The MRSA-import in ICUs is an important predictor for the occurrence of nosocomial MRSA cases.

Abstract

Nosocomial infections with methicillin-resistant Staphylococcus aureus (MRSA) account for increased morbidity, mortality and healthcare costs in critically ill patients worldwide. The intensive care unit (ICU) component of the German surveillance system for nosocomial infections (Krankenhaus-Infektions-Surveillance-System, KISS) has been supplemented with a module targeting the surveillance of multiresistant pathogens [Multiresistente Erreger (MRE)-KISS] in order to account for the increasing burden of antibiotic-resistant bacteria. The aim of this study was to assess the association between structural and organizational characteristics of ICUs and the number of nosocomial MRSA cases. Data were derived from routine data collected in the frame of the national surveillance system of nosocomial infections (ICU- and MRE-KISS) from January 2007 to December 2008 and from a questionnaire inquiring about structure and process parameters. One hundred and forty ICUs performing active screening have been included. Process parameters such as isolation of MRSA patients, decolonization procedures and introduction of MRSA alert systems have been implemented by the majority of the ICUs, whereas the application mode of screening procedures and pre-emptive isolation measures is heterogeneous. Multivariable analysis using negative binominal regression models shows that a stay on a medical ICU has a protective effect (incidence rate ratio, 0.42; 95% confidence interval, 0.24-0.74; p = 0.003), whereas the imported MRSA incidence is significantly associated with the number of nosocomial MRSA cases (incidence rate ratio, 1.74; 95% confidence interval, 1.23-2.45; p = 0.002). Structure and process parameters do not show any effect. ICU type and imported MRSA incidence should be considered for benchmarking between hospitals.

Full text

Page 1

Originally published as:

Schweickert, B., Geffers, C., Farragher, T., Gastmeier, P., Behnke, M., Eckmanns, T., Schwab, F.
The MRSA-import in ICUs is an important predictor for the occurrence of nosocomial MRSA
cases
(2011) Clinical Microbiology and Infection, 17 (6), pp. 901-906.

DOI: 10.1111/j.1469-0691.2010.03409.x


This is an author manuscript.
The definitive version is available at: http://onlinelibrary.wiley.com/

Page 2

The MRSA-import in ICUs is an important predictor
for the occurrence of nosocomial MRSA cases
B. Schweickert1, C. Geffers2, T. Farragher3, P. Gastmeier4, M. Behnke2, T. Eckmanns1, F. Schwab2
1 Robert Koch Institute, Berlin
2 Institute of Hygiene and Environmental Medicine, Charité University Medicine Berlin, Berlin, Germany
3 Centre for Medical Statistics and Health Evaluation, University of Liverpool, Liverpool, UK
4 National Reference Center for Nosocomial Infections, Institute of Hygiene and Environmental
Medicine, Charité University Medicine Berlin, Berlin, Germany
*Correspondence: Corresponding author: B. Schweickert, Robert Koch Institute, Abt. 3, FG
Surveillance, DGZ-Ring 1, D-13086 Berlin, Germany E-mail: schweickertb@rki.de

Abstract
Nosocomial infections with methicillin-resistant Staphylococcus aureus (MRSA) account for increased
morbidity, mortality and healthcare costs in critically ill patients worldwide. The intensive care unit
(ICU) component of the German surveillance system for nosocomial infections (Krankenhaus-
Infektions-Surveillance-System, KISS) has been supplemented with a module targeting the
surveillance of multiresistant pathogens [Multiresistente Erreger (MRE)-KISS] in order to account for
the increasing burden of antibiotic-resistant bacteria. The aim of this study was to assess the
association between structural and organizational characteristics of ICUs and the number of
nosocomial MRSA cases. Data were derived from routine data collected in the frame of the national
surveillance system of nosocomial infections (ICU- and MRE-KISS) from January 2007 to December
2008 and from a questionnaire inquiring about structure and process parameters. One hundred and
forty ICUs performing active screening have been included. Process parameters such as isolation of
MRSA patients, decolonization procedures and introduction of MRSA alert systems have been
implemented by the majority of the ICUs, whereas the application mode of screening procedures and
pre-emptive isolation measures is heterogeneous. Multivariable analysis using negative binominal
regression models shows that a stay on a medical ICU has a protective effect (incidence rate ratio,
0.42; 95% confidence interval, 0.24–0.74; p = 0.003), whereas the imported MRSA incidence is
significantly associated with the number of nosocomial MRSA cases (incidence rate ratio, 1.74; 95%
confidence interval, 1.23–2.45; p = 0.002). Structure and process parameters do not show any effect.
ICU type and imported MRSA incidence should be considered for benchmarking between hospitals.

Introduction
Nosocomial infections with multiresistant bacteria, particularly methicillin-resistant Staphylococcus
aureus (MRSA), play an increasing role in the hospital care of critically ill patients [1]. In Germany, the
percentage of MRSA in relation to Staphylococcus aureus isolates derived from blood cultures
increased from 9% in 1999 to 20% in 2002 [2]. However, in recent years the proportion of MRSA-
positive isolates appears to have stagnated in German hospitals, probably due to enhanced infection
control measures [3,4].
Cross-transmission by direct contact, such as hands of healthcare workers and visitors and a
contaminated environment (e.g. equipment), is an important and preventable cause of the spread of
MRSA in hospitals. In order to prevent and control nosocomial MRSA infections, a comprehensive
strategy comprising surveillance of nosocomial infections and the spread of MRSA, personal and
institutional hygiene measures, surveillance of antibiotic resistance and usage and measures to
ensure prudent antibiotic use (e.g. antibiotic stewardship programmes) is necessary [5].
In Germany, a national surveillance system for nosocomial infections has been introduced
(Krankenhaus-Infektions-Surveillance-System, KISS) in 1997 [6]. Nosocomial infections in ICUs were
recorded in the ICU component of KISS (ICU-KISS). In 2003, ICU-KISS has been supplemented by
the multiresistant microorganisms module [Multiresistente Erreger (MRE)-KISS] in order to account for

Page 3

the growing impact of multiresistant bacteria [7]. Besides infection control and prevention practices,
structural parameters of ICUs and hospitals, such as size and type of hospitals and ICUs, the
equipment in single bedrooms and ventilator places, may have an influence on the frequency of
nosocomial MRSA cases. So far, a systematic investigation of the effect of structural factors on
hospital MRSA acquisition has rarely been performed. The aim of this study is to obtain data on
structural and organizational characteristics of ICUs participating in ICU- and MRE-KISS and to
analyse the association of these factors with the number of nosocomial MRSA cases. Data are derived
from routine surveillance reports combined with a structured questionnaire from 140 German ICUs
participating in ICU- and MRE-KISS in 2007/2008.

Methods
Monthly recorded numbers of patients, patient days, device days (ventilator, central venous catheter
and urinary tract catheter) and patient-based MRSA cases were derived from routine data collected in
the frame of the national surveillance system of nosocomial infections (ICU- and MRE-KISS) from
January 2007 to December 2008.
In March 2008, a web-based questionnaire was sent out to all ICUs participating in ICU- and MRE-
KISS to collect information on structural characteristics and measures for prevention and infection
control. From 275 ICUs, 185 (67.3%) questionnaires were completed.
Only those 140 ICUs that perform active screening for MRSA have been included in the analysis, in
order to allow for the estimation of nosocomial MRSA acquisition.
The recorded data comprise the following.
 Structural type and size of hospitals and ICUs, geographical region, parameters: and number
of ventilator places, single bedrooms and ICUs per hospital.
Process length of stay, device utilization and MRSA prevention parameters: rate, employed
staff, proportion of short stayers (≤48 h), implementation of an MRSA alert system, screening
for MRSA on admission, type of screening performed (patients at high risk of MRSA, or all
patients), time point of screening, pre-emptive isolation, isolation of confirmed MRSA patients,
and decolonization measures.

The following definitions have been used.
 MRSA every patient in whom MRSA has been detected at any body site case: during his/her
stay on the ICU, irrespective of whether the patient presents with symptoms of infection or is
merely colonized.
Nosocomial detection of MRSA more than 48 h after admission to the ICU, MRSA case:
either as a result of the performance of screening cultures or the investigation of clinical
samples.
detection of MRSA within 48 h after admission to the ICU. Imported MRSA case:
number of MRSA cases/1000 patients. MRSA incidence density:
number of nosocomial MRSA cases/1000 patient-days. Nosocomial MRSA incidence density:
number of imported MRSA cases/100 patients. Imported MRSA incidence:
Device number of device days/100 patient-days. These rates utilization rate: are calculated
separately for invasive ventilation, central venous catheters (CVC) and urinary tract catheters
(UTC).
number of actually employed staff/patient/day/shift. Staff-patient ratio:








Data analysis
The results of descriptive analysis are presented as summary measures appropriate to the data type
and distribution. For categorial variables data are expressed in numbers and percentages, and for
continuous variables in means and standard deviation (SD) if the variables showed a normal
distribution or in medians with interquartile ranges (IQR) for non-normally distributed variables.

Page 4

Univariate and multivariable analysis has been performed to identify independent risk factors for
nosocomial MRSA cases. A negative binominal regression model has been used to estimate the
association of the number of nosocomial MRSA cases with the parameters. The log number of patient
days was treated as an offset in the model. As the magnitude of the variance of the outcome
parameter indicates overdispersion and the dispersion parameter alpha of the likelihood ratio test,
which differs significantly from zero, also confirms the presence of overdispersion, negative binominal
regression has been used instead of poisson regression. Comparison with a zero inflated model using
the Vuong test did not show any advantage over negative binominal regression.
For multivariable analysis a stepwise forward approach has been applied, which considered all
variables with a p-value of <0.25 in univariable analysis. The p-value for retention of a covariate in the
model was 0.05 to identify independent risk factors.
Comparisons between ICU types and between different hospital sizes have been performed by using
the Kruskal–Wallis test and the Wilcoxon rank sum test, respectively.
Calculations have been performed using Stata version 9.

Results
In 140 ICUS, coming from all major geographical regions in Germany, a total of 228 703 patients,
accounting for 852 835 patient days, have been treated in 2007/2008. From 4279 MRSA cases 3496
(81.7%) were contracted from outside the hospital and 783 (18.3%) were classified as nosocomial in
the ICU.

MRSA situation
The numbers and rates of MRSA in the ICUs are shown in Table 1. Stratification by ICU type shows
that medical ICUs exhibit significant (p = 0.02) lower numbers of nosocomial MRSA cases/month
(median, 0.07; IQR, 0–0.33) as compared with interdisciplinary (median, 0.17; IQR, 0.04–0.42) and
surgical ICUs (median,: 0.29; IQR, 0.10–0.54) and show a lower incidence density of nosocomial
MRSA cases (p = 0.006) as well, presenting with a median incidence of nosocomial MRSA/1000
patient-days of 0.24 (IQR, 0–0.82) in contrast to interdisciplinary and surgical ICUs, which show a
median value of 0.64 (IQR, 0.26–1.50) and 0.93 (IQR, 0.41–1.45), respectively. Comparison of the
MRSA parameters stratified by hospital size (hospitals ≤1000 beds and hospitals >1000 beds) reveals
a higher MRSA incidence density and a higher incidence of imported MRSA in large hospitals >1000
beds (Wilcoxon rank sum test, p <0.001 and p <0.001, respectively), whereas the nosocomial MRSA
incidence densities do not show any differences (Wilcoxon rank sum test, p = 0.34).

Structural and organizational characteristics
Structural and organizational characteristics of the ICUs are shown in Table 2. The largest proportion
of the participating hospitals was classified as academic teaching hospitals (50%). The hospitals were
broadly evenly distributed over the main geographical regions in Germany. Ninety-five (67.9%)
hospitals have 1000 beds or less and 65 (46.4%) participating ICUs are classified as interdisciplinary.
While 51 (36.4%) ICUs screen all patients, 89 (63.6%) ICUs restrict screening to patients at particular
risk, including previously known MRSA patients and contact patients. The majority (86.4%) of the ICUs
perform screening at the patient’s admission, while only a smaller proportion (27.9%) applied
screening at regular intervals during their stay. Twenty (14.3%) ICUs combine both strategies. From
the 51 ICUs screening all patients, only 14 (27.4%) perform prophylactic isolation measures. In
contrast, most ICUs arrange single room isolation of MRSA patients.

Page 5

Association of nosocomial MRSA cases with structure, process and MRSA-prevention
measures of the ICUs
The results of univariate analysis to assess the association between nosocomial MRSA cases and
structure, process and MRSA-prevention measures of the ICUs are shown in Table 2. Medical ICUs
showed a significantly lower number of nosocomial MRSA cases than interdisciplinary and surgical
ICUs. The incidence rate ratio (IRR) for nosocomial MRSA cases was 0.48 (95% confidence interval
(CI), 0.31–0.74) compared with interdisciplinary ICUs. Teaching hospitals and ‘other hospital types’
experience a lower number of nosocomial MRSA cases than university hospitals, but not statistically
significantly lower. Similiarly, hospital size does not appear to be related to the number of nosocomial
MRSA cases, while ICU sizes of ≥12 beds are associated with more nosocomial MRSA cases
(IRR = 1.49; 95% CI, 1.04–2.13). Increased application of devices such as UTCs was associated with
a higher rate of nosocomial MRSA cases. The higher incidence of imported MRSA (>0.94) is
associated with a 1.8-fold increase in the rate of nosocomial MRSA cases, compared with the lower
incidence of imported MRSA.
Multivariable analysis (Table 3) confirms that type of ICU and incidence of imported MRSA cases were
independent risk factors for nosocomial MRSA acquisition. The adjusted IRR for medical ICUs,
compared with interdisciplinary ICUs, was 0.42 (95% CI, 0.24–0.74) and for ICUs with an incidence of
imported MRSA cases greater than the median was 1.74 (95% CI, 1.23–2.46) as compared with those
with imported MRSA incidences lower than the median.

Discussion
Multivariable analysis shows that the incidence of imported MRSA is significantly associated with the
number of nosocomial MRSA cases in ICUs and that stay on medical ICUs has a protective effect.
The latter result does not contradict former studies, which found surgical ICUs to be at particular risk of
presenting with high MRSA infection rates [8,9]. The lower number of nosocomial MRSA cases in
medical ICUs is probably due to differences in patient characteristics between conservative and
surgical medicine with respect to gender, age, underlying diseases, severity of disease, frequency of
invasive procedures and antibiotic therapy. Patient-level risk factors have been extensively
investigated in former studies [10–12].
The incidence of imported MRSA cases showed a highly significant association with nosocomial
MRSA cases, which is plausible because with the increase of the MRSA-positive reservoir the
probability of transmission to other patients rises. This result is confirmed by several other studies,
which showed that colonization pressure is an independent risk factor for hospital MRSA acquisition
[13–15]. In order to account for this close relationship some authors proposed using adjusted MRSA
transmission rates for intra- and interhospital comparisons [16,17]. This has to be considered with
caution, as the measures characterizing colonization pressure might be strongly influenced by the
screening policy of the hospitals [18].
In multivariable analysis none of the structural parameters showed an association with the number of
nosocomial MRSA cases. Previous data on the relationship of structural properties of hospitals and
ICUs with MRSA acquisition rates are scarce. Grammatico-Guillon et al. [8] investigated the
relationship of MRSA prevalence and infection control indicators in French hospitals in 2005/2006.
They found that private for-profit hospitals present with the lowest MRSA prevalence, but this could be
explained by the specific patient population. Hospital ownership has not been considered in the
present study. Mears et al. [19] found a relationship between availability of single bedrooms and
nosocomial MRSA rates, which could not be confirmed in the present study. Another study shows
merely descriptive data [20]. The overall negative results imply that probably other factors, which have
not been considered in the present study, such as patient-level characteristics, may play a more
important role in determining the amount of nosocomial MRSA cases. Another reason might be that
hospitals participating in MRE-KISS are too homogenous to detect any differences.
Process-indicators of infection control and prevention measures also did not yield a significant effect in
multivariable analysis, which is mainly due to the design of the study, which enables an analysis of the
actual situation but does not allow estimating the temporal relationship of dependant and independent

Page 6

variables. Nevertheless, the data give insight into the current situation and provide a basis for the
evaluation of future developments.
Descriptive data show that a wide range of hospital and ICU types and sizes from all major regions of
Germany are represented. Most of the infection control and prevention measures are implemented in
the majority of the ICUs. Screening and pre-emptive isolation practices, which have been investigated
in more detail, provide a more heterogeneous picture. Regarding general organizational factors, such
as staff-patient ratio, portion of short-stayers and MRSA-specific infection control measures, no
significant differences between ICU types have been seen.
Despite the overall impression of a similar approach in combating the hospital spread of MRSA,
underlying heterogeneities can not be excluded. As standardized definitions are missing, the question
regarding screening of patients at risk does not specify which patients should be considered to be at
risk. As ICUs use individual risk definitions, a positive answer comprises different screening strategies.
Additionally, there is no information on other determinants of screening policy, such as the site of
screening, which influences the yield of MRSA cases [21,22]. Similarly, staff-patient ratio is an
important determinant of hospital staff policy, but reflects only one component of a multifaceted entity.
Several other factors, such as bed occupancy and workload, should be considered in order to provide
a more comprehensive picture [23].
Another limitation of the study regards the different screening policies, which might result in
misclassification. As only 27.9% of the ICUs screen for MRSA at regular intervals (e.g. two fixed days
a week), an underestimation of the nosocomial transmission rate has to be assumed. On the other
hand, those ICUs that do not screen at admission (13.6%) but perform screening at regular intervals
might overlook imported MRSA cases and misclassify imported MRSA cases as hospital acquired.
As the response rate in the study was 67.3% and after exclusion of the non-screeners only 50.9%
participants remained, selection bias has to be considered and representativity of the study population
for all screening ICUs participating in MRE-KISS can not be warranted. Furthermore, hospitals and
ICUs deciding to participate in ICU- and MRE-KISS might differ systematically from those that do not
take part in a surveillance system. Thus, the representativity for all German ICUs is difficult to assess.
Another important point to consider is the size of the study. As the study sample is relatively small, it
can be assumed that the power is not sufficient to detect small differences. For future studies efforts
should be made to enhance the participation rate.
Compliance with infection control procedures, one aspect known to have substantial influence on the
effectivity of infection control interventions, has not been addressed in the study [24,25]. Hence,
appraisal of the study results should take account of this problematic factor. In this context it should
also be kept in mind that the variables that have been included in the study represent only a proportion
of the whole spectrum of infection control measures aiming to prevent nosocomial infections as well as
the spread of multiresistant pathogens. Thus, there may be several confounding and/or interacting
factors, which may not have been considered but nevertheless make an essential contribution to the
outcome-measure.
In conclusion, multivariate analysis did not reveal any risk factors or protective effects originating from
the structural set-up of hospitals and ICUs and from organizational interventions targeting prevention
and control of the spread of MRSA. Nevertheless, it could be shown that medical ICUs experience
lower nosocomial MRSA case rates than the other ICU types and that the imported MRSA incidence is
si
gnificantly associated with higher numbers of nosocomial MRSA cases. This point should be
considered when benchmarking between hospitals.

Transparancy Declaration: The authors declare that they have no conflict of interests in relation to
this work.

Page 7

References

1. Wisplinghoff H, Bischoff T, Tallent SM, Seifert H, Wenzel RP, Edmond MB. Nosocomial
bloodstream infections in US hospitals: analysis of 24,179 cases from a prospective nationwide
surveillance study. Clin Infect Dis 2004; 39: 309–317.
2. Tiemersma EW, Bronzwaer SL, Lyytikäinen O et al. European antimicrobial resistance surveillance
system participants. Methicillin-resistant Staphylococcus aureus in Europe, 1999–2002. Emerg Infect
Dis 2004; 10: 1627–1634.
3. EARSS. EARSS annual report 2007. Bilthoven: European Antimicrobial Resistance Surveillance
System, 2008; 108–109.
4. Kresken M, Hafner D, Schmitz F-J, Wichelhaus T. PEG-Resistenzstudie 2007. Resistenzsituation
bei klinisch wichtigen Infektionserregern gegenüber Antibiotika in Deutschland und im
Mitteleuropäischen Raum. Paul Ehrlich Gesellschaft für Chemotherapie, Germany.
5. Humphreys H, Grundmann H, Skov R, Lucet JC, Cauda R. Prevention and control of methicillin-
resistant Staphylococcus aureus. Clin Microbiol Infect 2009; 15: 120–124.
6. Gastmeier P, Geffers C, Sohr D, Dettenkofer M, Daschner F, Rüden H. Five years working with the
German nosocomial infection surveillance system (Krankenhaus Infektions Surveillance System). Am
J Infect Control 2003; 31: 316–321.
7. Kohlenberg A, Schwab F, Meyer E, Behnke M, Geffers C, Gastmeier P. Regional trends in
multidrug-resistant infections in German intensive care units: a real-time model for epidemiological
monitoring and analysis. J Hosp Infect 2009; 7: 239–245.
8. Grammatico-Guillon L, Thiolet JM, Bernillon P, Coignard B, Khoshnood B, Desenclos JC.
Relationship between the prevalence of methicillin-resistant Staphylococcus aureus infection and
indicators of nosocomial infection control measures: a population-based study in French hospitals.
Infect Control Hosp Epidemiol 2009; 30: 861–869.
9. Gastmeier P, Schwab F, Geffers C, Rüden H. To isolate or not to isolate? Analysis of data from the
German Nosocomial Infection surveillance system regarding the lacement of patients with methicillin-
resistant Staphylococcus aureus in private rooms in intensive care units. Infect Control Hosp
Epidemiol 2004; 25: 109–113.
10. Graffunder EM, Venezia RA. Risk factors associated with nosocomial methicillin-resistant
Staphylococcus aureus (MRSA) infection including previous use of antimicrobials. J Antimicrob
Chemother 2002; 49: 999–1005.
11. Grundmann H, Hori S, Winter B, Tami A, Austin DJ. Risk faczors for the transmission of methicillin-
resistant Staphylococcus aureus in an adult intensive care unit: fitting a model to the data. J Infect Dis
2002; 15: 481–488.
12. Halwani M, Solaymani-Dodaran M, Grundmann H, Coupland C, Slack R. Cross – transmission of
nososcomial pathogens in an adult intensive care unit: incidence and risk factors. J Hosp Infect 2006;
63: 39–46.
13. Williams VR, Callery S, Vearncombe M, Simor AE. The role of colonization pressure in nosocomial
transmission of methicillin-resistant Staphylococcus aureus. Am J Infect Control 2009; 37: 106–110.
14. Merrer J, Santoli F, Appéré de Vecchi C, Tran B, De Jonghe B, Outin H. ‘‘Colonisation pressure’’
and risk of acquisition of methicillinresistant Staphylococcus aureus in a medical intensive care unit.
Infect Control Hosp Epidemiol 2000; 21: 718–723.
15. Bloemendaal AL, Fluit AC, Jansen WM et al. Acquisition and crosstransmission of Staphylococcus
aureus in European intensive care units. Infect Control Hosp Epidemiol 2009; 30: 117–124.
16. Chaberny IF, Sohr D, Rüden H, Gastmeier P. Development of a surveillance system for methicillin-
resistant Staphylococcus aureus in German hospitals. Infect Control Hosp Epidemiol 2007; 28: 446–
452.
17. Eveillard M, Lancien E, Hidri N et al. Estimation of methicillin-resistant Staphylococcus aureus
transmission by considering colonization pressure at the time of hospital admission. J Hosp Infect
2005; 60: 27–31.
18. Eveillard M, Lancien E, Barnaud G et al. Impact of screening for MRSA carriers at hospital
admission on risk-adjusted indicators according to the imported MRSA colonization pressure. J Hosp
Infect 2005; 59: 254–258.
19. Mears A, White A, Cookson B et al. Healthcare-associated infection in acute hospitals: which
interventions are effective? J Hosp Infect 2009; 71: 307–313.
20. Weber DJ, Hoffmann KK, Rutala WA, Pyatt DG. Control of health care associated Staphylococcus
aureus: survey of practices in North Carolina hospitals. Infect Control Hosp Epidemiol 2009; 30: 909–
911.

Page 8

21. Lautenbach E, Nachamkin I, Hu B et al. Surveillance cultures for detection of methicillin-resistant
Staphylococcus aureus: diagnostic yield of anatomic sites and comparison of provider- and patient-
collected samples. Infect Control Hosp Epidemiol 2009; 30: 380–382.
22. Eveillard M, de Lassence A, Lancien E, Barnaud G, Ricard JD, Joly-Guillou ML. Evaluation of a
strategy of screening multiple anatomical sites for methicillin-resistant Staphylococcus aureus at
admission to a teaching hospital. Infect Control Hosp Epidemiol 2006; 27: 181–184.
23. Clements A, Halton K, Graves N et al. Overcrowding and understaffing in modern health-care
systems: key determinants in meticillinresistant Staphylococcus aureus transmission. Lancet Infect Dis
2008; 8: 427–434.
24. Pittet D, Hugonnet S, Harbarth S et al. Effectiveness of a hospitalwide programme to improve
compliance with hand hygiene. Infection control programme. Lancet. 2000;356:1307–1312. Erratum
in: Lancet 2000;356(9248):2196.
25. Cromer AL, Latham SC, Bryant KG et al. Monitoring and feedback of hand hygiene compliance
and the impact on facility-acquired methicillin- resistant Staphylococcus aureus. Am J Infect Control
2008; 36: 672– 677.




Tables

Table 1. Cases and rates of methicillin-resistant Staphylococcus aureus (MRSA) in the 140 intensive
care units (ICUs), stratified by type of ICU



a Kruskal–Wallis rank test.
b Incidence density of MRSA, number of MRSA/1000 patient days.
c Incidence density of nosocomial MRSA, number of nosocomial MRSA/1000 patient days.
d Incidence of imported MRSA, number of imported MRSA/100 patients.

Page 9

Table 2. Summary of structure, process and MRSA parameters in 140 intensive care units (ICUs)
and their association with the number of nosocomial MRSA cases: univariate negative binomial
regression model




(Continued on next page)

Page 10

a CI, confidence interval.
b Staff-patient ratio, number of staff/patient/day/shift.
c CVC rate, central venous catheter-rate; number of CVC-days/100 patient-days, quartiles.
d UTC rate, urinary tract catheter rate; number of UTC-days/100 patient days, quartiles.
e Invasive ventilation rate, number of invasive ventilation-days/100 patient-days, quartiles.
f Incidence of imported MRSA, number of imported MRSA/100 patients.
g For example, patients with chronic wounds, previous stay in a healthcare facility and other risk factors.