Estimating the morbidity and mortality associated with infections due to multidrug-resistant bacteria (MDRB), France, 2012
Colomb-Cotinat et al. Antimicrobial Resistance and Infection Control
Estimating the morbidity and mortality associated with infections due to multidrug-resistant bacteria (MDRB), France, 2012
M. Colomb-Cotinat 2
J. Lacoste 2
C. Brun-Buisson 1
V. Jarlier 0
B. Coignard 2
S. Vaux 2
0 Sorbonne Universités, UPMC Univ Paris 06, Inserm, Centre d'Immunologie et des Maladies Infectieuses, UMR 1135 & APHP , CHU Pitié-Salpêtrière , Laboratoire de Bactériologie-Hygiène , F-75013 Paris , France
1 Assistance publique-hôpitaux de Paris , CHU Henri Mondor, F-94000 Créteil , France
2 Santé Publique France, The French Public Health Agency , F-94415 Saint-Maurice , France
Background: A study based on 2007 data estimated that 386,000 infections due to multidrug-resistant bacteria (MDRB) occurred in Europe that year and 25,000 patients died from these infections. Our objective was to estimate the morbidity and mortality associated with these infections in France. Methods: The MDRB considered were methicillin-resistant Staphylococcus aureus (MRSA), glycopeptide-resistant enterococci, third-generation cephalosporin-resistant (3GC-R) Escherichia coli and Klebsiella pneumoniae, carbapenemresistant Klebsiella pneumoniae, Acinetobacter spp. and Pseudomonas aeruginosa (CR P. aeruginosa). The number of invasive infections (infections with bacteria isolated from blood or cerebrospinal fluid) due to MDRB, as reported by France to EARS-Net in 2012, was corrected for the coverage of our surveillance network and extrapolated to other body sites using ratios from the French healthcare-associated infections point prevalence survey and the literature. Mortality associated with MDRB infection was estimated using proportions from the literature. Methods and parameters were reviewed by a panel of experts. Results: We estimate that 158,000 (127,000 to 245,000) infections due to MDRB occurred in 2012 in France (incidence: 1.48 to 2.85 per 1000 hospital days), including 16,000 invasive infections. MRSA, 3GC-R E. coli and K. pneumoniae were responsible for 120,000 (90,000 to 172,000) infections, i.e., 75% of the total. An estimated 12,500 (11,500 to 17,500) deaths were associated with these infections, including 2,700 associated with invasive infections. MRSA, 3GC-R E. coli and CR P. aeruginosa accounted for 88% of these deaths. Conclusion: These first estimates confirm that MRSA, 3GC-R Escherichia coli and Klebsiella pneumoniae account for the largest portion of the morbidity and mortality of infections due to MDRB in France. These results are not directly comparable with the European study because the methodology used differs in many respects. The differences identified between our study and previous studies underline the need to define a standardised protocol for international assessments of the morbidity and mortality of antibiotic resistance. Estimating morbidity and mortality will facilitate communication and awareness in order to reinforce adherence and support of healthcare professionals and policy-makers to MDRB prevention programs.
Antimicrobial resistance; Epidemiology; Morbidity; Mortality; Infection due to multidrug resistant bacteria; France
Antibiotic resistance is a constantly evolving phenomenon
and a threat to infection and disease control; it
complicates patient management and treatment strategy and
prolongs hospital stays. Nowadays, this international
public health problem is recognised as one of the scourges
of the 21st century .
Several studies have sought to estimate the morbidity
and mortality of infections due to multidrug-resistant
bacteria (MDRB). A joint report from the European
Centre for Disease Prevention and Control (ECDC) and
the European Medecines Agency (EMEA), published in
2009 and based on data from 2007 , estimated at
approximately 386,000 the annual number of infections
due to MDRB in Europe that year, including 42,500
cases (11%) of bloodstream infections. The number of
deaths associated with these infections was estimated at
more than 25,000. A report by the US Centers for
Disease Control and Prevention (CDC) from 2013 
provided an overview of the annual morbidity and
mortality of antibiotic-resistant infections in the United
States, estimating their number at approximately 2
million and the number of deaths associated with these
infections at 23,000. These two studies, although they
used different methods and did not consider the same
panel of microorganisms, both underlined the important
morbidity and mortality of antibiotic resistance on
No corresponding estimate has been so far available for
France, despite its wealth of antibiotic-resistance
surveillance networks. These networks focus on specific
pathogens and sometimes on specific sources of samples [4, 5].
The data they provide are useful for assessing trends and
detecting epidemics. Nonetheless, they do not provide an
overall view of the MDRB morbidity and mortality nor do
they permit simple communication on this topic.
The objective of this study conducted by Santé
publique France (the French national public health agency)
was to estimate for the first time the morbidity and
mortality (number of cases and number of attributable
deaths) of infections due to MDRB in France, in order to
advocate for strategies to control and prevent them, to
guide public health authorities in implementing these
policies, and to support communication towards
healthcare professionals and the general public.
– being associated with invasive infections, i.e.,
infections with a bacteria isolated from blood or
cerebrospinal fluid ;
– having a significant prevalence or having emerged
– being included in a surveillance network in France;
– being multidrug-resistant (MDR) as defined in the
European Antimicrobial Resistance Surveillance
Network (EARS-Net) protocol .
We thus selected eight bacteria–antibiotic combinations:
– Staphylococcus aureus resistant to methicillin
– Enterococcus faecium and E. faecalis resistant to
– Escherichia coli resistant to third-generation
cephalosporins (3GC-R E. Coli);
– Klebsiella pneumoniae resistant to third-generation
cephalosporins (3GC-R K. pneumoniae);
– Pseudomonas aeruginosa resistant to carbapenems
(CR P. aeruginosa);
– Klebsiella pneumoniae resistant to carbapenems
(CR K. pneumoniae);
– Acinetobacter spp. resistant to carbapenems
(CR Acinetobacter spp).
According to the EARS-Net protocol, antibiotics
considered for resistance were a) for MRSA: oxacillin,
methicillin, flucloxacillin, cloxacillin, dicloxacillin and
cefoxitin (and PCR mecA or PBP2a detection); b) for
GRE: vancomycin; c) for 3GC-R E.coli and 3GC-R K.
pneumoniae: cefotaxime, ceftriaxone, ceftazidime; d) for
CR P. aeruginosa; imipenem, meropenem; e) for CR
K. pneumoniae: imipenem, meropenem; f ) for CR
Acinetobacter spp: imipenem, meropenem, doripenem.
Streptococcus pneumoniae with reduced penicillin
susceptibility was not considered as MDRB in
accordance with generally accepted microbiological criteria .
Number of infections due to MDRB
To select the most frequently diagnosed infectious body
sites for each of these MDRB, data from the 2012 point
prevalence survey on healthcare-associated infections and
antimicrobial use in French hospitals (2012 PPS)  were
used. This survey, conducted every five years on a single
day in nearly all French healthcare facilities collects
standardised data about nosocomial infections. In 2012,
data were collected for 91% of French hospital beds.
Based on this data, the following infectious body sites
were selected, as there were the most frequent infections
in the 2012 PPS:
– for all MDRB: invasive infections (bacteria isolated
from blood or cerebrospinal fluid), urinary tract
infections, skin and soft tissue infections, and
surgical site infections;
– for all MDRB excluding GRE: respiratory tract
– for MRSA only: bone and joint infections (does not
include surgical site infections);
– for GRE only: gastrointestinal tract infections
(gastro-intestinal and intra-abdominal infections).
The number of invasive infections for each MDRB was
estimated from data transmitted by France to EARS-Net
. This European network collects resistance data of
bacterial strains isolated from invasive infections
(bacteria isolated from blood or cerebrospinal fluid). As
duplicate isolates for the same patients had already been
eliminated during data collection, the number of
incident cases of invasive infections due to MDRB
occurring each year in the network was directly
estimated from the number of resistant strains isolated. For
instance, there were 1005 strains of MRSA reported in
EARS-Net for France in 2012, so the number of incident
cases of invasive infections due to MRSA in 2012 in the
French EARS network was estimated as 1005.
The number of incident cases of invasive infections
was then corrected by taking into account the coverage
of the EARS-Net network for France, estimated from
the number of inpatient hospital days (HD) reported by
hospitals whose laboratories had transmitted data to the
network (secondary and tertiary care hospitals). The list
of these hospitals was obtained from the Epibac network
 and the total number of HD from the 2012 Health
Facilities Annual Statistics (SAE) report . The
EARS-Net coverage was estimated by dividing the number of
HD in the participating hospitals (N1) by the total number
of HD in all French secondary and tertiary care hospitals
that same year provided by the same source (N2). For
instance, the number of cases of invasive infections due to
MRSA in 2012 for France was estimated as 5574 (n1). The
incidence of patients with an invasive infection was also
expressed as the number of cases per 1000 HD.
In order to estimate the number of cases with
infections at other body sites, the ratio between the number
of cases with infections at a given body site and the
number of cases of invasive infections was calculated for
each MDRB and each body site using data from the
2012 PPS. For instance, to evaluate the number of cases
of urinary tract infections (UTI) due to MRSA in our
study, we first calculated a specific ratio between the
number of UTI due to MRSA reported in the 2012 PPS
(n = 103) divided by the number of invasive infections
due to MRSA reported in the 2012 PPS (n = 105). The
specific ratio for UTI due to MRSA was 103/105 = 0.98.
We then multiplied the number of cases with invasive
infections due to MRSA estimated in our study based on
EARS-Net data (n1 = 5574) by this ratio (0.98), to
estimate the number of cases with urinary tract infections
due to MRSA in France (n11 = 5468). The same approach
was used for other body sites and MDRB.
In order to have intervals of plausibility around
these estimates, we also estimated these ratios from a
non-systematic review of the literature targeting
French and European publications when available.
Plausibility intervals were then calculated using high
and low ratios estimated from the literature. Table 1
presents the ratios that were used, for each MDRB
and body sites, to estimate the incidence of cases of
non-invasive infections; ratios calculated from the
French PPS 2012 were used as target values, and
ratios excerpted from the literature as low and high
values for intervals of plausibility.
The total number of infections due to MDRB was
finally calculated by adding the number of infections
obtained for each MDRB species and each body site
Number of deaths associated with infections due to
For each infection due to MDRB and body site
considered, the number of deaths was estimated by applying to
the number of cases of infections estimated above the
proportion of deaths associated with infection due to
MDRB, based on a non-systematic review of the
literature (Table 2) that focused on publications reporting
proportions of deaths associated with infections due to
MDRB considered. Results from French or European
studies were used preferentially when available. The
mortality indicator used in our study was therefore the
number of deaths associated with MDRB infections as a
whole, but not the number specifically associated with
antibiotic resistance per se. Consequently, this study did
not estimate the extra deaths relative to infections due
to antibiotic-susceptible bacteria.
When a value for the associated mortality for a
body site could not be found in the literature, it was
estimated from the mortality associated with invasive
infections (mainly bloodstream infections) for each of
the MDRB studied, applying a correction factor from
a US publication .
The total number of deaths associated with an
infection due to MDRB was calculated by adding the number
of associated deaths obtained for each MDRB and each
body site studied.
We jointly validated the choices of the different
parameters used in this study (ratios used to estimate
the incidence of cases of non-invasive infections and
associated mortality proportion) after a critical review with
external experts, clinicians and microbiologists.
Figure 1 summarises the methodology used to estimate
morbidity and mortality associated with infections due
to MDRB in France.
Table 1 Ratios used to estimate the incidence of cases of non-invasive infections, France 2012
Ratios from the literature
No. MRSA urinary tract infections/No. MRSA invasive infections
No. MRSA respiratory infections/No. MRSA invasive infections
No. MRSA skin and soft tissue infections and surgical site
infections/No. MRSA invasive infections
No. MRSA bone and joint infections/No. MRSA invasive infections
No. GRE urinary tract infections/No. GRE invasive infections
No. GRE gastrointestinal tract infections /No. GRE invasive infections
No. GRE skin and soft tissue infections and surgical site
infections/No. GRE invasive infections
No. 3GC-R E. coli respiratory tract infections/No. 3GC-R E. coli
No. 3GC-R E. coli skin and soft tissue infections and surgical site
infections/No. 3GC-R E. coli invasive infections
No. 3GC-R K. pneumoniae urinary tract infections/No. 3GC-R
K. pneumoniae invasive infections
No. 3GC-R K. pneumoniae respiratory tract infections/No. 3GC-R
K. pneumoniae invasive infections
No. 3GC-R K. pneumoniae skin and soft tissue infections and
surgical site infections/No. 3GC-R K. pneumoniae invasive infections
No. CR K. pneumoniae urinary tract infections/No. CR
K. pneumoniae invasive infections
No. CR K. pneumoniae respiratory tract infections/No. CR
K. pneumoniae invasive infections
No. CR K. pneumoniae skin and soft tissue infections and surgical
site infections/No. CR K. pneumoniae invasive infections
No. CR P. aeruginosa urinary tract infections/No. CR P. aeruginosa
No. CR P. aeruginosa respiratory tract infections/No. CR
P. aeruginosa invasive infections
No. CR P. aeruginosa skin and soft tissue infections and surgical
site infections/No. CR P. aeruginosa invasive infections
No. CR Acinetobacter urinary tract infections/No. CR Acinetobacter
No. CR Acinetobacter respiratory tract infections/No. CR Acinetobacter
No. CR Acinetobacter skin and soft tissue infections and surgical site
infections/No. CR Acinetobacter invasive infections
Estimate of EARS-Net coverage, France
We estimate that in 2012 EARS-Net covered 18% of HD
in secondary and tertiary care hospitals in France.
Number of infections due to MDRB
The number of infections due to MDRB was estimated
at approximately 158,000 in France in 2012, for an
incidence of 1.83 per 1,000 HD. Data from the literature
enabled us to estimate a plausibility interval for the
annual number of cases from 127,000 to 245,000
(incidence: 1.48 to 2.85 per 1,000 HD).
Most cases were infections caused by MRSA (33%;
plausibility interval: 28 to 38%), 3GC-R E. coli (32%; 26
to 32%), CR P. aeruginosa (23%; 23 to 29%) or 3GC-R K.
pneumoniae (10%; 6 to 10%). The other infections due
Table 2 Proportion of mortality associated with infections due to MDRB, France 2012
MRSA urinary tract infections
MRSA respiratory tract infections
MRSA bone and joint infections
GRE invasive infections
GRE urinary tract infections
GRE gastrointestinal tract infections
MRSA skin and soft tissue infections and surgical site infections
GRE skin and soft tissue infections and surgical site infections
3GC-R E. coli invasive infections
3GC-R E. coli urinary tract infections
3GC-R E. coli respiratory tract infections
3GC-R E. coli skin and soft tissue infections and surgical site infections
3GC-R K. pneumoniae skin and soft tissue infections and surgical site infections
3GC-R K. pneumoniae skin and soft tissue infections and surgical site infections
3GC-R K. pneumoniae invasive infections
3GC-R K. pneumoniae urinary tract infections
3GC-R K. pneumoniae respiratory tract infections
CR K. pneumoniae invasive infections
CR K. pneumoniae urinary tract infections
CR K. pneumoniae respiratory tract infections
CR P. aeruginosa invasive infections
CR P. aeruginosa urinary tract infections
CR P. aeruginosa respiratory infections
CR Acinetobacter invasive infections
CR Acinetobacter urinary tract infections
CR Acinetobacter respiratory infections
CR P. aeruginosa skin and soft tissue infections and surgical site infections
CR Acinetobacter skin and soft tissue infections and surgical site infections
to MDRB considered accounted for less than 1% of all
cases. Most infections were due to Gram-negative
bacteria (67%; 62 to 71%).
Table 3 details the number and incidence rates of
infections according to whether or not the infection was
invasive. The total number of invasive infections due to
MDRB was estimated at approximately 16,000 in 2012
(incidence: 0.185 per 1,000 HD), or 10% (6 to 13%) of all
MDRB infections. Overall, 70% of the invasive infections
were due to MRSA or 3GC-R E. coli.
Infections caused by MRSA were mostly skin and soft
tissue and surgical site infections, which altogether
accounted for 56% (42 to 45%) of MRSA infections, with
an incidence of 0.318 (0.270 to 0.340) per 1,000 HD.
Urinary tract infections accounted for most of the
3GC-R E. coli infections (40%; range 40 to 61%), with an
incidence of 0.152 (0.364-0.364) per 1,000 HD, and
respiratory tract infections for most of the CR P.
aeruginosa infections (64%; range 48 to 64%), with an
incidence of 0.274 (0.274 to 0.398) per 1,000 HD.
Number of deaths associated with infections due to
The number of deaths associated with infections due to
MDRB in 2012 was estimated at 12,411 (11,422 to 17,470),
of which approximately one fourth (2,800) were due to
invasive infections. The corresponding incidence rate was
estimated at 0.144 death per 1,000 HD (0.133 to 0.203).
More than half the deaths were associated with CR P.
aeruginosa (53%; 54 to 58%) (Table 4). Infections caused
Fig. 1 Methodology for estimating the morbidity and mortality of infections due to MDRB in France, 2012
by MRSA and by 3GC-R E.coli were responsible for
respectively 16% (15 to 18%) and 18% (16 to 20%) of
deaths. Overall, Gram-negative bacteria infections
accounted for 83% (82 to 84%) of the estimated total
Among the deaths associated with infections due to
MDRB, 22% (16 to 24%) were due to invasive infections.
This proportion was 11% (7 to 11%) for the CR P.
aeruginosa infections, 24% (20 to 29%) for MRSA
infections, and 51% (29 to 51%) for the 3GC-R E. coli
Our study quantifies for the first time the morbidity and
mortality of infections due to MDRB in France. We thus
estimate at approximately 158,000 (127,000 to 245,000)
the number of infections due to MDRB occurring in
2012 in France, including nearly 16,000 (6 to 12%)
invasive infections (bloodstream infections and meningitis),
MRSA and the two major enterobacteriaceae species
resistant to third-generation cephalosporins account for
three quarters (70 to 75%) of the infections recorded.
The annual number of deaths associated with these
infections is estimated at 12,500 (11,500 to 17,500),
including 2,800 (22%) due to invasive infections; MRSA,
3GC-R E. coli and CR P. aeruginosa account for 88% (90
to 92%) of these deaths. Most of the infections
considered in this study are healthcare associated infections,
and 20 to 30% of them can be considered avoidable .
These results underestimate the morbidity and
mortality of infections due to antibiotic-resistant bacteria in
France. Although our study covers a large panel of
MDRB and the most frequent body sites, it does not
include all bacteria resistant to antibiotics and all infection
sites. For example, as only the most frequent infections
in the 2012 PPS data were selected for each MDRB,
gastro-intestinal tract infections due to Enterobacteriacae
were not taken into account in our study.
We did not take into account patients with more than
one MDRB. However, this bias in underestimating
MDRB infections should be very limited. Indeed,
according to data from the 2012 French PPS, only 39 patients
among a total of 15,180 infected patients (0.2%) had two
or more different MDRB (data not shown).
Furthermore, the extrapolation of EARS-Net data to
France was calculated on the number of inpatient
hospital days reported only by French secondary and
tertiary care hospitals, as the laboratories participating in
the EARS-Net network are located only in such institutions.
This may have underestimated the incidence of MDRB
infections, as cases occurring in other types of hospitals
have being ignored. Conversely, taking into account all
healthcare institutions in France would have greatly
overestimated the incidence since it is based on the assumption
that the burden of the antibiotic resistance is comparable in
The estimates produced in this study are consistent
with those from other French data sources, especially
Table 3 Annual number and incidence rate of infections due to MDRB, France 2012
MDRB Body site Total number of cases (%)
Incidence per 1,000 HD
PPS ratio Ratios from literature
Low value High value
0.065 0.065 0.065
0.538 0.498 0.741
0.603 0.563 0.806
<0.001 <0.001 <0.001
0.004 0.004 0.005
0,004 0,004 0,005
0.066 0.066 0.066
0.526 0.314 0.854
0.592 0.381 0.920
0.026 0.026 0.026
0.164 0.071 0.253
0,190 0.097 0.280
0.001 0.001 0.001
0.007 0.004 0.006
0.008 0.005 0.007
0.025 0.025 0.025
0.403 0.403 0.796
0.428 0.428 0.821
0.001 0.001 0.001
0.008 0.003 0.008
0.009 0.004 0.009
0.185 0.185 0.185
1.834 1.481 2.847
Table 4 Number of deaths associated with infections due to MDRB, France 2012
Subtotal Gram +
3GC-R K. pneumoniae
CR K. pneumoniae
CR P. aeruginosa
CR Acinetobacter spp
from the BMR-RAISIN network , where between
4,000 and 5,000 cases of MRSA bloodstream infections
and 4,000 to 9,000 cases of bloodstream infections due
to extended-spectrum beta-lactamase
(ESBL)-producing enterobacteriaceae — mostly with E. coli and K.
pneumoniae — were estimated to occur in France in 2013.
The results of our study should also be examined in
light of previous studies published in other countries.
Nonetheless they are not directly comparable because
the methodology used differs in many respects.
The European study  estimated there were more
than 386,000 annual infections due to MDRB in Europe,
and attributed 25,000 extra deaths to them. The
EARSNet data used for that European estimate date back to
2007, while the French study used more recent data
(2012). Trends in the epidemiology of MDRB between
2007 and 2012, especially major changes for MRSA and
3GC-R enterobacteriaceae , may thus explain some of
these differences. Indeed, Ears-Net data show the spread
of 3GC-R resistance in E. coli, with a four-fold increase
from 2.5% in 2007 to 10% in 2012.
In addition, the European study sought to examine the
morbidity and mortality of MDRB associated with
hospital-acquired infections only. This study considered
only a fraction of all bloodstream infections, by
correcting their incidence reported by EARS-Net by a factor
derived from national prevalence studies to take into
account only those of nosocomial origin, e.g., only 65.6%
of MRSA bloodstream infections and 58.9% of 3GC-R E.
coli bloodstream infections (ECDC, unpublished data).
Moreover, the panels of infections and MDRB species
considered in the present study were more complete
than those of the European study, which did not include
three MDRB from our panel (CR K. pneumoniae,
Acinetobacter spp. and GR Enterococcus faecalis) and
one infection site (bone and joint infections for MRSA).
Another major difference between the two studies is
related to the correction of the EARS-Net data to
estimate the number of cases for the entire country. The
European study estimated the incidence of invasive
infections due to resistant bacteria in France from the
median incidence of invasive infections observed in all
EU/EEA countries. The percentage of resistance in
species from the EARS-Net data was then applied (ECDC,
unpublished data). Our estimates are therefore more
precise because they are based on actual national
surveillance and hospital data.
Finally, the estimates of the number of deaths from
the European study cannot be compared to our results
because the associated mortality proportion used in our
study differs for some MDRB.
The 2013 CDC report  used surveillance data
collected between 2009 and 2011 to estimate that there
were more than 2 million infections due to
antibioticresistant microorganisms that year in the United States.
The same limitations raised for the European study
apply when comparing the CDC results with ours,
because the US considered more bacteria-antibiotic
combinations than our study did, e.g. infections with
Salmonella, Shigella, tuberculosis, and even Candida.
The CDC also used a different methodology to
calculate the number of deaths, which limits the
comparisons still further. It applied the same proportion of
deaths — 6.5% — to infections with
carbapenemaseproducing enterobacteriaceae, MDR Acinetobacter,
ESBL-producing enterobacteriaceae, GRE, and MDR
P. aeruginosa. This figure, derived from a study
published in 2009 , is an estimate of the overall
mortality associated with nosocomial infections due to
MDRB. Instead, we used MDR-specific and
sitespecific mortality ratios for each type of infection,
which appears more appropriate, in view of the great
variability of associated mortality rates between
different infections due to MDRB and body sites.
The specifications of the European study, the CDC
American study and our Santé publique France study have
been summarised in three additional tables (Additional file
1: Tables a, b, c). The differences in these studies underline
the need to define a standardised protocol for
international assessments of the morbidity and mortality of
antibiotic resistance. Only results from studies based on
such a protocol, which could be promoted for example by
the ECDC or WHO as part of the global action plan on
antimicrobial resistance recently adopted by the World
Health Assembly , will allow valid comparisons
We decided here to assess the overall mortality
associated with infections due to MDRB and not specifically
that associated with antibiotic resistance. Several
publications have already reported the excess mortality associated
with resistance alone. A study of two matched cohorts of
patients  showed that the 30-day mortality associated
with 3GC-R E. coli bloodstream infections was 2.5 times
greater (confidence interval: 0.9-6.8) than that associated
with susceptible strains. A meta-analysis of 16 studies 
also showed a significant increased risk in the mortality
associated with bloodstream infections due to
ESBLproducing Enterobacteriaceae, almost twice as high as that
observed for susceptible Enterobacteriaceae (pooled RR:
1.85; confidence interval: 1.39-2.47).
Several studies suggest that infections due to MDRB do
not replace those with susceptible bacteria, but add to
them (the so-called “Boyce effect”), thus increasing the
total morbidity and mortality of infection from a given
bacterium. This characteristic was initially described for
MRSA , and more recently for 3GC-R E. coli , two
bacteria that account for two thirds of the cases of
infections due to MDRB and more than one third of the deaths
estimated in our study. A substantial fraction — and if the
Boyce effect also applies to other MDRB, the majority —
of the deaths we estimate should therefore be considered
as excess mortality compared with those associated with
infections by susceptible bacteria.
Our estimates are accompanied by uncertainty. To
overcome this limit, for morbidity estimates, each
PPSbased ratio was assigned a lower and an upper
plausibility value, provided by the literature. The number of
deaths was calculated by multiplying this ratio with the
value of the morbidity estimates and their plausibility
intervals, allowing presenting a plausibility interval for
mortality estimates too. The numbers of deaths and
plausibility intervals for deaths were estimated with
these morbidity estimates and their plausibility intervals
multiplied with mortality ratio provided by the literature.
Part of uncertainty on the mortality estimates is taken
into account through these plausibility intervals.
However, uncertainty around proportions of deaths remain
because of lack of sufficient-sized studies leaded in
Europe on MDRB infections mortality.
The results of our study confirm that multidrug-resistance
to antibiotics is a major public health problem in France.
The morbidity and mortality associated with infections
due to MDRB is particularly important for MRSA and
3GC-R Enterobacteriaceae. Surveillance of resistance to
these bacteria therefore remains an important priority.
The number of infections due to emerging highly resistant
bacteria is still limited in France, probably due to the
measures implemented to prevent cross-transmission.
The number of cases of infections and the number of
deaths associated with infections due to MDRB
calculated in this study are indicators that should facilitate
communication and awareness of both healthcare
professionals and the general public. Repeat this study in
a few years with data from the next French PPS study,
which will occur in 2017, and EARS-Net data, would
provide trends estimations.
Additional studies remain to be conducted, for
example to assess the projected morbidity and mortality of
antibiotic resistance in France according to several
scenarios based on incidence rates and efficacy of
control programmes. It is equally important to assess
the economic costs (of medical care, of measures to
control dissemination in hospitals, and to the community)
of infections due to MDRB in order to reinforce the
adherence of healthcare professionals and policy-makers
to prevention programmes.
Additional file 1: Tables a, b, c. Table a, table b, table c have been
added as supplementary material. Table a presents the methodology,
parameters and main results of 3 studies estimating the morbidity and
mortality of multidrug-resistant bacteria infections : the European study,
the US study and our French study. Table b compares the ratios used for
estimating the incidence of non-invasive infections in the European study and
Santé publique France study. Table c compares the ratios used for estimating
mortality associated with infections due to MDRB in the European study and
Santé publique France study [31–34]. (PDF 285 kb)
3GC-R E. coli: Escherichia coli resistant to third-generation cephalosporins;
3GC-R K. pneumoniae: Klebsiella pneumoniae resistant to third-generation
cephalosporins; 3GC-R: Third-generation cephalosporin-resistant; CDC: US
Centers for disease control and prevention; CR Acinetobacter
spp: Acinetobacter spp. resistant to carbapenems; CR K. pneumoniae: K.
pneumoniae resistant to carbapenems; CR P. aeruginosa: Pseudomonas
aeruginosa resistant to carbapenems; EARS-Net: European antimicrobial
resistance surveillance network; ECDC: European centre for disease
prevention and control; ESBL: Extended-spectrum beta-lactamase producing
enterobacteriaceae; EU/EEA countries: European union and European economic
area countries; GRE: Enterococcus faecium and E. faecalis resistant to glycopeptides;
HD: Inpatient hospital days; MDR: Multidrug resistant; MDRB: Multidrug resistant
bacteria; MRSA: Methicillin-resistant Staphylococcus aureus; PPS 2012: Point
prevalence survey of healthcare-associated infections and antimicrobial use in
French hospitals in 2012
We thanks Jean Carlet, Lidia Kardas, Lucie Léon, Jean-Christophe Lucet, Camille
Pelat, Cécile Sommen, Dieter Van Cauteren, Yazdan Yazdanpanah for validation
of methodology and parameters used for morbidity and mortality estimations.
Availability of data and materials
The database of this project is held by Sante publique France.
All authors contributed to the interpretation of the results, the revision of the
draft manuscript and approved the final version. MCC conducted the data
analysis and wrote the manuscript; JL was involved in the data analysis; CBB
and VJ were involved in the validation of parameters used for morbidity and
mortality estimations; SV and BC conceived and designed the study and
were involved in the validation of parameters used for morbidity and
The authors declare that they have no competing interests.
Consent for publication
Ethics approval and consent to participate
This study does not constitute human research requiring Institutional Review
1. World Health Organization (WHO). Global action plan on antimicrobial resistance . WHO . 2015 . http://apps.who.int/gb/ebwha/pdf_files/ WHA68/ A68_ACONF1Rev1-en .pdf. Accessed 27 May 2016 .
2. European Centre for Disease prevention and Control (ECDC), European Medicine Agency (EMEA). The bacterial challenge: time to react . ECDC. 2009 . http://www.ecdc.europa.eu/en/publications/Publications/0909_TER_ The_Bacterial_Challenge_Time_to_ React.pdf. Accessed 27 May 2016 .
3. Centers for disease control and prevention (CDC). Antibiotic resistance threats in the United State . 2013 . http://www.cdc.gov/drugresistance/threatreport- 2013 /. Accessed 27 May 2016 .
4. Arnaud I , Jarlier V. The BMR-Raisin working group . Surveillance des bactéries multirésistantes dans les établissements de santé en France , Réseau BMRRaisin, résultats 2013 . Institut de veille sanitaire . 2015 . http://www.invs.sante. fr/Publications-et-outils/Rapports-et-syntheses/Maladies-infectieuses/2015/ Surveillance-des-bacteries-multiresistantes-dans-les-etablissements-de-santeen-France . Accessed 27 May 2016 .
5. European Centre for Disease prevention and Control (ECDC). Antimicrobial resistance surveillance in Europe, annual report of the European Antimicrobial Resistance Surveillance Network (Ears-Net) . ECDC . 2014 . http://ecdc.europa.eu/en/publications/_layouts/forms/ Publication_DispForm.aspx? List=4f55ad51-4aed-4d32-b960- af70113dbb90&ID=1205. Accessed 27 May 2016 .
6. Magiorakos AP , Srinivasan A , Carey RB , Carmeli Y , Falagas ME , Giske CG , et al. Multidrug-resistant, extensively drug-resistant and pandrugresistant bacteria: an international expert proposal for interim standard definitions for acquired resistance . Clin Infect Dis . 2012 ; doi: 10.1111/j. 1469- 0691 . 2011 .03570.x
7. Thiolet JM , Vaux S , Lamy M , Gauthier A , Barret AS , Leon L , et al. Enquête nationale de prévalence des infections nososcomiales et des traitements anti-infectieux en établissements de santé, France, mai-juin 2012. Institut de veille sanitaire . 2013 . http://www.invs.sante. fr/Publications-et-outils/Rapportset-syntheses/Maladies-infectieuses/2013/Enquete-nationale-de-prevalencedes-infections-nosocomiales-et-des-traitements-anti-infectieux-enetablissements-de-sante-France-mai-juin-2012 . Accessed 13 June 2016 .
8. Institut de Veille Sanitaire (InVS). Infections invasives d'origine bactérienne , Réseau EPIBAC. Institut de veille sanitaire . 2015 . http://www.invs.sante. fr/ Dossiers-thematiques/Maladies-infectieuses/Maladies-a-preventionvaccinale/Infections-invasives-d-origine-bacterienne-Reseau-EPIBAC/ Methodes-de-la-surveillance . Accessed 27 May 2016 .
9. Données administratives et statistiques sur les établissements de santé . Ministère des Affaires sociales et de la santé, Direction de la recherche , des études, de l' évaluation et des statistiques , Paris. 2015 . http://drees.socialsante.gouv. fr/etudes-et-statistiques/opendata/etablissements-de-santesociaux-et-medico-sociaux/article/la-statistique-annuelle-des-etablissementssae .
10. Martone WJ , Jarvis WR , Edwards JR , Culver DH , Haley RW . Incidence and nature of endemic and epidemic nosocomial infections . In: Benett J , Brachman P, editors. Hospitals infections. Philadelphia: Lippincott-Raven ; 1998 . p. 461 - 76 .
11. Gavalda L , Masuet C , Beltran J , Garcia M , Garcia D , Sirvent JM , et al. Comparative cost of selective screening to prevent transmission of methicillin-resistant Staphylococcus aureus (MRSA), compared with the attributable costs of MRSA infection . Infect Control Hosp Epidemiol . 2006 ; doi: 10.1086/507968.
12. Kallen AJ , Mu Y , Bulens S , Reingold A , Petit S , Gershman K , et al. Health care-associated invasive MRSA infections , 2005 - 2008 . JAMA. 2010 ; doi: 10.1001/jama.2010.1115.
13. Kanerva M , Ollgren J , Hakanen AJ , Lyytikainen O. Estimating the burden of healthcare-associated infections caused by selected multidrugresistant bacteria Finland , 2010 . Antimicrob Resist Infect Control . 2012 ; doi: 10.1186/ 2047 - 2994 - 1 - 33 .
14. Carmeli Y , Eliopoulos G , Mozaffari E , Samore M. Health and economic outcomes of vancomycin-resistant enterococci . Arch Intern Med . 2002 ; 162 : 2223 - 8 .
15. The Brooklyn Antibiotic Resistance Task Force . The cost of antibiotic resistance: effect of resistance among Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, and Pseudmonas aeruginosa on length of hospital stay . Infect Control Hosp Epidemiol . 2002 ; doi: 10.1086/502018.
16. Jain R , Walk ST , Aronoff DM , Young VB , Newton DW , Chenoweth CE , et al. Emergence of Carbapenemase producing Klebsiella Pneumoniae of Sequence type 258 in Michigan , USA. Infect Dis Rep . 2013 ; doi: 10.4081/idr. 2013 .e5.
17. Perez F , Endimiani A , Ray AJ , Decker BK , Wallace CJ , Hujer KM , et al. Carbapenem-resistant Acinetobacter baumannii and Klebsiella pneumoniae across a hospital system: impact of post-acute care facilities on dissemination . J Antimicrob Chemother . 2010 ; doi: 10.1093/jac/dkq191.
18. Livermore DM , Hill RL , Thomson H , Charlett A , Turton JF , Pike R , et al. Antimicrobial treatment and clinical outcome for infections with carbapenemand multiply-resistant Acinetobacter baumannii around London . Int J Antimicrob Agents . 2010 ; doi: 10.1016/j.ijantimicag. 2009 .09.014.
19. Cosgrove SE , Sakoulas G , Perencevich EN , Schwaber MJ , Karchmer AW , Carmeli Y. Comparison of mortality associated with methicillin-resistant and methicillin-susceptible Staphylococcus aureus bacteremia: a meta-analysis . Clin Infect Dis . 2003 ; doi: 10.1086/345476.
20. Marchaim D , Gottesman T , Schwartz O , Korem M , Maor Y , Rahav G , et al. National multicenter study of predictors and outcomes of bacteremia upon hospital admission caused by Enterobacteriaceae producing extendedspectrum beta-lactamases . Antimicrob Agents Chemother . 2010 ; doi: 10.1128/AAC.00565- 10 .
21. Ena J , Arjona F , Martinez-Peinado C , Lopez-Perezagua MM , Amador C. Epidemiology of urinary tract infections caused by extended-spectrum betalactamase-producing Escherichia coli . Urology . 2006 ; doi: 10.1016/j.urology. 2006 .08.1075.
22. Mouloudi E , Protonotariou E , Zagorianou A , Iosifidis E , Karapanagiotou A , Giasnetsova T , et al. Bloodstream infections caused by metallo-betalactamase/Klebsiella pneumoniae carbapenemase-producing K. pneumoniae among intensive care unit patients in Greece: risk factors for infection and impact of type of resistance on outcomes . Infect Control Hosp Epidemiol . 2010 ; doi: 10.1086/657135.
23. Suarez C , Pena C , Gavalda L , Tubau F , Manzur A , Dominguez MA , et al. Influence of carbapenem resistance on mortality and the dynamics of mortality in Pseudomonas aeruginosa bloodstream infection . Int J Infect Dis . 2010 ; doi: 10.1038/srep11715.
24. Grupper M , Sprecher H , Mashiach T , Finkelstein R. Attributable mortality of nosocomial Acinetobacter bacteremia . Infect Control Hosp Epidemiol . 2007 ; doi: 10.1086/512629.
25. Bonnal C , Mourvillier B , Bronchard R , de Paula D , Armand-Lefevre L , L'heriteau F , et al. Prospective assessment of hospital-acquired bloosdstream infections: how many may be preventable? Qual Saf Health Care . 2010 ; doi: 10.1136/qshc.2008.030296.
26. Roberts RR , Hota B , Ahmad I , Scott RD , Foster SD , Abbasi F , et al. Hospital and societal costs of antimicrobial-resistant infections in a Chicago teaching hospital: implications for antibiotic stewardship . Clin Infect Dis . 2009 ; doi: 10. 1086/605630.
27. de Kraker ME , Wolkewitz M , Davey PG , Koller W , Berger J , Nagler J , et al. Burden of antimicrobial resistance in European hospitals: excess mortality and length of hospital stay associated with bloodstream infections due to Escherichia coli resistant to third-generation cephalosporins . J Antimicrob Chemother . 2011 ; doi: 10.1093/jac/dkq412.
28. Schwaber MJ , Carmeli Y. Mortality and delay in effective therapy associated with extended-spectrum beta-lactamase production in Enterobacteriaceae bacteraemia: a systematic review and meta-analysis . J Antimicrob Chemother . 2007 ; doi: 10.1093/jac/dkm318.
29. Boyce JM , White RL , Spruill EY . Impact of methicillin-resistant Staphylococcus aureus on the incidence of nosocomial staphylococcal infections . J Infect Dis . 1983 ; 148 : 763 .
30. de Kraker ME , Jarlier V , Monen JC , Heuer OE , van de Sande N , Grundmann H. The changing epidemiology of bacteraemias in Europe: trends from the European Antimicrobial Resistance Surveillance System . Clin Microbiol Infect . 2013 ; doi: 10.1111/ 1469 - 0691 .12028.
31. Huang SS , Johnson KM , Ray GT , Wroe P , Lieu TA , Moore MR , et al. Healthcare utilization and cost of pneumococcal disease in the United States . Vaccine. 2011 ; 29 ( 18 ): 3398 - 412 . doi:10.1016/j.vaccine. 2011 .02.088.
32. Hoban DJ , Doern GV , Fluit AC , Roussel-Delvallez M , Jones RN . Worldwide prevalence of antimicrobial resistance in Streptococcus pneumoniae, Haemophilus influenzae, and Moraxella catarrhalis in the SENTRY antimicrobial surveillance program , 1997 - 1999 . Clin Infect Dis . 2001 ;32 Suppl 2: S81 - 93 . doi:10.1086/320181.
33. Schwaber MJ , Navon-Venezia S , Kaye KS , Ben-Ami R , Schwartz D , Carmeli Y. Clinical and economic impact of bacteremia with extended- spectrum-betalactamase-producing Enterobacteriaceae . Antimicrob Agents Chemother . 2006 ; 50 ( 4 ): 1257 - 62 . doi:10.1128/AAC.50.4. 1257 - 1262 . 2006 .
34. Carmeli Y , Troillet N , Karchmer AW , Samore MH . Health and economic outcomes of antibiotic resistance in Pseudomonas aeruginosa . Arch Intern Med . 1999 ; 159 ( 10 ): 1127 - 32 . doi:10.1001/archinte.159.10.1127.