Upload
others
View
3
Download
0
Embed Size (px)
Citation preview
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
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
General introduction | 6
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,
General introduction | 8
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.
General introduction | 10
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
General introduction | 12
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).
General introduction | 14
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
Netherlands and the north-west of Germany
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).
Cross-border comparison of antibiotic prescriptions among children and adolescents between the north of the | 20
Netherlands and the north-west of Germany
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)
Cross-border comparison of antibiotic prescriptions among children and adolescents between the north of the | 22
Netherlands and the north-west of Germany
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).
Cross-border comparison of antibiotic prescriptions among children and adolescents between the north of the | 24
Netherlands and the north-west of Germany
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%
Total of the 5 most prescribed 81.1% 68.2%
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%
Total of the 5 most prescribed 74.4% 62.5%
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%
Total of the 5 most prescribed 69.0% 53.6%
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-
Cross-border comparison of antibiotic prescriptions among children and adolescents between the north of the | 26
Netherlands and the north-west of Germany
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.
27
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.
An integrated stewardship model: Antimicrobial, Infection prevention, and Diagnostic (AID) | 32
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.
An integrated stewardship model: Antimicrobial, Infection prevention, and Diagnostic (AID) | 34
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-
An integrated stewardship model: Antimicrobial, Infection prevention, and Diagnostic (AID) | 36
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).
An integrated stewardship model: Antimicrobial, Infection prevention, and Diagnostic (AID) | 38
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.
An integrated stewardship model: Antimicrobial, Infection prevention, and Diagnostic (AID) | 40
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.
41
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
Measuring the impact of antimicrobial stewardship programs | 46
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
Measuring the impact of antimicrobial stewardship programs | 48
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
Measuring the impact of antimicrobial stewardship programs | 50
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
Measuring the impact of antimicrobial stewardship programs | 52
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.
Measuring the impact of antimicrobial stewardship programs | 54
55
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
Financial evaluations of antibiotic stewardship programs – a systematic review | 60
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.
Financial evaluations of antibiotic stewardship programs – a systematic review | 62
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%
Financial evaluations of antibiotic stewardship programs – a systematic review | 64
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
Financial evaluations of antibiotic stewardship programs – a systematic review | 66
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.
Financial evaluations of antibiotic stewardship programs – a systematic review | 68
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
Financial evaluations of antibiotic stewardship programs – a systematic review | 70
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
Financial evaluations of antibiotic stewardship programs – a systematic review | 72
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
73
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
Material and Methods
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.
Automatic day-2 intervention by a multidisciplinary Antimicrobial Stewardship-Team leads | 78
to multiple positive effects
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
Statistical analysis
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.
Automatic day-2 intervention by a multidisciplinary Antimicrobial Stewardship-Team leads | 80
to multiple positive effects
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.
Automatic day-2 intervention by a multidisciplinary Antimicrobial Stewardship-Team leads | 82
to multiple positive effects
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
Automatic day-2 intervention by a multidisciplinary Antimicrobial Stewardship-Team leads | 84
to multiple positive effects
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
Automatic day-2 intervention by a multidisciplinary Antimicrobial Stewardship-Team leads | 86
to multiple positive effects
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.
87
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
Material and Methods
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.
Cost-minimization model of a multidisciplinary Antibiotic Stewardship-Team based on a successful | 92
implementation on a urology ward of an academic hospital
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.
Cost-minimization model of a multidisciplinary Antibiotic Stewardship-Team based on a successful | 94
implementation on a urology ward of an academic hospital
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).
Statistical analysis
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).
Cost-minimization model of a multidisciplinary Antibiotic Stewardship-Team based on a successful | 96
implementation on a urology ward of an academic hospital
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.
Cost-minimization model of a multidisciplinary Antibiotic Stewardship-Team based on a successful | 98
implementation on a urology ward of an academic hospital
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.
Cost-minimization model of a multidisciplinary Antibiotic Stewardship-Team based on a successful | 100
implementation on a urology ward of an academic hospital
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.
Cost-minimization model of a multidisciplinary Antibiotic Stewardship-Team based on a successful | 102
implementation on a urology ward of an academic hospital
103
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.
Material and Methods
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.
Cost-analysis of seven nosocomial outbreaks in an academic hospital | 108
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.
Cost-analysis of seven nosocomial outbreaks in an academic hospital | 110
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.
Cost-analysis of seven nosocomial outbreaks in an academic hospital | 112
113
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.
Material and Methods
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).
Positive impact of infection prevention on the management of nosocomial outbreaks at an academic hospital | 118
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).
Positive impact of infection prevention on the management of nosocomial outbreaks at an academic hospital | 120
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
Positive impact of infection prevention on the management of nosocomial outbreaks at an academic hospital | 122
↑ 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).
Positive impact of infection prevention on the management of nosocomial outbreaks at an academic hospital | 124
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
Positive impact of infection prevention on the management of nosocomial outbreaks at an academic hospital | 126
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.
127
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.
Material and Methods
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
Performing timely blood cultures in patients receiving IV antibiotics is correlated with a shorter length of stay | 132
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.
Performing timely blood cultures in patients receiving IV antibiotics is correlated with a shorter length of stay | 134
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
Performing timely blood cultures in patients receiving IV antibiotics is correlated with a shorter length of stay | 136
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
Performing timely blood cultures in patients receiving IV antibiotics is correlated with a shorter length of stay | 138
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
Performing timely blood cultures in patients receiving IV antibiotics is correlated with a shorter length of stay | 140
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.
Performing timely blood cultures in patients receiving IV antibiotics is correlated with a shorter length of stay | 142
143
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.
General conclusion, discussion and recommendations | 148
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
General conclusion, discussion and recommendations | 150
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.
General conclusion, discussion and recommendations | 152
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
General conclusion, discussion and recommendations | 154
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.
155
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.
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
Antimicrobial Chemotherapy. 2006. 58;(1): 154-8.
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.
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
Antimicrobial Chemotherapy. 2015. 70;(4): 1245-55.
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.
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.
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.
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.
183
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.
189
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.
195
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
201
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
207
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.