on investment for this A-Team has already

been extensively evaluated separately, using the same patient groups, and found to be positive

(Dik, et al., 2015b). The clinical outcome measures in this study have been evaluated through

an interrupted time-series analysis as well as through a quasi-experimental set-up, thereby

providing an extensive evaluation of a set of clinically relevant parameters.

77 | Chapter 6


The study was performed at a 20-bed urology ward in a large 1339-bed academic hospital in

the northern part of the Netherlands from June 16th 2013 to June 16th 2014. Inclusion of

patients for the A-Team was done using a clinical rule program (Gaston, Medecs BV,

Eindhoven, the Netherlands). The applied algorithm selected patients who received 48 hours

of selected antimicrobials (Supplemental Table S6.1). These were chosen based upon in-house

evaluation of consumption at the ward and their respective risk for resistance development in

general. In total 72% of the prescriptions for this ward was covered, including all drugs

recommended in the applicable guidelines for the included patients. The clinical rule program

automatically sends an email alert to the A-Team members (day 2), which contains the patient

hospital ID, prescription details (e.g. administered antimicrobials, dosage and start date [day

0]). It also includes clinical chemistry data (ALAT, ASAT, CRP, leucocytes and creatinine).

Microbiological diagnostic reports were collected manually. The hypothesis was that in the

majority of the cases microbiologic diagnostic results will be (partly) available at day 2, and can

thus be used during the case-audit. Furthermore, all A-Team members had access to all

available microbiological data, including not yet finally authorized data (i.e. samples still being

processed). Patients whose antimicrobial therapy started within three days after admission were

included in the evaluation.

The A-Team is multi-disciplinary consisting of clinical microbiologists, infectious disease

physicians, and hospital pharmacists. It reports to the hospital’s Antimicrobials Committee and

Infection Prevention Committee, which have a mandate from the board of directors to

implement and run the ASP. It is an integral part of one of the “leading coalitions” within the

hospital to improve patient safety and quality of care. One A-Team member (clinical

microbiologist or infectious disease physician) visited the ward after being triggered by the

email alert and discussed antibiotic therapy with the bed-side physician(s) face-to-face.

Consensus on the continuation of the treatment is one of the main goals. During the case-

audit, the therapy was discussed and the available diagnostics were reviewed. Using the

expertise and experience of both the A-Team member and the bed-side physician a decision on

the continuation of the antimicrobial therapy was made. Decisions were always made based on

local guidelines for urological infections, which in turn are based on national and European

guidelines (Geerlings, et al., 2013; Grabe, et al., 2013). In the end, agreed-on interventions were

scored. The chosen interventions were based upon a previous pilot in the same hospital

(Supplemental Table S6.2) (Lo-Ten-Foe, et al., 2014).

Compliance was assessed at day 30. If the agreed-on intervention was followed by the

appropriate action within 24 hours, it was scored as compliant.



Evaluation of practice was the main goal of this study, focussing on a change in antimicrobial

consumption and a reduction in LOS. Based upon the pilot study (Lo-Ten-Foe, et al., 2014), a

reduction of at least one day was our pre-determined goal. A financial evaluation of the same

group of patients and using the same historic cohort was already performed (Dik, et al.,

2015b). Implementation of the A-Team is part of a local hospital-wide ASP, which is being

developed keeping in mind recommendations done by the IDSA/SHEA and the ESGAP

(Dellit, et al., 2007; Keuleyan and Gould, 2001).

Historic control cohort

For evaluation of the clinical effects, a frequency based historic control cohort was compiled.

The control cohort consisted of patients who stayed at the same ward in a 30-month period

prior to the intervention. Diagnosis Related Group codes (DRG) from the patients in the

intervention group and the consumption of >48 hours of the alert antimicrobials

(Supplemental Table S6.1) were used to filter the control cohort. DRG codes were assigned to

the patient after discharge for insurance purposes by a grouper, based upon scored procedures

(http://www.DBConderhoud.nl). Patients’ antimicrobial consumption was measured in

DDDs per 100 patient days, as stated by the WHO (http://www.whocc.no/atc_ddd_index).

The described case-audit was normal every day care implemented within the hospital and

approved by the Antimicrobials Committee, following national guidelines. This study was of an

observational nature, evaluating the newly implemented procedures. Data was collected

retrospectively from the hospital's data warehouse; it was anonymous, partly aggregated and

did not contain any directly or indirectly identifiable personal details. Following Dutch

legislation and guidelines of the local ethics committee, formal ethical approval was therefore

not required (http://ww.ccmo.nl/en).

Subgroup analysis

After explorative analysis of the data two subgroups were compiled to correct for the

modifying effect of the patients’ indication. The two groups were stratified by DRG codes.

The first group (Group 1) consisted of patients without severe underlying diseases and whose

infection was the most likely major problem and the main driver for LOS. The second group

(Group 2) consisted mainly of oncology and transplantation patients. Here the underlying

problem (e.g. cancer) was the most likely driver for LOS, rather than the infection.

79 | Chapter 6


For antibiotic consumption of the total ward, including patients without intervention(s), an

interrupted time-series analysis was performed. For subgroup analysis, unpaired t-tests, chi

square tests and Kaplan-Meier survival plots with a log-rank test were applied, as appropriate.

Significance threshold was p < 0.05. For subgroup analyses, a threshold of p < 0.025 was set in

order to account for possible family wise error rates. Analysis was done with IBM SPSS

Statistics 20 (IBM, Armonk NY, USA) after one year.

Results

Consulted patients

During the 1-year study period, 1298 patients were admitted to the urology ward. 850 received

at least one dose of antimicrobials. 114 alert patients were included in this study (61% male;

mean age 62 yrs male, 50 yrs female, see Table 6.1). They received a total of 126 case-audits

(including 12 follow-up consults), resulting in 166 interventions. Consensus was reached in

97.6% of the cases (n=123) and the compliance (i.e. action within 24 hrs) was 92.1% (n=116).

Case-audits took on average between 10 and 15 minutes, including administration time.

Table 6.1: Patient baseline characteristics. Included patients and the control group patients’ characteristics and their respective p-values. 95% Confidence intervals are shown in brackets, when applicable. Characteristics are given for the total group, as well as for the two subgroups. Group 1 consisted mainly of patients without severe underlying diseases, and Group 2 consisted of patients with severe underlying diseases. Intervention group Control group P-value

Total group N=114 N=357 Male 61% 69% 0.11a

Mean age (yrs) 57.51 (± 2.95) 61.52 (± 1.72) 0.01b,c

Group 1 N=70 (61%) N=209 (59%) Male 51% 60% 0.11a

Mean age (yrs) 55.25 (± 3.71 ) 59.35 (± 2.25) 0.046b

Group 2 N=44 (39%) N=148 (41%) Male 75% 84% 0.71a

Mean age (yrs) 61.12 (± 4.79) 64.57 (± 2.62) 0.14b

a) Chi square test b) Mann-Whitney U Test c) Total group was not used further in calculations. The difference in age and sex showed no significant influence on the outcome measures.



Results of microbiological diagnostics were mostly available on day 2

In 86.0% (n=98) of the alert patients microbiological diagnostics had been initiated, in 50.0%

(n=57) this was done on day 0 or 24 hours prior to starting antimicrobials. At the first case-

audit (day 2) results were (partly) available (gram staining, incomplete culture data) in 72.8%

(n=83) of the cases.

A large majority of the consulted patients received interventions

Of the patients who were consulted, there was an alteration of the therapy (any intervention

besides ‘continue’ at the first case-audit) in 74.7% of the patients. In 23.7% (n=27; 16.3% of

total interventions) treatment was stopped. A switch to oral treatment was performed in 23.7%

(n=27; 16.3% of total interventions). 21.9% (n=25; 15.1% of total interventions) received a

different antimicrobial, dosage was optimized in 4.4% (n=5; 3.0% of total interventions) and

treatment de-escalated in 15.8% (n=18; 10.8% of total interventions). For 8.8% (n=10; 6.0%

of total interventions) there another intervention (e.g. add an antimicrobial, perform extra

diagnostics) was performed (see Figure 6.1 for the stratification of interventions per subgroup).

Figure 6.1: Interventions performed. Distribution of the interventions performed for alert patients,

subdivided into the two Groups. Percentages of interventions refer to the total number done within the

75% of intervened patients, where one patient can receive multiple interventions.

81 | Chapter 6

Prescribing trends of the whole ward changed after implementation

Most notably, the positive effect transcended the target group on this ward. The trend of

antimicrobial consumption of all patients admitted to the ward (17.3% intervened and 82.7%

not intervened) changed after start of the intervention. Using an interrupted time-series

analysis there was an observed drop of 25.0% after 1 month (p=0.012), 23.6% at 6 months

(p=0.007), and 22.4% at 12 months (p=0.047), compared to expected usage, based upon the

extrapolated pre-intervention data (Figure 6.2A and 6.2C).

The mean percentage of antimicrobial recipients per month in relation to the total number of

patients dropped by 7.3% at 1 month (p=0.131), 10.4% at 6 months (p=0.018) and 12.8% at

12 months (p=0.024), compared to the expected percentage of recipients (Figure 6.2A and

6.2B).

Figure 6.2: A-Team effects on the whole ward. 6.2A: Trends of percentages of all patients on the

ward receiving antibiotics with and without intervention(s) and respective DDDs per 100 patient days.

Shown are two years before the intervention started (June 2013), until June 2014, including trend lines

and predicted trend lines. A second dotted trend line for the mean DDDs depicts the trend without the

outlier patient from April 2014 (*). 6.2B: Predicted and measured percentages of users at three different

time points with their respective p-values, calculated with an interrupted time series analysis. 2C:

Predicted and measured consumption with their respective p-values, on the same three time points with

the same interrupted time series analysis.

*) The peak in the month April is caused by a single patient who received correct but extensive small spectrum oral

antibiotic treatment for a deep-seated, complicated infection.



Length of stay was significantly reduced for Group 1 patients

Length of stay was evaluated for the two subset groups to take into account the modifying

effect of the patients’ indication. Group 1, without severe underlying diseases, showed an

average LOS reduction of 1.46 days compared to the control group (6.20 days; 95% CI: 5.59-

6.81 versus 7.57 days; 95% CI: 6.92-8.21; p=0.012) (Figure 6.3A). LOS for patients of Group

2, with severe underlying disease had a minor but non-significant increase compared to the

control group (8.36 days; 95% CI: 7.10-9.62; versus 8.10 days; 95% CI: 7.24-8.97; p=0.801)

(Figure 6.3B). Patients’ LOS remained the same for patients who stayed at the department and

did not receive an intervention (3.95 days; 95% CI: 3.75-4.16) compared to one year earlier

(3.96 days; 95% CI: 3.71-4.21; p=0.581).

Figure 6.3: Kaplan-Meier plots of length of stay. Percentages of patients’ days of discharge. Group 1

intervention patients compared to the historic cohort group 1 (6.3A) with Group 2 patients as insert

(6.3B). Significance was tested with a Log-Rank test (Mantel-Cox).

83 | Chapter 6

Antimicrobial consumption was lower for Group 1 patients

Overall antimicrobial consumption dropped by 2.24 DDD per patient for Group 1 (p=0.008).

There was a trend to lower intravenous administration (67%, compared to 69% in the control

cohort; p=0.099). For Group 2 there was a non-significant drop of 0.91 DDD per patient

(p=0.712) (Table 6.2).

Discussion

The main goal of this study was to evaluate an already implemented A-Team and the effects of

its case-audits, by looking both at LOS and DDDs. These parameters were taken as main

outcome measures, because they are known to have a major impact on the quality of care, and

they are affected by an ASP (Davey, et al., 2013). The A-Team reviewed antimicrobial therapy

on day 2, thus, making optimal use of available (microbiological) diagnostics. By means of the

automatic alert (which can be modified for specific groups of patients, departments, and

specific antimicrobials) face-to-face case-audit on the ward was encouraged and facilitated by

providing an easy overview of relevant patient information. Objective of the case-audit was to

reach a consensus-based agreement between the A-Team member and the physician at the

ward, using (local) guidelines, available diagnostics and the expertise and experience of both

physicians. This should optimize antimicrobial treatment. The case-audit focused on the

improvement of patient care through relatively easy to achieve improvements after only 48

hours of therapy, such as an early IV-PO switch, and stopping therapy when there was no

longer an indication. Furthermore, the bed-side physicians anticipated the A-Team visits.

These provided an extra opportunity for questions about appropriateness of therapy and

Table 6.2: Effects of the A-Team interventions on antibiotic use (DDD per patient). Antibiotic consumption compared between the intervention group and the historic control cohort for Group 1 and Group 2. Consumption is presented as mean DDDs per patient, the difference between the intervention and control in percentages and 95% CI in brackets. All tests were unpaired, two-tailed t-tests performed on log-transformed data. Intervention Group Control Group Difference P-value

Group 1

N=70

N=209

Overall 5.93 (± 0.90) 8.17 (± 1.07) -27.5% 0.008 PO 1.95 (± 0.52) 2.51 (± 0.44) -22.3% 0.849 IV 3.97 (± 0.93) 5.66 (± 1.05) -29.8% 0.099 Group 2

N=44

N=148

Overall 7.21 (± 1.47) 8.13 (± 1.11) -11.2% 0.712 PO 2.98 (± 0.92) 2.81 (± 0.59) 5.9% 0.931 IV 4.20 (± 1.36) 5.31 (± 0.95) -20.3% 0.738



requesting additional consultations for further patients on the ward. This resulted in 19

additional patients discussed during the evaluation period. These patients had not triggered an

alert, for example because therapy was still less than 48 hours or because therapy had not even

been started. 12 patients received more than one case-audit. These follow up audits were often

deemed necessary because culture results were not completely available.

Very unexpectedly, and unlike previously published in other studies, we found a considerable

significant additional positive effect on the antimicrobial consumption of the whole ward. With

consultation of 17.3% of the patients at the ward during the evaluation period, we saw a broad

effect on all patients, including those without any consultations or intervention(s). The drop in

the rate of patients receiving antimicrobials and the drop in DDDs per patient has been very

likely due to the continuing educational effect of the consensus-based case-audit where face-to-

face information exchange is taking place. Although antimicrobial consumption of the whole

ward would be directly affected by interventions, the percentage of recipients should not

change. This strengthens the conclusion that the A-Team presence at the ward had an effect

transcending the group of intervened patients. The knowledge that antimicrobial use is

monitored and evaluated will most likely contribute to a higher awareness by the prescribing

physicians, thereby possibly creating a kind of Hawthorne effect (Mangione Smith, et al.,

2002). Going to wards to discuss patients requires investment of staff time. However, it should

be noted that for this ward, on average only 10 to 15 minutes were spent per case-audit by an

A-Team member, including administration. It was not the goal to discuss every patient with

antimicrobial therapy. A large majority of patients received prophylactic therapy of which the

duration should be less than 48 hours (often just a single shot) and where an intervention

would not achieve a lot of effect. The quality of economic analyses of ASPs is often sub

optimal, due to insufficient outcome measures and performed methods (Dik, et al., 2015c).

Therefore, a more extensive and thorough economical analysis has been also undertaken for

the same patients and using the same historic cohort. Mainly due to the decrease in LOS for

the group of patients primarily admitted due to an infection, the implementation of the A-

Team had a positive direct return on investment (Dik, et al., 2015b).

Automatic email alerts can be easily adjusted to the specific needs of a hospital or ward,

depending on local challenges and required goals (e.g. IV-PO switch, review of only restricted

antimicrobials, only one ward or the whole hospital, children or adults, and at a timeframe of

choice), making it a relatively simple method to keep track of patients receiving (selected)

antimicrobials. We estimated that implementation of this specific program for the whole

hospital would require 2 to 3 fte A-Team specialists.

85 | Chapter 6

The large number of performed interventions shows that implementing an A-Team was highly

relevant. Indeed, frequent non-optimal antimicrobial treatment of urology patients has been

shown earlier (Hecker, et al., 2014). Antimicrobial treatment can be improved also in other

patient populations (Braykov, et al., 2015). In line with our results, an audit and feedback of

therapy after 48 hours has been recently shown to be fruitful in a hospital-wide setting (Liew,

et al., 2015). However, this study did not show a reduced LOS as a benefit. Although difficult

to compare due to the different setting, the higher availability of microbiological diagnostics in

our patients might account for the differences in LOS. Furthermore, patient it is important to

take the characteristics of the patient group into account.

Here, we observed that the interventions resulted in a reduced average LOS for a subgroup of

patients and a global drop in antimicrobial DDDs. This finding underscores that a substantial

proportion of patients can be switched earlier to oral administration or stopped completely, an

easily achievable target, the “low hanging fruits” (Goff, et al., 2012). The IV-PO switches also

explain why oral administration did not change significantly. However, the effects on oral

DDDs might even be larger, because 23% of the IV to PO switch patients was sent home

directly after the switch without getting inpatient oral therapy. Consequently, they did not

count for the calculation of mean DDDs for PO treatment.

By lowering LOS and DDDs the risks for hospital acquired infections, catheter-

related infections and resistance development should also be lower, thereby improving patient

care and safety (Chopra, et al., 2013; Rhame and Sudderth, 1981; Sevinç, et al., 1999).

However, the current time-frame of the study is too short to measure these effects and they

are thus to be investigated in the coming years. Other outcome measures, notably duration of

therapy in days and re-admissions rates were not significantly altered (data not shown). If an

ASP were implemented for more complex patients (e.g. on an ICU), it would also be important

to take morbidity and/or complications, and mortality into account, because these outcome

measures can be expected to affect mainly more complex patients.

It is important to note that LOS may not be taken as a universal outcome measure or quality

indicator for an ASP. As we conclusively show, a day 2 bundle will not lower the LOS for all

patients. Especially in an academic hospital there are many referral patients with severe

oncology or transplant surgery indications. If these patients subsequently develop an infection,

the driver for the LOS and/or their antimicrobial use is most likely the underlying disease and

a decrease in outcome measures such as LOS can thus not be expected. This fact could also

explain ambiguous results on patients’ LOS seen in the Cochrane Review (Davey, et al., 2013).

Consequently, this point should be taken into account for the design and analyses of future

ASP studies. Finally, it should be noted that the success of the program is not determined by a

decrease in LOS or DDDs. The main goal was to improve the quality of care by adjusting



therapy as quickly as possible according to the diagnostic results. This had an effect on LOS

and DDDs on a subgroup of patients. Of note, this does not imply that the intervention failed

in patients of group 2. Rather, this suggests that outcome measures to monitor success of an

ASP have to be chosen wisely. Possible (in)direct effects extend much further, such as a lower

risk of resistance and better quality of care for the patients due to a more optimal antimicrobial

treatment. However, these outcome measures are difficult to measure objectively, especially on

the short term in a retrospective set-up.

Our study has some limitations. Effects of the A-Team were evaluated for a urology ward in

an academic setting. To investigate effects in different settings reliably, further studies are

needed. Possible confounders with an effect on the LOS due to the quasi-experimental set-up

and the chosen cohort might have been present, rendering ASP studies complicated to

perform (Marwick and Nathwani, 2014). This study evaluated the effects of an already

implemented intervention. Performing a (randomized) controlled trial was therefore neither an

option nor a goal. Analyses were done within subgroups, to exclude modifying effects of the

underlying disease. Without such a sub analysis, effects would be averaged out and lost. During

the study period no other additional measures were performed to influence antimicrobial

prescriptions or reduce LOS (i.e. changes in formularia, additional education or changes in

restriction of antimicrobials). For antimicrobial consumption seasonal effects were ruled out by

looking back two years. The distribution of age and sex was not optimal, but no correlation or

effect was found on LOS or antibiotic use. The fact that the LOS of Group 2 did not change

significantly provides an extra internal negative control. Unfortunately, not all information was

available for evaluation (possible co-morbidities were not available as objective, measurable

data).

We conclude that ASP interventions should be further encouraged. Inappropriate use of

antimicrobials contributes to higher resistance rates, making infections even more difficult to

treat (Burke and Yeo, 2004). Recently a Dutch multicenter study showed that appropriateness

of antimicrobial therapy in UTI patients has a positive impact on the LOS (Spoorenberg, et al.,

2014), supporting our findings that the A-Team interventions successfully optimized therapy

thereby reducing the LOS. Even though A-Teams cost money to implement, the reduction in

LOS for some patients provides enough benefits to provide a positive return on investment

(Dik, et al., 2015b). Pro-active collaboration between the treating medical specialty, medical

microbiology, infectious diseases, and pharmacy departments via a day-2 evaluation of

antimicrobial therapies could be used to provide benefit for patients in other hospitals, as well.

88

Cost-minimization model of a multidisciplinary Antibiotic

Stewardship Team based on a successful implementation on a

urology ward of an academic hospital

Jan-Willem H. Dik, Ron Hendrix, Alex W. Friedrich, Jos Luttjeboer, Prashant Nannan Panday,

Kasper Wilting, Jerome R. Lo-Ten-Foe, Maarten J. Postma, Bhanu Sinha

PLoS One 2015; 10(5):e0126106

7

89 | Chapter 7

Abstract

In order to stimulate appropriate antimicrobial use and thereby lowering the

changes of resistance development, an Antibiotic Stewardship-Team (A-Team)

has been implemented at the University Medical Center Groningen, the

Netherlands. Focus of the A-Team was a pro-active day 2 case-audit which was

financially evaluated here to calculate the return on investment from a hospital

perspective.

Effects were evaluated by comparing audited patients with a historic cohort

with the same diagnosis-related groups. Based upon this evaluation a cost-

minimization model was created which can be used to predict the financial

effects of a day 2 case-audit. Sensitivity analyses were performed to deal with

uncertainties. Finally the model was used to financially evaluate the A-Team.

One whole year including 114 patients was evaluated. Implementation costs

were calculated to be €17,732, which represent total costs spent to implement

this A-Team. For this specific patient group admitted to a urology ward and

consulted on day 2 by the A-Team, the model estimated total savings of

€60,306 after one year for this single department, leading to a return on

investment of 5.9.

The implemented multi-disciplinary A-Team performing a day 2 case-audit in

the hospital had a positive return on investment caused by a reduced length of

stay due to a more appropriate antibiotic therapy. Based on the extensive data

analysis, a model of this intervention could be constructed. This model could be

used by other institutions, using their own data to estimate the effects of a day 2

case-audit in their hospital.

Cost-minimization model of a multidisciplinary Antibiotic Stewardship-Team based on a successful | 90

implementation on a urology ward of an academic hospital

Introduction

Inappropriate and inefficient use of antimicrobial therapy and consequent resistance

development can lead to unwanted clinical effects, such as toxicity, hospital acquired

infections, longer length of stay (Nowak, et al., 2012) and to high costs for hospitals and

society (Cosgrove, 2006; de Kraker, et al., 2011; Roberts, et al., 2009). In 2007, the extra in-

hospital costs for antibiotic resistance were estimated at 900 million Euros for the European

Union, and another 1.5 billion Euros societal cost (ECDC/EMEA Joint Working Group,

2009), although estimations on societal costs are thought to be underestimated (Smith and

Coast, 2013). Without appropriate action, resistance rates will rise, as will the subsequent costs

(World Health Organization, 2012). These unwanted financial consequences are becoming

more and more relevant in times of healthcare cost reductions and savings programs (Morgan

and Astolfi, 2014). To control these costs, it is critical to cut spending wisely by focusing on

inefficient services (World Health Organization, 2014b) such as inadequate antimicrobial

therapy. Improving this treatment can not only have a positive effect on patient care and

safety, but also be a potential contributor to wise budgetary savings.

One way to improve patients’ antimicrobial therapy and stimulate prudent use is through

Antimicrobial Stewardship Programs (ASPs) (Davey, et al., 2013; Dellit, et al., 2007; With de, et

al., 2013). This term covers a wide arrange of interventions, all done with the objective to

optimize patients’ antimicrobial therapy. Several papers discussed the financial effects of ASPs,

and although difficult to compare due to the wide arrange of different interventions in

different populations and because the quality of the evaluations is sub optimal, the consensus

seems to be that costs can be saved (Dik, et al., 2015c; Goff, et al., 2012; Gray, et al., 2012).

Part of an ASP is an Antibiotic Stewardship-Team (A-Team). At the University Medical Center

Groningen (UMCG) in the Netherlands, an A-Team has been implemented at a urology ward

to perform a case audit after 48 hours (day 2) of initiation of antibiotic therapy to improve the

quality of care. The aim of this study was to construct a cost-minimization model reflecting

direct costs and benefits on a hospital-wide level. This has been achieved by means of a

historic cohort study for this case-audit performed by an A-Team, with the focus on direct

return of investment from a hospital perspective. This study provides a framework, which can

be used for other hospitals when implementing an A-Team, giving an indication what the

direct return on investment of this specific ASP would be.

91 | Chapter 7


Antibiotic Stewardship Team

The day-2 case audits by the A-Team were implemented within a 1339-bed academic medical

center. It is a multidisciplinary team consisting of clinical microbiologists, infectious disease

specialists and hospital pharmacists. One of the team members performed ward visits and

discussed patients’ antibiotic therapy with the attending physician (fellow/resident). These

ward visits could include a bed-side consultation and examination depending on the patient’s

condition. During weekends, the consultations were either done by phone or on site during the

subsequent working day. Eligible patients were selected on the basis of an automatic e-mail

alert from the hospital’s pharmacy. An alert, generated from a clinical rule program (Gaston

Medecs BV, Eindhoven, the Netherlands) was sent when a patient used an antibiotic,

considered to be of specific relevance (flucloxacillin, amoxicillin/clavulanic acid,

piperacillin/tazobactam, cefuroxime, ceftriaxone, meropenem, clindamcyin, tobramycin,

ciprofloxacin, vancomycin and teicoplanin), beyond 48 hours after initiation. Together with the

attending physician, interventions were decided upon to improve the therapy. The case-audits

were implemented on a urology ward after a successful test on a trauma surgery ward (Lo-Ten-

Foe, et al., 2014).

Patients and historic cohort

Patients who stayed at the urology ward and received consultations by an A-Team member,

between June 2013 and June 2014 were evaluated. Patients of whom antibiotics were started

within three days of admission were included in this financial evaluation, resulting in data for

114 included patients. Financial effects related to length of stay, antibiotic consumption and

nursing time needed for intravenous treatment were compared to a frequency-based historic

cohort. The cohort included patients who stayed at the same ward within a period of 30

months before the intervention started. This cohort was filtered based on the frequency of the

Dutch equivalent of Diagnosis Related Group (DRG) codes of the intervention patients

(DBC, http://www.dbconderhoud.nl) and the same antibiotics for 48 hours consecutively.

Treatment policies based on national and local guidelines did not change within this period.

To account for the influence that patients’ indications can have, subgroups were analyzed on

the basis of their DRG codes. The first group (DRG Group 1) consisted of patients with

codes for infections and infections-related indications. The second group (DRG Group 2)

consisted of patients with codes for more severe underlying diseases (e.g. cancer), but who

developed a subsequent clinical infection and received antibiotics.



Economic data

All prices are given in Euros, 2013 level. Older prices were adjusted to 2013 using the Dutch

consumer price index (http://statline.cbs.nl). Costs for the implementation and running of the

A-Team were subdivided into personnel costs of the A-Team, costs made at the ward, medical

costs, research/evaluation costs, overhead and implementation costs. Personnel costs are

based upon Dutch gross salaries. Costs per hour for medical specialists in the A-Team, the

hospital pharmacist, attending physicians, nurses and investigators were based on gross salaries

and set at 120, 120, 35, 30 and 25 Euros respectively (Hakkaart-van Roijen, et al., 2010). The

A-Team member scored total time for every consultation, including administration; this time

has been used for the cost calculations. Because no new personnel were hired for this project,

fixed costs did not change. Following the principle of opportunity costing we therefore used

these measured times as indication for the opportunity costs that occurred for the A-Team

member and the attending physician.

Prices for the antibiotic medication were provided by the hospital pharmacy and

reflect the actual prices charged by pharmaceutical companies. For all data, 2013 prices were

used to account for possible changes over time.

Costs of one patient day in the hospital were based on the internal costing system

applied to the urology ward, based on the actual budgets of 2012, resulting in a total of €716

per patient per day. Notably, overhead costs (including building costs, maintenance,

equipment, personnel costs for daily care, etc) are included in these estimations; procedures are

not included in this figure because we believe that a reduction in LOS does not influence the

number of procedures substantially.

Analyses

Subsequently, total costs and benefits of this specific A-Team case-audit on day 2 were

compared via a model. Costs for the consultations were subtracted from the costs of the

chosen outcome measures, if they statistically significantly differed from the historic control

cohort. For the model, values were calculated using the historic cohort. The respective

variables are shown in brackets.

It was assumed that outcome measures of patients having a DRG for an infection related

problem can be influenced, but that patients with (severe) underlying causes will not see any

differences in the outcome measures. This ratio was therefore included in the model as the

percentage of patients with infection-related DRG codes .

93 | Chapter 7

A change in hospital days was calculated and multiplied by the abovementioned price

.

A switch form IV to oral administration is one of the interventions that was promoted by the

A-Team. Therefore, nursing time spent on the administration of IV treatment per patient

was calculated for this study, based on a survey on the ward. denotes the

percentages of patients that received IV treatment. This number was multiplied by the nursing

time and the gross salary of the nurse.

Case-audit costs were calculated based on time spent by the A-Team doctor for the

case-audit. For each consultation he/she scored the total time spent, including travel time and

administration. Costs were multiplied by the average number of audits per patient and

the gross salary. The total figure also included costs for the attending physician on the ward,

the hospital pharmacist, an investigator of the department and miscellaneous costs (e.g.

meetings of the A-Team members).

Lastly, the average costs of the antimicrobials per patient were evaluated between the

group that received consultations and the historic cohort group.

This resulted ultimately in the following formula for the costs and benefits of the A-Team:

( )

For the calculation of a return on investment (ROI), the total number is divided by the total

costs, giving a ratio for the direct return that can be expected for the hospital. Finally, for the

parameters with the most impact on the model, deterministic sensitivity analyses were

performed. Furthermore, for a set of parameters, a probabilistic sensitivity analysis (PSA) was

performed. For this purpose, 2500 repeat drawings were done.



Ethics statement

The described stewardship case-audit was normal every day care implemented within the

hospital and approved by the antimicrobials committee following national guidelines. ASPs are

mandatory by Dutch law for every hospital. Data was collected retrospectively from the

hospital's data warehouse. It was anonymous and did not contain any (in)directly identifiable

personal details. Following Dutch legislation and guidelines of the local ethics commission,

approval was therefore not required (http://www.ccmo.nl).


Depending on the type of data, appropriate tests were applied. Due to the analyses on two

subset groups, the significance threshold was set at p < 0.025 to account for possible

familywise error rates. Analysis was done with IBM SPSS Statistics 20 (IBM, Armonk NY,

USA).

Results

Firstly the implementation costs were calculated. To then examine the costs or benefits of the

ASP, all parameters for the model that were unknown were determined, based on the

comparison between the patients with interventions and the patients without intervention in

the historic cohort. To account for some uncertainties that the model might possess,

multivariate sensitivity analyses and a probabilistic sensitivity analysis were performed.

Ultimately the final number was calculated with the model.

Implementation costs

Before starting the A-Team there was time invested to develop the procedures and protocols

that were used, as well as the developing and testing of the clinical rule of the e-mail alerts.

Time was scored retrospectively based upon A-Team members’ personal time schedules. Total

costs came to €17,732. This includes implementation for other wards as well. These costs are

not used in the model, because they are independent of this specific ward.

Model parameters

All parameters in the model were calculated using the compared data of the patients with and

without interventions. Total compliance of the interventions was 92.1%. For the two DRG

95 | Chapter 7

groups, only in DRG Group 1 significant effects were found. For this ward, 61% of the

patients were part of this group . All measured effects were thus multiplied with this

percentage. Within DRG Group 1, LOS dropped from 7.57 (95% CI: ±0.65) to 6.20

(95% CI: ±0.61) (p=0.012). Nursing time needed for the administration of IV antibiotics

was based upon a survey at the specific urology ward. IV-treatment was assessed to

cost on average 64.83 minutes a day. Respective times were 10.5 minutes for preparation per

dosage; one time 7.5 minutes for catheter insertion; 10 minutes every 4 days for changing the

catheter; 10 minutes for control per day; one time 2.5 minutes for removal. On average, IV

antibiotics were given 2.27 times a day for 4.04 days. For DRG Group 1 the nursing time

dropped from 251.66 minutes (95% CI: ± 32.94) to 178.19 (95% CI: ±40.91) (p=0.016). 74%

of the audited patients received IV treatment . Within DRG Group 2 there were no

significant changes (Table 7.1).

Cost for 1 case-audit was on average €80.26 . 1 audit took on average 12.1 minutes

(95% CI: ± 0.77 min.), costing €24.03 (95% CI: ± €1.53). During the audit, the A-Team

member spoke with the attending physician, costing €7.01 (95% CI: ± €0.45). The A-Team’s

hospital pharmacist was responsible for the e-mail alerts and the clinical rules, including daily

maintenance. His costs were estimated at €35.95 per audit. An investigator of the department

evaluated the A-Team interventions for quality assurance, costing €2.08 per audit. This

amounts to a total of €69.07 of direct personnel costs of the A-Team. Furthermore there are

miscellaneous costs that need to be taken into account (e.g. meetings of A-Team members);

total costs of this are €11.19 per audit. On average, patients received 1.09 case-audits

per admission.

Mean costs per patient for antimicrobial medication showed a trend towards decreasing

costs compared to the historic control groups, however it was not significant. In DRG Group

1 mean costs decreased from €17.42 (95% CI: ± 3.33) to €13.34 (95% CI: ± 4.55) (p=0.030),

for DRG Group 2 costs went from €21.47 (95% CI: ± 6.56) to €15.76 (95% CI: ± 8.30)

(p=0.637). However, because p-values were higher than 0.025 effects on cost price were not

taken into account in the model (Table 7.1).



Table 7.1: Model parameters. Mean effects per patient, analyzed per subgroup. Age in years, length of stay (LOS) in days, IV nursing time in minutes and when applicable 95% CI or percentage is shown in brackets. P-values < 0.025 were considered significant to account for familywise error rates.

Intervened Control cohort P-value

DRG Group 1

Total patients 70 (61.40%) 209 (58.54%)

Age 55.25 (± 3.71) 59.35 (± 2.25) 0.046a,e

Male 51% 60% 0.112b

Mean LOS 6.20 (± 0.61) 7.57 (± 0.64) 0.012c IV patients 52 (74.29%) 168 (80.38%)

Mean IV nursing time 178.19 (± 40.91) 251.66 (± 32.94) 0.016d Antibiotic costs € 13.34 (± € 4.54) € 17.42 (± € 3.33) 0.030a

Mean number of consultations

1.11 -

DRG Group 2

Total patients 44 (38.60%) 148 (41.46%)

Age 61.12 (± 4.49) 64.57 (± 2.62) 0.139a

Male 75% 84% 0.708b

Mean LOS 8.36 (± 1.26) 8. 10 (± 0.87) 0.801c IV patients 32 (72.73%) 114 (77.07%)

Mean IV nursing time 206.70 (± 67.72) 249.66 (± 40.45) 0.550d Antibiotic costs € 15.76 (± € 8.30) € 21.47 (± € 6.56) 0.637a

Mean number of consultations

1.09 -

a) Mann-Whitney U test b) Chi square test

c) Log-Rank test (Mantel Cox)

d) Unpaired, two-sided t-test

e) No correlating influence of age was found

Patients’ main diagnosis and price of a hospital day had the largest impact on the total

benefits

For the parameters with the largest/highest effect on the end result, multivariate sensitivity

analyses were performed to visualize the effect they have on the return of investment. On the

surfaces plots (Figure 7.1) the ranges in Euros are shown. Besides the price for a hospital day,

the distribution of patients with or without severe underlying problems had the highest

impact on the financial profitability.

Probabilistic Sensitivity Analysis (PSA)

In the model for DRG Group 1, where significant results were found, a PSA was performed

for LOS, nursing time and antibiotic costs. Other parameters within the model had the

97 | Chapter 7

baseline value (Table 7.1). For percentage of patients primarily administered with an infection

baseline value was set at 100%. The PSA gave an expected total benefit between €32 and

€2,696, with a median of €877 and a 95% interval between €380 and €1,580 per patient (Figure

7.2).

The hospital saved considerable amounts of money with their A-Team

Costs and benefits for the hospital based on the implementation on the urology ward were

calculated with the model. Using all the baseline values as mentioned above, for one year of

interventions and 114 patients the total profits amounted to €60,306 (€529.05 per

consultation), which led to a direct return on investment (ROI) of 5.9.

Figure 7.1: Surface graphs showing different parameter ranges. All graphs show a range of the

hospital day price on the Y-axis with another parameter on the X-axis. The color represents the costs or

benefits ranging from minus €200 in dark red to €1400 in dark green; the dashed line indicates €0. The

total value for the UMCG, using the urology study values is shown in every plot. A: Range of baseline

average of LOS before starting with an A-Team. B: Range of percentage of patients primarily

administered with an infection. C: Range of costs of 1 consultation. D: Range of percentage of expected

LOS reduction through A-Team interventions.



Figure 7.2: Probabilistic Sensitivity Analysis

for DRG Group 1. Depicted is the probability

of the costs or benefits in Euros with a PSA

done for LOS, nursing time and antibiotic costs.

The model ran for 2500 repeats. The 95%

confidence interval is represented with the two

dashed lines.

Discussion

Here we show a model, which can be used to calculate the financial effects of an A-Team.

Using the data from the intervention study, the case-audit on day 2 performed by an A-Team

was cost-effective and saved more than 60,000 Euros with consulting just 114 patients in one

year on one ward. Revenues can be therefore expected to be much higher when a program like

the one presented here were implemented in the whole hospital while implementation costs

would remain relatively stable and acceptable.

Antimicrobial stewardship programs (ASP) are introduced in more and more hospitals to

improve quality of antimicrobial therapy and thereby controlling the growing resistance to

antibiotics. Consequently more studies are published about the effects of stewardship

programs, both with regard to clinical and financial aspects (Davey, et al., 2013). In our

hospital, an ASP was already in place and it has been expanded with an A-Team doing case-

audits on day 2.

This study looked at two outcome measures (LOS and antibiotic use). The effects on those

parameters will be greater for patients who were admitted primarily for an infection (DRG

group 1), which is clearly visible in the surface plots (Figure 7.1). A reduction of antimicrobial

costs has been reported as resulting in major savings in some studies (Niwa, et al., 2012; Perez,

et al., 2013). In this study however, these savings were much lower because consumption was

99 | Chapter 7

already relatively low and the antibiotics prescribed are mostly relatively inexpensive generic

products. A 23% reduction for DRG Group 1 was observed, which showed a trend to

significance, comprising 16% of the total costs (Table 7.1). Further studies with a financial

evaluation were published, but only a few studies included more than savings from direct

antimicrobial costs (Ansari, et al., 2003; Dik, et al., 2015c; Gross, et al., 2001b; Hamblin, et al.,

2012; Rüttimann, et al., 2004). It is therefore difficult to compare economical studies for the

effects of ASPs. Different parameters were taken into considerations, different methods and

interventions were used and different types of patients were targeted for intervention.

However, most intervention studies that included costs are reporting positive results (Ansari, et

al., 2003; Davey, et al., 2013; Dik, et al., 2015c; Hamblin, et al., 2012; Rüttimann, et al., 2004).

The monetary costs and benefits in this study are for the most part indirect costs and savings

because these are fixed costs in the hospital budget. Fixed costs are defined as costs that do

not vary with the quantity of output in the short run, and include e.g. capital, equipment and

staff (Drummond, et al., 2005). Personnel costs will not change on the short run because of

effects from this study as long as there is no mutation in staff. Thus direct benefits will be in

the reallocation of resources. The same holds true for the personnel costs of the A-Team. No

new staff was hired for the implementation of this study, but current staff members have been

prioritizing there available time differently. Because this was a pilot project, the time spent on

this project had been allocated from designated research time as a resource. There are thus

opportunity costs, which is why personnel costs should not be underestimated. Furthermore, if

the A-Team were to be implemented throughout the entire hospital, additional staff would be

needed. Current staff would have insufficient time to do consultations for the whole hospital

besides their normal duties. The mentioned savings in nursing time due to the IV-PO switch

are indirect savings as well. For IV treatment we saw a drop of 73.47 minutes per patient in

DRG Group 1, based upon an average time of around 64.83 minutes a nurse would need per

day per patient. This average is comparable to earlier Dutch studies, as well as international

studies (Handoko, et al., 2004; Tice, et al., 2007; van Zanten, et al., 2003). Total nursing time is

slightly higher in this study, because insertion, control, changing and removal of the IV

catheter are also taken into account. Since there was no change in the total nursing staff, the

budget did not change and current expenses remained the same. Reallocated resources is in

this case were more available time for treatment of other patients. It is very likely that the

quality of care, as well as infection prevention and control will benefit from that (Aiken, et al.,

2002; Clements, et al., 2008). The reallocation of nursing time could in itself lead to further

savings. This, however, falls outside the scope of this study. The results we see in the reduction

of length of stay are indirect, as well. The major parts of the high price of a bed-day are fixed

costs like equipment and housing. These costs do not change if patients stay less time in the

hospital. Freeing up beds for additional admissions, however, does mean these resources will

be available for additional patients and procedures. Thus, the benefits should create the

possibility for increasing the revenue-generating activities and shortening waiting lists (De

Angelis, et al., 2010). Effects might even be slightly higher when procedures were included.



Since we estimated that during the last days of admission these costs would be limited, we

therefore chose to be more conservative and leave them out of the equation. This study also

looked at the readmission rates between the intervention group and the historic cohort, but no

significant differences were observed between the two groups (DRG group 1: 17% vs 12%;

p=0.244; DRG group 2: 11% vs 20%; p=0.190). They were therefore not incorporated into

the model.

Unfortunately not all costs were known and could be incorporated into the model. For

example adverse reactions and complications are missing because they are not stored in an

objective, usable manner. Also missing are the indirect costs from a societal perspective.

Preferably a study should include all costs associated with the implementation of an ASP,

direct and indirect (Jönsson, 2009). However, indirect effects of the interventions and the costs

associated with these effects are difficult to determine correctly with just one ASP intervention

study on one ward. Inherent to the study and the model there are some uncertainties. DRG

group 1 has some sub-optimal baseline parameters. However extensive additional analyses

showed no influences of these parameters on the outcome measures. Uncertainties within the

model were made clear by using the different sensitivity analyses for five variables as well as

doing a probabilistic sensitivity analysis, which showed positive results within the whole 95%

confidence interval.

One of the main expected results of correct antibiotic usage is the lowering resistance rate in

the hospital. These lowered rates will most likely reduce the risk of acquiring nosocomial

infections in the hospital. This in turn will most likely contribute to reducing the risk of an

outbreak of such an infection, which will be accompanied by substantial costs (Roberts, et al.,

2009). As shown here, an A-Team on its own can already generate a positive revenue and ROI,

by improving the usage of antibiotics. Assuming that this would subsequently lead to lowering

the possible number of outbreaks, there are thus considerable additional indirect savings.

However, the time-frame of this study was too short to examine effects on resistance rates.

This is furthermore not only affected by an ASP, but also by for example hospital hygiene

measures and other infection prevention measures, not only from the targeted hospital, but

also from hospitals within the same health care network (Ciccolini, et al., 2013). Such networks

can exist regionally, but keeping in mind the European directive on international patient care

and differences between e.g. MRSA bacteremia in Germany and the Netherlands, international

borders should not be neglected (van Cleef, et al., 2012). Being an academic hospital, there are

more patient movements to and from our hospital, making the hospital more prone to receive

patients with (resistant) hospital acquired infections (Donker, et al., 2012). For a more

extensive financial evaluation of ASPs and supplemental infection prevention measures, it

would important that local and regional policies and measures are neglected. Including more

data would make a model more precise, also from an economical point of view (Donker, et al.,

101 | Chapter 7

2012). Further research should therefore be focused on the improvement of appropriate

models. However, using our model, it is already possible to estimate at least some of the

financial effects of an A-Team in any given hospital. Since this study is so far one of a few

studies looking beyond direct antimicrobial costs, it hopefully will contribute to more in depth

research into the financial effects of ASPs and A-Teams. For a more extensive financial

evaluation of ASPs and supplemental infection prevention measures, it would important that

local and regional policies and measures are neglected. Including more data would make a

model more precise, also from an economical point of view (Rudholm, 2002). Further research

should therefore be focused on the improvement of appropriate models. However, using our

model, it is already possible to estimate at least some of the financial effects of an A-Team in

any given hospital. Since this study is so far one of a few studies looking beyond direct

antimicrobial costs, it hopefully will contribute to more in depth research into the financial

effects of ASPs and A-Teams. For our hospital we can state that it is worth to invest in day-2

case-audits, because the reduction in LOS makes the program highly cost-effective.

104

Cost-analysis of seven nosocomial outbreaks

in an academic hospital

Jan-Willem H. Dik, Ariane G. Dinkelacker, Pepijn Vemer, Jerome R. Lo-Ten-Foe,

Mariëtte Lokate, Bhanu Sinha, Alex W. Friedrich, Maarten Postma

PLoS One 2016; 11(2):e:0149226

8

105 | Chapter 8

Abstract

Nosocomial outbreaks, especially with (multi-)resistant microorganisms, are a

major problem for health care institutions. They can cause morbidity and

mortality for patients and controlling these costs substantial amounts of funds

and resources. However, how much is unclear. This study sets out to provide a

comparable overview of the costs of multiple outbreaks in a single academic

hospital in the Netherlands.

Based on interviews with the involved staff, multiple databases and stored

records from the Infection Prevention Division all actions undertaken, extra

staff employment, use of resources, bed-occupancy rates, and other

miscellaneous cost drivers during different outbreaks were scored and

quantified into Euros. This led to total costs per outbreak and an estimated

average cost per positive patient per outbreak day.

Seven outbreaks that occurred between 2012 and 2014 in the hospital were

evaluated. Total costs for the hospital ranged between €10,778 and €356,754.

Costs per positive patient per outbreak day, ranged between €10 and €1,369

(95% CI: €49-€1,042), with a mean of €546 and a median of €519. Majority of

the costs (50%) were made because of closed beds.

This analysis is the first to give a comparable overview of various outbreaks,

caused by different microorganisms, in the same hospital and all analyzed with

the same method. It shows a large variation within the average costs due to

different factors (e.g. closure of wards, type of ward). All outbreaks however

cost considerable amounts of efforts and money (up to €356,754), including

missed revenue and control measures.

Cost-analysis of seven nosocomial outbreaks in an academic hospital | 106

Introduction

Nosocomial outbreaks are a major problem for health care institutions due to increased

morbidity and mortality for the affected patients. The containment and control of these

outbreaks costs substantial amounts of funds and resources, especially when left unnoticed or

untreated (van den Brink, 2013). Rising antimicrobial resistance levels further increase the

difficulty to treat nosocomial infections, incurring increasing costs (Roberts, et al., 2009; Stone,

et al., 2005; World Health Organization, 2012). Although it is known for some organisms what

the burden of disease is when a nosocomial infection occurs (Piednoir, et al., 2011; Roberts, et

al., 2009; Stone, et al., 2005), estimates of the exact costs for health care institutions during

outbreaks are scarce. Knowing the average cost of an outbreak per patient per day can help

decision makers to justify the necessary investments in infection prevention and control

measures, thus improving the decision making process (Perencevich, et al., 2007). This study

sets out to evaluate several nosocomial outbreaks within a single Dutch academic hospital with

an active Infection Prevention Unit, over a time period of three years.

Within the Netherlands there is proactive national infection prevention and control

collaboration through the Working group Infection Prevention (http://www.wip.nl). They

provide over 130 different guidelines on infection prevention, stating all the actions health care

institutions should perform and facilitate. The Search-and-Destroy policy for MRSA is one of

the success stories of the Dutch infection prevention approach (Bode, et al., 2011). In this

study we provide a transparent cost-analysis, describing in detail the costs that occur during the

control of an outbreak in a large Dutch academic hospital and related costs of missed revenues

due to closed beds. These data will give a comparable overview of outbreaks caused by

multiple microorganisms in one health care center, thus providing novel insights into

nosocomial outbreak costs.


All evaluated outbreaks occurred between 2012 and 2014 in a university medical center in the

north of the Netherlands with 1339 registered beds, including a separate rehabilitation center.

Outbreaks for which all data was available to perform an analysis were evaluated. Costing was

done from a hospital perspective. All identifiable extra costs that were made due to an

outbreak were taken into account (from the start of the outbreak until one year after the end of

the outbreak). An outbreak was defined as at least two patients who were tested positive as

indicator for colonization or infection for the same microorganism (bacterial or viral), with

some epidemiological link (e.g. same time-period, same ward). The duration of an outbreak

was counted in days and began on the day that the Infection Prevention Unit started measures

107 | Chapter 8

to control the outbreak until the day that they decided the risk of transmission was over and

additional control measures were not deemed necessary anymore. When an outbreak was

suspected, the Infection Prevention Unit provided assistance to the affected ward and advised

on extra surveillance cultures, extra cleaning, isolation of patients, and possible closure of the

ward if necessary. Actions are based on the local and national infection prevention guidelines.

Counted costs were divided into: microbiological diagnostics/surveillance costs; missed

revenue due to closed beds (based upon the difference in bed occupancy rates compared to the

two months before); additional cleaning costs; additional personnel (infection prevention,

nursing staff and clinicians); costs made for contact or strict isolation of patients and other

costs (e.g. purchase of extra materials, possible prolonged length of stay, extra medication). In

order to take into account all possible costs that occurred, a wide range of different sources

were used to ensure that no expenses were overlooked. Admission data came from the general

hospital database. Numbers of cultures came from the Medical Microbiology Database and the

prices for the diagnostics were internal cost prices (depending on type of diagnostic between

€25 and €200 per sample). Extra personnel and possible other costs were calculated based on

interviews with the, at that time, involved staff (i.e. head nurses and medical specialists),

together with detailed case reports from each outbreak made by the Infection Prevention Unit.

Bed day cost prices (ranging between €464.39 and €1829.78, depending on the type of ward

and excluding variable costs) and personnel costs (ranging between €19.23 and €75.54 per

hour) were based upon Dutch reference prices (Hakkaart-van Roijen, et al., 2010). It was

assumed that the hospital will not lay off any (temporary) personnel during closure of wards,

meaning personnel costs were considered fixed for this study. When calculating extra workload

for the personnel due to infection control measures, these costs were included as opportunity

costs. Isolation costs were calculated after internal evaluation (€25.14 for contact and €33.51

for strict isolation per day). Additional cleaning costs and the prices for those were calculated

based on interviews and stored records as provided by the department of technical - and

facility services (€25.38 per hour). When calculating the missed revenue, only the (fixed) bed

day cost price was taken into account. Possible opportunity costs for the (fixed) personnel

costs were left out in order to be more conservative in the calculation. All prices were

converted to 2015 Euro level, using Dutch consumer index figures (http://www.cbs.nl).

This analysis followed the CHEERS guideline and included all applicable items as

recommended when reporting economic evaluations (Husereau, et al., 2013b). The study was

purely observational and retrospective of nature and performed on outbreak level. The

anonymized data used for the analyses were collected by those authors functioning as treating

physicians, from the department’s own database. The collected data did not include any

(in)directly identifiable personal details and the analyzing authors had no access to those,

complying with the local data protection committee regarding clinical data processing.


Following Dutch legislation and guidelines of the local ethics commission approval was

therefore not required (http://www.ccmo.nl). Calculations were done with Microsoft Excel

(Microsoft, Redmond, WA, USA) and SPSS (IBM, Amonk, NY, USA). When statistics were

performed, a significance level of p < 0.05 was applied.

Results

Outbreak and patient characteristics

In total, seven different outbreaks could be financially evaluated. One outbreak caused by a

virus and seven bacterial outbreaks (see Table 8.1 for the responsible microorganisms). For

these outbreaks there were between 3 and 37 positive patients. The duration was between 16

and 86 days. Characteristics of the outbreaks can be found in Table 8.1. The Pantoea

transmission was treated as an outbreak because it occurred on a neonatal ICU.

Table 8.1: Outbreak and patient characteristics.

Microorganism Year Ward Positive persons

Average age (years)

Gender (% male)

Duration (days)

Pantoea spp. 2012 ICU 9 0.0 (25 days) 54% 36 S. aureus (MRSA)

2012 Nursing 3 56.9 75% 16

K. pneumonia (ESBL)

2012 Nursing 5 73.2 50% 24

K. pneumonia (ESBL)

2012 Rehabilitation 9 44.5 100% 17

E. faecium (VRE) 2013 Nursing 19 61.1 95% 50 Norovirus 2013 Rehabilitation 37 59.6 53% 28 S. marcescens 2014 ICU 8 0.1 50% 86

MRSA: Methicillin-resistant Staphylococcus aureus; ESBL: Extended-spectrum beta-lactamase; VRE: Vancomycin-resistant Enterococcus.

Cost-analysis

Total costs for the hospital ranged between €10,778 for the Norovirus outbreak and €356,754

for the 2014 S. marcescens outbreak. On average, costs per positive patient per outbreak day,

ranged between €10 for the Norovirus outbreak and €1,368 for the ESBL K. pneumonia on

the nursing ward (95% CI: €49-€1042). The mean of the total average costs per positive patient

per day comes to €546 and the median to €519.Within the average costs per positive patient

per outbreak day, the majority of the costs (50%) were made because there was the closure of

109 | Chapter 8

(multiple) ward(s) leading to missed revenue; 17% of the costs were for extra microbiological

diagnostics; 11% due to contact or strict isolation of patients; 10% for extra personnel; 7% for

other costs; and 5% due to extra cleaning on the affected wards (see Table 8.2). Interquartile

ranges of the costs per different categories are displayed in Figure 8.1. Binary regression

analysis showed that a bacterial outbreak is correlated with higher average costs per patient per

day compared to the single viral outbreak (p = 0.02).

Figure 8.1: Interquartile ranges of the costs (in Euros) per category. Medians are depicted by the X,

within the closed beds there was one outlier of €1144.

Table 8.2: Average costs per positive patient per outbreak day. Micro-organism

Total Diagnos-

tics Closed

bed Cleaning Personnel

Patient isolation

Other

Pantoea spp. €88.11 €53.50 €0.00 € 0.70 €8.80 € 24.40 €0.70 S. aureus (MRSA)

€657.08 €205.20 €221.79 €34.67 € 112.33 € 72.25 € 10.83

K. pneumonia (ESBL)

€1,368.92 €70.59 € 1,144.50 €17.76 €33.27 € 46.28 € 56.52

K. pneumonia (ESBL)

€980.51 €234.17 € 98.03 €150.93 € 154.94 € 227.55 € 114.90

E. faecium (VRE)

€197.26 € 64.45 € 69.27 € 3.74 €42.08 € 17.72 €0.00

Norovirus €10.40 € 5.21 €0.00 € 0.55 €2.21 €2.33 € 0.10 S. marcescens €518.54 € 19.34 €375.00 € 0.68 €15.59 € 25.67 € 82.27

MRSA: Methicillin-resistant Staphylococcus aureus; ESBL: Extended-spectrum beta-lactamase; VRE: Vancomycin-resistant Enterococcus.


Discussion

Seven outbreaks were evaluated and costs varied substantially. On average, the additional costs

due to an outbreak were €546 per positive patient per outbreak day. These average costs

ranged between €10 and €1,369. The most expensive outbreak per patient per outbreak day

was an ESBL producing K. pneumonia. The highest costs in this outbreak occurred due to a two

week closure of the ward, which not only led to a drop in admitted patients, but also to a

cancellation of scheduled procedures which could not be replaced with others due to the short

term of the cancellation. The lowest average costs were made for the Norovirus outbreak. This

was mainly due to the specifics of this infection, less diagnostics were necessary, because only

patients with Noro-like symptoms (e.g. watery diarrhea) were tested. Due to the short

incubation time, chances of undetected transmission are small. Furthermore, in this case

closure of the ward was deemed unnecessary and there was no excess morbidity or mortality

for the patients due to their Noro infection. In almost all cases of the evaluated outbreaks,

patients were colonized but not infected. During the Klebsiella outbreak in the rehabilitation

center there was excess morbidity in the form of one sepsis episode and subsequently all costs

made during this admission were taken into account and categorized under ‘Other’. For the S.

marcescens outbreak there was excess length of stay observed for two patients by the treating

medical specialist. Also in this case, these costs were taken into account and categorized under

‘Other’. Ergo, it seems that outbreak and microorganism specific characteristics cause large

variation in the total costs. This variation together with the relatively small number of

outbreaks also meant that correlation analyses on the data were impracticable. We therefore

chose to give a descriptive overview. Based upon this cost-analysis we hypothesize that on

average a viral outbreak is most likely to be less expensive than a bacterial one. This is mainly

due to easier and quicker detection, which reduces the duration and the microbiological costs.

For bacterial outbreaks we observed large variation in the costs. One of the biggest cost drivers

is the closure of wards and the subsequent drop in revenue, especially when this closure has

consequences for scheduled procedures. Cleaning costs are dependent on the type of ward,

with less additional costs on an ICU ward, because cleaning procedures are already strict and

higher costs in the rehabilitation center due to extra rooms (e.g. physiotherapy exercise rooms)

that had to be cleaned more strict than normal.

Although there are numerous studies on the (financial) burden of disease of resistant

organisms and nosocomial infections (Gandra, et al., 2014; Roberts, et al., 2009), there are just

a few cost-analyses published on nosocomial outbreaks. Notably, financial evaluations of

Norovirus outbreaks are published most (Fretz, et al., 2009; Johnston, et al., 2007; Navas, et

al., 2015; Sadique, et al., 2016; Zingg, et al., 2005). Besides Noro, there are publications on P.

aeruginosa, MRSA, Acinetobacter and VRE (Björholt and Haglind, 2004; Bou, et al., 2009;

Christiansen, et al., 2004; Jiang, et al., 2016). Although difficult to compare, because the

methods were not always similar, total costs seem comparable, but average costs per patient

per outbreak day seem slightly lower than those found here. The difference is likely to be

111 | Chapter 8

caused by different definitions for the duration of an outbreak. This study took the time during

which extra infection prevention measures on the ward were in place. Others choose to count

the days between the first positive culture of the index patient until discharge of the last

positive patient, which might be considerably longer, thereby lowering the average cost. The

lack of financial evaluations and consistency within those evaluations does however clearly

show the need for more studies and especially a more consistent methodological approach.

Strength of this study is the fact that multiple outbreaks caused by different microorganisms in

a single hospital were evaluated. This gives a comparable overview of how cost categories

differ between different outbreak situations. Limitations are that it is a retrospective analysis

and it might be that data are missing because of this. However, by using multiple sources for

data, this aspect is minimized. Still, preferable all data are collected prospectively. Depending

on national health care systems it will differ who will be bear the costs. This makes comparison

between different studies from different countries more difficult. By presenting the costs

within different categories we tried to make interpretation of the data more flexible and

adaptable to other settings. Finally, with only one viral outbreak, this category is under

represented, there were however no more suitable viral outbreaks to include during the

evaluated time-period.

Concluding, we present a cost-analysis of multiple outbreaks in an academic center. Although

costs differ between different outbreaks, due the microorganism or type of ward, the average

costs per patient per day seem substantial. Especially with ever rising antimicrobial resistance

levels, such outbreaks as described here are becoming continuously more difficult to treat and

costs are expected to rise even further in the future. Average costs of an outbreak per patient

per day as presented here can be used to further clarify costs and benefits within hospitals

related to infection prevention. Ultimately this should help decision makers to justify the

necessary investments in infection prevention and control measures. Our study may thus

contribute to more transparency in health care budgets and improve the decision-making

processes concerning where to invest. Notably, extra argumentation can be found in these

findings for investments into infection prevention and control measures to avert outbreaks or

to contain them swiftly.

114

Positive impact of eight years infection prevention on

nosocomial outbreak management at an academic hospital

Jan-Willem H. Dik, Mariëtte Lokate, Ariane G. Dinkelacker, Jerome R. Lo-Ten-Foe,

Bhanu Sinha, Maarten Postma, Alex W. Friedrich,

Future Microbiology. 2016; epub before print

9

115 | Chapter 9

Abstract

Infection prevention (IP) measures are vital to prevent (nosocomial) outbreaks.

Financial evaluations of these are scarce. An incremental cost-analysis for an

academic IP unit was performed.

On a yearly basis we evaluated: IP measures; costs thereof; numbers of patients

at risk for causing nosocomial outbreaks; predicted outbreak patients; and actual

outbreak patients.

IP costs rose on average yearly with €150,000; however more IP actions were

undertaken. Numbers of patients colonized with high-risk microorganisms

increased. The trend of actual outbreak patients remained stable. Predicted

prevented outbreak patients saved costs, leading to a positive return on

investment of 1.94.

This study shows that investments in IP can prevent outbreak cases thereby

saving enough money to earn back these investments.

Positive impact of infection prevention on the management of nosocomial outbreaks at an academic hospital | 116

Introduction

The unwanted spread of microorganisms within healthcare institutions can lead to major

problems, such as nosocomial infections and outbreaks, resulting in increased morbidity and

mortality (Eber, et al., 2010; Klevens, et al., 2007). Besides these highly undesirable clinical

effects, these problems also cost considerable amounts of finances and resources. (Eber, et al.,

2010; Graves, 2004; Plowman, et al., 2001). The worldwide problem of antimicrobial resistance

further increases the risk of difficult or even impossible to treat nosocomial outbreaks with

Multi Drug Resistant Organisms (MDROs). It is therefore vital to have a proactive infection

prevention department or infection control program that actively screens and acts upon

possible outbreak risks. Actions by these departments, such as hand hygiene measures or the

search-and-destroy policy for MRSA have already demonstrated positive clinical results that

can be achieved (Bode, et al., 2011; Grayson, et al., 2011; Stone, et al., 2012). All these

measures do, however, come at a price. Present-day studies on clinical effects of infection

prevention and/or control are still in need of improvement (Graves, 2014; Wolkewitz, et al.,

2014). Consequently, proper economic evaluations of these interventions, keeping in mind all

potential biases that are inherent to this field, are also scarce (Stone, et al., 2002). It is most

likely that infection prevention measures can be cost-effective and can yield substantial return

on their investment (Graves, 2004). However, some measures will be more cost-effective than

others, achieving the same clinical goals. Having the financial information on these measures

will improve the decision-making processes.

This study sets out to evaluate the Infection Prevention Unit and its budget at an academic

hospital over eight subsequent years. Because of the inherent difficulties in evaluating costs

occurring due to healthcare associated infections mentioned before, we focus purely on the

costs of nosocomial outbreaks, the prevention thereof and its impact on the whole infection

prevention budget, through an incremental cost-benefit analysis. Each year, the infection

prevention budget has been increased to cope with the rise in antimicrobial resistance. This

study evaluates the yearly incremental rise of the budget to calculate a yearly return on

investment (ROI). Such an approach eliminates the problems of having to estimate a baseline

situation without infection control measures based on numerous highly uncertain assumptions.

This study should be seen as a first step towards calculating a proper and comprehensive

financial analysis that incorporates all costs and benefits of infection prevention aspects.


The data within this study concerns eight subsequent years, from 2007 up to 2014 at a tertiary

academic center in the north of the Netherlands. During this period, national infection

prevention protocols changed due to updates, however the general approaches remained

117 | Chapter 9

similar (http://www.wip.nl). General hospital data such as number of admissions per year,

average length of stay (LOS) and number of beds per year were taken from the hospital’s

annual reports (http://www.desan.nl). Microbiological culture data were collected from the

database of the Department of Medical Microbiology. Data of consumables were based on the

purchase details from the hospital’s purchasing department. For the cost-analysis of the

outbreaks, data came from a previous cost-analysis, done within the same hospital on a subset

of the outbreaks investigated here (Dik, et al., 2016a). A hospital perspective was used with

2012 price levels.

Determining the size of the Infection Prevention Unit

For each year, the total number of infection prevention specialists employed by the hospital in

December of the respective year was taken in FTE (Full-Time Equivalent; based on 36 hours

per week). This number of FTEs was further increased by 10% of the total FTE of clinical

microbiologists to account for their pro rata infection prevention work, and for additional

related researchers and technicians in the field of infection prevention, including the next-

generation sequencing group of the department. The number of nursing staff for the whole

hospital was also evaluated over the same time period to control if possible changes was due to

a hospital-wide trend or an independent effect.

Evaluating different (indirect) infection prevention quality indicators

To evaluate the effect of the Infection Prevention Unit, different (indirect) quality indicators

were chosen that were objectively recorded over the years and that were available for data

analysis. Four quality indicators were used: i) the total amount of hand disinfection alcohol; ii)

the number of surveillance cultures performed; ii) the total amount of disposables; iv) the

results of point-prevalence studies for adherence to the dress codes. Disposables taken into

account were: non-sterile gloves; coats; caps; aprons; and mouth-nose masks. All disposables

are used mainly or exclusively for infection prevention and control measures as stated in the

hospital’s local guidelines. Surveillance cultures were defined as the following swabs: nose,

nose/throat, nose/perineum, nose/throat/perineum, throat/rectum, and rectum. These were

done in general for MRSA risk patients as defined by the Dutch MRSA guideline (Werkgroep

Infectie Preventie, 2012) and for certain other (resistant) microorganisms depending on the

ward and/or patient. The point-prevalence studies were hospital-wide studies performed five

times in 2012 and 2013. For each department it was scored how the personnel adhered to the

dress code guidelines. Six parameters were scored: no nail polish; correct hair-do; no jewellery;

closed coats; no long sleeves (protruding from under the short sleeves from white coats); and

correct outfit (long white coat or short white coat plus white trousers).


Incidence of risk microorganisms

The incidence per 1000 admissions of colonization of nine different (multi drug resistant)

microorganisms, for which the propensity for nosocomial outbreaks is known, was evaluated.

We choose to look at the resistant strains of the so-called ESKAPE organisms plus two extra

local additions: Methicillin-Resistant Staphylococcus aureus (MRSA); Extended-Spectrum ß-

lactamase (ESBL) and/or carbapenemase-producing Klebsiella pneumonia; multi-resistant

Pseudomonas aeruginosa; Vancomycin-Resistant Enterococcus faecium and Enterococcus faecalis (VRE);

Serratia marcescens; Acinetobacter baumanii; and Norovirus (requiring a positive PCR and

complaints). All positive cultures in the microbiological database were evaluated over the

respective eight years. Enterobacter spp. culture results were not available completely for the

whole eight years and could therefore not be included. Duplicate isolates for individual patients

were removed.

Determining the predicted and observed colonized patients during nosocomial

outbreaks

For the evaluated time period all outbreaks at the hospital as defined as such by the Infection

Prevention Unit were evaluated. An outbreak was defined as spread (i.e. colonization) of the

same strain of a microorganism (confirmed by either genotyping methods [e.g. MLVA, spa

typing] or Whole Genome Sequencing from 2012 on) among patients or personnel, with an

epidemiological link between the positive cases (e.g. same ward, same room), within a specific

time-frame (depending on the type of microorganism). All outbreaks were handled at the time

of occurrence by the Infection Prevention Department and total numbers of positive patients

were counted for all outbreaks.

Based on the number of risk microorganisms we calculated the incremental (yearly) rise/drop

in terms of percentage and used these to calculate the predicted (yearly) rise/drop for the

number of outbreak patients, giving a (yearly) predicted number of outbreak patients which

can be compared to the actual (yearly) found number of outbreak patients. Considering the

possible effect of the average LOS in the hospital on the incidence, the total number of

predicted outbreak patients has been corrected for a change in LOS.

Calculating a return on investment (ROI)

Using the data from nosocomial outbreaks at the same hospital, an average duration per

colonized patient and price per colonized patient per outbreak day was calculated. A cost-

analysis was performed for seven outbreaks with pathogenic microorganisms (e.g. MRSA,

ESBL Klebsiella pneumoniae, and VRE) that occurred between 2012-2014 (the included seven are

therefore also part of all investigated outbreaks in this study). The calculated median cost per

119 | Chapter 9

patient per outbreak day over the seven different outbreaks in 2012 price level was €499 and

the weighted mean duration was 36.9 days (Dik, et al., 2016a). To account for the variation

within this cost-analysis, we took the medians instead of the average to be on the conservative

side. These data were used for the rest of the financial evaluation.

ROI has been calculated for each year by taking the total amount of yearly

investments into infection prevention and dividing those by the yearly incremental costs or

benefits. This yielded a yearly ratio that represents the amount of costs or benefits for each

Euro invested. To calculate the ROI, the total budget of the Infection Prevention Unit was

determined at 1.5 million Euros for 2012. Due to the expected rise in risks and resistance, the

budget has been increased every year due to proactive investments into infection prevention.

Infection prevention is an integral part of the Department of Medical Microbiology within this

hospital and budgets are combined, making yearly detailed discrimination of costs (e.g.

overhead costs) not always possible. Therefore, an estimated budgetary increase was made of

€100,000 per year, taking into account the increased number of personnel. This is most likely

an overestimation, which implies that our incremental cost-benefit approach can be considered

conservative. Furthermore, the incremental change in costs of the consumables used for

infection prevention on the wards was added to the respective yearly investments.

Statistics and calculations

To analyze the trends over the years and calculate if there was a significant rise, decrease or no

change, we performed univariate linear regression analyses for each single variable against time

(eight years). Because numbers can fluctuate during years, we chose to evaluate a period of

eight subsequent years to level out potential outliers. To analyze a correlating effect between

the number of surveillance cultures and microorganisms found, binary logistic regression

analyses were performed. To correct for a possible ascertainment bias on the incidence of the

microorganisms in relation with the average LOS of the hospital, positive patients were

analyzed. Day of first positive culture was scored (thus taking into account the moment of

positivity) and a formula was plotted to calculate the cumulative incidence over time. For each

year the average LOS in the hospital was compared to the first year (2007) and with the

formula the difference in cumulative incidence was calculated and subtracted from the total

percentage. A univariate sensitivity analysis was performed to evaluate the impact of some of

the parameters. A single parameter was varied ±25% with the rest of the parameters at their

baseline level. A significance level of p ≤ 0.05 was applied. All calculations were done with

Microsoft Excel (Microsoft, Redmond, WA, USA) and SPSS (IBM, Amonk, NY, USA).


Results

The hospital had a growing Infection Prevention division

For the hospital, the total amount of infection prevention personnel was inventoried per year.

The number includes infection prevention specialists, clinicians and supporting personnel (e.g.

researchers). From 2007 to 2014 the unit saw a 65% increase (p = 0.025). The total number of

nursing staff per hospital bed remained stable (average of 46.03 FTE/1000 admissions), with

no significant change (p = 0.167). The average LOS showed a statistically significant

downward trend (p < 0.01) (see Table 9.1).

Coincidently infection prevention policy quality indicators improved

To evaluate the impact of the increased infection prevention staff on adherence to infection

prevention protocols within the hospital, four quality indicators were measured: the number of

surveillance cultures; the amount of hand disinfection alcohol; use of disposables (i.e. gloves,

coats, caps, aprons and masks); and the adherence to the dress codes. All quality indicators

showed a significant change over time (Figure 9.1A-9.1C and Table 9.1). Compared to 2007, in

2014 use of hand alcohol went up by 43%, surveillance cultures by 131%, and total amount of

disposables by 69% (only the use of caps decreased, the rest of the disposable increased in

use). The adherence to the dress codes was measured five times in a hospital-wide point-

prevalence study in the last two years and adherence went up by 25%. Six factors were looked

at: no nail polish; correct hair-do; no jewellery; closed coats; no long sleeves (protruding from

under the short sleeves from white coats); and correct coat (Figure 9.1D and Table 9.1).

The hospital faced an increase in risk microorganisms

All cultures within the evaluated period of eight years were taken into account and positive

cultures were counted for risk microorganisms, whereby duplicate isolates from individual

patients were excluded. Over the eight years, there was an increased incidence for all

microorganisms within the hospital (Figure 9.2 and Table 9.1). Only P. aeruginosa and S.

marcescens were found to correlate significantly with the total number of surveillance cultures (p

= 0.007 and p = 0.037, respectively). The total number of positive cultures was corrected for

this correlation. Considering the significant reduction in average LOS in the hospital that was

seen over the eight years, the total number of risk microorganisms was corrected for this drop.

121 | Chapter 9

p-v

alu

e

N/

A

P =

0.1

0

P <

0.0

1

P =

0.1

7

P =

0.0

3

P =

0.0

2

P <

0.0

1

P <

0.0

1

P =

0.0

3

P <

0.0

1

P <

0.0

1

P =

0.0

1

N/

A

N/

A

N/

A

N/

A

N/

A

N/

A

N/

A

N/

A

N/

A

N/

A

N/

A

a) D

iffe

rence

bet

wee

n f

irst

an

d las

t ye

ar.

b)

Per

1000 a

dm

issi

on

s

c) D

ue

to u

nav

aila

bili

ty o

f d

ata

fro

m 2

007 a

nd

2008 d

iffe

ren

ce p

rese

nte

d h

ere

is b

etw

een

2009 a

nd

2014.

d)

Aver

age

per

centa

ge p

er y

ear

bas

ed o

n m

ult

iple

po

int-

pre

val

ence

aud

its.

e) M

DR

is

resi

stan

ce f

or

at lea

st t

hre

e of

the

follo

win

g: c

efta

zid

ime,

mer

op

enem

, ci

pro

flo

xac

in, p

iper

acill

in-t

azo

bac

tam

, ge

nta

mic

in.

ESB

L: E

xte

nd

ed s

pec

trum

bet

a-la

ctam

ase;

KP

C: K

lebs

iella

pro

duci

ng

carb

apen

amas

e; L

OS: L

engt

h o

f st

ay; M

DR

: M

ult

iple

dru

g re

sist

ant;

N/

A: N

ot

app

licab

le; V

RE

:

Van

com

ycin

-res

ista

nt

Ent

eroc

occu

s.

Δ 2

007-2

014

a

0%

6%

-11%

10%

65%

43%

130%

80%

-26%

75%

33%

132%

-

39%

1112%

157%

∞

-11%

∞

48%

48%

-39%

Tab

le 9

.1:

Su

mm

ary

of

all

data

cate

go

rized

per

year.

2014

1339

34,6

71

8.8

7

50.6

8

0.4

2

410

4214

8140

2909

10055

6608

498

79%

1.7

3

1.8

5

0.5

7

2.3

4

3.1

4

0.0

3

4.4

1

2.3

0

0.9

5

2013

1339

37,2

49

8.2

1

46.4

4

0.3

1

361

2802

7599

2493

7928

5973

367

62%

1.4

2

2.0

7

0.6

6

0.7

2

3.0

2

0.0

0

4.7

5

2.0

5

1.9

6

2012

1339

36,6

95

8.3

4

45.7

6

0.2

8

329

2603

6549

2327

8170

5136

329

-

1.5

0

2.6

7

0.6

7

0.7

9

3.8

6

0.0

0

3.1

3

2.0

5

1.1

2

2011

1339

36,8

92

8.7

1

43.9

4

0.2

6

312

2513

6749

2953

6562

4993

288

-

1.1

9

2.3

3

0.6

8

2.6

0

3.1

6

0.0

0

4.0

1

2.3

2

0.9

5

2010

1339

35,8

42

8.9

6

44.9

9

0.2

5

288

2144

5595

2741

6796

4581

308

-

1.0

6

1.5

3

0.4

6

3.1

2

3.7

5

0.0

0

4.3

8

2.4

0

4.0

2

2009

1339

35,4

12

9.1

8

44.6

4

0.2

3

288

2036

4520

2909

6676

4909

270

-

0.9

9

1.4

7

0.4

7

1.3

3

3.1

9

0.0

3

2.8

8

1.8

9

0.9

6

2008

1339

34,4

11

9.5

8

45.7

6

0.2

6

277

1872

-

3982

6048

4653

193

-

1.2

8

0.6

1

0.5

3

0.0

0

3.6

1

0.0

3

3.0

7

1.7

1

0.2

6

2007

1339

32,8

31

9.9

4

46.0

8

0.2

5

287

1835

-

3911

5744

4975

214

-

1.2

5

0.1

5

0.2

2

0.0

0

3.5

4

0.0

0

2.9

8

1.5

5

1.5

5

Ou

tco

me m

easu

res

Ho

spit

al b

eds

(n)

Ad

mis

sio

ns

(n)

Aver

age

LO

S (

day

s)

Nurs

ing

per

son

nel

(n

)b

IP p

erso

nn

el (

n)b

IP q

uali

ty i

nd

icato

rsb

Han

d a

lco

ho

l (l

)

Glo

ves

(n

)

Co

ats

(n)c

Cap

s (n

)

Ap

ron

s (n

)

Mas

ks

(n)

Surv

eilla

nce

cult

ure

s (n

)

Dre

ss c

od

e ad

her

ence

d

Incid

en

ce r

isk

org

an

ism

sb

Sta

phyl

ococ

cus

aure

us (

MR

SA

)

Kle

bsie

lla p

neum

onia

(E

SB

L a

nd K

PC

)

Pse

udom

onas

aer

ugin

osa

(MD

R)e

Ent

eroc

occu

s fa

ecia

um/E

nter

ococ

cus

faec

alis

(V

RE

)

Ser

ratia

mar

cesc

ens

Acine

toba

cter

bau

man

ii

No

rovir

us

Pre

dic

ted

outb

reak

pat

ien

tsb

Ob

serv

ed o

utb

reak

pat

ien

tsb


↑ Figure 9.1: Quality indicators of infection

prevention guidelines. 10.1A: the yearly amount

of purchased disposables, whereby the coats were

only available from 2009 onwards. 10.1B: the

yearly amount of purchased hand disinfection

alcohol in liters. 10.1C: the yearly number of

performed surveillance cultures. 10.1D: the results

of 5 point prevalence studies on the adherence to

the dress codes. Data from A, B and C are

depicted as numbers per 1000 admissions.

← Figure 9.2: Incidence of risk micro-

organisms. Yearly incidence from 2007 until

2014 of the risk microorganisms as defined within

the graph. Data is plotted as colonized patients

per 1000 admissions. Most upper line represents

the total amount; the dotted line represents the

total amount of multi drug resistant organisms.

NB: colonized in the case of norovirus means

having a positive PCR and complaints; the peak in

2010 is caused by a VRE outbreak, which was

considered unrepresentatively large. MRSA:

Methicillin resistant Staphylococcus aureus; ESBL:

extended spectrum beta-lactamase; KPC: Klebsiella

producing carbapenamase; CFTA: ceftazidime;

MERO: meropenem; CIPR: ciprofloxacin; PITA:

piperacillin-tazobactam; Genta: gentamicin; VRE:

vancomycin resistant Enterococcus.

123 | Chapter 9

Figure 9.3: Univariate sensitivity analysis. Three parameters were varied with 25% and the average

yearly ROI is shown centered around the average yearly ROI found with the baseline values (6.26). The

assumed €100,000 yearly increase in infection prevention costs, the €525 costs of one outbreak day per

patient and the number of predicted outbreak patients were varied. All of them gave positive average

yearly ROI. Colored bars are +25%, open bars -25%.

The number of involved patients during nosocomial outbreaks remained at a stable level

For all nosocomial outbreaks within the hospital, the number of colonized patients were

counted and evaluated. The trend remained constant over the eight evaluated years in contrast

to the yearly incremental increase in risk microorganisms (see Table 9.1).

This difference in expected positive patients and identified positive outbreak patients

prevented substantial costs

The incremental yearly difference in risk microorganisms over the last eight years was used to

predict the number of outbreak patients and compared to the actual observed outbreak

patients. Using the previously calculated median cost of one extra outbreak day per positive

patient (€499) (Dik, et al., 2016a), the number of prevented incremental outbreak patients was

quantified in a monetary value, giving a yearly value of incremental costs or savings (Table 9.2).

On average this gave a yearly ROI of 1.94 (median: 3.87) ranging from -8.74 to 6.77 (Table

9.2). The univariate sensitivity analysis showed, with one exception, only positive average yearly

ROIs, ranging from -0.41 to 4.28 (Figure 9.3).


Discussion and conclusions

This study investigated the Infection

Prevention Unit of a Dutch academic

hospital retrospectively during an eight-

year period from 2007 to 2014. During

these years the number of resistant

microorganisms rose within the hospital,

most likely due to increased admission of

patients carrying MDRO. This trend is

seen in the Netherlands as well in Europe

and the rest of the world (European

Centre for Disease Prevention and

Control, 2013). This pressure is further

increased by a high connectivity between

hospitals, with academic centers (such as

this one) often acting as a central hub

(Donker, et al., 2010). Partly as a

response to this rising pressure, there was

a significant rise in infection prevention

personnel. Coincidently with this rise,

several (indirect) quality parameters for

infection prevention protocols also saw a

rise. We hypothesize that more infection

prevention personnel leads to increased

awareness and better adherence to the

different guidelines. Proving a direct

correlation is difficult. However, the rise

in the eight different quality parameters

does show undoubtedly that more

protocol actions are performed. The

question remains if they were performed

correctly. During the eight years there

was also a considerable drop in the

observed outbreak patients when

comparing them with the predicted

numbers. Also here, proving a direct

correlation is difficult, but by looking at

multiple parameters and correcting for

several confounders, we tried to Tab

le 9

.2:

Incre

men

tal

co

sts

an

d b

en

efi

ts.

RO

I

-

5.8

4

2.1

0

-8.7

4

3.8

7

4.4

7

-0.7

3

6.7

7

R

OI:

Ret

run o

n in

ves

tmen

t.

Dif

fere

nce

infe

cti

on

pre

ven

tio

n

co

sts

(€)

-

782,3

41

409,9

12

-1,2

07,2

57

742,5

49

515,8

73

-162,0

72

748,3

33

Incre

men

tal

rise

in

fecti

on

pre

ven

tio

n

bu

dg

et

(€)

-

134,0

49

195,5

15

138,1

30

191,8

43

115,4

91

222,0

46

110,5

26

Dif

fere

nce

ou

tbre

ak

s

co

sts

(€)

-

958,9

12

633,5

20

-1,1

18,7

36

977,7

49

660,6

60

62,7

56

898,7

11

Dif

fere

nce i

n

ou

tbre

ak

s

pati

en

t d

ays

-

1826

1207

-2131

1862

1258

120

1712

Ob

serv

ed

ave

rag

e

pati

en

t d

ays

1872

330

1248

5285

1285

1505

2679

1211

Pre

dic

ted

ou

tbre

ak

pati

en

t d

ays

1872

2157

2455

3154

3147

2763

2799

2923

Year

2007

2008

2009

2010

2011

2012

2013

2014

125 | Chapter 9

minimize the risk of a biased conclusion. The observed drop in outbreak patients can be

quantified financially. In an earlier study, looking at seven different outbreaks (using the same

definition as in this study) in the same hospital, a weighted mean duration of an outbreak and a

median price per patient per day was calculated (Dik, et al., 2016a). Using these data, this study

shows, as one of the first, the positive incremental bundle effect of an Infection Prevention

Unit at an academic hospital. Nosocomial outbreak patients were prevented and the yearly

extra investments that were done to keep up with the rise in high-risk microorganisms as well

as the rise in antimicrobial resistance levels had an overall positive return on investment (ROI).

This was evaluated over a period of eight subsequent years, to rule out large influences of

potential positive or negative outliers (i.e. years).

The Netherlands has a good infection prevention track record. High screening rates and the

proactive search-and-destroy policy for MRSA are good examples of this (Bode, et al., 2011).

In total there are 78 different national infection prevention protocols (http://www.wip.nl) that

are keeping prevalence of risk microorganisms low. Even so, the rise in resistant

microorganisms worldwide is reflected in the Dutch situation as well (European Centre for

Disease Prevention and Control, 2013). It is therefore important to anticipate for higher

prevalence of MDROs by investing in infection prevention. For this hospital, the present

number of staff per bed seems to be sufficient for a low incidence country and the rise in

infection prevention specialists seems to be able to keep up with the rise in MDROs. Although

outbreaks did occur more frequently, the number of positive patients per outbreak dropped,

possibly indicating that outbreaks are more quickly contained (data not shown). This appears

to be case nationwide in the Netherlands (van der Bij, et al., 2015) and corroborates that the

(national and regional) approach to infection prevention is the main contributor. Of course,

these results can only be obtained by having the proper facilities and staff in a healthcare

institution (Dik, et al., 2016b) and proper harmonized education (Zingg, et al., 2015).

Up to 2007, the Dutch national norm for hospitals was 1 FTE of infection prevention

specialist per 250 beds (5.4 FTE for this hospital) (VHIG, 2006). This changed to 1 FTE per

5000 admission from 2007 onwards (6.6 FTE for this hospital) (van den Broek, et al., 2007).

This is comparable to Australia (Mitchell, et al., 2015), and slightly lower than US data (Stone,

et al., 2014). The updated Dutch norm in 2012 did not increase number of personnel, but

focused more on quality of their work and correct implementation of different guidelines

(NVMM, 2012a; Spijkerman, et al., 2012a; Spijkerman, et al., 2012b). This hospital employed

even more personnel (from 2011 on), although the definition of infection prevention

personnel used in this study is somewhat broader (including supporting staff (e.g. mathematical

modellers) as well; but no laboratory technicians and medical specialists mainly active in

diagnostic and/or antimicrobial stewardship). Besides infection prevention personnel, it is

known that the number of nurses per hospital bed and their workload also influences patient


safety and infection rates (Hugonnet, et al., 2004; Rogowski, et al., 2013; Stone, et al., 2007). A

change in the nurse per admission ratio could thus be a confounding factor. It was therefore

taken into account in this study. There was a stable rate over the eight years and National

Dutch numbers for academic hospitals showed similar stable trends

(http://www.dutchhospitaldata.nl). Infection rates most likely increase when length of stay

(LOS) increases, a so-called ascertainment bias if not taken into account. Therefore also this

factor was evaluated. Average LOS in the hospital showed significant drop over the years,

most likely due to other improvements in healthcare and logistics (although this was not

examined in detail as it was not part of the study), and the predicted outbreak patient number

was consequently corrected for this drop. Indirect measurement of infection prevention

through the amount of hand alcohol or disposables (product volume measurement [PVM]) is

done before (Bittner and Rich, 1998; Chakravarthy, et al., 2011). It is a relatively easy method

and does not require labour intensive audits. It is unfortunately not ideal because it does not

provide information on the way these consumables are used. However, audits are not always

an option, certainly not when analysing data retrospectively. We therefore feel that by

evaluating not just hand alcohol, but a set of parameters, that all showed similar trends, the

causal relation we hypothesize is plausible. Finally, to account for the possible effect that a

change in assumed parameters has, a sensitivity analysis was performed.

We show that savings can be achieved by investing in the prevention of outbreaks and thus

preventing specific measures or actions such as closed wards (reducing the revenue

opportunities) and usage of personnel (creating extra costs by hiring temporary staff or

opportunity costs by redistributing existing staff). Savings mentioned in this study are based

upon the previous cost-analysis of outbreaks within the same hospital. Depending on the

outbreak, different savings will be achieved, but in general it can be expected that savings are

more or less similar distributed as within the cited study (i.e. 50% is saved on prevention of

bed/ward closure, 17% on microbiological diagnostics, 11% on isolation, 10% on extra

personnel, 5% on cleaning and 7% on other expenditures) (Dik, et al., 2016a). Part of these

savings are direct (e.g. diagnostics, extra personnel) and part are indirect such as preventing the

closure of beds or opportunity costs due to redistributing personnel, which in turn can lead to

an increase in revenue and improved quality of care. In both studies, infection rates were low

and most patients were only colonized, making the impact on medication costs small. Included

parameters are however not a complete overview of all costs and benefits. Considering the

numerous other positive effects of an infection prevention department on nosocomial

infections, other small non-outbreak situations and antimicrobial resistance rates, potential

benefits might be substantially higher. Furthermore, expected benefits from a societal

perspective are even bigger, especially if looking to the increasing problems of antimicrobial

resistance. Taken together, it is a bundle of capable and sufficient infection prevention

personnel, and proper protocols that are correctly followed. By doing so, the hospital can

prevent outbreak patients thereby improving patient safety. In conclusion, infection prevention

will usually save a sufficient amount of resources to become highly cost beneficial.

128

Performing timely blood cultures in patients receiving

IV antibiotics is correlated with a shorter length of stay

Jan-Willem H. Dik, Alex W. Friedrich, Jerome R. Lo-Ten-Foe, Job van der Palen,

Maarten J. Postma, Sander van Assen, Ron Hendrix* & Bhanu Sinha*

*Contributed equally

Submitted

10

129 | Chapter 10

Abstract

Performing blood cultures is part of correct antimicrobial use and mentioned in

all antimicrobial stewardship guidelines. However, precise effects are unknown.

This study used a large patient cohort of almost 3000 patients over a 5-year time

period to diminish this knowledge gap.

Patients receiving more than 2 days of intravenous antibiotics started at

admission to a Dutch academic hospital were selected. We evaluated if

performing blood cultures around start of therapy correlated with our primary

outcome measure, length of stay (LOS). Effects on therapy and on costs were

also evaluated. Finally, a similar analysis was done for a community hospital.

Of 2997 included patients, 48% (1441) had blood cultures performed around

the start of antibiotic therapy. The group with blood cultures had a significantly

shorter LOS compared to patients without (13.0 vs. 14.0 days; p=0.017). In a

multivariate regression model, this positive correlation of blood cultures

remained. There was also a strong correlation with clinical chemistry orders,

suggesting a bundle effect, as well as a shorter duration of antibiotic therapy.

Effects within the community hospital were less pronounced, most likely due to

an already low baseline LOS.

Patients with blood cultures performed, as part of a diagnostic bundle, had

significantly shorter duration of therapy and LOS with consequent financial

benefits. Increasing the rate of blood cultures taken prior to starting antibiotic is

a useful target for antimicrobial stewardship programs in order to improve

quality of patient care, patient safety and can be financially attractive.

Performing timely blood cultures in patients receiving IV antibiotics is correlated with a shorter length of stay | 130

Introduction

Antimicrobial resistance is a worldwide health problem and appropriately prescribing of

antimicrobials is crucial to curb resistance development. In order to provide optimal therapy,

performing timely diagnostics is crucial. Diagnostics are therefore mentioned in all

antimicrobial stewardship guidelines (Barlam, et al., 2016; de With, et al., 2016; van den Bosch,

et al., 2015) and are a cornerstone of management of patients with severe infections. Two (or

more) sets of blood cultures prior to starting antimicrobial therapy are the diagnostic standard.

Besides antimicrobial stewardship guidelines, also protocols for patients with infections

invariably ask for (multiple) sets of blood cultures besides other cultures and other diagnostic

modalities such as clinical chemistry tests (a diagnostic bundle), before starting antimicrobial

therapy (Dellinger, et al., 2013; Habib, et al., 2015; Tunkel, et al., 2004; Woodhead, et al., 2011).

However, as part of a bundle, effects of blood cultures on patients who receive antimicrobial

therapy are difficult to quantify resulting in a lack of data and a knowledge gap (Schuts, et al.,

2016). Therefore, as a first step towards better (prospective) evaluation studies, we comprised a

large retrospective observational cohort study looking at effects of blood cultures performed in

patients receiving multiple days of intravenous antimicrobial therapy.

Performing appropriate and timely (microbiological) diagnostics not only helps making a timely

correct diagnosis, but can also improve antibiotic therapy (Buehler, et al., 2016). This supports

patient-tailored decisions on appropriate therapy, which in turn reduces overall collateral

damage, such as toxicity and resistance development (Dik, et al., 2016; Goossens, 2009). Due

to faster results, innovative rapid diagnostics can improve quality indicators such as length of

stay (LOS) and costs (Barlam, et al., 2016; Huang, et al., 2013; Perez, et al., 2014). Recently, we

have shown that even preliminary results of microbiological diagnostics after 48 hours are of

great value in a bedside case-audit model for patients receiving intravenous antibiotics. These

case-audits were very effective in reducing length of stay in a major proportion of patients, and

the microbiological diagnostics (positive or negative) were considered a crucial part of the

diagnostic work-up (Dik, et al., 2015). Thus, data seems to support the hypothesis that blood

cultures can have a positive impact on patients receiving antimicrobial therapy, although this

still has to be evaluated.

Despite the requirement of blood cultures for patients receiving antimicrobial therapy in all the

different guidelines and protocols, there are numerous studies showing that adherence to

guidelines is limited: in only 50% of the cases blood cultures were obtained for community

acquired pneumonia (CAP) in the UK (Collini, et al., 2007) and the EU (Reissig, et al., 2013),

and for bacterial meningitis in the US (Chia, et al., 2015). In the Netherlands non-adherence in

general to guidelines for antibiotic treatment occurs frequently and results in a higher than

necessary use of broad-spectrum antibiotics (van der Velden, et al., 2012).

131 | Chapter 10

Lack of data and low guideline adherence clearly show the need for investigating the effects of

diagnostics. Here, we show the results of a large data set consisting of almost 3000 patients

over a 5-year time period, where we looked at effects of blood cultures on a subset of outcome

measures as well as factors that contribute to the drawing of blood cultures. A second group of

patients was admitted to a neighbouring community hospital in the Netherlands, and a

comparable analysis was performed. This study provides an excellent starting point for actions

targeted to improve the quality of antimicrobial usage (e.g. in an Antimicrobial or Diagnostic

Stewardship Program), hospital efficiency and patient safety.


The study was performed retrospectively in a large 1339-bed academic tertiary referral hospital

in the north of the Netherlands for five years (2009-2013), and at a 284-bed community

hospital within the same region of the country where four years of data (2010-2013) was

available. Data were retrieved from the hospitals’ electronic databases. These databases

provided clinical data (admission dates, gender, date of birth, hospital department, ICD9

codes, mortality date [if applicable] and discharge date), pharmacy data (antimicrobial

prescriptions), clinical chemistry data (C-reactive protein [CRP], leucocytes, and estimated

glomerular filtration rate [eGFR]), and microbiological data (blood cultures).

From these data all patients admitted to the hospital receiving intravenous antibiotics

for more than 48 hours (thus excluding single day prophylactic use) starting at the day of

admission (within a timeframe of 24 hours) were selected. The following antibiotic therapies

were included as they were the most frequently prescribed for therapeutic purposes within the

hospital during the evaluated years: amoxicillin/clavulanic acid, cefuroxime, ceftriaxone,

piperacillin/tazobactam, ciprofloxacin, clindamycin, amoxicillin, amoxicillin/clavulanic acid +

ciprofloxacin, and meropenem. Prophylactic therapy was excluded. Patients under 18 years and

those admitted to haematology or otorhinolaryngology were excluded, because these groups

have specific protocols regarding blood cultures and/or antibiotic treatment. Included patients

were stratified further into two groups: those where blood cultures were obtained for the first

time on the day of admission (within a timeframe of 24 hours) and those without blood

cultures during this timeframe. Performance of clinical chemistry data (CRP, leucocytes, and

eGFR) was evaluated using the same criteria as for blood cultures for both groups.

To rule out any potential biases that would result into non-comparable groups of patients (e.g.

a more severe group with cultures compared to a less severe group without) we looked at in-

hospital mortality, treating speciality, the ward of admission and the values of the clinical

chemistry data at admission. Furthermore, it should be noted that the main evaluation of this


study regards tertiary referral patients, which are in general already of a relatively high

complexity.

Primary outcome measure was length of stay (LOS) and potential correlating factors. Besides

LOS we looked at duration of antimicrobial therapy and at costs. For the latter, a hospital

perspective was used and all prices were converted to 2013 price level using the Dutch

consumer price index. Dutch reference price for a bed day (€640) was used (Hakkaart-van

Roijen, et al., 2010). We assumed that 2 sets of blood cultures per patient had been taken at a

price of € 25 per set, which comprises all in costs (personnel, materials, depreciation etc.). We

assumed a positivity rate of 10% of blood cultures resulting in an additional €30 for

determination and susceptibility testing.

Statistics were performed with SPSS version 22 (IBM, Armonk, NY, USA). P-values <0.05

were considered significant. For inclusion into the multivariate regression model, we adopted a

p-value of <0.1. First, a Kaplan-Meier test was performed to examine an initial difference in

LOS between the patients with and without blood cultures, whereby patients who died within

the hospital were censored. Second, we examined which parameters correlated with drawing of

blood for cultures. For nominal parameters this was tested with a Chi Square test and for

continuous variables with a Student’s t-test (if normally distributed) or with a Mann-Whitney U

test (if not normally distributed). Third, we examined which further factors correlated with

LOS. For each included nominal parameter a Kaplan-Meier test was performed with LOS,

while in the case of continuous variables an ANOVA was performed. Furthermore,

proportional risks were visually checked for the different variables. Finally, all parameters that

had a significant correlation with drawing blood cultures and with LOS were included in a Cox

Regression Model as possible confounders, whereby in-hospital mortality was censored again.

Subsequently, variables were removed, one by one, based on their significance and taking into

account a change in -2 log likelihood, until a parsimonious model was obtained, or until the

coefficient for drawing of blood for cultures changed by > 10%. To further look into which

factors correlated with drawing of blood cultures, a multiple logistic regression analysis was

performed, which included all parameters that showed a significant correlation with drawing of

blood cultures.

A comparable analysis was performed in a similar cohort of patients receiving intravenous

antibiotics during admission at the community hospital with similar inclusion criteria. Included

antibiotics were: amoxicillin/clavulanic acid, amoxicillin/clavulanic acid + ciprofloxacin,

ciprofloxacin, piperacillin/tazobactam, clindamycin, ciprofloxacin + clindamycin, levofloxacin,

cefuroxime and ceftriaxone.

133 | Chapter 10

Data was merged based upon patient numbers. Personal details such as names and dates of

birth were not available to guarantee anonymity. The local ethics committee waived consent,

due to the retrospective analytic nature of the study without active intervention (Ref: 2014/530

METc UMCG).

Results

Patients and their characteristics

The group of evaluated patients in the academic hospital consisted of 2997 patients receiving

IV antibiotics starting at the day of admission and lasting for multiple days. 1441 (48%)

patients had blood cultures taken during admission versus 1556 (52%) patients who had not.

In-hospital mortality was similar between those two groups (3.3% vs. 2.3%; p=0.114). Clinical

chemistry data showed statistically significant but minor differences, and mean values of both

groups were such, that the clinical relevance is disputable. See Table 10.1 for detailed

characteristics.

Length of stay (LOS) was significantly shorter for the group with blood cultures

As primary outcome measure, LOS was evaluated for the stratified groups (with vs. without

blood cultures taken). The mean LOS for the patients with blood cultures was significantly

lower than for the group without (13.0 vs. 14.0 days; p=0.017). This was further sub-stratified

into LOS per medical specialty to examine possible differences within the two groups (Table

10.1). The most positive effect of blood culturing on LOS was seen in patients of Internal

Medicine (13.0 days [1035 patients] vs. 15.8 days [593 patients]; p<0.001).

Performing blood cultures had a significant effect on reducing LOS in a multivariate

model

Variables that significantly correlated with drawing of blood cultures and significantly

correlated with the LOS were included in the multivariate Cox Regression as possible

confounders. Those were, besides having blood cultures or not: route of admission; antibiotic;

medical specialty, week or weekend admission, age, gender and measuring leucocytes. All of

them showed significant associations with LOS within the final multivariate Cox regression

model (Table 10.2). Table 10.3 shows the variables associated with drawing of blood for

cultures within a multivariate logistic regression model.


A diagnostic bundle of blood cultures and clinical chemistry diagnostics had the highest

effect on LOS

The multiple regression model on drawing blood for cultures showed a strong association with

clinical chemistry diagnostics (CRP, leucocytes, eGFR), suggesting a bundle effect of blood

cultures and clinical chemistry diagnostics (Table 10.3). We therefore compared patients with

both blood cultures and all three clinical chemistry diagnostics to patients who had just one of

both components, and examined the effect on LOS. Patients receiving a bundle had the

shortest LOS (12.6 days [95%CI:11.3-13.9]), compared to blood cultures only (13.4 days [95%

CI:9.7-17.1]) and clinical chemistry only (14.4 days [95%CI:13.3-15.5]). Only the difference

between the first and the last was significant (p=0.001).

Table 10.1: Patient’s characteristics for the academic hospital. Possible confounding variables were tested and included in the final multivariate Cox Regression model. With blood

cultures Without blood

cultures p-value

Patients [n] 1441 (48%) 1556 (52%) Medical discipline <0.001a

Surgery 283 (20%) 700 (45%) Internal medicine 1035 (72%) 593 (38%)

Other 123 (9%) 263 (17%) Median age [yrs] (IQR) 62.40 (42.4-86.1) 58.60 (35.6-85.4) <0.001b

Percentage males 58% 55% 0.070a

In-hospital mortality 3.3% 2.3% 0.114a

Median CRP [mg/L] (IQR) 92 (40-178) 73 (22-146) <0.001b

Median eGFR [m/m 1.73] (IQR) 74 (50-99) 89 (62-112) <0.001b Median leucocyte count [10E9/L] (IQR) 12 (8-16) 11 (8-15) 0.021b

Mean LOS [d] (95% CI) 13.0 (12.2-13.7) 14.0 (13.3-14.7) 0.017c

Surgery 12.2 (11.4-13.1) 12.1 (11.2-12.9) 0.732c

Internal medicine 13.0 (12.2-13.9) 15.8 (14.5-17.1) <0.001c

Other

12.5 (10.7-14.4)

15.0 (13.2-16.8)

0.113c

a) chi-square test

b) Mann-Whitney U test c) log-rank test CRP: C-Reactive Protein; eGFR: estimated Glomular Filtration Rate; LOS: Length of Stay; IQR: Inter-Quartile Ranges

135 | Chapter 10

Table 10.2: Multivariate Cox regression results. Multivariate regression analysis for relation with length of stay. A positive Hazard Ratio (HR) corresponds with earlier discharge. Factor Hazard Ratio 95% CI p-value

Age 0.99 0.99 - 1.00 <0.001

Males 0.86 0.80 - 0.93 <0.001

Blood cultures 1.11 1.02 - 1.21 0.011

Leucocytes 0.86 0.77 - 0.96 0.006

Weekend admission 1.20 1.08 - 1.31 <0.001

Route of admission to the hospital <0.001

via ER (n=990) 1 (Reference)

via GP (n=1510) 0.99 0.91 - 1.08 0.824

via out-patient clinic (n=202) 0.84 0.71 - 0.95 0.031

via unknown route (n=64) 0.87 0.67 - 1.12 0.285

via transfer from other hospital (n=231) 0.57 0.49 - 0.66 <0.001

Type of antibiotic 0.011

Co-amoxiclav (n=858) 1 (Reference)

Cefuroxime (n=724) 0.97 0.88 - 1.08 0.626

Ceftriaxone (n=294) 0.90 0.78 - 1.03 0.130

Piperacillin/Tazobactam (n=546) 0.85 0.76 - 0.95 0.005

Ciprofloxacin (n=130) 0.77 0.63 - 0.93 0.007

Clindamycin (n=130) 0.79 0.66 - 0.96 0.016

Amoxicillin (n=87) 0.90 0.72 - 1.14 0.382

Meropenem (n=150) 0.83 0.70 - 0.99 0.043

Co-amoxiclav + Ciprofloxacin (n=73) 0.83 0.64 - 1.06 0.128

Medical specialty <0.001

Surgery (n=983) 1 (Reference)

Internal medicine (n=1628) 0.80 0.73 - 0.87 <0.001

Other (n=386) 0.83 0.73 - 0.94 0.003

ER: Emergency Room; GP: General Practitioner


Table 10.3: Multiple logistic regression analysis on the relationship with blood cultures. A positive Odds Ratio (HR) corresponds with taking blood cultures at admission. Factor Odds Ratio 95% CI p-value

Age 1.01 1.00 - 1.01 0.047

Males 1.22 1.04 - 1.45 0.018

CRP 8.13 4.07 - 16.22 <0.001

eGFR 0.83 0.69 - 0.98 0.032

Leucocytes 0.52 0.26 - 1.04 0.064

Weekend admission 1.21 0.98 - 1.50 0.077

Broad spectrum 2.33 1.46 - 3.73 <0.001

Route of admission to the hospital <0.001

via ER (n=990) 1 (Reference)

via GP (n=1510) 0.72 0.60 - 0.87 <0.001

via out-patient clinic (n=202) 0.43 0.30 - 0.62 <0.001

via unknown route (n=64) 0.62 0.35 - 1.10 0.104

via transfer from other hospital (n=231) 0.17 0.11 - 0.25 <0.001

Type of antibiotic <0.001

Co-amoxiclav (n=858) 1 (Reference)

Cefuroxime (n=724) 2.28 1.81 - 2.86 <0.001

Ceftriaxone (n=294) 1.07 0.79 - 1.46 0.648

Piperacillin/Tazobactam (n=546) 1.90 1.49 - 2.43 <0.001

Ciprofloxacin (n=130) 2.41 1.59 - 3.65 <0.001

Clindamycin (n=130) 1.41 0.77 - 2.57 0.269

Amoxicillin (n=87) 1.77 1.07 - 2.93 0.028

Meropenem (n=150) 3.11 2.07 - 4.67 <0.001

Co-amoxiclav + Ciprofloxacin (n=73) 2.10 1.10 - 4.02 0.025

Medical specialty <0.001

Surgery (n=983) 1 (Reference)

Internal medicine (n=1628) 3.74 3.07 - 4.55 <0.001

Other (n=386) 1.41 1.06 - 1.88 0.019

ER: Emergency Room; GP: General Practitioner

137 | Chapter 10

Duration of antibiotic therapy was lower in the blood culture group

Total duration of antibiotic therapy was significantly lower for the group of patients who

received blood cultures at admission compared to patients without (9.8 days [95%CI:9.3-10.3]

vs. 11.0 days [95%CI:10.4-11.6]; p=0.030). Total antibiotic consumption in DDDs did not

differ significantly (18.01 [95% CI:16.68-19.34] vs. 20.46 [95% CI:18.89-22.03]; p=0.915).

Costs were lower in the blood culture group

Mean antibiotic costs were calculated for the two groups of patients. The group with blood

cultures had significant lower mean costs than patients without blood cultures (€61.58

[95%CI:€46.54 - €76.52] vs. €74.75 [95% CI:€53.55-€95.95]; p=0.002). Regarding LOS, the

1441 patients with blood cultures spent 1.01 day shorter in the hospital. Multiplying these

numbers with the Dutch reference bed price (€640) and adding the costs for the blood cultures

and antibiotics gave a total cost for the blood culture group of €12,140,227.08 and for the

group without blood cultures €14,018,237.40; a difference of €1,914,010.32 over five years (i.e.

a difference of €382,802.06 per year).

Table 10.4: Patient characteristics for the community hospital.

With blood cultures Without blood cultures p-value

Patients [n] 944 (57%) 720 (43%)

Medical specialty <0.001a

Surgery 325 (34%) 447 (62%)

Internal medicine 578 (61%) 209 (29%)

Other 41 (4%) 64 (9%)

Median age [yrs] (IQ range) 71.0 (48.3-93.7) 68.0 (43.4-92.6) 0.019b

Percentage males 51% 51% 0.927a

Mean LOS [d] (95% CI) 8.7 (8.3-9.2) 8.9 (8.3-9.5) 0.700c

Surgery 8.3 (7.5-9.1) 7.9 (7.3-8.6) 0.516c

Internal medicine 8.7 (8.2-9.3) 10.3 (9.1-11.5) 0.006c

Other 12.9 (9.2-16.6) 11.5 (8.8-14.3) 0.568c

a) chi-square test b) Mann-Whitney U test c) log-rank test


Effects within a community hospital showed similar trends, albeit with smaller effects

In order to investigate effects in a different setting, data from 1664 patients from the

community hospital during a four-year period (2010-2013) were evaluated similarly. 57% of

patients had blood cultures drawn at admission, whereas 43% had not (Table 10.4). Although

LOS was shorter for the group of patients with blood cultures, this difference was not

statistically significant (p=0.700). Therefore, a regression model analysis was not performed.

The sub-stratification for LOS per medical specialty (with blood cultures vs. without) showed

similar trends as for the academic centre with the biggest positive differences in LOS for

Internal Medicine (8.7 days vs. 10.3 days; p=0.006) compared to surgical and other disciplines

(Table 10.4). Mortality data was unfortunately not available.

Discussion

This study evaluates, to the best of our knowledge as one of the first, effects of obtaining

blood cultures for patients receiving multiple days IV antimicrobial therapy at hospital

admission. A suspected infection is the most probable presumptive diagnosis in these patients.

Analysis was performed retrospectively in an observational manner using a large cohort of

patients. Blood for cultures were drawn in only 48% of patients who received antibiotics

during admission, which is in accordance with other studies (Chia, et al., 2015; Collini, et al.,

2007; Reissig, et al., 2013). Length of stay (LOS) was one day shorter (-7%) for patients with

blood cultures compared to those without (p=0.017). Within a multivariate Cox regression

model, taking into account all known and available relevant confounders, the drawing of blood

cultures was still a significant variable associated with a reduced LOS. Patients with timely

blood cultures can be predicted to be discharged 1.11 times earlier than those without

(p=0.011). A reduction in duration of antimicrobial therapy was also seen (-11%; p=0.030),

however DDDs remained the same (p=0.915), suggesting that dosage increased for those

patients. Especially interesting is the fact that clinical chemistry diagnostics correlated

significantly with blood cultures, suggesting a bundle effect of diagnostics thereby supporting

medical decision making. We could corroborate this by looking at the LOS of patients with

blood cultures and CRP, leucocytes and eGFR (a bundle). These patients had significantly

shorter LOS compared to those in whom only clinical chemistry diagnostics were performed,

indicating that a bundle that included blood cultures was an important driver for reduced LOS.

We hypothesize that physicians who obtained blood cultures on time, according to existing

guidelines, are more likely to perform other relevant diagnostics also in a more appropriate

manner, leading to a more adequate clinical management for the patients due to better

diagnostics, therefore impacting LOS positively. Taking blood cultures on time might therefore

also be an easy to measure indicator for correct diagnostics during admission of patients with a

suspected infection.

139 | Chapter 10

Based on national and international guidelines regarding infections and antibiotic use, blood

cultures should have been drawn for all of these the patients (Barlam, et al., 2016; de With, et

al., 2016; Dellinger, et al., 2013; Habib, et al., 2015; Tunkel, et al., 2004; van den Bosch, et al.,

2015; Woodhead, et al., 2011). Thus, guidelines were presumably not properly followed for all

patients. Although comparable to other studies, the reasons behind this needs further

exploration to also adequately tackle this problem and improve compliance. The excess LOS

for patients without blood cultures suggests that an improvement in quality and efficiency can

be achieved, resulting in a better quality of care, which in turn leads to saved bed days. This is a

possibility for hospitals to increase revenue. In our case this could lead to possible yearly

benefits of nearly €200,000, already taking into account the extra costs of diagnostics. Costs for

clinical chemistry were not considered due to the large number of diagnostic tests and variation

between applied protocols, which might lead to an overestimation of financial benefits.

However, looking at the average price of approximately €2 per test, the financial impact of

these additional diagnostic tests would be rather small compared to the costs of excess LOS.

Possible revenue could be even greater when fewer procedures are performed due to earlier

diagnosis, making the return on investment of blood cultures even more prominent. This issue

should be subject to further investigation in a proper (prospective) economic evaluation.

This study is especially relevant for local and regional antimicrobial and/or diagnostic

stewardship programs since these results can be used as indicators for possible interventions,

stratified per ward, medical specialty, or type of antibiotic. The required contents of

stewardship programs are often not precisely specified. Indeed, depending on type of hospital,

antibiotic use, local/regional resistance levels, and other factors, different interventions for

different settings are required (although microbiological diagnostics are always mentioned).

Analyses such as the one performed here, are a relatively easy method for healthcare centres to

evaluate local characteristics and tailor their stewardship program in collaboration with all

clinical departments. The data shown here exemplifies that these programs should also focus

on start of antimicrobial therapy ensuring that correct diagnostics are performed. A simple

pop-up on the computer when ordering antibiotics, alerting the physician that (blood) cultures

should be obtained prior to starting antimicrobial therapy, is an example of a relative easy and

cheap way of doing so. In addition, electronic order entry could support syndrome-based

diagnostic bundles via one-click orders.

The analysis performed for the community hospital showed similar trends. However, no

significant general effect on LOS between suspected infectious patients with and without

blood cultures at admission were identified. Also in this setting, the largest effect of blood

cultures on LOS was visible for Internal Medicine. A possible cause of the lower difference is

that the average overall LOS at that hospital is already low in general (4.1 days compared to 8.7

days at the academic centre). Thus, possible room for improvement overall is more limited. It

appears that the minimum of LOS for the current conditions has already been achieved. Also


within the community hospital around 50% of the patients did not have any blood cultures

taken on admission, making directed antibiotic therapy impossible.

A limitation of this study, as in many studies on adequate antimicrobial therapy, is that

indications for the prescription (i.e. the presumptive diagnosis) were unknown. This could lead

to a bias in the inclusion, with more complex patients in one group compared to the other.

However, the vast majority of the patients are tertiary referral (having an already high

complexity) and furthermore, we observed similarities in mortality, treating specialty, wards,

and clinical chemistry values. Although the latter did differ on a significance level between the

two groups, clinical relevance due to the small differences in diagnostic values is disputable. All

in all, we could not find any indicators suggesting a bias. It is impossible to rule this out

completely. However, if there was such a bias, one would presume that group with blood

cultures which has a higher age and higher inflammation parameters would have a tendency

towards a longer LOS (which is also partly confirmed by the Cox regression). This clearly

contrasts the results observed here, where the cohort with blood cultures has a shorter LOS,

suggesting that the effect of blood cultures might even be dampened. Results from the cultures

were not taken into account, because both positive and negative results are clinically relevant in

the treatment of the patient. Also, both results could impact infection management and the

different outcome measures.

For certain indications the effects of correct (and timely) diagnostics according to protocol are

well studied. For blood stream infections (BSIs) for example, positive blood cultures are a

proven indicator (Dellinger, et al., 2013; Weinstein, et al., 1997). A review on community-

acquired pneumonia (CAP) showed that, following the guideline (which includes diagnostics),

improved patient outcomes and cost-effectiveness (Nathwani, et al., 2001). Furthermore,

patients with septic shock suffer increased mortality when standard protocols are not followed

(Gao, et al., 2005). Using checklists can help to improve patient care (Wolff, et al., 2004).

Dutch and international guidelines stipulate therefore that blood cultures should be drawn

from patients when sepsis is suspected or when a patient is diagnosed with severe CAP.

However, guidelines can differ for other indications and taking blood cultures for some

patients (e.g. pyelonephritis) is under discussion (Coburn, et al., 2012; McMurray, et al., 1997).

Furthermore, costs can be substantially higher due to the diagnostics performed extra and

other additional charges (Bates, et al., 1991; McMurray, et al., 1997). Our data suggest that even

for patients needing IV antibiotics in general, there is a clear beneficial impact on LOS from

correct on-time diagnostics.

Concluding, this study observed that patients receiving multiple days of IV antimicrobial

therapy on admission in which timely blood cultures were drawn have a reduced LOS

141 | Chapter 10

compared to patients with antimicrobial therapy but without those cultures. This is most likely

due to a bundle effect of both blood cultures and additional diagnostics. Blood cultures appear

to be an essential part of the bundle. Having these diagnostics available during treatment can

improve clinical management, makes it possible to better streamline therapy and therefore can

lead to reduced LOS. Furthermore, we corroborated that for almost half of the patients, blood

cultures were not performed. Considering the focus on optimal antimicrobial therapy to reduce

risks of side effects and resistance development, this should be addressed by local diagnostic

and antimicrobial stewardship programs. Correct diagnostics help guide therapy and contribute

to optimized use of antimicrobials. Finally, blood cultures as part of a diagnostic bundle can

have considerable financially benefits estimated almost €200,000 per year in this academic

centre. This effect makes performing blood cultures a highly cost-efficient intervention and an

interesting target to improve patient care, patient safety and hospital expenditures.

144

General conclusion, discussion and recommendations

11

145 | Chapter 11

Conclusions and discussion

With a strong political focus in the Netherlands on controlling healthcare costs and providing

efficient healthcare, as well as the competitive pressure for healthcare institutions to be more

cost-efficient due the relatively open market in the Dutch healthcare system, there is a clear

need for more impact analyses on provided healthcare (Brook, 2011; NZa, 2015; Schippers and

van Rijn, 2013; Ubbink, et al., 2014). Studies that evaluate current practices, compare them,

and make recommendations. Only with such studies, it will be possible to make (financially

and clinically) founded decisions on new and current interventions, but also create the

justification towards insurance companies, patients, and hospital boards of directors/managers.

The research described in this thesis looked at the current and new practices of the

Department of Medical Microbiology and Infection Prevention of the University Medical

Center Groningen (UMCG). Data from these studies and the resulting publications provide a

comprehensive overview and evaluation on a clinical and a financial level. In this final chapter,

findings and conclusions from the previous chapters will be brought forward leading to a

general conclusion and answer to the main question: What is the clinical and financial impact of the

combined activities of Medical Microbiology and Infection Prevention on relevant outcome measures, at an

academic hospital such as the UMCG? followed by some recommendations for the future.

For a healthcare institution and especially for a Medical Microbiology laboratory, it is

important to know the patient population, their pattern of transfer(s) between healthcare

providers and the antimicrobial resistance rates in the healthcare region (Ciccolini, et al., 2013;

Donker, et al., 2015). Data on antimicrobial use (that directly affects resistance development

(Goossens, 2009)) is therefore also of importance. Previously performed studies showed that

antimicrobial use in Germany is higher than in the Netherlands (European Centre for Disease

Prevention and Control, 2013; Holstiege and Garbe, 2013; Holstiege, et al., 2014). However,

these were always studies done on a national level with incomplete data. Due to the large,

cross-border catchment area of the UMCG, it is especially important to know antimicrobial

use among patients along the border region. These German patients living in the border region

can visit Dutch hospitals in accordance with the new EU directive 2011/24/EU on cross-

border patient care (The European Parliament and the Council of the European Union, 2011).

They are therefore also part of the healthcare region of the UMCG. It is therefore wise to take

these cross-border patients into account as UMCG and to be prepared for a possible increase

in numbers for the coming years. With regard to infection prevention it is therefore highly

valuable to have more in-depth information on these patients, their antimicrobial use and local

antimicrobial resistance patterns. Data like this can be used to adapt local UMCG guidelines in

the best-fitting manner. In Chapter 2, for the first time, data is shown on antimicrobial use of

patients across the Dutch-German border. And indeed, also in the border region antimicrobial

use among pediatric outpatients is higher in Germany compared to the Dutch border region.

General conclusion, discussion and recommendations | 146

Especially the differences in second-generation cephalosporin use in the first line are

something to take into account when looking at infection prevention. The Euregio can and

should work together on this topic to improve the usage of antimicrobials. Dutch experiences

and guidelines can be of great help to German health care professionals and Dutch health care

professionals should take these baseline differences into account when German patients are

visiting a Dutch hospital. Indeed, this is one of the factors that we propose to be part of a

more integral infection management approach within Chapter 3.

When patients with an infectious problem are visiting the hospital, it is vital that all involved

specialists from different medical disciplines work together in an integrated, interdisciplinary

manner. The work of the Department of Medical Microbiology is structured into three

different, but integrative focus points and we translated that into three stewardships aspects:

Antimicrobial, Infection Prevention and Diagnostic (AID) Stewardships (see figure 11.1). It is

of no use to implement interventions to improve antimicrobial use, if there are no appropriate

diagnostics performed or if they are performed too late (i.e. after starting of empirical therapy).

Without diagnostics it is impossible to know what the causative pathogen of the patient is,

making guided therapy in the correct form and dosage also impossible. Similarly, without

infection prevention, uncontrolled spread of resistant microorganisms will impact

antimicrobial use and consequently also resistance development. Thus all three aspects should

be in place within a healthcare institution and within the healthcare region, to complement

each other, thereby providing an integrative infection management approach. This focuses on

the correct diagnostics to adequately identify the pathogenic microorganism(s) in a timely

manner, correct therapy using the results from the diagnostics and finally infection prevention

measures to ensure that colonized patients are not transferring their microorganisms to other

patients and resistance development is curbed.

It also recognizes that different patients require different attention, with more

complex patients requiring the input and experience of several more experienced specialists

that should work together on infection management. These specialists are supported by a

network of specially trained doctors and nurses (i.e. link-docs and link-nurses). Less complex

patients can be covered for a large part by this network; thereby making sure that the limited

time of the more experienced specialists is used in the most efficient manner. This patient-

specific approach should lead in the near future to a so-called theragnostic approach where

therapy and diagnostics are integrated, limiting the use of broad empiric substances further and

improving quality of healthcare. Finally, the model matches supply and demand of numbers of

patients and available staff.

Antimicrobial Stewardship Programs (ASPs) target antimicrobial therapy to make sure that

patients are most effectively treated (thereby also limiting possible side-effects) and at the same

147 | Chapter 11

time resistance development is kept at an as low as possible level. Many different interventions

are implemented worldwide to ensure those goals and it is important to evaluate these

interventions because results and effects depend on the setting (Davey, et al., 2013). There is

no universal approach and continuous evaluation is thus necessary. We provide some ways and

methods to measure ASP interventions in Chapter 4 and evaluate the quality of financial

outcomes and methods in Chapter 5. Different methods exist to evaluate an ASP and these

evaluations can be difficult due to missing negative controls or bundle interventions. It is

therefore important to think about the method of evaluation that one choses and the specific

pros and cons of each approach. Regarding financial evaluations that are published,

improvements can and should be made. In general, not all (relevant) costs and benefits are

looked at, giving an incomplete picture, which also leads to possible erroneous conclusions to

be drawn. Also, methods are often not explained properly making it difficult or even

impossible to generalize conclusions to other settings or to repeat the analysis. Results that

were found, mainly focus on direct costs of antimicrobials and the conclusion seem to be that

(especially in high use countries) costs can be saved due to an ASP. These conclusions were

also drawn in a similar review published at the same time, strengthening our own findings and

conclusions (Coulter, et al., 2015).

Figure 11.1: Multi Stakeholder Platform of the AID Stewardship model. Pyramid platform showing

the interdisciplinary stakeholder connections between the Antimicrobial Stewardship Program (ASP),

Infection Prevention Stewardship Program (ISP) and Diagnostic Stewardship Program (DSP) as was

presented in Chapter 3 of this thesis.


Keeping these results in mind, we set out to evaluate the ASP in the UMCG. Chapter

6 describes the clinical outcomes and Chapter 7 the financial outcomes of an ASP on a urology

ward within the UMCG. The program consisted of an Antimicrobial Stewardship-Team (A-

Team) that visited the ward after a patient received 48 hours of antimicrobial therapy. The A-

Team member discussed the patient’s therapy with the attending physician. Based on the

available diagnostics and present guidelines, a decision was made regarding continuation of

antimicrobial therapy. Such an audit and feedback intervention was described earlier, although

focused on 72 hours of therapy instead of the 48 hours that was chosen at the UMCG (Pulcini,

et al., 2008). Due to the presence of the microbiological laboratory on the premises of the

hospital leading to relatively fast turn-around-times of the diagnostics, it was chosen to

perform the intervention after 48 hours instead of 72 hours. We show that the audit and

feedback led to an increased switch from intravenous administered antimicrobials to oral

antimicrobials in one out of four patients. Furthermore, in one out of four patients therapy

could be stopped after 48 hours, because there was no (longer) an indication to warrant the

prescription of antimicrobials. This led to a substantial and significant decrease in length of

stay for a subgroup of the intervened patients. Finally, a decrease in the overall numbers of

patients receiving antimicrobial therapy was observed, indicating that the intervention also had

a more systemic effect beyond the patients that were consulted by the A-Team. These results

not only show the large effects of an audit and feedback program, but also that this is possible

to perform successfully after 48 hours instead of 72 hours. In the separate economic evaluation

(Chapter 7), all costs and benefits of the A-Team were calculated for the same two cohorts as

the clinical evaluation. Considering all outcomes from a hospital perspective, this intervention

was considered (highly) cost-efficient, primarily due to the decreased length of stay of the

intervened patients. The quality of the intervention and its financial evaluation was recognized

and therefore chosen as one of the three best practices in Dutch hospitals as presented at the

European Ministerial Conference on Antibiotic Resistance organized by the Dutch

government and published in the European best practice report (Oberjé, et al., 2016).

The Infection Prevention Stewardship Program (ISP) of the Department of Medical

Microbiology and Infection Prevention is the overarching name for all infection prevention

and infection control related activities. The unit of Infection Prevention mainly coordinates

these. Examples are the screening of (risk) patients within the hospital; educating and

promoting hygiene measures such as hand hygiene; auditing, reporting and improving current

practices; and actions to control events with colonized patients. All actions are done to ensure

a safe environment and to minimize the deleterious effects that infections due to (resistant)

microorganisms can have on patients but also on hospital staff and visitors. All these actions

are done in close collaboration with regional partners, and this collaboration should intensify

even more in the nearby future (Donker, et al., 2015). For the impact analysis as was

performed for this thesis, two projects were undertaken.

149 | Chapter 11

Firstly, it was investigated which costs are associated with a nosocomial outbreak

within and the hospital and how high these costs are. In general, there is little data published

on costs and infection prevention. As with antimicrobial stewardship, the same difficulties are

present when performing research. Because an outbreak within the hospital is the biggest

impact that (nosocomial) pathogens can have, this was chosen to be the main topic. In 2011,

the Netherlands saw the enormous impact of uncontrolled spread of resistant bacteria. From

2010 to 2011, the Maasstad Hospital in Rotterdam had one of the biggest nosocomial

outbreaks ever seen in the Netherlands (Externe onderzoekscommissie MSZ, 2012). This

outbreak exemplifies two important aspects: the devastating effect of a not correctly

functioning ISP (and ASP and DSP); and the subsequent impact this has on patients and costs.

Costs were reported to be millions of Euros and several patients died directly due to their

infection with the highly resistant Klebsiella pneumonia (van den Brink, 2013). The UMCG never

had an outbreak of this magnitude, but outbreaks do occur. Chapter 8 describes the different

actions and measures that were undertaken during seven different outbreaks that occurred

between 2012 and 2014. By using different databases and performing interviews with the, at

that time, involved professionals, it was possible to collect a complete overview of all the

different actions that were performed in order to control the respective outbreak. All these

actions could then be quantified into Euros, making it possible to calculate the average cost per

patient per day for the UMCG during such an event. Depending on the type of organism,

ward, number and characteristics of affected patients, these costs will differ. Unfortunately the

dataset was too small to determine the different effects of these variables precisely and

conclusively. However, the average daily cost of €546 per patient is a good indication of the

large financial impact and the first time that a figure was calculated in this comparable manner.

Using this number, it was also possible to do further research into the financial effects

of Infection Prevention. Because there was never situation without infection prevention in the

past, it is also debatable if such a situation should be taken as a baseline level when looking at

costs (Graves, 2004). We therefore choose to do an incremental cost-analysis, looking at the

extra investments done each year and the effects of those investments to see if they would be

financially beneficial or not. Thus, effectively comparing each year with the previous year. As

with an ASP, it is of course patient safety that should be the main driver to invest in infection

prevention, but when there are multiple ways to achieve that goal, it is important to know

which is the most cost-efficient to keep healthcare costs under control. In Chapter 9, we

describe this incremental cost-analysis over a time-period of eight years within the UMCG

(2007-2014). It was observed that the number of patients colonized with high-risk

microorganisms that are known to cause outbreaks is rising each year (e.g. MRSA, ESBL K.

pneumoniae and VRE). With more colonized patients, the risk of spread to other patients thus

increases and with that the risk on outbreaks increases as well. To keep up with this growing

risk, more money is spent each year on infection prevention personnel. The effects of these

extra infection prevention facilities were measured by looking at a subset of indirect indicators

to confirm that the investments had an effect. And indeed, an increase in the use of utensils,

hand disinfection and surveillance cultures was observed over the eight years. Based on the

increased number of high-risk patients, it was calculated how many outbreak patients were to


be expected and these number were compared to the actual found numbers of patients. These

found numbers were lower than the expected ones. This entails that savings were achieved,

because less money had to be spent on the control of outbreaks. These savings were higher

than the yearly investments, giving a return on investment (ROI) of 1.9. This is just one aspect

where infection prevention has an effect and the beneficial financial effects are therefore

expected to be even higher.

Regarding the Diagnostic Stewardship Program (DSP), it was chosen to look at one of the

most frequently performed diagnostic tools by the department: the blood cultures. With a

database of five years of UMCG admissions, it was determined what the differences in

outcomes were when looking at antimicrobial users with blood cultures during start of therapy

and those without. The results presented in Chapter 10 suggest a beneficial effect of

performing blood cultures for patients receiving multiple days of IV antimicrobials (i.e. with

infectious problems). Of all investigated patients, almost half of the patients that received IV

therapy started at admission, did so without having blood cultures taken at the same time

(ideally prior to start). This was also seen in a nearby community hospital. These numbers

alone are worrisome, because these patients are in a sense treated ‘blindly’. Without proper

diagnostics, the causative agent will remain unknown, and adjustment of therapy therefore will

be impossible. This not only affects the patient due to potentially suboptimal and toxic

therapy, but also the general population because unnecessary broad-spectrum use drives

antimicrobial resistance. Furthermore, the group with blood cultures had also a positive

correlation with a reduced LOS, suggesting that therapy might have been more effective,

and/or physicians felt more confident to stop therapy earlier. In parallel, duration of

antimicrobial therapy was also shorter in the group with blood cultures. Due to the high

correlation of performing blood cultures and performing other clinical chemical diagnostics

(i.e. CRP, eGFR and leucocyte counts), we assume a bundle effect in which patients who

received an integral approach of diagnostics have the better outcomes. An ASP (or DSP)

should therefore also focus on timely performing diagnostics in patients who receive

antimicrobials.

All these studies done on the three focus points of the department finally come together in a

single conclusion and an answer to the question posed in the beginning of the thesis: What is

the clinical and financial impact of the combined activities of Medical Microbiology and Infection Prevention on

relevant outcome measures, at an academic hospital such as the UMCG? Impact is defined here as

clinical and financial effects of different interventions done by the department and the

outcome measures are defined depending on the interventions as described in the previous 9

chapters. We can conclude that all interventions that were looked at, improved patient care in

one way or another. The A-Team as implemented within the hospital improved antimicrobial

therapy (less antimicrobial usage and users, less IV antimicrobials, shorter duration) and

151 | Chapter 11

reduced length of stay and therefore also costs for a subgroup of patients, earning back the

investments. Infection prevention measures improved patient safety by lowering the risk for

outbreaks and thereby preventing outbreak patients and again saving enough money to earn

back more than the yearly investments. Finally, looking at the diagnostic interventions, effects

of blood cultures are also found on length of stay, improving patient care and safety and

earning back the costs of the diagnostics. Overall, all these interventions came to a potential of

yearly savings of €538,931. This figure takes into account an A-Team on just one ward as

described in Chapter 6 and 7. If the A-Team were to be implemented hospital-wide an

extrapolation can be calculated keeping in mind the different wards and patients, their

respective antimicrobial use and the costs to visit them. Per year, another €1,212,016 could be

saved due to more than 3400 saved bed days, making the hospital more efficient, more

competitive and creating a huge potential to increase revenue (or reduce capacity of the

hospital). Freeing up beds also means waiting times will be reduced. Waiting times is one of the

main healthcare quality outcome measures that is looked at nowadays and especially from a

patient perspective an important aspect. Complications, readmissions and effects from a

societal perspective are not even (fully) included in these analyses, making the potential

benefits even greater.

The importance of knowing the impact of Medical Microbiology is partly driven by an

imperfect financial system within this hospital. Cost prices for microbiological diagnostics are

higher than the actual cost for performing them. This is due to an overhead of almost 25%,

which covers, among others, the cost for infection prevention. It makes the department less

competitive in an open healthcare market of the Netherlands that is based on unit cost prices

for diagnostics. Therefore, diagnostics can be an easy target for budget cuts and they are

subject of discussions if they should be priced lower (because for example, competitors in

Germany are much cheaper) (VGZ, 2013). This thesis concludes that both diagnostics and

infection prevention paid for by the diagnostics’ overhead have a highly positive impact both

clinically and financially. Consequently, cutting back globally on diagnostics in general (and

thus on the budget of the whole department as well) is not advisable. Out-sourcing laboratory-

based diagnostic work to cheaper competitors in for example Germany or Belgium could be an

option, and promises to save the hospital money on the short term. The question is however,

what these diagnostics cover. Is for example consulting included (on appropriate diagnostics,

antimicrobial therapy and/or infection prevention)? Furthermore, what will be the turn-

around-time to deliver results? With a laboratory located further away from the premises of the

hospital, turn-around-time until results will be longer due to transportation time. Furthermore,

it is important to take into account that cutting back on diagnostics, in the case of the UMCG

also means cutting back on the budget of infection prevention and consulting. These tasks will

need to be covered either by the selected cheaper competitors (which does not match reality),

or by the hospital itself. If costs within the hospital organization and especially within the

Department of Medical Microbiology were more transparent, it would most likely also lead to a

better understanding and a more sustainable situation with less quick, unfounded cut backs.


Because this thesis shows quite clearly that interventions performed by the department are

highly cost-beneficial on the long term, we would also advise highly against such cutbacks.

We propose therefore a more transparent financial system for Medical Microbiology and

Infection Prevention, whereby diagnostics, consulting and infection prevention are

disconnected from each other financially and infection prevention is paid for by departments

through an insurance-based system. This system should prevent such quick, unfounded budget

cuts on diagnostics (and at the same time on infection prevention), secondly it should provide

a financial incentive for the rest of the hospital to improve their infection prevention and

thirdly it should make the diagnostics of the department more competitive on a financial level.

Under the current Dutch system of reimbursed bundle payments, the DBC/DOT

system, infection prevention and control procedures are not covered within the bundles as

specific actions that can be reimbursed. Hospitals are therefore left to finance these units or

departments as they see fit, independently of reimbursement by insurance companies or other

external revenues. This makes these budgets easy targets for cutbacks, even though in general

they compose just 1-2% of the total hospitals’ budgets. One solution is to make infection

prevention part of the overhead of microbiological diagnostics (which are covered as

reimbursable actions within some bundles). The UMCG has chosen for this solution as

discussed before. A consequence is, that these diagnostics become more expensive, making it

more difficult to compete with commercial laboratories (this is similar for some Dutch

commercial laboratories that also perform infection prevention and need to compete with

foreign [e.g. German] commercial laboratories). These higher prices also fuel the discussion

that microbiological diagnostics are too expensive and that they can be cheaper, as the case is

in Germany. Again, important to note is that the unit cost price comparison usually purely

comprises technical diagnostics (i.e. technical analysis and reports), but no

consulting/Stewardship.

A more sustainable solution proposed here, would be an insurance-based finance

system for Infection Prevention. This system entails that each ward/specialty/department

becomes more financially responsible for their infection prevention and control measures.

Proper infection prevention starts at the ward itself. Actions of physicians and nurses that see

patients on a daily basis are the key components within the whole healthcare process. These

are completed by actions performed by an infection prevention team, such as identification of

risk patients. Correct hand hygiene for example is a vital part of infection prevention. If wards

are performing suboptimal, microorganisms can spread among patients and outbreaks can

occur. And when an outbreak indeed occurs, it costs hundreds of Euros per patient per day to

control and eradicate the outbreak (as we clearly demonstrated in chapter 9). Part of these

costs during outbreak situations, fall upon the unit of infection prevention who has to free up

people, perform environmental surveillance and start to follow-up all patients at risk (up to

26% of the total outbreak costs). A more fair solution would be that all wards and specialties

153 | Chapter 11

bear the costs for infection prevention together. Through a monthly insurance policy for

example. This fee will cover all infection prevention and control measures that are needed

when an event occurs that requires actions and can cover the daily work of the infection

prevention department. To provide a financial incentive and stimulate wards to improve their

infection prevention, wards pay according to their relative risk, meaning that those that are

performing better will pay less than the suboptimal performing wards. Such a system would

make the cash flows more transparent and creates a ‘polluter-pays’ principle.

Based upon the infection prevention databases and using the results from the study

described in chapter 9, an estimate can be made on the total cost per year that fell upon the

UMCG because of outbreaks and smaller isolated events (“epi events”). This total amount

should be at least be covered by newly implemented insurance fees, but needs to account for

possible extra costs due to unforeseeably bigger outbreaks or other situation (e.g. events like

the 2015 Ebola outbreak or the 2010 H1N1 pandemic), we would advise to include an extra

margin for emergency coverage. If this extra coverage has not been used, it could for example

be applied to implement innovative tools for the hospital, cover extra educational courses or it

can be returned to the wards in the form of a discount/cash back on next year’s fee. This

would correspond to the payment models implemented for classic insurances (e.g. liability,

health care, life insurance). Using a set of transparent quality indicators such as number of

events during the last years, total scores on the different quality audits (e.g. on adherence to

dress code protocols), and antibiotic use, a relative risk of each ward can be calculated to score

them objectively. By publishing these scoring lists and updating these on a regular basis, wards

can follow their performance, compare their score to other wards and try to improve it. Wards

that score the best are generally assumed to have fewer problems with spread of

microorganisms and will consequently also bear less of the infection prevention costs.

Eventually this should stimulate all healthcare professionals to actively improve infection

prevention, making the whole hospital a safer environment for patients and a more cost-

efficient institution creating therefore a win-win situation.

Concluding, we show here that investments lead to improved quality of care, and it thus pays

off to invest in Medical Microbiology and Infection Prevention. Interventions performed by a

department such as the UMCG Medical Microbiology and Infection Prevention, improve

patient care and overall healthcare quality. By reducing LOS, it creates potential to increase

revenue by making healthcare processes more efficient and at the same time lowering potential

waiting lists. We recommend a proactive ASP, ISP and DSP for each healthcare institute to

prevent, diagnose, and treat infectious problems in an integrated way. We recommend

evaluating antimicrobial therapy after 48 hours, as this will lead to improvement of therapy and

will save costs. It should however be noted that these results are highly depended on the

patient’s characteristics. A general evaluation of practices without adjusting outcomes to

specific patient groups is therefore not recommended. Infection Prevention departments or

units should have enough funds to keep up with the growing resistance problem and the risks


that this brings with it. Outbreaks are costly events and prevention or quicker control of these

can easily be cost-efficient. Focus should therefore be on a preventative approach from a

regional perspective. Blood cultures (and other diagnostics) are still not performed for all

indicated patients and this should be a focus for A-Teams. Patients with blood cultures taken

can be treated more effectively, thereby creating again potential improve patient care, reduce

length of stay, and save costs. ASPs or DSPs should focus on these aspects. Finally, the current

financial system is not adequate anymore to address the urgent challenges that we face.

Growing antimicrobial resistance rates creates more risks of difficult to treat nosocomial

infections. Prevention and control therefore requires a sustainable budget. An insurance-like

financial system following the ‘polluter-pays’ principle could be a solution to tackle this issue.

Ultimately, implementing these recommendations should lead to situation where antimicrobial

resistance and the increasing risk on nosocomial outbreaks is tackled, the importance of

correct and timely diagnostics is recognized and most importantly: patient care and overall

healthcare quality is improved in a sustainable manner.

156

Appendix I: Bibliography

I

157 | Appendix I

AT. [Urinary tract chemotherapeutic Nitrofuration - Organ damaging and outdated.]. Arznei

Telegramm. 1993. 11;128.

Abu Dabrh A, Firwana B, Cowl C, Steinkraus L, Prokop L and Murad M Health assessment of

commercial drivers: a meta-narrative systematic review. BMJ Open. 2014. 4;(3): e003434-.

Aiken L, Clarke S and Sloane D Hospital staffing, organization, and quality of care: Cross-

national findings. Nursing Outlook. 2002. 50;(5): 187-94.

Al Eidan FA, McElnay JC, Scott MG, Kearney MP, Corrigan J and McConnell JB Use of a

treatment protocol in the management of community-acquired lower respiratory tract

infection. Journal of Antimicrobial Chemotherapy. 2000. 45;(3): 387-94.

Ansari F, Gray K, Nathwani D, Phillips G, Ogston S, Ramsay C, et al. Outcomes of an

intervention to improve hospital antibiotic prescribing: interrupted time series with segmented

regression analysis. Journal of Antimicrobial Chemotherapy. 2003. 52;(5): 842-8.

Arcilla MS, van Hattem J, Matamoros S, Melles DC, Penders J, de Jong MD, et al.

Dissemination of the mcr-1 colistin resistance gene. The Lancet Infectious Diseases. 2016. 16;(2):

147-9.

Banerjee R, Teng CB, Cunningham SA, Ihde SM, Steckelberg JM, Moriarty JP, et al.

Randomized Trial of Rapid Multiplex Polymerase Chain Reaction-Based Blood Culture

Identification and Susceptibility Testing. Clinical Infectious Diseases. 2015. 61;(7): 1071-80.

Barenfanger J, Drake C, Leon N, Mueller T and Troutt T Clinical and financial benefits of

rapid detection of respiratory viruses: an outcomes study. Journal of Clinical Microbiology. 2000.

38;(8): 2824-8.

Barlam TF, Cosgrove SE, Abbo LM, MacDougall C, Schuetz AN, Septimus EJ, et al.

Implementing an Antibiotic Stewardship Program: Guidelines by the Infectious Diseases

Society of America and the Society for Healthcare Epidemiology of America. Clinical Infectious

Diseases. 2016.

Bartlett J, Gilbert D and Spellberg B Seven ways to preserve the miracle of antibiotics. Clinical

Infectious Diseases. 2013. 56;(10): 1445-50.

Bates DW, Goldman L and Lee TH Contaminant blood cultures and resource utilization. The

true consequences of false-positive results. Journal of the American Medical Association. 1991.

265;(3): 365-9.

Bauer K, West J, Balada Llasat JM, Pancholi P, Stevenson K and Goff D An antimicrobial

stewardship program's impact with rapid polymerase chain reaction methicillin-resistant

Staphylococcus aureus/S. aureus blood culture test in patients with S. aureus bacteremia.

Clinical Infectious Diseases. 2010a. 51;(9): 1074-80.

Bibliography | 158

Bauer K, West J, Balada Llasat JM, Pancholi P, Stevenson K and Goff D An antimicrobial

stewardship program's impact with rapid polymerase chain reaction methicillin-resistant

Staphylococcus aureus/S. aureus blood culture test in patients with S. aureus bacteremia.

Clinical Infectious Diseases. 2010b. 51;(9): 1074-80.

Beaulieu J, Fortin R, Palmisciano L, Elsaid K and Collins C Enhancing clinical decision

support to improve appropriate antimicrobial use. American Journal of Health-System Pharmacy.

2013. 70;(13): 1103-4.

Beerlage-de Jong N, Eikelenboom-Boskamp A, Voss A, Sanderman R and van Gemert-Pijnen

JEWC Combining user-centered design with the persuasive systems design model: The

development process of a web-based registration and monitoring system for healthcare-

associated infections in nursing homes. International Journal on Advances in Life Sciences. 2014.

6;(4): 262-71.

Bettiol E and Harbarth S Development of new antibiotics: taking off finally?. Swiss Medical

Weekly. 2015. 145;

Bittner MJ and Rich EC Surveillance of handwashing episodes in adult intensive-care units by

measuring an index of soap and paper towel consumption. Clinical Performance and Quality Health

Care. 1998. 6;(4): 179-82.

Björholt I and Haglind E Cost-savings achieved by eradication of epidemic methicillin-resistant

Staphylococcus aureus (EMRSA)-16 from a large teaching hospital. European Journal of Clinical

Microbiology and Infectious Diseases. 2004. 23;(9): 688-95.

Björnberg A Euro Health Consumer Index 2014. 2015. 1-96. Health Consumer Powerhouse.

Täby, Sweden. Available at: http://www.healthpowerhouse.com/

files/EHCI_2014/EHCI_2014_report.pdf. Accessed at: November 1, 2015.

Bode LG, Wertheim HF, Kluytmans JA, Bogaers-Hofman D, Vandenbroucke-Grauls CM,

Roosendaal R, et al. Sustained low prevalence of meticillin-resistant Staphylococcus aureus

upon admission to hospital in The Netherlands. Journal of Hospital Infection. 2011. 79;(3): 198-

201.

Bonten MJ [Medical microbiology laboratories in Dutch hospitals: essential for safe patient

care]. Nederlands Tijdschrift Voor Geneeskunde. 2008. 152;(49): 2650-2.

Bonten MJ, Friedrich AW, Kluytmans JA, Vandenbroucke-Grauls CM, Voss A and Vos MC

[Prevention of infections with highly resistant mircoorganisms: maximizing transparency by

using outcome indicators]. Nederlands Tijdschrift Voor Geneeskunde. 2015. 159;A8588.

Bonten MJ, Friedrich AW, Kluytmans JA, Vandenbroucke-Grauls CM, Voss A and Vos MC

[Infection prevention in Dutch hospitals; results say more than process indicators]. Nederlands

Tijdschrift Voor Geneeskunde. 2014. 158;A7395.

http://www.healthpowerhouse.com/files/EHCI_2014/EHCI_2014_report.pdf

http://www.healthpowerhouse.com/files/EHCI_2014/EHCI_2014_report.pdf

159 | Appendix I

Bou R, Lorente L, Aquilar A, Parpiñán J, Ramos P, Peris M, et al. Hospital economic impact of

an outbreak of Pseudomonas aeruginosa infections. Journal of Hospital Infection. 2009. 71;(2):

138-42.

Bouadma L, Luyt CE, Tubach F, Cracco C, Alvarez A, Schwebel C, et al. Use of procalcitonin

to reduce patients' exposure to antibiotics in intensive care units (PRORATA trial): a

multicentre randomised controlled trial. Lancet, The. 2010. 375;(9713): 463-74.

Brauer R, Ruigómez A, Downey G, Bate A, Garcia Rodriguiz LA, Huerta C, et al. Prevalence

of antibiotic use: a comparison across various European health care data sources.

Pharmacoepidemiology and Drug Safety. 2015.

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 study.

Lancet Infectious Diseases. 2015. 14;(12): 1220-7.

Brook RH The role of physicians in controlling medical care costs and reducing waste. JAMA:

The Journal of the American Medical Association. 2011. 306;(6): 650-1.

Brookes-Howell L, Wood F, Verheij T, Prout H, Cooper L, Hood K, et al. Trust, openness

and continuity of care influence acceptance of antibiotics for children with respiratory tract

infections: a four country qualitative study. Family Practice. 2014. 31;(1): 102-10.

Brüggemann RJ and Aarnoutse RE Fundament and Prerequisites for the Application of an

Antifungal TDM Service. Current Fungal Infection Reports. 2015. 9;(2): 122-9.

Buehler SS, Madison B, Snyder SR, Derzon JH, Cornish NE, Saubolle MA, et al. Effectiveness

of Practices To Increase Timeliness of Providing Targeted Therapy for Inpatients with

Bloodstream Infections: a Laboratory Medicine Best Practices Systematic Review and Meta-

analysis. Clinical Microbiology Reviews. 2016. 29;(1): 59-103.

Burke J and Yeo T. Nosocomial urinary tract infections. In: Mayhall C ed. 3rd ed.Lippincott

Williams & Wilkins. Philadelphia. 2004.

Caliendo AM, Gilbert DN, Ginocchio CC, Hanson KE, May L, Quinn TC, et al. Better tests,

better care: improved diagnostics for infectious diseases. Clinical Infectious Diseases. 2013. 57;139-

70.

Carlet J, Collignon P, Goldmann D, Goossens H, Gyssens I, Harbarth S, et al. Society's failure

to protect a precious resource: antibiotics. Lancet. 2011. 378;(9788): 369-71.

Caro JJ, Briggs A, Siebert U and Kuntz K Modeling good research practices--overview: a

report of the ISPOR-SMDM Modeling Good Research Practices Task Force-1. Medical Decision

Making. 2012. 32;(5): 667-77.

Bibliography | 160

Cars H and Håkansson A Prescriptions of antibiotics for children. Prescribing habits of

district, hospital, and private physicians. Scandinavian Journal of Primary Health Care. 1997. 15;(1):

22-5.

Chakravarthy M, Adhikary R, Gokul B and Pushparaj L Hospital acquired infection is inversely

related to utilization of isopropyl alcohol and tissue paper pulls--a prospective observational

study. Journal of the Association of Physicians of India. 2011. 59;548-50.

Chen S Assessing costs in health care. Current Surgery. 2004. 61;(2): 155-60.

Chia D, Yavari Y, Kirsanov E, Aronin S and Sadigh M Adherence to Standard of Care in the

Diagnosis and Treatment of Suspected Bacterial Meningitis. American Journal of Medical Quality.

2015. 30;(6): 539-42.

Chopra V, O'Horo J, Rogers MAM, Maki D and Safdar N The risk of bloodstream infection

associated with peripherally inserted central catheters compared with central venous catheters

in adults: a systematic review and meta-analysis. Infection Control and Hospital Epidemiology. 2013.

34;(9): 908-18.

Christiansen KJ, Tibbett PA, Beresford W, Pearman JW, Lee RC, Coombs GW, et al.

Eradication of a large outbreak of a single strain of vanB vancomycin-resistant Enterococcus

faecium at a major Australian teaching hospital. Infection Control and Hospital Epidemiology. 2004.

25;(5): 384-90.

Ciccolini M, Donker T, Köck RR, Mielke M, Hendrix R, Jurke A, et al. Infection prevention in

a connected world: The case for a regional approach. International Journal of Medical Microbiology.

2013. 303;(6-7): 380-7.

Cisneros JM, Neth O, Gil-Navarro MV, Lepe JA, Jiménez-Parrilla F, Cordero E, et al. Global

impact of an educational antimicrobial stewardship programme on prescribing practice in a

tertiary hospital centre. Clinical Microbiology and Infection. 2014. 20;(1): 82-8.

Clements A, Halton K, Graves N, Pettitt A, Morton A, Looke D, et al. Overcrowding and

understaffing in modern health-care systems: key determinants in meticillin-resistant

Staphylococcus aureus transmission. The Lancet Infectious Diseases. 2008. 8;(7): 427-34.

Clerc O, Prod'hom G, Senn L, Jaton K, Zanetti G, Calandra T, et al. Matrix-assisted laser

desorption ionization time-of-flight mass spectrometry and PCR-based rapid diagnosis of

Staphylococcus aureus bacteraemia. Clinical Microbiology and Infection. 2014. 20;(4): 355-60.

Coburn B, Morris A, Tomlinson G and Detsky A Does this adult patient with suspected

bacteremia require blood cultures?. Journal of the American Medical Association. 2012. 308;(5): 502-

11.

161 | Appendix I

Cohen A, Bont L, Engelhard D, Moore E, Fernández D, Kreisberg-Greenblatt R, et al. A

multifaceted 'omics' approach for addressing the challenge of antimicrobial resistance. Future

Microbiology. 2015. 10;(3): 365-76.

Collini P, Beadsworth M, Anson J, Neal T, Burnham P, Deegan P, et al. Community-acquired

pneumonia: doctors do not follow national guidelines. Postgraduate Medical Journal. 2007.

83;(982): 552-5.

Cooke FJ and Holmes AH The missing care bundle: antibiotic prescribing in hospitals.

International Journal of Antimicrobial Agents. 2007. 30;(1): 25-9.

Cosgrove S The relationship between antimicrobial resistance and patient outcomes: mortality,

length of hospital stay, and health care costs. Clinical Infectious Diseases. 2006. 42 (Suppl 2);S82-9.

Costelloe C, Metcalfe C, Lovering A, Mant D and Hay AD Effect of antibiotic prescribing in

primary care on antimicrobial resistance in individual patients: systematic review and meta-

analysis. Bmj. 2010. 340;

Coulter S, Merollini K, Roberts JA, Graves N and Halton K The need for cost-effectiveness

analyses of antimicrobial stewardship programmes: A structured review. International Journal of

Antimicrobial Agents. 2015. 46;(2): 140-9.

Covvey JR, Johnson BF, Elliott V, Malcolm W and Mullen AB An association between

socioeconomic deprivation and primary care antibiotic prescribing in Scotland. Journal of

Antimicrobial Chemotherapy. 2014. 69;(3): 835-41.

Czaplewski L, Bax R, Clokie M, Dawson M, Fairhead H, Fischetti VA, et al. Alternatives to

antibiotics-a pipeline portfolio review. Lancet Infectious Diseases. 2016. 2;239-51.

Daneman N, Sarwar S, Fowler RA, Cuthbertson BH and SuDDICU Canadian Study Group

Effect of selective decontamination on antimicrobial resistance in intensive care units: a

systematic review and meta-analysis. Lancet Infectious Diseases. 2013. 13;(4): 328-41.

Davey P, Brown E, Charani E, Fenelon L, Gould I, Holmes A, et al. Interventions to improve

antibiotic prescribing practices for hospital inpatients. Cochrane Database of Systematic Reviews.

2013. 4;CD003543.

De Angelis G, Murthy A, Beyersmann J and Harbarth S Estimating the impact of healthcare-

associated infections on length of stay and costs. Clinical Microbiology and Infection. 2010. 16;(12):

1729-35.

de Jong E, van Oers JA, Beishuizen A, Vos P, Vermeijden WJ, Haas LE, et al. Efficacy and

safety of procalcitonin guidance in reducing the duration of antibiotic treatment in critically ill

patients: a randomised, controlled, open-label trial. Lancet Infectious Diseases. 2016. epub ehead of

print;

Bibliography | 162

de Jong J, van den Berg PB, de Vries TW and de Jong-van den Berg LTW Antibiotic drug use

of children in the Netherlands from 1999 till 2005. European Journal of Clinical Pharmacology.

2008. 64;(9): 913-9.

de Kraker ME, Davey PG, Grundmann H and BURDEN study group Mortality and hospital

stay associated with resistant Staphylococcus aureus and Escherichia coli bacteremia:

estimating the burden of antibiotic resistance in Europe. PLoS Medicine. 2011. 8;(10): e1001104.

de Kraker MEA, Jarlier V, Monen JCM, Heuer OE, van de Sande N and Grundmann H The

changing epidemiology of bacteraemias in Europe: trends from the European Antimicrobial

Resistance Surveillance System. Clinical Microbiology and Infection. 2013. 19;(9): 860-8.

de With K, Allerberger F, Amann S, Apfalter P, Brodt HR, Eckmanns T, et al. Strategies to

enhance rational use of antibiotics in hospital: a guideline by the German Society for Infectious

Diseases. Infection. 2016. epub ahead of print;

de With K, Bestehorn H, Steib Bauert M and Kern WV Comparison of defined versus

recommended versus prescribed daily doses for measuring hospital antibiotic consumption.

Infection. 2009. 37;(4): 349-52.

DEGAM. [Acute coughing]. DEGAM. 2014a. Available from: http://www.degam.de/

files/Inhalte/Leitlinien-Inhalte/Dokumente/DEGAM-S3-Leitlinien/Kurzversion_Akuter%20

Husten_20140320.pdf. Accessed at: May 15, 2015.

DEGAM. [Otitis media]. DEGAM. 2014b. Available from: http://www.degam.de/

files/Inhalte/Leitlinien-Inhalte/Dokumente/DEGAM-S3-Leitlinien/LL-

07_Kurzversion_Ohrenschmerzen_AWMF%20053_009.pdf. Accessed at: May 15, 2015.

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.

Intensive Care Medicine. 2013. 39;(2): 165-228.

Dellit T, Owens R, McGowan J, Gerding D, Weinstein R, Burke J, 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. Clinical Infectious

Diseases. 2007. 44;(2): 159-77.

DePestel DD, Eiland 3rd EH, Lusardi K, Destache CJ, Mercier RC, McDaneld PM, et al.

Assessing appropriateness of antimicrobial therapy: in the eye of the interpreter. Clinical

Infectious Diseases. 2014. 59;154-61.

Deurenberg RH, Nulens E, Valvatne H, Sebastian S, Driessen C, Craeghs J, et al. Cross-border

dissemination of methicillin-resistant Staphylococcus aureus, Euregio Meuse-Rhin region.

Emerging Infectious Diseases. 2009. 15;(5): 727-34.

163 | Appendix I

Dik JH, Dinkeracker AG, Vemer P, Lo-Ten-Foe JR, Lokate M, Sinha B, et al. Cost-analysis of

seven nosocomial outbreaks in an academic hospital. PLoS ONE. 2016a. 11;(2): e0149226.

Dik JH, Hendrix R, Lo-Ten-Foe JR, Wilting KR, Panday PN, van Gemert-Pijnen JEWC, et al.

Automatic day-2 intervention by a multidisciplinary antimicrobial stewardship-team leads to

multiple positive effects. Frontiers in Microbiology. 2015a. 6;(546):

Dik JH, Poelman R, Friedrich AW, Nannan Panday P, Lo-Ten-Foe JR, van Assen S, et al. An

integrated stewardship model: antimicrobial, infection prevention and diagnostic (AID). Future

Microbiology. 2016b. 11;(1): 93-102.

Dik JH, Sinha B, Lo-Ten-Foe JR, Hendrix R, Nannan Panday P, Postma MJ, et al. Cost-

minimization model of a multidisciplinary Antibiotic Stewardship-Team based on a successful

implementation on a urology ward of an academic hospital. PLoS ONE. 2015b. 10;(5):

e0126106.

Dik JH, Vemer P, Friedrich AW, Hendrix R, Lo-Ten-Foe JR, Sinha B, et al. Financial

evaluations of antibiotic stewardship programs - a systematic review. Frontiers in Microbiology.

2015c. 6;(317):

Donker T, Ciccolini M, Wallinga J, Kluytmans JA, Grundmann H and Friedrich AW [Analysis

of patient flows: basis for regional control of antibiotic resistance]. Nederlands Tijdschrift Voor

Geneeskunde. 2015. 159;A8468.

Donker T, Wallinga J and Grundmann H Patient referral patterns and the spread of hospital-

acquired infections through national health care networks. PLoS Computational Biology. 2010.

6;(3): e1000715.

Donker T, Wallinga J, Slack R and Grundmann H Hospital networks and the dispersal of

hospital-acquired pathogens by patient transfer. PLoS ONE. 2012. 7;(4): e35002.

Draenert R, Seybold U, Grützner E and Bogner JR Novel antibiotics: are we still in the pre-

post-antibiotic era?. Infection. 2015. 43;(2): 145-51.

Drummond MF, Sculpher MJ, Torrance GW, O'Brien BJ and Stoddart GL. Cost Analysis. In:

Methods of the Economic Evaluation of Health Care. Drummond MF ed. Third ed.Oxford

University Press. Oxford. 2005.

Earnshaw S, Mancarella G, Mendez A, Todorova B, Magriorakos AP, Possenti E, et al.

European Antibiotic Awareness Day: a five-year perspective of Europe-wide actions to

promote prudent use of antibiotics. Eurosurveillance. 2014. 19;(41):

Eber MR, Laxminarayan R, Perencevich E and Malani A Clinical and economic outcomes

attributable to health care-associated sepsis and pneumonia. Archives of Internal Medicine. 2010.

170;(4): 347-53.

Bibliography | 164

ECDC/EMEA Joint Working Group The bacterial challenge: time to react. 2009.

EMEA/576176/2009;13-42. ECDC. Stockholm.

EFPIA MID3 Workgroup, Marshall SF, Burghaus R, Cosson V, Cheung S, Chenel M, et al.

Good Practices in Model-Informed Drug Discovery and Development: Practice, Application,

and Documentation. CPT: Pharmacometrics and Systems Pharmacology. 2016. 5;(3): 93-122.

European Centre for Disease Prevention and Control. Antimicrobial resistance interactive database

(EARS-Net). 2015. Available from: http://ecdc.europa.eu/en/healthtopics/

antimicrobial_resistance/database/Pages/database.aspx. Accessed at: December 1, 2015.

European Centre for Disease Prevention and Control Surveillance of antimicrobial

consumption in Europe 2012. 2014. TQ-AT-14-002-EN-N; ECDC. Stockholm, Sweden.

European Centre for Disease Prevention and Control Antimicrobial resistance surveillance in

Europe 2012. Annual Report of the European Antimicrobial Resistance Surveillance Network

(EARS-Net). 2013. TQ-AM-13-001-EN-C; ECDC. Stockholm, Sweden.

Evans SR, Hujer AM, Jiang H, Hujer KM, Hall T, Marzan C, et al. Rapid Molecular

Diagnostics, Antibiotic Treatment Decisions, and Developing Approaches to Inform Empiric

Therapy: PRIMERS I and II. Clinical Infectious Diseases. 2016. 62;(2): 181-9.

Evans SR, Rubin D, Follmann D, Pennello G, Huskins WC, Powers JH, et al. Desirability of

Outcome Ranking (DOOR) and Response Adjusted for Duration of Antibiotic Risk

(RADAR). Clinical Infectious Diseases. 2015. 61;(5): 800-6.

Evers S, Goossens M, de Vet H, van Tulder M and Ament A Criteria list for assessment of

methodological quality of economic evaluations: Consensus on Health Economic Criteria.

International Journal of Technology Assessment in Health Care. 2005. 21;(2): 240-5.

Externe onderzoekscommissie MSZ. Oog voor het onzichtbare. 2012. Rotterdam,

Netherlands.

Faber MS, Heckenbach K, Velasco E and Eckmanns T Antibiotics for the common cold:

expectations of Germany's general population. Eurosurveillance. 2010. 15;(35):

FDA. Generic Competition and Drug Prices. 2010. Available from: http://www.fda.gov/

aboutfda/centersoffices/officeofmedicalproductsandtobacco/cder/ucm129385.htm

Fleet E, Gopal Rao G, Patel B, Cookson B, Charlett A, Bowman C, et al. Impact of

implementation of a novel antimicrobial stewardship tool on antibiotic use in nursing homes: a

prospective cluster randomized control pilot study. Journal of Antimicrobial Chemotherapy. 2014.

69;(8): 2265-73.

165 | Appendix I

Fleming A. Penicillin. Nobel Lecture, December 11, 1945. Elsevier Publishing Company.

Amsterdam, Netherlands. 1964.

Fleming A On the antibacterial action of cultures of a penicillium, with special reference to

their use in the isolation of B. influenzae. British Journal of Experimental Pathology. 1929. 10;226-36.

Forrest G, Mehta S, Weekes E, Lincalis D, Johnson J and Venezia R Impact of rapid in situ

hybridization testing on coagulase-negative staphylococci positive blood cultures. Journal of


Fortin E, Fontela PS, Manges AR, Platt RW, Buckeridge DL and Quach C Measuring

antimicrobial use in hospitalized patients: a systematic review of available measures applicable

to paediatrics. Journal of Antimicrobial Chemotherapy. 2014. 69;(6): 1447-56.

Fowler T, Walker D and Davies SC The risk/benefit of predicting a post-antibiotic era: is the

alarm working?. Annals of the New York Academy of Sciences. 2014. 1323;1-10.

Fretz R, Schmid D, Jelovcan S, Tschertou R, Krassnitzer E, Schirmer M, et al. An outbreak of

norovirus gastroenteritis in an Austrian hospital, winter 2006-2007. Wiener Klinische

Wochenschrift. 2009. 121;(3-4): 137-43.

Friedrich AW, Daniels Haardt I, Köck RR, Verhoeven F, Mellmann A, Harmsen D, et al.

EUREGIO MRSA-net Twente/Münsterland--a Dutch-German cross-border network for the

prevention and control of infections caused by methicillin-resistant Staphylococcus aureus.

Eurosurveillance. 2008a. 13;(35): pii: 18965.

Friedrich AW, Daniels-Haardt I, Köck R, Verhoeven F, Mellmann A, Harmsen D, et al.

EUREGIO MRSA-net Twente/Münsterland--a Dutch-German cross-border network for the

prevention and control of infections caused by methicillin-resistant Staphylococcus aureus.

Eurosurveillance. 2008b. 13;(35):

Frighetto L, Marra CA, Stiver HG, Bryce EA and Jewesson PJ Economic impact of

standardized orders for antimicrobial prophylaxis program. The Annals of Pharmacotherapy. 2000.

34;(2): 154-60.

G7 Germany. Meeting of G7 health ministers. Limiting the use antibiotics worldwide. Berlin, Germany:

The Federal German Government. 2015. Available from: https://www.g7germany.de/

Content/EN/Artikel/2015/10_en/2015-10-09-g7-gesundheitsministertreffen_en.html.

Accessed at: December 1, 2015.

Gagliotti C, Ricchizzi E, Buttazzi R, Tumietto F, Resi D and Moro ML Hospital statistics for

antibiotics: defined versus prescribed daily dose. Infection. 2014. 42;(5): 869-73.

Gallagher J. Antibiotic resistance: World on cusp of 'post-antibiotic era'. BBC News. 2015,

November 19, 2015.

Bibliography | 166

Gandra SS, Barter DM and Laxminarayan R Economic burden of antibiotic resistance: how

much do we really know?. Clinical Microbiology and Infection. 2014. 20;(10): 973-80.

Gao F, Melody T, Daniels D, Giles S and Fox S The impact of compliance with 6-hour and

24-hour sepsis bundles on hospital mortality in patients with severe sepsis: a prospective

observational study. Critical Care. 2005. 9;(6): R764-70.

Geerlings SE, van Nieuwkoop C, van Haarst E, van Buren M, Knottnerus BJ, Stobberingh

EE, de Groot CJ and Prins JM SWAB guidelines for antimicrobial therapy of complicated

urinary tract infections in adults. 2013.1-94. SWAB. Utrecht, Netherlands. Available at:

http://www.swab.nl/swab/cms3.nsf/uploads/41949F6BD9ED10EDC1257B7F00212560/$F

ILE/revised%20uti%20guideline%20FINAL%20010413.pdf. Accessed at: November 17,

2014.

Goff D, Bauer K, Reed E, Stevenson K, Taylor J and West J Is the "low-hanging fruit" worth

picking for antimicrobial stewardship programs?. Clinical Infectious Diseases. 2012. 55;(4): 587-92.

Goossens H Antibiotic consumption and link to resistance. Clinical Microbiology and Infection.

2009. 15 Suppl 3;12-5.

Grabe M, Bjerklund Johansen TE, Botto H, Çek M, Naber KG and Tenke P Guidelines on

urological infections. 2013. European Association of Urology. Arnhem, Netherlands. Available

at: http://uroweb.org/gls/pdf/18_Urological%20infections_LR.pdf. Accessed at: November

17, 2014.

Graves N How costs change with infection prevention efforts. Current Opinion in Infectious

Diseases. 2014. 27;(4): 390-3.

Graves N Economics and preventing hospital-acquired infection. Emerging Infectious Diseases.

2004. 10;(4): 561-6.

Gray A, Dryden M and Charos A Antibiotic management and early discharge from hospital: an

economic analysis. Journal of Antimicrobial Chemotherapy. 2012. 67;(9): 2297-302.

Grayson ML, Russo PL, Cruickshank M, Bear JL, Gee CA, Hughes CF, et al. Outcomes from

the first 2 years of the Australian National Hand Hygiene Initiative. Medical Journal of Australia.

2011. 195;(10): 615-9.

Grobe TG, Dörning H and Schwartz F [BARMER GEK Physician's report 2011]. 2012.

Asgard-Verlag. St. Augustin.

Gross R, Morgan AS, Kinky DE, Weiner M, Gibson GA and Fishman NO Impact of a

hospital-based antimicrobial management program on clinical and economic outcomes. Clinical

Infectious Diseases. 2001a. 33;(3): 289-95.

http://www.swab.nl/swab/cms3.nsf/uploads/41949F6BD9ED10EDC1257B7F00212560/$FILE/revised%20uti%20guideline%20FINAL%20010413.pdf

http://www.swab.nl/swab/cms3.nsf/uploads/41949F6BD9ED10EDC1257B7F00212560/$FILE/revised%20uti%20guideline%20FINAL%20010413.pdf

http://uroweb.org/gls/pdf/18_Urological%20infections_LR.pdf

167 | Appendix I

Gross R, Morgan AS, Kinky DE, Weiner M, Gibson GA and Fishman NO Impact of a

hospital-based antimicrobial management program on clinical and economic outcomes. Clinical

Infectious Diseases. 2001b. 33;(3): 289-95.

Habib G, Lancellootti P, Antunes MJ, Bongiorni MG, Casalta JP, Zotti FD, et al. 2015 ESC

Guidelines for the management of infective endocarditis. European Heart Journal. 2015. 36;(44):

3075-128.

Hakkaart-van Roijen L, Tan SS and Bouwmans CAM Handleiding voor kostenonderzoek,

methoden en standaard kostprijzen voor economische evaluaties in de gezondheidszorg. 2010.

CVZ. Diemen.

Hamblin S, Rumbaugh K and Miller R Prevention of adverse drug events and cost savings

associated with PharmD interventions in an academic Level I trauma center: an evidence-based

approach. The Journal of Trauma and Acute Care Surgery. 2012. 73;(6): 1484-90.

Handoko KB, van Asselt GJ and Overdiek JW [Preventing prolonged antibiotic therapy by

active implementation of switch guidelines]. Nederlands Tijdschrift Voor Geneeskunde. 2004.

148;(5): 222-6.

Harbarth S, Theuretzbacher U, Hackett J and DRIVE-AB consortium Antibiotic research and

development: business as usual?. The Journal of Antimicrobial Chemotherapy. 2015. 70;(6): 1604-7.

Hawkings NJ, Butler CC and Wood F Antibiotics in the community: a typology of user

behaviours. Patient Education and Counseling. 2008. 73;(1): 146-52.

Health Council of the Netherlands Antibiotics in hospitals: prophylaxix and antibiotic

stewardship. 2015. 2015/12;1-64. Health Council of the Netherlands. The Hague, Netherlands.

Available at: http://www.gezondheidsraad.nl/sites/default/files/201512_antibiotica_in_

ziekenhuizen.pdf. Accessed at: January 14, 2016.

Hecker M, Fox C, Son A, Cydulka R, Siff J, Emerman C, et al. Effect of a stewardship

intervention on adherence to uncomplicated cystitis and pyelonephritis guidelines in an

emergency department setting. PLoS ONE. 2014. 9;(2): e87899.

Hjelmgren J, Berggren F and Andersson F Health economic guidelines--similarities, differences

and some implications. Value in Health. 2001. 4;(3): 225-50.

Hoffmann F and Bachmann CJ [Differences in sociodemographic characteristics, health, and

health service use of children and adolescents according to their health insurance funds].

Bundesgesundheitsblatt - Gesundheitsforschung - Gesundheitsschutz. 2014. 57;(4): 455-63.

Hoffmann F and Icks A [Structural differences between health insurance funds and their

impact on health services research: results from the Bertelsmann Health-Care Monitor].

Gesundheitswesen. 2012. 74;(5): 291-7.

Bibliography | 168

Hoffmann F, Petermann F, Glaeske G and Bachmann CJ Prevalence and comorbidities of

adolescent depression in Germany. An analysis of Health Insurance Data. Zeitschrift Für Kinder-

Und Jugendpsychiatrie Und Psychotherapie. 2012. 40;(6): 399-404.

Holstiege J and Garbe E Systemic antibiotic use among children and adolescents in Germany:

a population-based study. European Journal of Pediatrics. 2013. 172;(6): 787-95.

Holstiege J, Schink T, Molokhia M, Mazzaglia G, Innocenti F, Oteri A, et al. Systemic

antibiotic prescribing to paediatric outpatients in 5 European countries: a population-based

cohort study. BMC Pediatrics. 2014. 14;

Hop MJ, Polinder S, van der Vlies CH, Middelkoop E and van Baar M Costs of burn care: a

systematic review. Wound Repair and Regeneration. 2014. 22;(4): 436-50.

Howard P, Pulcini C, Levy Hara G, West RM, Gould IM, Harbarth S, et al. An international

cross-sectional survey of antimicrobial stewardship programmes in hospitals. Journal of


Huang AM, Newton D, Kunapuli A, Gandhi TN, Washer LL, Isip J, et al. Impact of rapid

organism identification via matrix-assisted laser desorption/ionization time-of-flight combined

with antimicrobial stewardship team intervention in adult patients with bacteremia and

candidemia. Clinical Infectious Diseases. 2013. 57;(9): 1237-45.

Hugonnet S, Harbarth S, Sax H, Duncan RA and Pittet D Nursing resources: a major

determinant of nosocomial infection?. Current Opinion in Infectious Diseases. 2004. 17;(4): 329-33.

Husereau D, Drummond MF, Petrou S, Carswell C, Moher D, Greenberg D, et al.

Consolidated Health Economic Evaluation Reporting Standards (CHEERS) statement.

European Journal of Health Economics. 2013a. 14;(3): 367-72.

Husereau D, Drummond MF, Petrou S, Carswell C, Moher D, Greenberg D, et al.

Consolidated Health Economic Evaluation Reporting Standards (CHEERS) statement. The

European Journal of Health Economics. 2013b. 14;(3): 367-72.

Janes CR, Corbett KK, Jones JH and Trostle J Emerging infectious diseases: the role of social

sciences. Lancet. 2012. 380;(9857): 1884-6.

Jiang Y, Resch S, Liu X, Rogers jr SO, Askari R, Klompas M, et al. The Cost of Responding to

an Acinetobacter Outbreak in Critically Ill Surgical Patients. Surgical Infections. 2016. 17;(1): 58-

64.

Johannsson B, Beekmann S, Srinivasan A, Hersh A, Laxminarayan R and Polgreen P

Improving antimicrobial stewardship: the evolution of programmatic strategies and barriers.

Infection Control and Hospital Epidemiology. 2011. 32;(4): 367-74.

169 | Appendix I

Johnston CP, Qiu H, Ticehurst JR, Dickson C, Rosenbaum P, Lawson P, et al. Outbreak

management and implications of a nosocomial norovirus outbreak. Clinical Infectious Diseases.

2007. 45;(5): 534-40.

Jönsson B Ten arguments for a societal perspective in the economic evaluation of medical

innovations. The European Journal of Health Economics. 2009. 10;(4): 357-9.

Jurke A, Köck RR, Becker K, Thole S, Hendrix R, Rossen JW, et al. Reduction of the

nosocomial meticillin-resistant Staphylococcus aureus incidence density by a region-wide

search and follow-strategy in forty German hospitals of the EUREGIO, 2009 to 2011.

Eurosurveillance. 2013. 18;(36): pii= 20860.

Kaan JA [Experiences with practical microbiology in Germany; what happens when the clinical

microbiologist is put at a distance. An interview with W.E. Silvis, clinical microbiologist in

Twente.]. Nederlands Tijdschrift Voor Medische Microbiologie. 2015. 23;(3): 95-100.

Kawai K, Preaud E, Baron Papillon F, Largeron N and Acosta C Cost-effectiveness of

vaccination against herpes zoster and postherpetic neuralgia: a critical review. Vaccine. 2014.

32;(15): 1645-53.

Kern WV, de With K, Nink K, Steib Bauert M and Schröder H Regional variation in

outpatient antibiotic prescribing in Germany. Infection. 2006. 34;(5): 269-73.

Keuleyan E and Gould M Key issues in developing antibiotic policies: from an institutional

level to Europe-wide. European Study Group on Antibiotic Policy (ESGAP), Subgroup III.

Clinical Microbiology and Infection. 2001. 7 Suppl 6;16-21.

Klevens RM, Edwards JR, Richards Jr CL, Horan TC, Gaynes RP, Pollock DA, et al.

Estimating health care-associated infections and deaths in U.S. hospitals, 2002. Public Health

Reports. 2007. 122;(2): 160-6.

Koch R. Die Ätiologie der Milzbrand-Krankheit, begründet auf die Entwicklungsgeschichte

des Bacillus anthracis. Beitrage Zur Biologie Der Pflanzen. 1876. 2;(2): 277-310.

Köck RR, Becker K, Cookson B, van Gemert-Pijnen JEWC, Harbarth S, Kluytmans JAJW, et

al. Systematic literature analysis and review of targeted preventive measures to limit healthcare-

associated infections by meticillin-resistant Staphylococcus aureus. Eurosurveillance. 2014.

19;(37): pii: 20902.

Köck RR, Brakensiek L, Mellmann A, Kipp F, Henderikx M, Harmsen D, et al. Cross-border

comparison of the admission prevalence and clonal structure of meticillin-resistant

Staphylococcus aureus. Journal of Hospital Infection. 2009. 71;(4): 320-6.

Koenders AG State of the European Union 2015 - The Presidency Edition. 2015. 1-24.

Ministerie van Buitenlandse Zaken. Den Haag, Netherlands.

Bibliography | 170

Koller D, Hoffmann F, Maier W, Tholen K, Windt R and Glaeske G Variation in antibiotic

prescriptions: is area deprivation an explanation? Analysis of 1.2 million children in Germany.

Infection. 2013. 41;(1): 121-7.

Kusters R, Pronk C and Kluytmans JA. Efficiency van labtesten niet op waarde geschat; Stop

met ongenuanceerd bezuinigen op laboratoriumdiagnostiek. De zorg heeft er op alle fronten

baat bij. Het Financieele Dagblad. 2014.

Lammie SL and Hughes JM Antimicrobial Resistance, Food Safety, and One Health: The

Need for Convergence. Annual Review of Food Science and Technology. 2016.

Laxminarayan R, Duse A, Wattal C, Zaidi AK, Wertheim HF, Sumpradit N, et al. Antibiotic

resistance-the need for global solutions. Lancet Infectious Diseases. 2013. 13;(12): 1057-98.

Lesprit P, Landelle C and Brun-Buisson C Clinical impact of unsolicited post-prescription

antibiotic review in surgical and medical wards: a randomized controlled trial. Clinical

Microbiology and Infection. 2013. 19;(2): 91-7.

Liew YX, Lee W, Tay D, Tang SS, Chua NG, Zhou Y, et al. Prospective audit and feedback in

antimicrobial stewardship: Is there value in early reviewing within 48h of antibiotic

prescription?. International Journal of Antimicrobial Agents. 2015. 45;(2): 168-73.

Lo-Ten-Foe JR, Sinha B, Wilting KR, Veenstra-Kyuchukova Y, Panday PN and Hendrix R

[Bedside consultation by a multidisciplinary antibiotics team: an Antibiotic Stewardship

Programme at UMCG]. Nederlands Tijdschrift Voor Geneeskunde. 2014. 158;(5): A6795-.

Lui YY, Wang Y, Walsh TR, Yi LX, Zhang R, Spencer J, et al. Emergence of plasmid-mediated

colistin resistance mechanism MCR-1 in animals and human beings in China: a microbiological

and molecular biological study. Lancet Infectious Diseases. 2015. epub before print;

Maddox M, DeBoer E and Hammerquist R Administration of Extended Infusion Piperacillin-

TazobactamWith the Use of Smart Pump Technology. Hospital Pharmacy. 2014. 49;(5): 444-8.

Malhotra-Kumar S, Britto Xavier B, Das AJ, Lammens C, Butaye P and Goossens H Colistin

resistance gene mcr-1 harboured on a multidrug resistant plasmid. The Lancet Infectious Diseases.

2016.

Mangione Smith R, Elliott M, McDonald L and McGlynn E An observational study of

antibiotic prescribing behavior and the Hawthorne effect. Health Services Research. 2002. 37;(6):

1603-23.

Marshall BM and Levy SB Food animals and antimicrobials: impacts on human health. Clinical

Microbiology Reviews. 2011. 24;(4): 718-33.

171 | Appendix I

Marwick CA and Nathwani D Linking process measures to outcome for patients with

complicated urinary tract infection: it's complicated. Clinical Infectious Diseases. 2014. 58;170-2.

Mather CA, Werth BJ, Sivagnanam S, SenGupta DJ and Butler-Wu SM Rapid Detection of

Vancomycin-Intermediate Staphylococcus aureus by Matrix-Assisted Laser Desorption

Ionization-Time of Flight Mass Spectrometry. Journal of Clinical Microbiology. 2016. 54;(4): 883-

90.

Matteelli A, Roggi A and Carvalho AC Extensively drug-resistant tuberculosis: epidemiology

and management. Clinical Epidemiology. 2014. 6;111-8.

McGowan J Antimicrobial stewardship--the state of the art in 2011: focus on outcome and

methods. Infection Control and Hospital Epidemiology. 2012. 33;(4): 331-7.

McGowan JE and Gerding DN Does antibiotic restriction prevent resistance?. New Horizons.

1996. 4;(3): 370-6.

McMurray BR, Wrenn KD and Wright SW Usefulness of blood cultures in pyelonephritis. The

American Journal of Emergency Medicine. 1997. 15;(2): 137-40.

Mitchell BG, Hall L, MacBeth D, Gardner A and Halton K Hospital infection control units:

Staffing, costs, and priorities. American Journal of Infection Control. 2015. 43;(6): 612-6.

Moher D, Liberati A, Tetzlaff J and Altman D Preferred reporting items for systematic reviews

and meta-analyses: the PRISMA statement. PLoS Medicine. 2009. 6;(7): e1000097.

Morgan D and Astolfi R Health spending continues to stagnate in many OECD countries.

2014. OECD Health Working Papers, No. 68;25. OECD Publishing. Paris, France.

Morris AM Antimicrobial Stewardship Programs: Appropriate Measures and Metrics to Study

their Impact. Current Treatment Options in Infectious Diseases. 2014. 6;(2): 101-12.

Müller J, Voss A, Köck RR, Sinha B, Rossen JW, Kaase M, et al. Cross-border comparison of

the Dutch and German guidelines on multidrug-resistant Gram-negative microorganisms.

Antimicrobial Resistance and Infection Control. 2015. 4;(7):

Muñoz P, Valerio M, Vena A and Bouza E Antifungal stewardship in daily practice and health

economic implications. Mycoses. 2015. 58;14-25.

Nathwani D, Lawson W, Dryden M, Stephens J, Corman S, Solem C, et al. Implementing

criteria-based early switch/early discharge programmes: a European perspective. Clinical

Microbiology and Infection. 2015. 21;47-55.

Bibliography | 172

Nathwani D, Rubinstein E, Barlow G and Davey P Do guidelines for community-acquired

pneumonia improve the cost-effectiveness of hospital care?. Clinical Infectious Diseases. 2001.

32;(5): 728-41.

Nathwani D, Sneddon J, Patton A and Malcolm W Antimicrobial stewardship in Scotland:

impact of a national programme. Antimicrobial Resistance and Infection Control. 2012. 1;(1): 7.

Navas E, Torner N, Broner S, Godoy P, Martínez A, Bartolomé R, et al. Economic costs of

outbreaks of acute viral gastroenteritis due to norovirus in Catalonia (Spain), 2010-2011. BMC

Public Health. 2015. 15;999.

NHG. [Acute paediatric otitis media ]. NHG. 2014. Available from: https://www.nhg.org/

standaarden/samenvatting/otitis-media-acuta-bij-kinderen. Accessed at: May 15, 2015.

NHG. [Acute coughing]. NHG. , 2013. Available from: https://www.nhg.org/standaarden/

samenvatting/acuut-hoesten. Accessed at: May 15, 2015.

Niwa T, Shinoda Y, Suzuki A, Ohmori T, Yasuda M, Ohta H, et al. Outcome measurement of

extensive implementation of antimicrobial stewardship in patients receiving intravenous

antibiotics in a Japanese university hospital. International Journal of Clinical Practice. 2012. 66;(10):

999-1008.

Nobre V, Harbarth S, Graf JD, Rohner P and Pugin J Use of procalcitonin to shorten

antibiotic treatment duration in septic patients: a randomized trial. American Journal of Respiratory

and Critical Care Medicine. 2008. 177;(5): 498-505.

Nowak MN, Breidenbach R, Thompson J and Carson P Clinical and economic outcomes of a

prospective antimicrobial stewardship program. American Journal of Health-System Pharmacy. 2012.

69;(17): 1500-8.

NVMM. [Norm infection prevention specialists 2012]. Utrecht, Netherlands: NVMM. , 2012a.

Available from: http://www.nvmm.nl/system/files/KRIZ%20-%20norm%20versie

%202%200.pdf. Accessed at: November 20, 2015.

NVMM. Tussen laboratorium en kliniek. 2012b.1-23. NVMM. Utrecht, Netherlands.

NZa. Stand van de zorgmarkten 2015. 2015.1-54. NZa. Utrecht, Netherlands.

Oberjé E, Tanke M and Jeurissen P Cost-Effectiveness of Policies to Limit Antimicrobial

Resistance in Dutch Healthcare Organisations. 2016. Celsius. Nijmegen, Netherlands.

Oosterheert J, van Loon A, Schuurman R, Hoepelman AIM, Hak E, Thijsen S, et al. Impact of

rapid detection of viral and atypical bacterial pathogens by real-time polymerase chain reaction

for patients with lower respiratory tract infection. Clinical Infectious Diseases. 2005. 41;(10): 1438-

44.

173 | Appendix I

Oppong R, Jit M, Smith R, Butler C, Melbye H, Mölstad S, et al. Cost-effectiveness of point-

of-care C-reactive protein testing to inform antibiotic prescribing decisions. British Journal of

General Practice. 2013. 63;(612): e465-71.

Otters HBM, van der Wouden JC, Schellevis F, van Suijlekom-Smit LWA and Koes B Trends

in prescribing antibiotics for children in Dutch general practice. Journal of Antimicrobial

Chemotherapy. 2004. 53;(2): 361-6.

Page K, Graves N, Halton K and Barnett AG Humans, 'things' and space: costing hospital

infection control interventions. Journal of Hospital Infection. 2013. 84;(3): 200-5.

Pasteur L Translation of an adress on the Germ Theory. The Lancet. 1881. 118;(3024): 271-2.

Pene F, Courtine E, Cariou A and Mira JP Toward theragnostics. Critical Care Medicine. 2009.

37;(1 Suppl): S50-58.

Perencevich E, Stone PW, Wright S, Carmeli Y, Fisman D and Cosgrove SE Raising standards

while watching the bottom line: making a business case for infection control. Infection Control

and Hospital Epidemiology. 2007. 28;(10): 1121-33.

Perez K, Olsen R, Musick W, Cernoch P, Davis J, Land G, et al. Integrating rapid pathogen

identification and antimicrobial stewardship significantly decreases hospital costs. Archives of

Pathology & Laboratory Medicine. 2013. 137;(9): 1247-54.

Perez K, Olsen R, Musick W, Cernoch P, Davis J, Peterson L, et al. Integrating rapid

diagnostics and antimicrobial stewardship improves outcomes in patients with antibiotic-

resistant Gram-negative bacteremia. The Journal of Infection. 2014. 69;(3): 216-25.

Philips H, Huibers L, Holm Hansen E, Bondo Christensen M, Leutgeb R, Klemenc-Ketis Z, et

al. Guidelines adherence to lower urinary tract infection treatment in out-of-hours primary care

in European countries. Quality in Primary Care. 2014. 22;(4): 221-31.

Piednoir E, Thibon P, Borderan GC, Godde F, Borgey F, Le Coutour X, et al. Long-term

clinical and economic benefits associated with the management of a nosocomial outbreak

resulting from extended-spectrum beta-lactamase-producing Klebsiella pneumoniae. Critical

Care Medicine. 2011. 39;(12): 2672-7.

Plantinga NL and Bonten MJ Selective decontamination and antibiotic resistance in ICUs.

Critical Care. 2015. 19;259.

Plowman R, Graves N, Griffin MA, Roberts JA, Swan AV, Cookson B, et al. The rate and cost

of hospital-acquired infections occurring in patients admitted to selected specialties of a district

general hospital in England and the national burden imposed. Journal of Hospital Infection. 2001.

47;(3): 198-209.

Bibliography | 174

Poelman R, Schölvinck EH, Borger R, Niesters HG and van Leer-Buter C The emergence of

enterovirus D68 in a Dutch University Medical Center and the necessity for routinely screening

for respiratory viruses. Journal of Clinical Virology. 2015. 62;1-5.

Polk RE, Fox C, Mahoney A and Letcavage JM,C. Measurement of adult antibacterial drug use

in 130 US hospitals: comparison of defined daily dose and days of therapy. Clinical Infectious

Diseases. 2007. 44;(5): 664-70.

Popowitch EB, O'Neill SS and Miller MB Comparison of the Biofire FilmArray RP, Genmark

eSensor RVP, Luminex xTAG RVPv1, and Luminex xTAG RVP fast multiplex assays for

detection of respiratory viruses. Journal of Clinical Microbiology. 2013. 51;(5): 1528-33.

Pulcini C, Defres S, Aggarwal I, Nathwani D and Davey P Design of a 'day 3 bundle' to

improve the reassessment of inpatient empirical antibiotic prescriptions. Journal of Antimicrobial

Chemotherapy. 2008a. 61;(6): 1384-8.

Pulcini C, Defres S, Aggarwal I, Nathwani D and Davey P Design of a 'day 3 bundle' to

improve the reassessment of inpatient empirical antibiotic prescriptions. Journal of Antimicrobial

Chemotherapy. 2008b. 61;(6): 1384-8.

Rahamat-Langendoen JC, Lokate M, Schölvinck EH, Friedrich AW and Niesters HG Rapid

detection of a norovirus pseudo-outbreak by using real-time sequence based information.

Journal of Clinical Virology. 2013. 58;(1): 245-8.

Rauh S, Wadsworth E and Weeks W The fixed-cost dilemma: what counts when counting

cost-reduction efforts?. Healthcare Financial Management. 2010. 64;(3): 60-3.

Reissig A, Mempel C, Schumacher U, Copetti R, Gross F and Aliberti S Microbiological

diagnosis and antibiotic therapy in patients with community-acquired pneumonia and acute

COPD exacerbation in daily clinical practice: comparison to current guidelines. Lung. 2013.

191;(3): 239-46.

Renwick MJ, Simpkin V and Mossialos E International and European Initiatives Targeting

Innovation in Antibiotic Drug Discovery and Development. The Need of a One Health - One

Europe - One World Framework. 2016.1-93. Dutch Ministry of Health. Den Haag,

Netherlands. Available at: http://www.eu2016.nl/binaries/eu2016/documenten/

publicaties/2016/02/10/2016-report-on-antibiotic-rd-initiatives/2016-report-on-antibiotic-rd-

initiaitives.pdf.

Review on Antimicrobial Resistance Tackling drug-resistant infection globally: Final report and

recommendations. 2016. AMR Review. Londen, United Kingdom. Available at: http://amr-

review.org/sites/default/files/160525_Final%20paper_with%20cover.pdf. Accessed at: May

27, 2016.

http://amr-review.org/sites/default/files/160525_Final%20paper_with%20cover.pdf

http://amr-review.org/sites/default/files/160525_Final%20paper_with%20cover.pdf

175 | Appendix I

Rhame FS and Sudderth WD Incidence and prevalence as used in the analysis of the

occurrence of nosocomial infections. American Journal of Epidemiology. 1981. 113;(1): 1-11.

Rimawi R, Cook P, Gooch M, Kabchi B, Ashraf M, Rimawi B, et al. The impact of penicillin

skin testing on clinical practice and antimicrobial stewardship. Journal of Hospital Medicine. 2013.

8;(6): 341-5.

Roberts R, Frutos PW, Ciavarella GG, Gussow LM, Mensah EK, Kampe LM, et al.

Distribution of variable vs fixed costs of hospital care. Journal of the American Medical Association.

1999. 281;(7): 644-9.

Roberts R, Hota B, Ahmad I, Scott RD, Foster S, Abbasi F, et al. Hospital and societal costs of

antimicrobial-resistant infections in a Chicago teaching hospital: implications for antibiotic

stewardship. Clinical Infectious Diseases. 2009. 49;(8): 1175-84.

Rogers BB, Shankar P, Jerris RC, Kotzbauer D, Anderson EJ, Watson JR, et al. Impact of a

rapid respiratory panel test on patient outcomes. Archives of Pathology and Laboratory Medicine.

2015. 139;(5): 636-41.

Rogowski JA, Staiger D, Patrick T, Horbar J, Kenny M and Lake ET Nurse staffing and NICU

infection rates. JAMA Pediatrics. 2013. 167;(5): 444-50.

Rudholm N Economic implications of antibiotic resistance in a global economy. Journal of

Health Economics. 2002. 21;(6): 1071-83.

Rüttimann S, Keck B, Hartmeier C, Maetzel A and Bucher H Long-term antibiotic cost savings

from a comprehensive intervention program in a medical department of a university-affiliated

teaching hospital. Clinical Infectious Diseases. 2004. 38;(3): 348-56.

Sadique Z, Lopman B, Cooper B and Edmunds WJ Cost-effectiveness of Ward Closure to

Control Outbreaks of Norovirus Infection in United Kingdom National Health Service

Hospitals. Journal of Infectious Diseases. 2016. 213;19-26.

Schippers EI and van Rijn MJ Geamenlijke agenda VWS 'Van systemen naar mensen'. 2013.

MEVA/AEB-3155166;1-25. Ministerie van Volksgezond, Welzijn en Sport. Den Haag,

Netherlands. Available at: https://www.rijksoverheid.nl/binaries/rijksoverheid/

documenten/kamerstukken/2013/02/08/kamerbrief-gezamenlijke-agenda-vws-van-systemen-

naar-mensen/meva-3155166a.pdf. Accessed at: November 3, 2015.

Schuetz P, Müller B, Christ-Crain M, Stolz D, Tamm M, Bouadma L, et al. Procalcitonin to

initiate or discontinue antibiotics in acute respiratory tract infections. Cochrane Library, The.

2012. 9;CD007498.

Bibliography | 176

Schulz LT, Fox BC and Polk RE Can the antibiogram be used to assess microbiologic

outcomes after antimicrobial stewardship interventions? A critical review of the literature.

Pharmacotherapy. 2012. 32;(8): 668-76.

Schuts EC, Hulscher MEJL, Mouton JW, Verduin CM, Cohen Stuart JWT, Overdiek JW, et al.

A systematic review and meta-analysis of current evidence on hospital Antimicrobial

Stewardship objectives. Lancet Infectious Diseases. 2016.

Scott RD, Solomon SL and McGowan JE Applying economic principles to health care.

Emerging Infectious Diseases. 2001. 7;(2): 282-5.

Senn L, Burnand B, Francioli P and Zanetti G Improving appropriateness of antibiotic

therapy: randomized trial of an intervention to foster reassessment of prescription after 3 days.

Journal of Antimicrobial Chemotherapy. 2004. 53;(6): 1062-7.

Sevinç F, Prins JM, Koopmans RP, Langendijk PNJ, Dankert J and Speelman P Vroege

omzetting van intraveneuze naar orale antibiotica: 'switchtherapie'. Nederlands Tijdschrift Voor

Geneeskunde. 1999. 143;(47): 2364-5.

Sick A, Lehmann C, Tamma P, Lee CKK and Agwu A Sustained savings from a longitudinal

cost analysis of an internet-based preapproval antimicrobial stewardship program. Infection

Control and Hospital Epidemiology. 2013. 34;(6): 573-80.

Slimings C and Riley TV Antibiotics and hospital-acquired Clostridium difficile infection:

update of systematic review and meta-analysis. Journal of Antimicrobial Chemotherapy. 2014. 69;(4):

881-91.

Smith R and Coast J The true cost of antimicrobial resistance. British Medical Journal. 2013.

346;f1493.

Society for Healthcare Epidemiology of America, Infectious Diseases Society of America and

Pediatric Infectious Diseases Society Policy statement on antimicrobial stewardship by the

Society for Healthcare Epidemiology of America (SHEA), the Infectious Diseases Society of

America (IDSA), and the Pediatric Infectious Diseases Society (PIDS). Infection Control and

Hospital Epidemiology. 2012. 33;(4): 322-7.

Spijkerman I, Ruijs G and Kluytmans JA [The future of infection prevention. Part 1 guidelines

and implementation.]. Nederlands Tijdschrift Voor Medische Microbiologie. 2012a. 20;(2): 70-2.

Spijkerman I, Ruijs G and Kluytmans JA [The future of infection prevention. Part 2 Norm,

collaboration, education and funding.]. Nederlands Tijdschrift Voor Medische Microbiologie. 2012b.

20;(4): 149-52.

177 | Appendix I

Spoorenberg V, Hulscher MEJL, Akkermans R, Prins J and Geerlings S Appropriate antibiotic

use for patients with urinary tract infections reduces length of hospital stay. Clinical Infectious

Diseases. 2014. 58;(2): 164-9.

Stevenson K, Balada Llasat JM, Bauer K, Deutscher M, Goff D, Lustberg M, et al. The

economics of antimicrobial stewardship: the current state of the art and applying the business

case model. Infection Control and Hospital Epidemiology. 2012. 33;(4): 389-97.

Stewardson A, Harbarth S and Graves N Valuation of hospital bed-days released by infection

control programs: a comparison of methods. Infection Control and Hospital Epidemiology. 2014.

35;(10): 1294-7.

Stone PW, Braccia D and Larson E Systematic review of economic analyses of health care-

associated infections. American Journal of Infection Control. 2005. 33;(9): 501-9.

Stone PW, Larson E and Kawar L A systematic audit of economic evidence linking

nosocomial infections and infection control interventions: 1990-2000. American Journal of

Infection Control. 2002. 30;(3): 145-52.

Stone PW, Mooney-Kane C, Larson EL, Horan T, Glance LG, Zwanziger J, et al. Nurse

working conditions and patient safety outcomes. Medical Care. 2007. 45;(6): 571-8.

Stone PW, Pogorzelska-Maziarz M, Herzig CT, Weiner LM, Furuya EY, Dick A, et al. State of

infection prevention in US hospitals enrolled in the National Health and Safety Network.

American Journal of Infection Control. 2014. 42;(2): 94-9.

Stone SP, Fuller C, Savage J, Cookson B, Hayward A, Cooper B, et al. Evaluation of the

national Cleanyourhands campaign to reduce Staphylococcus aureus bacteraemia and

Clostridium difficile infection in hospitals in England and Wales by improved hand hygiene:

four year, prospective, ecological, interrupted time series study. BMJ: British Medical Journal.

2012. 344;

SWAB De kwaliteit van het antibioticabeleid in Nederland. 2012.

Tacconelli E, Cataldo MA, Paul M, Leibovici L, Kluytmans JA, Schröder W, et al. STROBE-

AMS: recommendations to optimise reporting of epidemiological studies on antimicrobial

resistance and informing improvement in antimicrobial stewardship. BMJ Open. 2016. 6;(2):

Tacconelli E, De Angelis G, Cataldo MA, Mantengoli E, Spanu T, Pan A, et al. Antibiotic

Usage and Risk of Colonization and Infection with Antibiotic-Resistant Bacteria: a Hospital

Population-Based Study. Antimicrobial Agents and Chemotherapy. 2009a. 53;(10): 4264-9.

Tacconelli E, De Angelis G, De Waure C, Cataldo MA, La Torre G and Cauda R Rapid

screening tests for meticillin-resistant Staphylococcus aureus at hospital admission: systematic

review and meta-analysis. Lancet Infectious Diseases. 2009b. 9;(9): 546-54.

Bibliography | 178

Taheri PA, Butz DA and Greenfield LJ Length of stay has minimal impact on the cost of

hospital admission. Journal of the American College of Surgeons. 2000. 191;(2): 123-30.

Tan C, Ritchie M, Alldred J and Daneman N Validating hospital antibiotic purchasing data as a

metric of inpatient antibiotic use. Journal of Antimicrobial Chemotherapy. 2016. 71;(2): 547-53.

Task Force for Combating Antibiotic-Resistance Bacteria National Action Plan for Combating

Antibiotic-Resistant Bacteria. 2015.1-63. The White House. Washington, DC, USA. Available

at: https://www.whitehouse.gov/sites/default/files/docs/national_action_plan_for_

combating_antibotic-resistant_bacteria.pdf. Accessed at: December 21, 2015.

Ternhag A, Grünewald M, Nauclér P and Wisell KT Antibiotic consumption in relation to

socio-demographic factors, co-morbidity, and accessibility of primary health care. Scandinavian

Journal of Infectious Diseases. 2014. 46;(12): 888-96.

Tersmette T, Ruijs G. and Friedrich AW. Verkeerde zuinigheid duur betaald;

Ongecontroleerde verspreiding van multiresistentie bacteriën kan miljoenen kosten. Het

Financieele Dagblad. 2012, 6.

The European Parliament and the Council of the European Union. Directive 2011/24/EU on

the application of patients' rights in cross-border healthcare. 2011.

Tice A, Turpin R, Hoey C, Lipsky B, Wu J and Abramson M Comparative costs of ertapenem

and piperacillin-tazobactam in the treatment of diabetic foot infections. American Journal of

Health-System Pharmacy. 2007. 64;(10): 1080-6.

Tunkel AR, Hartman BJ, Kaplan SL, Kaufman BA, Roos KL, Scheld WM, et al. Practice

guidelines for the management of bacterial meningitis. Clinical Infectious Diseases. 2004. 39;(9):

1267-84.

Ubbink DT, Papadopoulos NE and Legemate DA Slimmere ziekenhuiszorg. Nederlands

Tijdschrift Voor Geneeskunde. 2014. 158;A6957.

Valerio M, Muñoz P, Rodriguez CG, Caliz B, Paidlla B, Fernández-Cruz A, et al. Antifungal

stewardship in a tertiary-care institution: a bedside intervention. Clinical Microbiology and Infection.

2015. 21;(5): 1-9.

van Cleef BAGL, Kluytmans JAJW, van Benthem BHB, Haenen A, Monen J, Daniels Haardt

I, et al. Cross border comparison of MRSA bacteraemia between The Netherlands and North

Rhine-Westphalia (Germany): a cross-sectional study. PLoS ONE. 2012. 7;(8): e42787.

van den Bosch CM, Geerlings SE, Natsch S, Prins JM and Hulscher ME Quality indicators to

measure appropriate antibiotic use in hospitalized adults. Clinical Infectious Diseases. 2015. 60;(2):

281-91.

179 | Appendix I

van den Brink R. [The human and financial costs]. In: Het einde van de antibiotica. Van den Brink

R ed. De Geus. Breda, the Netherlands. 2013.

van den Broek PJ, Kluytmans JA, Ummels LC, Voss A and Vandenbroucke-Grauls CM How

many infection control staff do we need in hospitals?. Journal of Hospital Infection. 2007. 65;(2):

108-11.

van der Bij AK, Kardamanidis K, Frakking FN, Bonten MJ and Signaleringsoverleg

Ziekenhuisinfecties en Antimicrobiële Resistentie [Nosocomial outbreaks and resistant

microorganisms]. Nederlands Tijdschrift Voor Geneeskunde. 2015. 159;

van der Linden MW, van Suijlekom-Smit LWA, Schellevis FG and van der Wouden JC

[Second nationwide study on diseases and procedures at general practioners]. 2005.172-32.

NIVEL. Utrecht.

van der Velden LBJ, Tromp M, Bleeker Rovers CP, Hulscher M, Kullberg BJ, Mouton JW, et

al. Non-adherence to antimicrobial treatment guidelines results in more broad-spectrum but

not more appropriate therapy. European Journal of Clinical Microbiology & Infectious Diseases. 2012.

31;(7): 1561-8.

van Gemert-Pijnen JE, Peters O and Ossebaard HC. Improving eHealth. Eleven International

Publishing. Den Haag, the Netherlands. 2013.

van Gemert-Pijnen JEWC, Nijland N, van Limburg M, Ossebaard HC, Kelders SM,

Eysenbach G, et al. A holistic framework to improve the uptake and impact of eHealth

technologies. Journal of Medical Internet Research. 2011. 13;(4): e111.

van Leeuwenhoeck A Observations, communicated to the publisher by Mr. Antony van

Leeuwenhoeck, in a Dutch letter of the 9th of Octob. 1676. Here English'd: Concerning little

animals by him observed in rain- well- sea- and snow water; as also water wherein pepper had

lain infused. Philosophical Transactions. 1677. 12;821-31.

van Limburg M, Sinha B, Lo-Ten-Foe JR and van Gemert-Pijnen JEWC Evaluation of early

implementations of antibiotic stewardship program initiatives in nine Dutch hospitals.

Antimicrobial Resistance and Infection Control. 2014. 3;(1): 33.

van Limburg M, van Gemert-Pijnen JEWC, Nijland N, Ossebaard HC, Hendrix R and Seydel

ER Why business modeling is crucial in the development of eHealth technologies. Journal of

Medical Internet Research. 2011. 13;(4): e124.

van Zanten ARH, Engelfriet P, van Dillen K, van Veen M, Nuijten MJC and Polderman K

Importance of nondrug costs of intravenous antibiotic therapy. Critical Care. 2003. 7;(6): R184-

90.

Bibliography | 180

Verhoeven F, Steehouder MF, Hendrix RMG and van Gemert-Pijnen JEWC How nurses seek

and evaluate clinical guidelines on the Internet. Journal of Advanced Nursing. 2010-1. 66;(1): 114-

27.

Verhoeven F, Van Gemert-Pijnen JEWC, Hendrix R, Friedrich AW and Daniels-Haardt I.

Development of a web-based learning tool to enhance healthcare workers’ knowledge, attitude,

and risk perception about safe work practices concerning meticillin-resistant Staphylococcus

aureus. 18th European Congress of Clinical Microbiology and Infectious Diseases. Barcelona, 2008.

VGZ. Diagnostiek 60 mln. besparing en hogere kwaliteit. VGZ. 2013. Available from:

https://www.cooperatievgz.nl/newsroom/diagnostiek-60-mln.-besparing-en-hogere-kwaliteit

VHIG. [Norm infection prevention specialists]. Naarden, Netherlands: VHIG. 2006. Available from:

http://www.vhig.nl/images/documenten_besloten/Normen/2006_norm_

deskundige_infectiepreventie.pdf. Accessed at: November 20, 2015.

Visser S, Schuiling-Veninga CCM, Bos JHJ, de Jong-van den Berg LTW and Postma MJ The

population-based prescription database IADB.nl: its development, usefulness in outcomes

research and challenges. Expert Review of Pharmacoeconomics & Outcomes Research. 2013. 13;(3):

285-92.

Wagenlehner FME, Bartoletti R, Cek M, Grabe M, Kahlmeter G, Pickard R, et al. Antibiotic

stewardship: a call for action by the urologic community. European Urology. 2013. 64;(3): 358-60.

Wassenberg M, Kluytmans JA, Erdkamp S, Bosboom R, Buiting A, van Elzakker E, et al.

Costs and benefits of rapid screening of methicillin-resistant Staphylococcus aureus carriage in

intensive care units: a prospective multicenter study. Critical Care. 2012. 16;(1):

Weinstein MP, Towns ML, Quartey SM, Mirrett S, Reimer LG, Parmigiani G, et al. The clinical

significance of positive blood cultures in the 1990s: a prospective comprehensive evaluation of

the microbiology, epidemiology, and outcome of bacteremia and fungemia in adults. Clinical

Infectious Diseases. 1997. 24;(4): 584-602.

Wentzel J, van Velzen L, van Limburg M, de Jong N, Karreman J, Hendrix R, et al.

Participatory eHealth development to support nurses in antimicrobial stewardship. BMC

Medical Informatics and Decision Making. 2014. 14;(1): 45.

Wentzel MJ, de Jong N, Karreman J and van Gemert-Pijnen JEWC. Implementation of MRSA

infection prevention and control measures. What works in practice? In: Sudhaker C ed. InTech. Reijka,

Croatia. 2012.

Werkgroep Infectie Preventie. [MRSA guideline Hospitals.]. Utrecht, the Netherlands: WIP. ,

2012. Available from: http://www.rivm.nl/dsresource?objectid=rivmp:46603&type=

org&disposition=inline&ns_nc=1. Accessed at: November 20, 2015.

https://www.cooperatievgz.nl/newsroom/diagnostiek-60-mln.-besparing-en-hogere-kwaliteit

181 | Appendix I

Wilson JW and Tsukayama DT Extensively Drug-Resistant Tuberculosis: Principles of

Resistance, Diagnosis, and Management. Mayo Clinic Proceedings. 2016. 91;(4): 482-95.

With de K, Allerberger F, Amann S, Apfalter P, Brodt H, Eckmanns T, Fellhauer M, Geiss

HK, Janata O, Krause R, Lemmen S, Meyer E, Mittermayer H, Porsche U, Presterl E, Reuter

S, Sinha B, Strauß R, Wechsler-Fördös A, Wenisch C and Kern WV S3-Leitlinie. Strategien zur

Sicherung rationaler Antibiotika-Anwendung im Krankenhaus. 2013. 092/001; AWMF.

Düsseldorf. Available at: http://www.awmf.org/uploads/tx_szleitlinien/092-

001l_S3_Antibiotika_Anwendung_im_Krankenhaus_2013-12.pdf.

Wolff A, Taylor S and McCabe J Using checklists and reminders in clinical pathways to

improve hospital inpatient care. Medical Journal of Australia. 2004. 181;(8): 428-31.

Wolkewitz M, Barnett A, Palomar Martinez M, Frank U, Schumacher M and IMPLEMENT

Study Group Interventions to control nosocomial infections: study designs and statistical

issues. Journal of Hospital Infection. 2014. 86;(2): 77-82.

Wood F, Phillips C, Brookes Howell L, Hood K, Verheij T, Coenen S, et al. Primary care

clinicians' perceptions of antibiotic resistance: a multi-country qualitative interview study.

Journal of Antimicrobial Chemotherapy. 2013. 68;(1): 237-43.

Woodhead M, Blasi F, Ewig S, Garau J, Huchon G, Ieven M, et al. Guidelines for the

management of adult lower respiratory tract infections--summary. Clinical Microbiology and

Infection. 2011. 17;1-24.

World Health Organization. Antimicrobial resistance: global report on surveillance 2014. 2014a

.1-232. WHO. Geneva, Switzerland. Available at:

http://apps.who.int/iris/bitstream/10665/112642/1/9789241564748_eng.pdf?ua=1.

Accessed at: December 21, 2015.

World Health Organization. Health systems in times of global economic crisis: An update of

the situation in the WHO - European Region. 2014b.27-37. WHO. Geneva, Switzerland.

World Health Organization. The evolving threat of antimicrobial resistance - Options for

action. 2012. Geneva, Switzerland.

Yen YH, Chen HY, Wuan Jin L, Lin YM, Shen W and Cheng KJ Clinical and economic

impact of a pharmacist-managed i.v.-to-p.o. conversion service for levofloxacin in Taiwan.

International Journal of Clinical Pharmacology and Therapeutics. 2012. 50;(2): 136-41.

Zhou K, Ferdous M, de Boer RF, Kooistra-Smid AMD, Grundmann H, Friedrich AW, et al.

The mosaic genome structure and phylogeny of Shiga toxin-producing Escherichia coli O104:H4

is driven by short term adaptation. Clinical Microbiology and Infection. 2015.

http://www.awmf.org/uploads/tx_szleitlinien/092-001l_S3_Antibiotika_Anwendung_im_Krankenhaus_2013-12.pdf

http://www.awmf.org/uploads/tx_szleitlinien/092-001l_S3_Antibiotika_Anwendung_im_Krankenhaus_2013-12.pdf

http://apps.who.int/iris/bitstream/10665/112642/1/9789241564748_eng.pdf?ua=1

Bibliography | 182

Zingg W, Colombo C, Jucker T and Bossart WR,C. Impact of an outbreak of norovirus

infection on hospital resources. Infection Control and Hospital Epidemiology. 2005. 26;(3): 263-7.

Zingg W, Mutters NT, Harbarth S and Friedrich AW Education in infection control: A need

for European certification. Clinical Microbiology and Infection. 2015. 21;(12): 1052-6.

184

Appendix II: English summary

II

185 | Appendix II

Healthcare expenditures are an important topic. As a country we aim at maintaining one of the

best healthcare systems in the world, where everyone can be provided with the best quality of

care. Such a system has considerable costs and with a growing elderly population, it is expected

that these costs will rise further in the future. One can spend each Euro only once. Thus is

important to do so with the highest return. In other words, the relationship between costs and

desired outcome should be as optimal as possible. This way one is ensured of efficient

spending. It is thus of utmost importance to know what the effects of treatments and

interventions are. How is the budget of a hospital department spent for example? What are the

effects of a treatment? In politics, at insurance companies, but also at healthcare providers

such as hospitals such questions are asked more and more. The research of this thesis covers

these questions regarding the Department of Medical Microbiology and Infection Prevention

(MMB) of the University Medical Center Groningen (UMCG), the Netherlands.

A hospital’s Department of Medical Microbiology is in place to make sure that several

questions that patients (and physicians) have are answered. What is the cause of an infection?

How should the infection be treated? How is spread of infections prevented? To answer these

questions, clinical microbiologists, infectious disease specialists, infection prevention nurses

and many laboratory analysts are day in day out busy with culturing and detecting

microorganisms, interpreting results of these cultures, giving treatment advice based on these

results and preventing spread of microorganisms within the hospital. This thesis covers mainly

the treatment and prevention of bacterial infections. Treatment of bacteria is becoming more

and more difficult, because they are becoming more and more resistant against the treatment

that physicians prescribe, antimicrobials, or more specifically in this thesis, antibiotics.

Widespread resistance against antimicrobials exists already for decades in clinical medicine,

basically as long as antimicrobials are used in a healthcare setting. This is partly caused by

inappropriate use. Initially, there were always new classes of antibiotics found and developed

that could be used if old ones were not efficient anymore due to resistance. However, during

the last twenty years, no new classes were found, just some new antibiotics within the existing

classes. It can be expected that also for the coming years, no new classes will be discovered.

This means we have to make do with what we have available right now. However, clinically

available antibiotics are less and less effective. To make sure that they do not completely lose

clinical effectiveness, it is important to curb resistance development. One way to do so is to

ensure that all antimicrobial treatment given to patients is with appropriate drugs, dosage and

duration. Treating too long, with too low doses or with the wrong antibiotic should be

prevented as much as possible. The Department of Medical Microbiology plays an important

role in appropriate infection management.

English summary | 186

Firstly, it is important to know what kind of patients are cared for in one’s hospital. Do they

carry resistant bacteria with them? Are they from another country with a different

epidemiology? How much antibiotics are prescribed within the region surrounding the

hospital? This last point is important because this is a contributor and indicator for resistance

(development). The research therefore started with a study looking at this. How much

antibiotics are prescribed in the north of the Netherlands. But also, how much is prescribed in

the bordering region of northwestern Germany. These are all patients that eventually might

end up in the UMCG. The study showed that there are big differences between the

Netherlands and Germany. Especially for the use of oral second-generation cephalosporins, an

antibiotic class that should be used with limitation, prescriptions are much more common in

Germany. Even though we are comparable, neighboring countries.

If one knows the patients living in the healthcare region of the hospital, you can adjust policies

and treatment to these people. It is important that the whole process of treatment is properly

documented: performing appropriate diagnostics, prescribing the appropriate therapy and

having proper infection prevention in place. The UMCG calls this the “AID Stewardship

Model”, comprising Antimicrobial Stewardship, Infection Prevention Stewardship and

Diagnostic Stewardship. It is based on the principle of Antimicrobial Stewardship, which is

currently implemented and rolled-out worldwide in healthcare institutions. The goal is to

provide optimal antimicrobial therapy. We argue that this is not enough, but that optimal

diagnostics and infection prevention are also essential. Without diagnostics it is impossible to

provide appropriate therapy and without infection prevention microorganisms can spread

uncontrollably. This thesis follows from here on this model, covering evaluations of all these

three subjects.

An Antimicrobial Stewardship Program has as focus the improvement of antimicrobial therapy

to provide the optimal antibiotic, dosage and duration. You can reach this goal by different

means, depending on the country, the setting, the healthcare institution, the patient groups, etc.

With different interventions and multiple ways of evaluation it is important to first consider

these factors. Next, one should establish the kind of intervention, the outcome measures and

the way of measuring these. It is also important to know what has been published already.

Financial evaluations are of a questionable quality so far, making comparison nearly impossible.

Keeping this in mind, we evaluated the UMCG ASP. A program that evaluates antimicrobial

therapy after two days to see if adjustments can be made based on diagnostic results. The

program was implemented at the urology ward and after one year patients were compared to

patients from the same ward before implementation of the ASP. This was done on a clinical

and financial level. Results were positive in all aspects investigated. Patients used less

antibiotics, especially intravenously, and less complex patients could be sent home earlier.

Financial benefits outweighed costs six times.

187 | Appendix II

An Infection Prevention Stewardship Program (ISP) focuses on the prevention of infections

due to spread within a hospital. Furthermore, it is important, in cases of spread (such as an

outbreak), to ensure swift control. However, if there is an outbreak, it is not clear how much

this the hospital costs. This was evaluated and the conclusion was, that per patient per day

there are extra costs of €519, a considerable figure. This shows that not just the clinical

consequences are major, but also the financial ones. It is also highly likely, that it is financially

beneficial to spend extra to prevent outbreaks. In the second ISP study we showed indeed that

this is the case. Over a time period of eight years a lot of extra money was invested in infection

prevention in the UMCG, e.g. for extra disposable clothing, disinfection alcohol, and more

surveillance cultures. Even though there were more patients entering the UCMG carrying

outbreak-causing bacteria, we detected no rise in the number of outbreak patients. It seems,

that the extra spending for infection prevention pays off by having less outbreak patients (and

their associated costs).

Finally, the Diagnostic Stewardship Program (DSP) had been addressed. Here, we looked at

one of the most performed diagnostic test in Clinical Microbiology: blood cultures. These

cultures are prescribed for almost all patients receiving antibiotics and the results can be used

to optimize therapy. For a group of nearly 3000 patients who received antibiotics at start of

their admission it was evaluated whether blood cultures were performed. This was the case in

only 48% of the patients. Alarmingly low, but unfortunately comparable with other studies in

Europe. We analyzed which factors were associated with the drawing of blood for cultures, as

well as which factors were associated with the length of stay of the two patient groups.

Performing blood cultures was associated with other diagnostics (such as CRP), suggesting a

bundle approach. Patients with blood cultures had a significantly shorter length of stay and

having blood cultures was positively associated with this. These data should trigger more

elaborate (prospective) studies looking at the effects of microbiological diagnostics.

We can conclude that the Department of Medical Microbiology has a highly positive financial

impact on a hospital-wide level. Interventions performed to improve antimicrobial therapy,

infection prevention and diagnostics all ensure better treatment of patients, improved patient

safety and eventually lower costs. This in turn creates room for increased efficiency and

revenues. The recommendation of this thesis therefore is to firmly embed costs for Medical

Microbiology and Infection Prevention and not to see these costs as easily-accessible short-

term targets for budget cuts. A solution for embedding would be to distribute costs and

benefits of infection prevention hospital-wide through an insurance model. Such a model

should provide a financial incentive for individual departments to improve infection

management and at the same time provide more clarity for all stakeholders regarding

distribution of budgets. Ultimately, prevention is the key to lower costs in the future, and

English summary | 188

improved quality of care by an optimal, high quality process-oriented approach will be highly

cost-effective.

190

Appendix III: Nederlandse samenvatting

III

191 | Appendix III

Kosten in de zorg zijn een belangrijk onderwerp. Als land willen we het beste zorgstelsel ter

wereld, waarin iedereen geholpen kan worden met de beste kwaliteit zorg. Hieraan zitten

echter kosten verbonden en zeker met de ouder wordende bevolking is de verwachting dat

zorgkosten de komende decennia nog aanzienlijk zullen stijgen. Elke euro in de zorg kun je

maar één keer uitgeven en het is daarom belangrijk dat je deze euro zo uitgeeft, dat je hiervoor

het meeste terug krijgt. Met andere woorden, de verhouding tussen de kosten en de gewenste

uitkomst moet zo optimaal mogelijk zijn. Op die manier zorg je dat er zo efficiënt mogelijk

wordt omgegaan met de beschikbare middelen. Het is dus van groot belang om te weten wat

de effecten zijn van behandelingen en interventies. Wat gebeurt het met het geld van een

afdeling in een ziekenhuis bijvoorbeeld? Wat is het effect van een behandeling? Dit soort

vragen worden steeds vaker gesteld. In de politiek, bij zorgverzekeraars, maar ook bij

zorgverleners zoals de ziekenhuizen. Het onderzoek dat behandeld wordt in deze thesis heeft

deze vragen gepoogd te beantwoorden voor de Afdeling Medische Microbiologie en

Infectiepreventie (MMB) van het Universitair Medisch Centrum Groningen (UMCG).

Een Afdeling Medische Microbiologie in een ziekenhuis is er om te zorgen dat een aantal

vragen van patiënten (en hun behandelaars) te beantwoorden. Wat is de oorzaak van een

infectie? Hoe moet je deze infectie behandelen? Hoe voorkom je dat besmettingen

plaatsvinden in het ziekenhuis? Om deze vragen te kunnen beantwoorden zijn medisch

microbiologen, infectiologen, deskundigen infectiepreventie en heel veel analisten elke dag

bezig met het kweken van micro-organismen, het interpreteren van uitslagen van deze kweken,

het geven van behandeladvies op basis hiervan en het voorkomen van verspreiding van micro-

organismen binnen het ziekenhuis. Deze thesis behandeld met name de behandeling en

preventie van bacteriële infecties. De behandeling van infecties veroorzaakt door bacteriën is

steeds vaker, steeds moeilijker. Dit komt omdat bacteriën steeds vaker resistent zijn voor de

behandeling die ziekenhuizen voorschrijven, antimicrobiële middelen oftewel antibiotica.

Wijdverbreide resistentie tegen antimicrobiële middelen komt al decennia lang voor, al zolang

er antimicrobiële middelen gebruikt worden in de gezondheidszorg. Dit wordt mede

veroorzaakt door verkeerd gebruik van deze middelen. Er werden echter ook altijd weer

nieuwe klassen antibiotica gevonden die gebruikt konden worden als de oude niet meer

efficiënt waren door opgekomen resistentie. De laatste twintig jaar zijn er echter geen nieuwe

klassen meer gevonden, enkel nog nieuwe antibiotica binnen de al bekende klassen. Het is de

verwachting dat ook de komende jaren er geen nieuwe klassen antibiotica gevonden gaan

worden. Dit betekent dat we het moeten doen met wat we nu hebben. En wat we nu hebben is

steeds vaker niet meer effectief. Om te voorkomen dat er straks geen enkel middel meer

effectief is, moet er worden gezorgd dat de ontwikkeling van resistentie zo veel mogelijk terug

gedrongen wordt. Dit kan onder meer door te zorgen dat de gegeven antibiotica aan patiënten

correct is, in de juiste dosis en voor de juiste duur. Te veel, te weinig of het verkeerde type

Nederlandse samenvatting | 192

moet zo veel mogelijk vermeden worden. De Afdeling Medische Microbiologie speelt hierin

een belangrijke rol.

Allereerst is het belangrijk om te weten wat voor patiënten er in je ziekenhuis liggen. Dragen

deze al resistentie bacteriën bij zich? Komen ze uit het buitenland met een andere

epidemiologie? Wordt er veel of weinig antibiotica voorgeschreven in de regio rondom het

ziekenhuis? Dit laatste is belangrijk, omdat dit een oorzaak, en dus ook indicator is voor

resistentie (ontwikkeling). Het onderzoek begint daarom met een studie die hier naar kijkt. Hoe

veel antibiotica wordt er voorgeschreven in Noord-Nederland. Maar tevens ook hoe veel

wordt er voorgeschreven in de aangrenzende regio in Noordwest-Duistland. Dit zijn allen

mensen die uiteindelijk in het UMCG terecht kunnen komen. De studie laat zien dat er grote

verschillen zijn tussen Nederland en Duitsland. Met name het gebruik van oraal gegeven

tweede generatie cefalosporines, een middel dat idealiter zo min mogelijk gebruikt wordt, ligt in

Duitsland veel hoger. En dat terwijl we vergelijkbare buurlanden zijn.

Als je weet wat voor patiënten in de regio van je ziekenhuis wonen, kun je beleid en

behandeling aanpassen op deze mensen. Hiervoor is het ook van belang dat het hele proces

van behandelen goed beschreven is: het doen van de juiste diagnostiek, het toepassen van de

juiste therapie en het hebben van goede infectiepreventie. Het UMCG noemt dit het AID

Stewardship Model: Antimicrobial Stewardship, Infection Prevention Stewardship en

Diagnostic Stewardship. Dit is gebaseerd op het principe van Antimicrobial Stewardship wat

inmiddels wereldwijd op verschillende manieren wordt geïmplementeerd in zorginstellingen.

Het doel hiervan is het geven van correcte antimicrobiële therapie. Wij betogen dat alleen dit

niet genoeg is, maar dat correcte diagnostiek en infectiepreventie ook van essentieel belang is.

Zonder diagnostiek is het onmogelijk om correcte therapie te geven en zonder

infectiepreventie vindt er ongebreidelde verspreiding van micro-organismen plaats. De thesis

volgt verder vanaf hier ook dit model, waarbij evaluaties van elke van deze drie onderdelen

afzonderlijk behandeld worden.

Een Antimicrobial Stewardship Program (ASP) gaat uit van stimulatie tot het geven van de

meest optimale antimicrobiële therapie. Dit kan op allerlei verschillende manieren en is

afhankelijk van het land, de instelling, de patiënt etc. Omdat er verschillende interventies zijn,

die je op verschillende manieren kunt evalueren, is het belangrijk daar eerst over na te denken.

Je moet vaststellen wat voor interventie er wordt geïmplementeerd, wat de uitkomstmaten zijn

er en hoe je deze meet. En ook belangrijk, wat is er al beschreven in de literatuur. Financiële

evaluaties zijn vaak van een matige kwaliteit en dit maakt vergelijkingen erg lastig. Dat in

ogenschouw nemend, is het ASP van het UMCG geëvalueerd. Een programma waarbij op dag

twee van antimicrobiële therapie, deze therapie geëvalueerd wordt om te kijken of deze

193 | Appendix III

wellicht aangepast kan worden op basis van gedane diagnostiek. Dit is geïmplementeerd op de

urologie afdeling en na een jaar tijd zijn deze patiënten vergeleken met patiënten die op

dezelfde afdeling lagen voordat het ASP geïmplementeerd werd. Dit is zowel op een klinisch

niveau als een financieel niveau geëvalueerd. De resultaten waren heel positief. Patiënten

gebruikten minder antibiotica, met name minder antibiotica die intraveneus wordt gegeven en

de minder complexe patiënten konden daardoor eerder naar huis toe. De financiële

opbrengsten van deze interventie waren zes keer groter dan de kosten van het ASP.

Een Infection Prevention Stewardship Program (ISP) richt zich op het voorkómen van

infecties door verspreiding in het ziekenhuis. Daarnaast is het belangrijk om, in het geval van

verspreiding (zoals bijvoorbeeld een uitbraak), te zorgen dat dit zo snel mogelijk tot controle

wordt gebracht. Maar als er nou een uitbraak is, wat kost dit het ziekenhuis dan? Dit is

onderzocht en de conclusie is dat het per positieve patiënt €519 per dag kost. Een aanzienlijk

bedrag dus. Wat laat zien dat niet alleen de klinische consequenties van een uitbraak groot zijn,

maar ook de financiële. Dit maakt, dat het waarschijnlijk financieel ook loont om kosten te

maken om zo uitbraken te voorkomen. In het tweede ISP onderzoek laten we dat zien.

Gedurende acht jaren is er veel extra geld geïnvesteerd in het UMCG aan infectiepreventie,

bijvoorbeeld aan extra wegwerpkleding, desinfectie alcohol en meer surveillance kweken. Maar,

ondanks dat er in het UMCG steeds meer patiënten binnenkomen met bacteriën die uitbraken

kunnen veroorzaken, zien we geen stijging in het aantal uitbraakpatiënten. Het lijkt er dus op,

dat de extra kosten gemaakt voor infectiepreventie zich uit betalen in minder uitbraakpatiënten

(en de geassocieerde kosten van deze patiënten).

Tenslotte, het Diagnostic Stewardship Program (DSP). Hier hebben we gekeken naar een van

de meest uitgevoerde diagnostische testen binnen de microbiologie: de bloedkweek. Deze

kweek wordt voorgeschreven voor bijna iedereen die antibiotica krijgt en de uitslag kan

worden gebruikt in het stroomlijnen van de therapie. Voor een groep van bijna 3000 patiënten

die antibiotica kregen bij de start bij hun opname, is er gekeken of er ook gelijktijdig

bloedkweken zijn afgenomen. Dit was maar in 48% van de gevallen gedaan. Schrikbarend laag,

maar helaas overeenkomstig met andere studies in Europa. Daarna is er gekeken wat voor

factoren associëren met het afnemen van bloedkweken en nog interessanter, welke factoren

associëren met de ligduur van beide groepen patiënten. Het blijkt dat het doen van

bloedkweken een associatie vertoont met andere diagnostiek (zoals CRP) wat suggereert dat er

een bundel van diagnostiek wordt gedaan. De groep met kweken lag significant korter in het

ziekenhuis en het doen van bloedkweken vertoonde een positieve associatie. Dit soort

onderzoek biedt perspectief voor nieuwe grotere studies waar nog preciezer wordt gekeken

naar de effecten van microbiologische diagnostiek.

Nederlandse samenvatting | 194

De conclusie is dan ook dat de afdeling Medische Microbiologie in het UMCG een zeer

positieve financiële impact heeft. Interventies die gedaan worden in het kader van therapie

verbeteren, infectiepreventie en diagnostiek zorgen elk voor betere behandeling van patiënten,

grotere patiëntveiligheid en uiteindelijk voor lagere kosten en/of potentie tot hogere efficiëntie

en omzet. Het advies voortvloeiend uit dit onderzoek is dan ook om de kosten van een

afdeling als Medische Microbiologie al dan niet inclusief Infectiepreventie goed in te bedden en

dit zeker niet te zien als makkelijke korte termijn bezuinigingspost. Een oplossing voor betere

inbedding zou kunnen zijn dat kosten en opbrengsten van infectiepreventie ziekenhuis breed

worden verdeeld via een verzekeringsmodel. Zo’n model moet een financiële prikkel geven om

correct om te gaan met infectiemanagement en tegelijkertijd de onduidelijke geldstromen van

tegenwoordig verhelderen voor alle stakeholders. Uiteindelijk geldt dat preventie de sleutel is

tot lagere kosten in de toekomst en dat verbeterde kwaliteit van zorg dankzij optimale, proces-

georiënteerde aanpak waarschijnlijk zeer kosteneffectief is.

196

Appendix IV: Publication lists

IV

197 | Appendix IV

Publications included in this thesis

Dik JW, Vemer P, Friedrich AW, Hendrix R, Lo-Ten-Foe JR, Sinha B, Postma MJ. Financial

evaluations of antibiotic stewardship programs - a systematic review. Front Microbiol. 2015

6:317.

Dik JW, Hendrix R, Friedrich AW, Luttjeboer J, Panday PN, Wilting KR, Lo-Ten-Foe JR,

Postma MJ, Sinha B. Cost-minimization model of a multidisciplinary antibiotic stewardship

team based on a successful implementation on a urology ward of an academic hospital. PLoS

One. 2015 10(5):e0126106.

Dik JW, Hendrix R, Lo-Ten-Foe JR, Wilting KR, Panday PN, van Gemert-Pijnen LE, Leliveld

AM, van der Palen J, Friedrich AW, Sinha B. Automatic day-2 intervention by a

multidisciplinary antimicrobial stewardship-team leads to multiple positive effects. Front

Microbiol. 2015 6:546.

Dik JW, Poelman R, Friedrich AW, Panday PN, Lo-Ten-Foe JR, Assen SV, van Gemert-Pijnen

JE, Niesters HG, Hendrix R, Sinha B. An integrated stewardship model: antimicrobial,

infection prevention and diagnostic (AID). Future Microbiol. 2016 11:93-102.

Dik JW, Poelman R, Niesters HG, Sinha B, Friedrich AW. Measuring the impact of an

antimicrobial stewardship program. Exp Rev Anti-Infect Therapy. 2016 14(6):569-575.

Dik JW, Dinkelacker AG, Vemer P, Lo-Ten-Foe JR, Lokate M, Sinha B, Friedrich AW,

Postma MJ. Cost-analysis of seven nosocomial outbreaks in an academic hospital. PLoS One.

2016 11(2):e0149226.

Dik JW, Sinha B, Friedrich AW, Lo-Ten-Foe JR, Hendrix R, Köck R, Bijker B, Postma MJ,

Freitag MH, Glaeske G, Hoffmann F. Cross-border comparison of antibiotic prescriptions

among children and adolescents between the north of the Netherlands and the north-west of

Germany. Antimicrob Resist Infect Control. 2016 5:14.

Dik JW, Lokate M, Dinkelacker AG, Lo-Ten-Foe JR, Sinha B, Postma MJ, Friedrich AW.

Positive impact of infection prevention on the management of nosocomial outbreaks at an

academic hospital. Future Microbiol. 2016. Epub ahead of print.

Dik JW, Hendrix R, van der Palen J, Lo-Ten-Foe JR, Friedrich AW, Sinha B. Performing

timely blood cultures in patients receiving IV antibiotics is correlated with a shorter length of

stay. 2016. Submitted

Publication lists | 198

Other publications

Dik JW, Rosema S, Geutjes M. Ziekenhuisuitbraken en diagnostiek. Analyse. 2016 3:100-103.

Dik JW, MJ Postma, B Sinha. Challenges for a sustainable financial foundation for

Antimicrobial Stewardship. Infect Dis Rep. 2016 submitted.

Oral presentations

Dik JW. Costs and benefits of an A-Team: Preliminary results and future perspectives.

Infection Prevention by deploying A-teams: Alleviation or Obligation? October 11, 2013.

Groningen, the Netherlands.

Lo-Ten-Foe JR, Sinha B, Wilting KR, Kyuchikova Y, Nannan Panday N, Dik JW, Hendrix R,

Friedrich AW. Multidisciplinair Antimicrobial Stewardship in het UMCG. Najaarsvergadering

NVMM. November 21, 2013. Utrecht, the Netherlands. Ned Tijdschr Med Microbiol. 2013

21(4):162.

Dik JW. A-Team in een academisch ziekenhuis, een financiële evaluatie. BD Diagnostics HAI

event. December 3, 2013. Rosmalen, the Netherlands.

Dik JW, Lo-Ten-Foe JR, Sinha B, Wilting KR, Kyuchikova Y, Nannan Panday P, Hendrix R,

Friedrich AW. Financial evaluation of a multidisciplinary Antimicrobial Stewardship-Team

using an automatic day-2 intervention. ECCMID 2014. May 10, 2014. Barcelona, Spain.

Sinha B & Dik JW. Implementation of an A-Team. Antibiotic Stewardship Symposium ‐

Optimal use of antimicrobial therapy. November 10, 2014. Groningen, the Netherlands.

Dik JW, Lokate M, Dinkelacker AG, Hendrix R, Lo-Ten-Foe JR, Postma MJ, Sinha B,

Friedrich AW. Positieve klinische en financiële impact van infectiepreventie met betrekking tot

nosocomiale uitbraken. Najaarsvergadering NVMM. November 19, 2015. Amersfoort, the

Netherlands. Ned Tijdschr Med Microbiol. 2015 23(4):166.

Dik JW. An integrated stewardship model: antimicrobial, infection prevention and diagnostic

(AID). EWMA Patient Outcomes Group meeting. March 9, 2016. Amsterdam, the

Netherlands.

Dik JW. Clinical and financial impact of medical microbiology. Annual meeting of RBSLM.

October 14, 2016. Val-Saint-Lambert, Belgium.

199 | Appendix IV

Poster presentations

Dik JW, Lo-Ten-Foe JR, Wilting KR, Nannan Panday P, Hendrix R, Postma MJ, Friedrich

AW, Sinha B. Financial model based on a successful implementation of a day-2 intervention by

a multidisciplinary Antimicrobial Stewardship-Team on a urology ward. ECCMID 2015. April

26, 2015. Copenhagen, Denmark.

Dik JW, Lo-Ten-Foe JR, Sinha B, Nannan Panday P, Hendrix R, Postma MJ, Friedrich AW,.

Performing diagnostics, especially blood cultures, on-time for infectious patients reduces

length of stay and costs. ECCMID 2015. April 26, 2015. Copenhagen, Denmark.

Dik JW, Hoffmann F, Lo-Ten-Foe, JR, Bijker B, Saß D, Postma MJ, Hendrix R, Sinha B,

Glaeske G, Friedrich AW. Cross border comparison of antibiotic prescriptions among children

and adolescents between the north of the Netherlands and the north-west of Germany.

ECCMID 2015. April 25-28, 2015. Copenhagen, Denmark.

Dik JW, Kirste A, Begeman A, Lo-Ten-Foe JR, Vemer P, Postma MJ, Sinha B, Lokate M,

Friedrich AW. Cost analysis of nosocomial outbreaks in an academic hospital setting ICAAC

2015. September 18, 2015. San Diego CA, USA.

Dik JW, Hendrix R, Lo-Ten-Foe JR, Friedrich AW, Sinha B. Blood cultures performed in

patients with a suspected infection are an essential part of a diagnostic bundle and reduce

length of stay (LOS). ECCMID 2016. April 11, 2016. Amsterdam, the Netherlands.

Dik JW, Lokate M, Dinkelacker AG, Hendrix R, Lo-Ten-Foe JR, Postma MJ, Sinha B,

Friedrich AW. Assessing the impact of infection prevention – an incremental cost-benefit

analysis. ECCMID 2016. April 11, 2016. Amsterdam, the Netherlands.

Publication lists | 200

202

Appendix V: Acknowledgements / Dankwoord

V

203 | Appendix V

All good things come to an end, also PhD projects. And when the time is there, I have to

acknowledge all the fantastic people that made this project possible. First and foremost that

has to be my first promoter, prof. dr. Friedrich. Dear Alex, we first met in October 2012. I

graduated in the summer of that year and was looking for a job. Someone suggested talking to

you. Even though you started working in Groningen just 1 year earlier, you were already

famous for your network. After listening to my talk, instead of suggesting which biomedical

company might be interesting, you offered me a PhD position. Something I did not expect at

all. Although I was not specifically looking for a research position, your hugely inspirational

talk about antimicrobial resistance, differences between Germany and the Netherlands and the

financial consequences quickly convinced me I should grasp this opportunity with both hands.

During my whole PhD you remained this visionary and inspirational leader. Sometimes more

on the background, sometimes more in the forefront, but always present. I cannot thank you

enough for the opportunity you gave me.

Secondly, prof. dr. dr. Sinha, my clinical microbiological conscience and second

promoter. Dear Bhanu, during my project, the focus shifted more and more towards

antimicrobial stewardship and the implementation of the A-Teams. As UMCG we are on the

forefront in the Netherlands when it came to A-Teams, not just the implementation but also

the evaluation. It seemed only natural that you should become my second promoter of my

project and during the years you became my direct supervisor. Whenever I needed some

explanation of the clinical processes, I could always go to you, and you where more than

willing to explain everything in detail. Although I had to plan such questions carefully, because

it might just cost me the rest of the afternoon. Over the years we gave dozens of presentations

all over the world advocating the UMCG A-Team and ASP. Being chosen as a best-practice by

the Dutch government was fantastic recognition of the work.

Due to the broad scope of the project, I am blessed (and sometimes cursed) with

another two extra (co-)promoters. As this was a financial project as well, I had the pleasure of

having prof. dr. Postma as my last promoter. Dear Maarten, I saw you perhaps less frequent as

my other (co-)promoters, being in another department and another field of research. However,

you were always available when I needed help in the financial evaluations. It quickly became

clear that you had the same enthusiasm about the project as Alex had, which made it a joy to

discuss our work. Furthermore, you arranged perfect substitution for all the times you were

unavailable, in the form of Pepijn Vemer.

Finally, besides a clinical microbiological promoter from the UMCG, I also had one

co-promoter from Certe Laboratories, dr. Hendrix. Dear Ron, as close collaborator and friend

of Alex you were right there from the beginning of my project. Especially the financial focus is

something that can spark you interest and enthusiasm, especially when it deals with your pet-

project of stewardship and A-Teams. I think it is safe to say I had to get used to your direct but

also lighthearted way of communicating. Eventually that worked out more than fine, and I

always thoroughly enjoyed our talks and meetings. It also gave me to opportunity to ‘escape’

Acknowledgements / Dankwoord | 204

the UMCG once in a while and cycle to the other side of Groningen, to get some much

needed fresh air and exercise.

Off course I have to thank my defense committee for their willingness to read, comment and

question my thesis. Prof. dr. Degener, prof. dr. Kluytmans and prof. dr. Voss. As well as the

extended committee, prof. dr. Von Eiff, dr. ir. Goettsch, prof. dr. Kern, prof. dr. De Smet, and

prof. dr. Werker.

Dear Jerome, even though I already had plenty of (co-)promoters, the more daily supervision

was often your task. When all of them were once again abroad or buried in other tasks, I could

always depend on you (and your Nespresso machine). Thank you so much for all the help, but

even more for all the coffee breaks and discussions on cars, cloths and other non-related work

stuff. Besides that, you were also always available for questions and arranging contacts in the

UMCG, which made my work so much easier.

My paranimfs, Corien and Anne. Thank you so much of the support during these last couple

of months as paranimf. Corien, we have known each other already 20 odd years and during

those years we had fantastic times travelling together throughout Europe (and even Africa),

cooking, drinking or just meeting up. Off course you had to be part of this project in one way

or another. Anne, when you and Peter got together, I was more or less included as friend and

roommate. Luckily for us, that worked out more than fine, especially considering the fact that

the three of us ended up living together when all other roommates moved out. When I started

my PhD, you were working in the UMCG as well doing your research internship, which meant

lots of coffee meetings. You then left to finish your MD, but in my final year, you came back

to the UMCG. Now as fellow PhD student which meant we could continue our coffee

meetings just as before. Thank you both!

All the people from the department of Medical Microbiology, thank you so much for all the

help and support I got during my project. In non-particular order (and with the danger of

forgetting someone), thank you: Piet and Johan, Randy, Anne, Kasper, Jan, Rik, John, Hajo

and Corinna, Erik, Greetje, Bert, Jet, Dirk, Mariëtte, Jan, Lenny, Adriana, Ineke, Martijn, Jan-

Maarten, Hermie, Anke, all the AIOS’s and many more. We work closely together with the

Pharmacy department and of the people there I need to thank at least Prashant Nannan

Panday, the best pharmacist an A-Team could wish for. I changed rooms three times and that

means many roommates to thank as well: Henna, Matty, René, Henk, Anja, Mehdi, Mariano,

Tjibbe, Kai, Silvia, Mithilla, Linda, Sigrid, Jelte, and Ruud. The fantastic secretary staff of our

205 | Appendix V

department: Ank, Marchiene, Carolien, Jolanda, Judith, and Bianca. And finally all my students:

Tessa, Marlon, Anja, Ties, Rizki and although not really a student, Ariane, thank you all for

participating in my project.

This whole project relied largely on huge datasets. It was therefore impossible to do

all this work without the support and fantastic help of Tjibbe and Igor who performed dozens

of queries for me to collect all the data from the UMCG data warehouse and who were always

willing to provide advice.

As department we have a close collaboration with the University of Twente. We were

both EurSafety Health-net partners and also in Enschede PhD projects were done within the

EurSafety project. Thank you Lisette, Jobke, Nienke, Maarten and Noreen for the enjoyable

collaboration.

With nine publications, comes a long list of co-authors. Most of them are mentioned

above. There are however more to thank: dr. Annemarie Leliveld, dr. Sander van Assen, Jos

Luttjeboer, Bert Bijker, prof. dr. Job van der Palen for his enthusiastic help with the statistics,

and all the German collaborators in Bremen and Oldenbrug, prof. dr. Glaeske, prof. dr.

Hoffmann, prof. dr. Freitag, and prof. dr. Köck.

Erik, besides being a great friend, thanks for the fantastic design of the cover!

Lieve familie, Pap en Mam, Jasper en Berber, dank voor het altijd klaarstaan voor mij, zonder

jullie had ik het niet gekund. En uiteindelijk, last but not least, lieve Chris, de eerste week van

januari 2013 was een bijzondere week. Niet alleen begon ik aan dit avontuur, maar ik leerde

twee dagen nadat ik was begonnen jou kennen. De afgelopen 3,5 jaar heb je talloze keren mijn

gezeur en gedweep moeten aanhoren over dit ‘boekje’, over stomme tijdschriften die mij niet

begrepen en elke keer maar weer dezelfde grafiekjes, posters, presentaties en tekstjes. En nog

steeds weet je niet precies wat ik nou heb gedaan. Maar ook nu geldt, zonder jou had dit boekje

er heel anders uitgezien. Jij stimuleert mij het beste in mijzelf naar boven te halen en staat altijd

voor mij klaar. Voor dat en nog veel meer, duizendmaal dank!

“Het was unaniem een belachelijk groot succes!”

Groningen, 15 september 2016.

Acknowledgements / Dankwoord | 206

Jan-Willem Hendrik Dik was born on November the 14th, 1986 in Groningen, the

Netherlands. After graduating secondary school at the Dr. Nassau College in Assen in 2005, he

took a gap-year and spent six months teaching underprivileged children in Accra, Ghana for

the NGO Glona. In September 2006 he started his bachelor study Biology at the University of

Groningen majoring in Biomedical Sciences. He continued in this direction for his masters,

also at the University of Groningen. During his masters he did his first research internship

under the supervision of dr. Ben Giepmans on the EpCAM protein at the department of Cell

Biology and the UMCG Microscopy and Imaging Center. The second phase of his masters

focused on bridging science and society within the so-called M-variant. This led to a second

research internship done at IMENz Bioengineering, where he performed a feasibility study on

novel food antimicrobials. He obtained his master’s degree in Biomedical Sciences in 2012.

From January 1st 2013, Jan-Willem started working at the Department of Medical Microbiology

at the University Medical Center Groningen on his PhD project on the clinical and financial

impact of Medical Microbiology. This project was part of the EurSafety Health-net project,

funded by the European Commission and Dutch and German provinces. The PhD project

was under the supervision of prof. Alex Friedrich and prof. Bhanu Sinha, together with dr.

Ron Hendrix from Certe Laboratory for Infectious Diseases and prof. Maarten Postma from

the University of Groningen and finished in June 2016, with a defence on November 7th. Since

May 2016 he is working at the National Health Care Institute (Zorginstituut Nederland) as

Business Intelligence Officer, as well as keeping a post-doc position at the Department of

Medical Microbiology of the UMCG.

University of Groningen Impact of medical …...Paranimfen Corien Kuiper Anne Loohuis The work...

Documents

Transcript of University of Groningen Impact of medical …...Paranimfen Corien Kuiper Anne Loohuis The work...