Current Medical Research and Opinion® 0300-7995

Vol. 24, No. 3, 2008, 737–751 doi:10.1185/030079908X273336

© 2008 Informa UK Ltd All rights reserved: reproduction in whole or part not permitted

ORIGINAL ARTICLE

A cost-effectiveness analysis
of antimicrobial treatment
of community-acquired
pneumonia taking into account
resistance in Belgium
Curr Med Res Opin Downloaded from informahealthcare.com by University Library Utrecht on 09/15/13

Monique Martin a, Lee Moore a, Sibilia Quilici a, Marc

Decramer b and Steven Simoens c
a 
i3 Innovus, Uxbridge, UK

td
b 
Respiratory Division, University Hospitals Leuven, Belgium
c 
Research Centre for Pharmaceutical Care and Pharmaco-economics,
K L
Katholieke Universiteit Leuven, Belgium
a U
o rm House, Cricket Field Road,
Address for correspondence:  Monique Martin, i3 Innovus, Beaufort
Uxbridge UB8 1QG, UK. Tel.: +44 1895 455380 ext. 383. f
For personal use only.

In
Fax: +44 1895 520039;
monique.martin@i3innovus.com
0 8
20
Key words:  Antimicrobial resistance – Community-acquired pneumonia – Cost-effectiveness analysis
– Decision analytic model – Moxifloxacin
©
ABSTRACT

Objective: This article assesses the cost-
effectiveness of outpatient antimicrobial treatment
of community-acquired pneumonia (CAP) taking
into account resistance in Belgium.
Research design and methods: Our decision
analytic model focused on mild to moderate
CAP, but did not consider severe CAP. Treatment
pathways reflected empirical treatment initiated
in the absence of data on CAP aetiology.
First-line treatment consisted of moxifloxacin,
co-amoxiclav, cefuroxime or clarithromycin. If
first-line treatment was unsuccessful, patients
were either hospitalised or second-line treatment
with a different antimicrobial was initiated. Clinical
failure rates were obtained from the published
literature or expert opinion. Costs were calculated
using published sources from the third-party payer
perspective.
Main outcome measures: Effectiveness
measures included first-line clinical failure
avoided, second-line treatment avoided,
hospitalisation avoided and death avoided.
Healthcare costs were included, but costs of
productivity loss were not considered.
Results: Costs of treating a CAP episode
amounted to 144€ with moxifloxacin/co-amoxiclav;
222€ with co-amoxiclav/clarithromycin; 211€
with cefuroxime/moxifloxacin; and 193€ with
clarithromycin/moxifloxacin. The rate of first-line
failure was 5%, 16%, 19% and 18% for these
four treatment strategies, respectively. The rate
of second-line treatment amounted to 4%, 13%,
16% and 15%, respectively. The hospitalisation
rate was 1%, 4%, 4% and 4%, respectively. The
death rate was 0.01%, 0.04%, 0.03% and 0.03%,
respectively. Sensitivity analyses supported the
dominance of moxifloxacin/co-amoxiclav in nearly
all scenarios.
Conclusions: First-line treatment of CAP
patients with moxifloxacin followed by
co-amoxiclav or hospitalisation if required was
more effective and less costly as compared
with first-line treatment with co-amoxiclav,
cefuroxime or clarithromycin.

Paper 4358 737

 Introduction
The term ‘community-acquired pneumonia’ (CAP)
covers those infections of the lung parenchyma that are
not acquired in hospital or a long-term care facility. The
incidence, clinical and economic burden of CAP are
significant, although Belgian data about these aspects
are lacking. CAP is a common infectious disease, with
annual incidence rates of 1.6 per 1000 adults in Spain1
and 11.6 per 1000 adults in Finland2. CAP is associated
with significant morbidity, is the leading cause of
death due to infection in developed countries3 and is
the sixth leading cause of death in the United States4.
The economic burden of CAP is substantial. In the
United States, CAP accounts each year for 10 million
physician visits 5. It is a primary driver of hospital
Curr Med Res Opin Downloaded from informahealthcare.com by University Library Utrecht on 09/15/13
admissions, resulting in about 1 million hospitalisations
in the United States per year at an estimated cost of $9
billion6,7. The key component of inpatient costs is the
length of hospitalisation8.
Patients suffering from CAP generally present with
a cough (more than 90% of patients), dyspnea (66%),
sputum production (66%) and pleuritic chest pain
(50%)9. Other symptoms may include headache and
myalgia. Specific aetiologies, such as Legionella spp.,
may be associated with gastro-intestinal symptoms.
For personal use only.
Chest X-rays, examination of the blood and sputum
for infectious micro-organisms, and blood tests are
commonly used to diagnose individuals with suspected
CAP based upon symptoms and physical examination.
Several decision rules are available to categorise
patients as suffering from either mild, moderate or
severe CAP, such as the Pneumonia Severity Index10
and the CURB-65 score11. These decision rules have
also been developed to guide treatment.
Patients suffering from mild to moderate CAP
are generally treated with oral antimicrobials in the
community 12. Hospitalisation may be required for
elderly patients, patients who have underlying chronic
illnesses or patients with more serious disease. In the
majority of patients 4, treatment is empirical in that
treatment is initiated in the absence of information
about the causative pathogen involved in CAP and the
antibiotic susceptibility pattern of the isolated organism.
The treatment approach to CAP is challenging due to
the expanding spectrum of causative pathogens, the
rising prevalence of resistance to antimicrobial agents
and the increasing pressure to reduce the length of
hospitalisation13.
These trends have led to a debate about the choice
of antimicrobial in the management of CAP. The
European Respiratory Society has published clinical
guidelines to guide antimicrobial treatment of CAP14.
With respect to management of CAP in the community,
amoxicillin or tetracyclines are recommended. In
738  Cost-effectiveness of antimicrobial treatment of CAP in Belgium © 2008 Informa UK Ltd – Curr Med Res Opin 2008; 24(3)
case of hypersensitivity, newer macrolides, such as
clarithromycin, are alternatives in countries with
low pneumococcal macrolide resistance. If there
is clinically relevant resistance against first-choice
agents, treatment with moxifloxacin or levofloxacin
may be considered. With respect to management of
CAP in hospital, preferred treatment strategies in
regions with low resistance rates include penicillin
G, aminopenicillin, co-amoxiclav, cephalosporin II or
III; each ± a macrolide. In countries with increased
resistance rates, levofloxacin or moxifloxacin may
be alternatives. In patients with severe CAP and
no risk factors for Pneumococcal aeruginosa, a non-
antipseudomonal cephalosporin III + a macrolide or
moxifloxacin/levofloxacin is preferred. In patients with
severe CAP and risk factors for P. aeruginosa, guidelines
recommend an antipseudomonal cephalosporin, an
aclyureidopenicillin/β‑lactamase inhibitor or carbapen­
em; each with ciprofloxacin.
The guidelines of the European Respiratory Society
also stated that the selection of antimicrobial needs to
reflect local patterns of microbial resistance in addition
to the severity of illness, the frequency of specific
pathogens and drug safety profiles 14. With respect
to resistance, data from the Belgian Pneumococcal
Reference Laboratory indicate that in vitro resistance
of Streptococcus pneumoniae to macrolides has increased
from 5.2% in 1986 to 36.1% in 2003 in Belgium.
Similarly, resistance has grown from 16.4% to 30.2%
in the case of tetracyclines and from 2% to 13% in the
case of penicillin. On the other hand, pneumococcal
resistance was low for fluoroquinolones, remaining
below 1% for levofloxacin and being 0% for moxifloxacin
over the 1995–2005 period15,16. Antimicrobial resistance
can have a substantial impact on outcomes and costs of
CAP treatment. There is evidence that CAP patients
with pneumococcal resistance may be at greater risk of
poor outcomes17. Also, if first-line treatment fails due to
resistance, additional costs are incurred due to the need
for second-line treatment or hospitalisation, or both.
The aim of this article is to assess the cost-
effectiveness of empirical outpatient treatment
of mild to moderate CAP with moxifloxacin or
other antimicrobials (co-amoxiclav, cefuroxime or
clarithromycin) in Belgium from the perspective of
the third-party payer. Treatment strategies involve oral
antimicrobials, are recommended by Belgian guidelines
and reflect prevailing treatment pathways in Belgium18.
The analysis takes into account the presence of
antimicrobial resistance in Belgium. In particular, the
health economic model investigates whether the higher
drug acquisition costs of moxifloxacin are balanced by a
lower clinical failure rate and reduced costs of second-
line treatment and hospitalisation as a result of a lower
resistance rate and a higher clinical success rate.
 Methods
A decision analytic model was used to assess the
cost-effectiveness of empirical treatment with oral
antimicrobials in a population of CAP patients. The
model was originally developed for France, Germany
and the United States 19 , but the current study
adapted the model to reflect empirical antimicrobial
treatment of CAP in Belgium. The model adaptation
involved the selection of prevalent treatment
strategies, antimicrobial resistance data, resource
use data and unit cost data pertaining to Belgium.
The adapted model adds to the evidence base by
presenting, to the best of the authors’ knowledge,
the first results on the cost-effectiveness of current
antimicrobial treatment pathways in Belgium. Also,
Curr Med Res Opin Downloaded from informahealthcare.com by University Library Utrecht on 09/15/13
the model provides results for a country which has
an intermediate level of antimicrobial resistance in
CAP pathogens20 as compared with other European
countries such as Germany which has a low level
of resistance, or France which has a high level of
resistance21.
Decision analytic model
Figure 1 represents the decision analytic model of
For personal use only.
empirical antimicrobial treatment of CAP in Belgium.
For each antimicrobial treatment option, the decision
tree consists of three arms reflecting the different
pathogens, S. pneumoniae, Haemophilus influenzae
and the atypical pathogens. For each pathogen,
the tree contains two arms, namely susceptible or
non-susceptible (resistant) to first-line treatment.
Resistance rates vary by the pathogens and treatments
considered. If first-line treatment fails, patients are
either hospitalised or receive second-line treatment
in the community. This is identical for treatment
failures occurring in susceptible and resistant isolates.
Second-line treatment leads to treatment success
or failure. If second-line treatment fails, all patients
are hospitalised. Following hospitalisation, patients
experience either treatment success (patient is
discharged alive) or treatment failure (patient dies
in hospital). As the time scale of the model equals
the treatment duration of a CAP episode and is,
thus, shorter than 1 year, no discounting of costs or
outcomes was carried out. Each of the steps of the
decision analytic model is described in more detail in
the following sections.
Study population
A hypothetical population was considered consisting
of patients suffering from mild to moderate CAP (i.e.
Fine risk category I–III10) and who had not undergone
© 2008 Informa UK Ltd – Curr Med Res Opin 2008; 24(3) Cost-effectiveness of antimicrobial treatment of CAP in Belgium  Martin et al.  739
antimicrobial treatment in the previous 3  months.
These patients are generally treated empirically in
the community12. The relative incidence of Fine risk
classes in an outpatient population has been reported
to be 62%, 26% and 8% for categories I, II and III,
respectively10. Recalibrating these percentages to add
up to 100%, the model applied a relative incidence
of 65%, 27% and 8% for categories I, II and III,
respectively.
Pathogens
The most common causative microbial pathogens
involved in CAP [S. pneumoniae, H. influenzae and
the atypical pathogens (Mycoplasma pneumoniae,
Chlamydia pneumoniae, Legionella spp.)] were
considered in the model at their approximate relative
isolation frequencies. Estimates were based on
observed prevalences across German CAP patients
treated initially in the community, reporting 40% for S.
pneumonia, 8% for H. influenzae and 12% for atypical
pathogens22, confirming prevalence estimates from the
published literature23. Normalising these prevalences
to 100% for the purpose of the model generated
frequencies of 67% for S. pneumoniae, 10% for H.
influenzae and 23% for the atypical pathogens.
Antimicrobial treatment
Antimicrobial treatments represented current
treatment pathways in Belgium and were based on
Belgian guidelines relating to treatment of CAP in
the community 18. First-line treatment consisted of
moxifloxacin (a fluoroquinolone), co-amoxiclav
(a beta-lactam), cefuroxime (a beta-lactam) or
clarithromycin (a macrolide). If first-line treatment was
unsuccessful as determined at a follow-up visit, some
patients were hospitalised or a second-line treatment
was initiated. A hospitalisation rate following first-line
treatment failure of 15.4% was used24. Second-line
treatments were co-amoxiclav (if first-line treatment
with moxifloxacin failed), clarithromycin (if first-line
treatment with co-amoxiclav failed) and moxifloxacin
(if first-line treatment with clarithromycin or
cefuroxime failed). Regimens of first-line and second-
line treatment amounted to a daily dose of 400 mg of
moxifloxacin during 10 days; a daily dose of 2625 mg
of co-amoxiclav during 10  days; a daily dose of
1500   mg of cefuroxime during 10  days and a daily
dose of 1000 mg of clarithromycin during 5 days18. If
second-line treatment was unsuccessful, all patients
were hospitalised. Antimicrobial treatment in hospital
was assumed to be identical for all patients in order to
avoid introducing variability as the model focuses on
patients treated in the community.
 Curr Med Res Opin Downloaded from informahealthcare.com by University Library Utrecht on 09/15/13
For personal use only.

Live
Treatment success
Die

Live
Treatment success
Die
ICU
Live
Susceptible Treatment failure
Die
Hospitalisation
Live
Treatment success
Die
Regular ward

740  Cost-effectiveness of antimicrobial treatment of CAP in Belgium

Live
Treatment failure
Treatment failure
Die

Live
S. pneumonia
Treatment success
Die

Outpatient Live
ICU
Die
Treatment failure / Hospitalisation
1st line empirical treatment of CAP Live
Regular ward
Die

Non-susceptible
[+]

H. Influenza
[+]

Atypicals
[+]

Figure 1.  Decision analytic model to determine the cost-effectiveness of antimicrobial treatment of CAP in the community

© 2008 Informa UK Ltd – Curr Med Res Opin 2008; 24(3)

 Clinical failure rates
The model is based on the premise that clinical failure
can occur due to two main reasons: lack of response to
treatment in patients with susceptible pathogens and
failure due to the presence of antimicrobial-resistant
pathogens.
The failure rate in susceptible pathogens was
estimated on the basis of antimicrobial success rates
from published clinical trials in CAP. Meta-analyses
were not suitable for use in the model as they did not
focus on the population considered. A wide range
of efficacy data was found in the literature due to
differences in trial design, treatment setting, patient
population, duration of treatment, trial size, outcome
measures used, etc. Our estimates were based on those
Curr Med Res Opin Downloaded from informahealthcare.com by University Library Utrecht on 09/15/13
clinical trials that could be matched as far as possible
to the patient population, causative pathogens, time
horizon, dosage and outcome measures considered
in our analysis. The antimicrobial success rate was
calculated as the simple average of selected studies.
The obtained antimicrobial success rates were 95%
for moxifloxacin25–28, 90% for co-amoxiclav29, 90% for
cefuroxime30 and 93% for clarithromycin25,31,32.
The failure rate in antimicrobial-resistant pathogens
was estimated on the basis of antimicrobial resistance
For personal use only.
data from published surveillance studies. The majority
of samples considered in surveillance studies relate
to hospitalised patients. However, a recent study
showed that the prevalence of resistant isolates in
a hospitalised and community setting is similar for
community-acquired pathogens33, thus enabling us to
generalise surveillance data to the community setting.
Antimicrobial resistance data related to Belgium or
were derived from published sources in the absence of
Belgian data.
For S. pneumoniae, the Clinical and Laboratory
Standards Institute (CLSI) breakpoints were used to
define susceptibility and resistance of a pathogen as
these breakpoints represent the ‘gold standard’ and
are used internationally in trials. For H. influenzae,
pharmacokinetic/pharmacodynamic (PK/PD) break­
points were used, except for moxifloxacin, for which
CLSI breakpoints were applied34. The use of PK/PD
rather than CLSI breakpoints had a major effect on
the resistance rate in H. influenzae. PK/PD breakpoints
for H. influenzae were used in the base case analysis
because they are thought to better reflect clinical
success/failure containing information on both in vitro
and in vivo on time and eradication and because they
have been validated in animal models and clinical
studies in otitis media and sinusitis 34–36. No CLSI
breakpoints have been developed for the atypical
pathogens as no resistance has been observed in these
pathogens.
© 2008 Informa UK Ltd – Curr Med Res Opin 2008; 24(3) Cost-effectiveness of antimicrobial treatment of CAP in Belgium  Martin et al.  741
In Belgium, surveillance studies from 1995 to
2005 have consistently found 0% resistance of S.
pneumoniae to moxifloxacin 15,16 . There were no
reports of moxifloxacin resistance from the Alexander
Project (1998–2000)34 and from the PROTEKT study
(1999–2000)37 in H. influenzae isolates. Atypical isolates
have not demonstrated any resistance to moxifloxacin
and are assumed to be 100% susceptible.
Beta-lactams are among the most effective agents for
the treatment of S. pneumoniae, with > 90% of isolates
in Belgium being susceptible. Resistance rates of 0%
and 9.2% have been reported to co-amoxiclav and
cefuroxime, respectively, for Belgium38. In H. influenza,
the model used resistance rates of 0% to co-amoxiclav
and 22.1% to cefuroxime34. As the beta-lactams are
not active against the atypical pathogens, the model
assumed a 100% resistance rate of atypical pathogens
to co-amoxiclav and cefuroxime.
For macrolides, there are two main resistance
mechanisms in S. pneumoniae; the mefA genotype
was believed to confer mainly low-level resistance
(MICs 1–32  mg/L), whereas the ermB genotype was
expected to result in mainly high-level resistance
(MICs ≥ 8 mg/L)27. Until recently, it was believed that
macrolides would be expected to retain efficacy against
a proportion of macrolide-resistant strains with mefA
and low MICs (1–4 mg/L), but be ineffective against
ermB strains28. However, recent evidence has indicated
that both mechanisms can result in a high level of
resistance. The model used estimates of resistance
rates to clarithromycin of 23.7% for S. pneumoniae and
100% for H. influenzae34,39. For atypical pathogens, no
resistance has been reported to clarithromycin.
Rates of clinical failure due to resistance were
estimated for each pathogen. As the published
literature did not provide quantitative evidence on
this relationship, a number of assumptions were
made based on the opinion of clinical experts. For
moxifloxacin, co-amoxiclav and cefuroxime, the
model assumed that 50% of resistant isolates of S.
pneumoniae resulted in clinical failure. Also, for
clarithromycin, a 50% clinical failure rate was assumed
for both mefA resistance and for strains with ermB
resistance (± mefA). This was based on the premise
that even in case of treatment failure, host defence
mechanisms can overcome resistance and still result
in clinical success. Clinical failure rates for resistant
isolates of H. influenzae were assumed to be 50% for
all antimicrobials considered. As atypical pathogens
are not resistant to moxifloxacin and clarithromycin,
the clinical failure rate due to resistance was assumed
to be 0%. Finally, a spontaneous cure rate of 65%
(or a clinical failure rate of 35%) was assumed when
co-amoxiclav or cefuroxime was used to treat an
atypical pathogen.
 Mortality
Data from the PORT study indicated that the mortality
rate of Fine category I, II and III patients treated for
CAP in the community was 0.1%10. As the mortality
rate in the community is very small, the mortality rate
was considered to be zero. For hospitalised patients,
the mortality rate of Fine category I, II and III patients
of 0.9%, as obtained from the PORT study, was used
in the model 10. The PORT study also provided an
estimate of the mortality rate in the intensive care unit
of 37%10.
Resource use and costs
Resource use was derived from the literature. Unit
Curr Med Res Opin Downloaded from informahealthcare.com by University Library Utrecht on 09/15/13
costs were obtained from RIZIV/INAMI, the national
insurance company which acts as the Belgian third-
party payer40, or from published sources in the absence
of Belgian data. Where needed, costs were inflated to
2006.
Daily reimbursed costs of antimicrobial treatment
amounted to 4.12€ for moxifloxacin 400 mg; 2.47€ for
co-amoxiclav 2625 mg; 2.94€ for cefuroxime 1500 mg
and 2.75€ for clarithromycin 1000 mg40. Antimicrobial
treatments were prescribed at the first consultation
For personal use only.
for a period of 10 days for moxifloxacin, co-amoxiclav
and cefuroxime, and for a period of 5  days for
clarithromycin 18. Evaluation of treatment success
took place after 3  days. If first-line treatment was
failing, patients were either hospitalised or second-line
treatment with a different antimicrobial was initiated.
As the costs of the first-line treatment had been
incurred, the full 10-day or 5-day cost of treatment
was included in the model for patients failing first-
line treatment. Similarly, patients failing second-line
treatment also incurred the full cost of this treatment
prior to hospitalisation.
Patients who sought medical advice for CAP incurred
a general practitioner (GP) office visit (at a reimbursed
cost of 16.71€), a GP home visit (19.95€), a specialist
office visit (20.41€) or an emergency room visit
(16.71€)40. The frequency of initial visits was estimated
at 61% for GP office visits, 19% for GP home visits, 18%
for specialist office visits and 2% for emergency room
visits41. The initial visit was followed by one follow-up
visit after 3 days to assess the efficacy of treatment. The
same frequency of the different visit types was assumed
for this follow-up visit, with the exception of emergency
room visits which were assumed to be followed by a GP
office visit. This generated a frequency of follow-up visits
of 63% for GP office visits, 19% for GP home visits, 18%
for specialist office visits and 0% for emergency room
visits41. In case of treatment failure, the costs of another
visit were included to assess the efficacy of second-line
742  Cost-effectiveness of antimicrobial treatment of CAP in Belgium © 2008 Informa UK Ltd – Curr Med Res Opin 2008; 24(3)
treatment. If second-line treatment failed, no further GP
visits were included prior to hospitalisation.
CAP guidelines advise the use of an initial chest
X-ray (at a reimbursed cost of 9.96€) to diagnose CAP
and a follow-up chest X-ray (11.95€) to ascertain that
the infection has disappeared following treatment18,40.
Initial chest X-rays were performed in 39% of patients
and follow-up chest X-rays were conducted in 35% of
patients24. A white blood cell count was assumed for
all patients failing treatment (at a reimbursed cost of
20€)40. Susceptibility testing is not generally carried
out in the community setting and was included in total
hospital costs14.
The reimbursed hospitalisation cost for CAP
patients admitted to the regular ward included all
costs related to treatment and diagnostic tests and
amounted to 3150.64€40. Based on a study comparing
costs of patients surviving and dying in hospital, costs
of surviving patients were multiplied by 1.45 to obtain
hospitalisation costs for patients dying in hospital 42.
Similarly, based on French data comparing costs of
patients admitted to a regular ward or the intensive
care unit, a multiplier of 1.63 was used to calculate
costs of patients admitted to the intensive care unit43.
Costs related to adverse events were not included
due to the low incidence rates of adverse events in
antimicrobial treatment44. As a result, the impact of
these costs on the cost-effectiveness of antimicrobial
treatment strategies is expected to be limited.
Base case analysis
The base case analysis calculated incremental cost-
effectiveness ratios (ICERs) using the following equation:
ICER = (C1–C0)/(E1–E0)
where C1 is the cost of the intervention; C0 is the
cost of the alternative with which the intervention is
compared; E1 and E0 are the respective health outcomes
of intervention and comparator.
The intervention was a strategy consisting of first-line
treatment with moxifloxacin and second-line treatment
with co-amoxiclav if required. The three comparator
strategies were co-amoxiclav/clarithromycin, cefur­
oxime/moxifloxacin, and clarithromycin/moxifloxacin.
Effectiveness was assessed in terms of the rate of first-
line clinical failures, of second-line treatments required,
of hospitalisations required and of mortality. A strategy
is said to dominate the comparator when it is both less
expensive and more effective.
Sensitivity analysis
One-way deterministic sensitivity analyses were carried
out to ascertain the robustness of ICERs by varying key
 input parameters over plausible ranges. The sensitivity
analyses set all antimicrobial success rates at 100%
with a view to isolating the impact of antimicrobial
resistance on ICERs. To assess the impact of changes in
resistance level, resistance rates for S. pneumoniae and
H. influenzae were changed from –50% to +50% of base
case rates for each of the antimicrobials considered. The
clinical failure rate for resistant isolates was reduced
from 50% to 20%. CLSI resistance rates were used for
H. influenzae instead of PK/PD resistance rates.
To explore the impact of hospitalisation,
hospitalisation costs were either increased or decreased
by 50%. Also, the base case analysis assumed
no admissions to the intensive care unit. In the
deterministic sensitivity analysis, a rate of admittance
to the intensive care unit for hospitalised patients of
Curr Med Res Opin Downloaded from informahealthcare.com by University Library Utrecht on 09/15/13
A probabilistic sensitivity analysis based on a
10 000-iteration Monte Carlo simulation was performed
in Microsoft Excel using the add-on programme @RISK.
Such an analysis requires that a probability distribution
is assigned to each input parameter. Hospitalisation
unit costs and antimicrobial success rates were assumed
to have a triangular distribution. The beta distribution
was chosen for resistance rates. Spontaneous cure
rates and clinical failure rates for resistant isolates
were assumed to be distributed as uniform random
variables. For each iteration, the simulation drew
input parameters at random from their statistical
distributions and calculated cost and effectiveness pairs.
At the end of the 1000 iterations, the joint statistical
distribution for costs and effectiveness was represented
as a cloud of points on the cost-effectiveness plane.

4.9% was used, based on the Fine category mix of Cost-effectiveness acceptability curves representing
the population9. Furthermore, the analysis explored the probability that the moxifloxacin/co-amoxiclav
the impact of first-line treatment failure resulting in strategy is cost-effective for a range of cost-effectiveness
no hospitalisation, but always leading to second-line thresholds were also drawn.
treatment in the community.
Viruses were not considered in the base case analysis
as the main focus is on antimicrobial resistance and Results
viruses are insensitive to antimicrobial treatment. The
deterministic sensitivity analysis considered viruses as
these are often mistaken for microbial pathogens and Base case analysis
For personal use only.

are treated empirically with antimicrobials. Normalised
frequencies used in the sensitivity analysis were 20%
for viruses, 54% for S. pneumoniae, 8% for H. influenzae
and 18% for the atypical pathogens 23. Finally, to
equalise the effect of second-line treatment, an analysis
was performed in which all second-line treatments
were changed to moxifloxacin. A similar analysis was
conducted with co-amoxiclav as second-line treatment
for all treatment strategies.
Table 1 presents the cost-effectiveness results of the
base case analysis. A strategy consisting of first-line
treatment with moxifloxacin and second-line treatment
with co-amoxiclav dominated each of the three
comparator strategies (co-amoxiclav/clarithromycin,
cefur­o xime/moxifloxacin and clarithromycin/
moxifloxacin). The cost of treating a CAP episode with
moxifloxacin/co-amoxiclav was 35%, 32% and 26%
less than treatment with co-amoxiclav/clarithromycin,

Table 1.  Cost-effectiveness of moxifloxacin/co-amoxiclav treatment strategy versus other treatment strategies

Moxifloxacin/ Co-amoxiclav/ Cefuroxime/ Clarithromycin/

co-amoxiclav -clarithromycin moxifloxacin moxifloxacin
Costs Costs (€) 143.53 221.97 211.16 192.79
Effectiveness First-line failure 5.00% 15.65% 19.00% 18.13%
Second-line treatment 4.23% 13.24% 16.08% 15.34%
Hospitalisation 1.43% 4.28% 3.73% 3.56%
Deaths 0.01% 0.04% 0.03% 0.03%
ICERs Cost per first-line Dominant Dominant Dominant
failure avoided
Cost per second-line Dominant Dominant Dominant
treatment avoided
Cost per Dominant Dominant Dominant
hospitalisation avoided
Cost per death Dominant Dominant Dominant
avoided
Note: ICER = incremental cost-effectiveness ratio

© 2008 Informa UK Ltd – Curr Med Res Opin 2008; 24(3) Cost-effectiveness of antimicrobial treatment of CAP in Belgium  Martin et al.  743
 Curr Med Res Opin Downloaded from informahealthcare.com by University Library Utrecht on 09/15/13
For personal use only.

Table 2.  Deterministic sensitivity analysis in terms of first-line clinical failures avoided

Description Input value Costs (€) First-line failure ICERs

Moxifloxacin/ Co-amoxiclav/ Cefuroxime/ Clarithromycin/ Moxifloxacin/ Co-amoxiclav/ Cefuroxime/ Clarithromycin/ Co-amoxiclav/ Cefuroxime/ Clarithromycin/
co-amoxiclav clarithromycin moxifloxacin moxifloxacin co-amoxiclav clarithromycin moxifloxacin moxifloxacin clarithromycin moxifloxacin moxifloxacin
Effect of resistance All clinical success rates set 95.25 121.41 146.93 139.63 0.00% 7.95% 12.14% 12.94% Dominant Dominant Dominant
at 100%
Range of resistance Moxifloxacin 143.53 221.97 211.16 192.79 5.00% 15.65% 19.00% 18.13% Dominant Dominant Dominant
for –50% resistance
S. pneumoniae Moxifloxacin 143.53 221.97 211.16 192.79 5.00% 15.65% 19.00% 18.13% Dominant Dominant Dominant
+50% resistance
H. influenzae Clarithromycin 143.53 207.04 211.16 154.43 5.00% 15.65% 19.00% 12.56% Dominant Dominant Dominant
–50% resistance
Clarithromycin 143.53 236.91 211.16 231.16 5.00% 15.65% 19.00% 23.69% Dominant Dominant Dominant
+50% resistance
Co-amoxiclav/cefuroxime 143.53 221.97 199.46 192.79 5.00% 15.65% 17.33% 18.13% Dominant Dominant Dominant
–50% resistance

744  Cost-effectiveness of antimicrobial treatment of CAP in Belgium

Co-amoxiclav/cefuroxime 143.53 221.97 222.85 192.79 5.00% 15.65% 20.68% 18.13% Dominant Dominant Dominant
+50% resistance
Effect of clinical Moxifloxacin 143.53 221.97 211.16 192.79 5.00% 15.65% 19.00% 18.13% Dominant Dominant Dominant
failure rates 20% clinical failure rate
for resistant isolates Clarithromycin 143.53 201.13 211.16 139.27 5.00% 15.65% 19.00% 10.36% Dominant Dominant 79.44
20% clinical failure rate
Co-amoxiclav/cefuroxime 143.53 221.97 193.61 192.79 5.00% 15.65% 16.49% 18.13% Dominant Dominant Dominant
20% clinical failure rate
Effect of choice of CLSI resistance rates 143.53 216.83 205.12 178.12 5.00% 15.70% 18.14% 16.00% Dominant Dominant Dominant
resistance rates for
H. influenzae
Effect of Costs decreased by 50% 120.88 154.29 152.15 136.51 5.00% 15.65% 19.00% 18.13% Dominant Dominant Dominant
hospitalisation
Costs increased by 50% 166.18 289.66 270.16 249.08 5.00% 15.65% 19.00% 18.13% Dominant Dominant Dominant
Intensive care unit 145.52 227.91 216.33 197.93 5.00% 15.65% 19.00% 18.13% Dominant Dominant Dominant
admission rate 4.9%
Hospitalisation rate 0% 123.43 157.61 125.37 110.83 5.00% 15.65% 19.00% 18.13% Dominant Dominant 96.03
following first-line failure
Viruses 20% viruses as causative 131.87 191.33 183.61 165.80 4.00% 12.52% 15.20% 14.50% Dominant Dominant Dominant
pathogen
First-line treatment All moxifloxacin second- 129.92 187.33 211.16 192.79 5.00% 15.65% 19.00% 18.13% Dominant Dominant Dominant
line
comparison All co-amoxiclav second- 143.53 258.65 286.51 225.28 5.00% 15.65% 19.00% 18.13% Dominant Dominant Dominant
line

Note: CLSI = Clinical and Laboratory Standards Institute; ICER = incremental cost-effectiveness ratio

© 2008 Informa UK Ltd – Curr Med Res Opin 2008; 24(3)

 Curr Med Res Opin Downloaded from informahealthcare.com by University Library Utrecht on 09/15/13
For personal use only.

Table 3.  Deterministic sensitivity analysis in terms of hospitalisations avoided

Description Input value Costs (€) Hospitalisation rate ICERs

Moxifloxacin/ Co-amoxiclav/ Cefuroxime/ Clarithromycin/ Moxifloxacin/ Co-amoxiclav/ Cefuroxime/ Clarithromycin/ Co-amoxiclav/ Cefuroxime/ Clarithromycin/
co-amoxiclav clarithromycin moxifloxacin moxifloxacin co-amoxiclav clarithromycin moxifloxacin moxifloxacin clarithromycin moxifloxacin moxifloxacin
Effect of resistance All clinical success rates set 95.25 121.41 146.93 139.63 0.00% 1.22% 1.87% 1.99% Dominant Dominant Dominant
at 100%
Range of resistance Moxifloxacin 143.53 221.97 211.16 192.79 1.43% 4.28% 3.73% 3.56% Dominant Dominant Dominant
for –50% resistance
S. pneumoniae and Moxifloxacin 143.53 221.97 211.16 192.79 1.43% 4.28% 3.73% 3.56% Dominant Dominant Dominant
+50% resistance
H. influenzae Clarithromycin 143.53 207.04 211.16 154.43 1.43% 3.81% 3.73% 2.47% Dominant Dominant Dominant

© 2008 Informa UK Ltd – Curr Med Res Opin 2008; 24(3)

–50% resistance
Clarithromycin 143.53 236.91 211.16 231.16 1.43% 4.75% 3.73% 4.65% Dominant Dominant Dominant
+50% resistance
Co-amoxiclav/cefuroxime 143.53 221.97 199.46 192.79 1.43% 4.28% 3.40% 3.56% Dominant Dominant Dominant
–50% resistance
Co-amoxiclav/cefuroxime 143.53 221.97 222.85 192.79 1.43% 4.28% 4.06% 3.56% Dominant Dominant Dominant
+50% resistance
Effect of clinical Moxifloxacin 143.53 221.97 211.16 192.79 1.43% 4.28% 3.73% 3.56% Dominant Dominant Dominant
failure rates 20% clinical failure rate
for resistant isolates Clarithromycin 143.53 201.13 211.16 139.27 1.43% 3.62% 3.73% 2.03% Dominant Dominant 707.45
20% clinical failure rate
Co-amoxiclav/cefuroxime 143.53 221.97 193.61 192.79 1.43% 4.28% 3.24% 3.56% Dominant Dominant Dominant
20% clinical failure rate
Effect of choice of CLSI resistance rates 143.53 216.83 205.12 178.12 1.43% 4.12% 3.56% 3.14% Dominant Dominant Dominant
resistance rates for
H. influenzae
Effect of Costs decreased by 50% 120.88 154.29 152.15 136.51 1.43% 4.28% 3.73% 3.56% Dominant Dominant Dominant
hospitalization
Costs increased by 50% 166.18 289.66 270.16 249.08 1.43% 4.28% 3.73% 3.56% Dominant Dominant Dominant
Intensive care unit 145.52 227.91 216.33 197.73 1.43% 4.28% 3.73% 3.56% Dominant Dominant Dominant
admission rate 4.9%
Hospitalisation rate 0% 123.43 157.61 125.37 110.83 0.78% 2.21% 0.95% 0.91% Dominant Dominant 10,192.01
following first-line failure
Viruses 20% viruses as causative 131.87 191.33 183.61 165.80 1.15% 3.42% 2.98% 2.85% Dominant Dominant Dominant
pathogen
First-line treatment All moxifloxacin second- 129.92 187.33 211.16 192.79 0.98% 3.07% 3.73% 3.56% Dominant Dominant Dominant
line
comparison All co-amoxiclav second- 143.53 258.65 286.51 225.28 1.43% 5.39% 6.19% 4.66% Dominant Dominant Dominant
line

Note: CLSI = Clinical and Laboratory Standards Institute; ICER = incremental cost-effectiveness ratio

Cost-effectiveness of antimicrobial treatment of CAP in Belgium  Martin et al.  745

 cefuroxime/moxifloxacin or clarithro­m ycin/moxi­
floxacin, respectively. The moxifloxacin/co-amoxiclav
strategy offered lower rates of first-line clinical failures,
of second-line treatments required, of hospitalisations
required and of mortality, than the three comparator
strategies.
Deterministic sensitivity analysis
Deterministic one-way sensitivity analyses were carried
out for the effectiveness measures of first-line clinical
failures avoided (see Table 2) and hospitalisations
avoided (see Table 3). Results are extremely robust to
change, with moxifloxacin/co-amoxiclav remaining the
dominant strategy in nearly all scenarios. This is driven
primarily by the absence of resistance to moxifloxacin
Curr Med Res Opin Downloaded from informahealthcare.com by University Library Utrecht on 09/15/13
and the high clinical success rate associated with
moxifloxacin. There are only two scenarios for which
moxifloxacin/co-amoxiclav is not the dominant
strategy: (a) when the clarithromycin clinical failure
rate is reduced from 50% to 20% while keeping the
failure rate for moxifloxacin unchanged and (b)
when first-line treatment failure did not result in any
hospitalisation, but initiated only second-line treatment
in the community.
The first scenario decreased clinical failure rates from
For personal use only.
50% to 20% using 5% decrements (i.e. the following
values were tested for each antimicrobial: 50%, 45%,
40%, 35%, 30%, 25% and 20%. Only the results
for 20% are presented in Tables 2 and 3). Adjusting
clinical failure rates of moxifloxacin, co-amoxiclav and
cefuroxime incrementally did not change any results
significantly, nor did reducing clarithromycin clinical
failure rates incrementally from 50% to 25%. However,
once clarithromycin clinical failure rates were decreased
to 20%, the clarithromycin/moxifloxacin strategy
became less costly than the moxifloxacin/co-amoxiclav
strategy (139.27€ vs. 143.53€). This is due to an
improved clinical profile for clarithromycin combined
with the lower cost of treatment with clarithromycin
therapy in Belgium (13.76€ vs. 41.20€ for treatment
with moxifloxacin). Effectiveness measures remain
favourable for the moxifloxacin/co-amoxiclav strategy,
resulting in ICERs of 79.44€ per first-line failure avoided
and 707.45€ per hospitalisation avoided. However, it
may be considered unrealistic to assume that the clinical
failure rate for clarithromycin is 20% whereas the clinical
failure rate for moxifloxacin is 50%.
The second scenario for which moxifloxacin/
co-amoxiclav lost its dominant position is in the
unlikely event of no hospitalisations following first-line
treatment failure. In this situation, the clarithromycin/
moxifloxacin strategy became less costly than the
moxifloxacin/co-amoxiclav strategy (110.83€ vs.
123.43€) since the higher level of clarithromycin
746  Cost-effectiveness of antimicrobial treatment of CAP in Belgium © 2008 Informa UK Ltd – Curr Med Res Opin 2008; 24(3)
resistance can no longer result in direct hospitalisation
after a first regimen of clarithromycin without an
outpatient second-line treatment of moxifloxacin first
administered. Effectiveness measures remained in
favour of the moxifloxacin/co-amoxiclav strategy. This
resulted in ICERs of moxifloxacin/co-amoxiclav as
compared with clarithromycin/moxifloxacin of 96.03€
per first-line clinical failure avoided and 10  192.01€
per hospitalisation avoided.
There was little impact on results of setting all
antimicrobial success rates at 100%. This was due
to the low prevailing levels of resistance. Adjusting
resistance rates by 50% had little effect on results,
in particular those of moxifloxacin or co-amoxiclav
treatments as both have S. pneumoniae resistance rates
of 0%. Results were also robust to the use of CLSI
rather than PK/PD resistance rates for H. influenzae,
50% increases and decreases to hospitalisation costs,
the allowance of admissions to the intensive care unit,
the inclusion of viruses as a ‘causative’ pathogen, and
the standardisation of second-line treatments to either
moxifloxacin or co-amoxiclav.
Probabilistic sensitivity analysis
Probabilistic sensitivity analyses were carried out of
incremental cost-effectiveness results expressed as
cost per first-line clinical failure avoided. Figure 2
shows the scatter plot of 1000 ICER results comparing
moxifloxacin/co-amoxiclav with co-amoxiclav/
clarithromycin on the cost-effectiveness plane
and a 95% confidence ellipse. The moxifloxacin/
co-amoxiclav strategy has a 99.4% probability of
dominating co-amoxiclav/clarithromycin. Similarly,
the moxifloxacin/co-amoxiclav strategy has a 99.9%
probability of dominating cefuroxime/moxifloxacin
(see Figure 3) and a 92.6% probability of dominating
clarithromycin/moxifloxacin (see Figure 4). If decision
makers are willing to pay a maximum amount of, for
instance, 200€ per first-line clinical failure avoided, the
cost-effectiveness acceptability curves demonstrate that
the probability of moxifloxacin/co-amoxiclav being
cost-effective is 99.5%, 100% and 98.5% as compared
with co-amoxiclav/clarithromycin (see Figure 5), cefur­
oxime/moxifloxacin (see Figure 6) and clarithro­mycin/
moxifloxacin (see Figure 7), respectively.
Discussion
This study assessed the cost-effectiveness of various
empirical antimicrobial strategies to treat CAP in the
community in Belgium. The base case analysis showed
that first-line/second-line treatment with moxi­
floxacin/co-amoxiclav dominated all other strategies
 50

Incremental Costs
0
-5.00% 0.00% 5.00% 10.00% 15.00% 20.00% 25.00%
-50

-100

-150

-200

-250
Incremental Outcomes

Figure 2.  Cost-effectiveness plane of moxifloxacin/co-amoxiclav versus co-amoxiclav/clarithromycin in terms of cost per
first-line clinical failure avoided

50
Incremental Costs
Curr Med Res Opin Downloaded from informahealthcare.com by University Library Utrecht on 09/15/13

0
-5.00% 0.00% 5.00% 10.00% 15.00% 20.00% 25.00%
-50

-100

-150

-200

-250
Incremental Outcomes

Figure 3.  Cost-effectiveness plane of moxifloxacin/co-amoxiclav versus cefuroxime/moxifloxacin in terms of cost per first-line
For personal use only.

clinical failure avoided

40
20
Incremental Costs

0
0.00%
-20 5.00% 10.00% 15.00% 20.00% 25.00%

-40
-60
-80
-100
-120
Incremental Outcomes

Figure 4.  Cost-effectiveness plane of moxifloxacin/co-amoxiclav versus clarithromycin/moxifloxacin in terms of cost per
first-line clinical failure avoided

under all effectiveness measures considered (first-
line clinical failure avoided, second-line treatment
avoided, hospitalisation avoided and deaths avoided).
The deterministic and probabilistic sensitivity analyses
demonstrated that results are extremely robust to
change, with moxifloxacin/co-amoxiclav remaining
the dominant strategy in nearly all scenarios. From a
clinical perspective, these results indicate that the
absence of resistance to moxifloxacin and the high
clinical success rate associated with moxifloxacin leads
to moxifloxacin/co-amoxiclav being the most effective
and least expensive option in most cases.
This model was developed to assess the impact of
resistance for CAP patients treated in the community.
Treatment guidelines are available in most countries,
though these are not always followed in general
practice 45. This model demonstrated the economic
and clinical effects of a possible policy change, as
moxifloxacin may currently be under-utilised in
Belgium for the treatment of CAP. The analysis was
based on three commonly prescribed and less costly
comparators, substantially penalising moxifloxacin,
which is 67%, 40% and 199% more costly than
co-amoxiclav, cefuroxime and clarithromycin,

© 2008 Informa UK Ltd – Curr Med Res Opin 2008; 24(3) Cost-effectiveness of antimicrobial treatment of CAP in Belgium  Martin et al.  747
 1.00
0.90

co-amoxiclav versus co-amoxiclav/

clarithromycin being cost-effective
Probability of moxifloxacin /
0.80
0.70
0.60
0.50
0.40
0.30
0.20

0.00
0.00 200.00 400.00 600.00 800.00 1000.00
Cost-effectiveness threshold (€ per first-line clinical failure avoided)

Figure 5.  Cost-effectiveness acceptability curve of moxifloxacin/co-amoxiclav versus co-amoxiclav/clarithromycin in terms of

Curr Med Res Opin Downloaded from informahealthcare.com by University Library Utrecht on 09/15/13

cost per first-line clinical failure avoided.

1.00
0.90
co-amoxiclav versus cefuroxime /
moxifloxacin being cost-effective
Probability of moxifloxacin /

0.80
0.70
0.60
0.50
For personal use only.

0.40
0.30
0.20
0.10
0.00
0.00 200.00 400.00 600.00 800.00 1000.00

Cost-effectiveness threshold (€ per first-line clinical failure avoided)

Figure 6.  Cost-effectiveness acceptability curve of moxifloxacin/co-amoxiclav versus cefuroxime/moxifloxacin in terms of cost
per first-line clinical failure avoided.

1.00
co-amoxiclav versus clarithromycin /

0.90
moxifloxacin being cost-effective
Probability of moxifloxacin /

0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.00
0.00 200.00 400.00 600.00 800.00 1000.00
Cost-effectiveness threshold (€ per first-line clinical failure avoided)

Figure 7.  Cost-effectiveness acceptability curve of moxifloxacin/co-amoxiclav versus clarithromycin/moxifloxacin in terms of

cost per first-line clinical failure avoided.

748  Cost-effectiveness of antimicrobial treatment of CAP in Belgium © 2008 Informa UK Ltd – Curr Med Res Opin 2008; 24(3)
 respectively. In spite of this, these higher drug costs
are offset by the increased clinical benefits offered
by moxifloxacin in terms of reduced number of first-
line clinical failures and hospitalisations, making this a
viable treatment option for CAP in Belgium.
Few studies have examined clinical outcomes and
costs of empirical antimicrobial treatment of CAP.
Nevertheless, there is some evidence supporting the
validity of our model. Our study found hospitalisation
rates varying from 1.43% to 4.28%, depending on the
antimicrobial treatment strategy. As in our model, two
studies have calculated the rate of hospitalisation for
CAP in the presence of antimicrobial resistance. A
French study observed hospitalisation rates of 13–14%
of patients diagnosed with CAP 46. These higher
hospitalisation rates can be attributed to the inclusion
Curr Med Res Opin Downloaded from informahealthcare.com by University Library Utrecht on 09/15/13
of patients suffering from mild to severe CAP (i.e.
Fine category I–V) in the French study, whereas our
study was limited to patients suffering from mild to
moderate CAP (i.e. Fine category I–III). Hospitalisation
rates in a Spanish study amounted to 8–14%. These
hospitalisation rates exceeded the rates observed in
our study because the Spanish model assumed higher
hospitalisation rates following first-line treatment
failure 47. For example, the Spanish study assumed
that treatment failure with moxifloxacin is always
For personal use only.
followed by hospitalisation. This does not reflect
current treatment pathways in Belgium, where around
85% of patients failing treatment with moxifloxacin
start second-line treatment in the community and
approximately 15% of patients are hospitalised.
An economic evaluation examined the cost-
effectiveness of antimicrobial treatment of CAP in
France, United States and Germany 19. Our Belgian
study supports the findings of this international study,
where first-line treatment with moxifloxacin followed
by co-amoxiclav dominated all other treatments for all
outcome measures in the three countries considered.
Another economic evaluation, using a different model,
was set in Spain. The authors found that first-line
treatment with moxifloxacin dominated treatment with
clarithromycin, telithromycin or amoxicillin in that
moxifloxacin treatment was less expensive and more
effective in terms of the number of patients without
complications and the number of patients hospitalised47.
This model suffers from a number of limitations.
As the aim of the study was to examine the cost-
effectiveness of initial antimicrobial treatment in the
community, the model included patients suffering from
mild-to-moderate CAP. By limiting the analysis to Fine
categories I–III, the patient population is relatively
healthy. As a consequence, our cost-effectiveness
results cannot be extrapolated to patient populations
that have a different mix of Fine categories. For
instance, patients suffering from severe CAP (i.e. Fine
© 2008 Informa UK Ltd – Curr Med Res Opin 2008; 24(3) Cost-effectiveness of antimicrobial treatment of CAP in Belgium  Martin et al.  749
category IV–V) are likely to have a higher mortality rate
in the community, hospitalisation rate and in-hospital
mortality rate, influencing the cost-effectiveness of
antimicrobial treatment strategies.
The majority of patients classified as Fine category
I–III are treated in the community with some
requiring hospitalisation. In our model, all patients are
hospitalised following second-line treatment failure.
This is the case for the majority of patients, although
some patients may still receive a third treatment in the
community. The effect of changes in the hospitalisation
rate was also tested in the sensitivity analysis, with little
effect on the ICERs.
Our model of treatment pathways is necessarily
simplistic and reflects empirical treatment in Belgium.
Antimicrobial treatment in the community is often
empirical4. The empirical approach is also the prevailing
method of drug selection 48. However, empirical
treatment does not always reflect real clinical practice
where data on the pathogen and the susceptibility may
be available following first-line treatment failure. If the
physician can adopt, with a high degree of reliability,
the antimicrobial treatment strategy according to the
cause, the most efficient treatment strategy may vary.
Estimates of several input parameters in Belgium were
unknown and, in these cases, international data were
relied upon. With respect to antimicrobial resistance
data, the model used Belgian data when available
and used international resistance data in the absence
of Belgian data. The deterministic and probabilistic
sensitivity analyses showed that cost-effectiveness results
were robust to changes in resistance rates. With respect
to resource utilisation, the analysis drew on French data
on the frequency of physician visits, the frequency of
diagnostic tests and the hospitalisation rate for CAP
patients. For instance, the hospitalisation rate following
first-line treatment failure of 15.4% for France was used in
our model24. This was because the French estimate related
to a similar patient population to that considered in our
model (i.e. patients initially treated in the community)
and because of similar treatment patterns between
Belgium and France. Nevertheless, although the French
health care system may be similar in many aspects to the
Belgian system, the results may not present an accurate
picture of Belgian resource utilisation and costs for CAP.
Rates of clinical failure due to resistance were
assumed for each CAP pathogen. There is no published
literature documenting the link between resistance and
clinical failure. Therefore, the calculations of failure
rates were based on the opinion of clinical experts.
A previous version of this model, developed for the
United States, was compared to empirical data on
treatment failure and results were very comparable,
thus validating the assumption that only 50% of
resistant pathogens results in clinical failure.
 Indirect costs due to productivity loss were not
included in this model as the perspective was that of
the Belgian third-party payer. The inclusion of indirect
costs is expected to result in even more positive
outcomes for the moxifloxacin/co-amoxiclav strategy.
The patient population considered in the model
includes a large percentage of the working population
and as moxifloxacin treatment entails fewer clinical
consequences, this would result in less time off-work
and, hence, less productivity loss.
Despite its limitations, this study reports on one
of the few models in CAP that assess both costs and
outcomes associated with antimicrobial treatment
taking into account bacterial resistance. Its results can
aid local decision makers to allocate scarce health care
resources. Future modelling work in this domain would
Curr Med Res Opin Downloaded from informahealthcare.com by University Library Utrecht on 09/15/13
benefit from more recent data on resistance. There is
also a need for studies that examine the relationship
between resistance and clinical failure.
Conclusions
First-line/second-line empirical treatment of CAP
patients with moxifloxacin/co-amoxiclav is a dominant
strategy as compared with other current treatment
For personal use only.
strategies in Belgium. Treatment with moxifloxacin
is more effective in terms of first-line clinical failure,
need for second-line treatment and hospitalisation as
compared with first-line treatment with co-amoxiclav,
cefuroxime or clarithromycin. As a consequence,
total healthcare costs of treating a CAP episode with
moxifloxacin/co-amoxiclav are lower than costs of
comparator strategies, despite the higher drug cost
of moxifloxacin. These results demonstrate that it is
more cost-effective to select an antimicrobial as first-
line treatment that is more effective, i.e. with lower
resistance and a higher clinical success rate.
Acknowledements
Declaration of interest: This study was carried out
by i3 Innovus under an unrestricted grant from Bayer
Healthcare. The authors have no conflicts of interest that
are directly relevant to the content of this manuscript.
References
1. Almirall J, Bolibar I, Vidal J, et al. Epidemiology of community-
acquired pneumonia in adults: a population-based study. Eur
Respir J 2000;15:757-63
2. Jokinen C, Heiskanen L, Juvonen H, et al. Incidence of
community-acquired pneumonia in the population of
four municipalities in eastern Finland. Am J Epidemiol
1993;137:977-88
750  Cost-effectiveness of antimicrobial treatment of CAP in Belgium © 2008 Informa UK Ltd – Curr Med Res Opin 2008; 24(3)
3. Anderson RN. Deaths: leading causes for 2000. Natl Vital Stat
Rep 2002;50:1-86
4. Talwar A, Lee H, Fein A. Community-acquired pneumonia:
what is relevant and what is not? Curr Opin Pulm Med
2007;13:177-85
5. Grossman RF, Rotschafer JC, Tan JS. Antimicrobial treatment
of lower respiratory tract infections in the hospital setting. Am
J Med 2005;118(Suppl 7A);29-38
6. Niederman MS, McCombs JS, Unger AN, et al. The cost
of treating community-acquired pneumonia. Clin Ther
1998;20:820-37
7. Lave JR, Lin CJ, Fine MJ, et al. The cost of treating patients
with community-acquired pneumonia. Semin Respir Crit Care
Med 1999;20:189-97
8. Fine MJ, Pratt HM, Obrosky DS, et al. Relation between length
of hospital stay and costs of care for patients with community-
acquired pneumonia. Am J Med 2000;109:378-85
9. Fine MJ, Stone RA, Singer DE, et al. Processes and outcomes of
care for patients with community-acquired pneumonia: results
from the Pneumonia Patient Outcomes Research Team (PORT)
cohort study. Arch Intern Med 1999;159:970-80
10. Fine MJ, Auble TE, Yealy DM, et al. A prediction rule to
identify low-risk patients with community-acquired pneumonia.
New Engl J Med 1997;336:243-50
11. Lim WS, van der Eerden MM, Laing R, et al. Defining
community acquired pneumonia severity on presentation to
hospital: an international derivation and validation study.
Thorax 2003;58:377-82
12. Aliyu ZY, Aliyu MH, McCormick K. Determinants for
hospitalization in ‘low-risk’ community acquired pneumonia.
BMC Infect Dis 2003;3:11
13. Carratala J, Martin-Herrero JE, Mykietiuk A, et al. Clinical
experience in the management of community-acquired
pneumonia: lessons from the use of fluoroquinolones. Clin
Microbiol Infect 2006;12:2-11
14. Woodhead M, Blasi F, Ewig S, et al. Guidelines for the
management of adult lower respiratory tract infections. Eur
Respir J 2005;26:1138-80
15. Simoens S, Verhaegen J, Laekeman G, Peetermans WE. Treating
respiratory tract infections in ambulatory care in Belgium:
fluoroquinolone consumption and resistance development. Int J
Antimicrob Agents 2005;26:62-8
16. Institut Pasteur. Evolution of fluoroquinolone resistance and
detection of resistance mechanisms in clinical isolates of
Streptococcus pneumoniae collected in Belgium in 1995–2005.
Institut Pasteur. Brussels, 2006. Available at www.pasteur.
be/pasteur_fr/index5930.html?page=antibiotic_research [last
accessed 4 January 2008]
17. File Jr TM. Streptococcus pneumoniae and community-acquired
pneumonia: a cause for concern. Am J Med 2004;117(Suppl
3A);39-50
18. Belgian Antibiotic Policy Coordination Committee. Belgian
guide for anti-infectious treatment in ambulatory practice.
Brussels: Belgian Antibiotic Policy Coordination Committee;
2006
19. Martin M, Quilici S, File T, et al. Cost-effectiveness of
empirical prescribing of antimicrobials in community-acquired
pneumonia in three countries in the presence of resistance.
J Antimicrob Chemother 2007;59:977-89
20. Valcke Y. CAP guidelines: why differences between Belgian and
American/Canadian guidelines? Louvain: Université Catholique
de Louvain; 2001. Available at: http://www.farm.ucl.ac.be/
seminfec//Atelier-Mandell-23-11-01/Valcke/Valcke.pdf [last
accessed 4 January 2008]
21. Jones ME, Blosser-Middleton RS, Critchley IA, et al. In
vitro susceptibility of Streptococcus pneumoniae, Haemophilus
influenzae and Moraxella catarrhalis: a European multicenter
study during 2000–2001. Clin Microbiol Infect 2003;9:590-9
22. Welte T, Marre R, Suttorp N. What is new in the treatment
of community-acquired pneumonia? Med Klin (Munich)
2006;101:313-20
23. Bartlett JG, Dowell SF, Mandell LA, et al. Practice guidelines
for the management of community-acquired pneumonia in
adults [Infectious Diseases Society of America]. Clin Infect Dis
2000;31:347-82
 24. Laurichesse H, Gerbaud L, Baud O, et al. Hospitalization
decision for ambulatory patients with community-acquired
pneumonia: a prospective study with general practitioners in
France. Infection 2001;29:320-5
25. Torres A, Muir JF, Corris P, et al. Effectiveness of oral
moxifloxacin in standard first-line therapy in community-
acquired pneumonia. Eur Respir J 2003;21:135-43
26. Petitpretz P, Arvis P, Marel M, et al. Oral moxifloxacin vs
high-dosage amoxicillin in the treatment of mild-to-moderate,
community-acquired, suspected pneumococcal pneumonia in
adults. Chest 2001;119:185-95
27. Hoeffken G, Meyer HP, Winter J, Verhoef L. The efficacy
and safety of two oral moxifloxacin regimens compared to
oral clarithromycin in the treatment of community-acquired
pneumonia. Respir Med 2001;95:553-64
28. Faich GA, Morganroth J, Whitehouse AB, et al. Clinical
experience with moxifloxacin in patients with respiratory tract
infections. Ann Pharmacother 2004;38:749-54
29. Leophonte P, File T, Feldman C. Gemifloxacin once daily for
7 days compared to amoxicillin/clavulanic acid thrice daily for
10 days for the treatment of community-acquired pneumonia of
suspected pneumococcal origin. Respir Med 2004;98:708-20
Curr Med Res Opin Downloaded from informahealthcare.com by University Library Utrecht on 09/15/13
36. Dagan R. Achieving bacterial eradication using pharmacokinetic/
pharmacodynamic principles. Int J Infect Dis 2003;7(Suppl
1):S21-S26
37. Hoban D, Felmingham D. The PROTEKT surveillance study:
antimicrobial susceptibility of Haemophilus influenzae and
Moraxella catarrhalis from community-acquired respiratory
tract infections. J Antimicrob Chemother 2002;50(Suppl
S1):49-59
38. Institut Pasteur. Presence of macrolide resistance mechanisms in
clinical isolates of Streptococcus pneumoniae collected in Belgium
in 1995–2005. Institut PasteurP: Brussels; 2006. Available at
www.pasteur.be/pasteur_fr/index5930.html?page=antibiotic_
research [last accessed 30 January 2007]
39. Reinert RR, Simic S, Al Lahham A, et al. Antimicrobial resistance
of Streptococcus pneumoniae recovered from outpatients with
respiratory tract infections in Germany from 1998 to 1999:
results of a national surveillance study. J Clin Microbiol
2001;39:1187-9
40. Institut National d’Assurance Maladie-Invalidité (INAMI).
Brussels, 2006. Available at: www.inami.fgov.be [last accessed
30 January 2007]
41. Fantin B, Aubert JP, Unger P, et al. Clinical evaluation of the

30. Higuera F, Hidalgo H, Feris J, et al. Comparison of oral
cefuroxime axetil and oral amoxycillin/clavulanate in the
treatment of community-acquired pneumonia. J Antimicrob
Chemother 1996;37:555-64
31. Mathers DL, Hassman J, Tellier G. Efficacy and tolerability of
once-daily oral telithromycin compared with clarithromycin
for the treatment of community-acquired pneumonia in adults.
Clin Ther 2004;26:48-62
32. Tellier G, Niederman MS, Nusrat R, et al. Clinical and bacteriological
efficacy and safety of 5 and 7 day regimens of telithromycin once
daily compared with a 10  day regimen of clarithromycin twice
daily in patients with mild to moderate community-acquired
pneumonia. J Antimicrob Chemother 2004;54:515-23
33. Gordon KA, Biedenbach DJ, Jones RN. Comparison
For personal use only.
management of community-acquired pneumonia by general
practitioners in France. Chest 2001;120:185-92
42. Colice GL, Morley MA, Asche C, Birnbaum HG. Treatment
costs of community-acquired pneumonia in an employed
population. Chest 2004;125:2140-5
43. Programme de Médicalisation des Systèmes d’Information
(PMSI). Paris, 2006. Available at: www.pmsi.fr [last accessed
30 January 2007]
44. Williams D, Hopkins S. Safety of trovafloxacin in treatment of
lower respiratory tract infections. Eur J Clin Microbiol InfectDis
1998;17:454-8
45. Flanders SA, Halm EA. Guidelines for community-acquired
pneumonia: are they reflected in practice? Treat Respir Med
2004;3:67-77

of Streptococcus pneumoniae and Haemophilus influenzae
susceptibilities from community-acquired respiratory tract
infections and hospitalized patients with pneumonia: five-year
results for the SENTRY Antimicrobial Surveillance Program.
Diagn Microbiol Infect Dis 2003;46:285-9
34. Jacobs MR, Felmingham D, Appelbaum PC, et al. The
Alexander Project 1998–2000: susceptibility of pathogens
isolated from community-acquired respiratory tract infection to
commonly used antimicrobial agents. J Antimicrob Chemother
2003;52:229-46
35. Jacobs MR. How can we predict bacterial eradication? Int J
Infect Dis 2003;7(Suppl 1):S13-S20
46. Singer ME, Harding I, Jacobs MR, Jaffe DH. Impact of
antimicrobial resistance on health outcomes in the out-
patient treatment of adult community-acquired pneumonia: a
probability model. J Antimicrob Chemother 2003;51:1269-82
47. Sabes-Figuera R, Segu JL, Puig-Junoy J, Torres A. Influence of
bacterial resistances on the efficiency of antibiotic treatments
for community-acquired pneumonia. Eur J Health Econ
2007: published online 13 January 2007, doi: 10.1007/
s10198-006-0019-0
48. Kuti JL, Capitano B, Nicolau DP. Cost-effective approaches to
the treatment of community-acquired pneumonia in the era of
resistance. Pharmacoeconomics 2002;20:513-28

CrossRef links are available in the online published version of this article:
http://www.cmrojournal.com
Paper CMRO-4358_2, 09:03-19.02.08
Accepted for publication: 11 January 2008
Published Online: 28 January 2008
doi:10.1185/030079908X273336

© 2008 Informa UK Ltd – Curr Med Res Opin 2008; 24(3) Cost-effectiveness of antimicrobial treatment of CAP in Belgium  Martin et al.  751