Carbapenem-Resistant Enterobacteriaceae Infection Deaths
Carbapenem-Resistant Enterobacteriaceae Infection Deaths
We performed a systematic search in the PubMed (http://www.ncbi.nlm.nih.gov/pubmed/) and Scopus (http://www.scopus.com/home.url?zone=header&origin=searchbasic) databases on April 9, 2012, by using the following search terms: carbapenem-resistant or carbapenemase-producing or KPC and outcome or mortality. We also conducted a hands-on search of the reference lists of relevant studies to identify additional studies. Articles published in languages other than English, French, German, Italian, Spanish, or Greek were not evaluated.
Any article that compared death rates between CRE-infected patients and CSE-infected patients was considered eligible for inclusion in the review. Studies that reported only on carbapenem-resistant isolates (without comparison with susceptible isolates) were excluded, as were studies that compared patients who had carbapenem-resistant infections with patients who were not infected. Other excluded studies were those that did not distinguish the outcomes for infected patients from those for colonized patients and studies that reported on isolates resistant to a carbapenem other than imipenem, meropenem, or doripenem. Studies that compared infection-related but not all-cause deaths among CRE-infected patients with those among CSE-infected patients were excluded because of homogeneity of the outcome analysis. Unpublished studies presented as abstracts at scientific conferences were not eligible for inclusion because of the low quantity of information provided in these types of articles.
Literature search, study selection, and data extraction were performed independently by 2 of the authors (G.S.T. and K.Z.V.). Any disagreement was resolved by consensus in meetings with all investigators and by reviewing the original articles to assess validity of the abstracted data. Extracted data included study characteristics (author, design, country, period, number of patients) and patient characteristics (type of infection, causative pathogen, and concurrent condition or severity of illness score at admission). We also recorded the all-cause deaths in each group of patients (CRE and CSE), deaths attributable to carbapenem resistance, and the independent predictors of all-cause deaths evaluated in the total population of each study.
For studies in which analyses were performed for the individual patient groups (CRE and CSE) rather than the study population as a whole, we could not conclude whether carbapenem resistance predicted death. Thus, we did not extract results from such studies.
We compared 2 groups of patients: CRE-infected and CSE-infected patients. The primary outcome of our analysis was the comparison of all-cause deaths between CRE and CSE groups with the same type of infection (i.e., bacteremia or pneumonia) caused by the same species (i.e., K. pneumoniae). The secondary outcome was deaths attributable to carbapenem resistance in Enterobacteriaceae infections. Attributable death was defined as the difference in all-cause deaths between the 2 compared groups.
Carbapenem resistance was defined as the resistance of a pathogen to imipenem, meropenem, or doripenem, according to the susceptibility breakpoints that had been applied by the investigators of each study. Carbapenemase production was not considered as carbapenem resistance if the MIC of an antibiotic was within the susceptible range according to those breakpoints.
We calculated pooled risk ratios (RRs) and 95% CIs regarding deaths. The statistical heterogeneity between studies was assessed by using the χ test (p<0.10 was defined to indicate the presence of heterogeneity) and the I index (for assessing the degree of heterogeneity). The random effects model was applied because we considered the nonrandomized, comparative studies that we analyzed to be heterogeneous by definition. We used RevMan 5.1 software (Nordic Cochrane Centre of the Cochrane Collaboration, Copenhagen, Denmark) to perform the metaanalysis.
Methods
Literature Search
We performed a systematic search in the PubMed (http://www.ncbi.nlm.nih.gov/pubmed/) and Scopus (http://www.scopus.com/home.url?zone=header&origin=searchbasic) databases on April 9, 2012, by using the following search terms: carbapenem-resistant or carbapenemase-producing or KPC and outcome or mortality. We also conducted a hands-on search of the reference lists of relevant studies to identify additional studies. Articles published in languages other than English, French, German, Italian, Spanish, or Greek were not evaluated.
Study Selection Criteria
Any article that compared death rates between CRE-infected patients and CSE-infected patients was considered eligible for inclusion in the review. Studies that reported only on carbapenem-resistant isolates (without comparison with susceptible isolates) were excluded, as were studies that compared patients who had carbapenem-resistant infections with patients who were not infected. Other excluded studies were those that did not distinguish the outcomes for infected patients from those for colonized patients and studies that reported on isolates resistant to a carbapenem other than imipenem, meropenem, or doripenem. Studies that compared infection-related but not all-cause deaths among CRE-infected patients with those among CSE-infected patients were excluded because of homogeneity of the outcome analysis. Unpublished studies presented as abstracts at scientific conferences were not eligible for inclusion because of the low quantity of information provided in these types of articles.
Data Extraction
Literature search, study selection, and data extraction were performed independently by 2 of the authors (G.S.T. and K.Z.V.). Any disagreement was resolved by consensus in meetings with all investigators and by reviewing the original articles to assess validity of the abstracted data. Extracted data included study characteristics (author, design, country, period, number of patients) and patient characteristics (type of infection, causative pathogen, and concurrent condition or severity of illness score at admission). We also recorded the all-cause deaths in each group of patients (CRE and CSE), deaths attributable to carbapenem resistance, and the independent predictors of all-cause deaths evaluated in the total population of each study.
For studies in which analyses were performed for the individual patient groups (CRE and CSE) rather than the study population as a whole, we could not conclude whether carbapenem resistance predicted death. Thus, we did not extract results from such studies.
Definitions and Outcomes
We compared 2 groups of patients: CRE-infected and CSE-infected patients. The primary outcome of our analysis was the comparison of all-cause deaths between CRE and CSE groups with the same type of infection (i.e., bacteremia or pneumonia) caused by the same species (i.e., K. pneumoniae). The secondary outcome was deaths attributable to carbapenem resistance in Enterobacteriaceae infections. Attributable death was defined as the difference in all-cause deaths between the 2 compared groups.
Carbapenem resistance was defined as the resistance of a pathogen to imipenem, meropenem, or doripenem, according to the susceptibility breakpoints that had been applied by the investigators of each study. Carbapenemase production was not considered as carbapenem resistance if the MIC of an antibiotic was within the susceptible range according to those breakpoints.
Statistical Analysis
We calculated pooled risk ratios (RRs) and 95% CIs regarding deaths. The statistical heterogeneity between studies was assessed by using the χ test (p<0.10 was defined to indicate the presence of heterogeneity) and the I index (for assessing the degree of heterogeneity). The random effects model was applied because we considered the nonrandomized, comparative studies that we analyzed to be heterogeneous by definition. We used RevMan 5.1 software (Nordic Cochrane Centre of the Cochrane Collaboration, Copenhagen, Denmark) to perform the metaanalysis.
