To assess the efficacy of topical chlorhexidine for prevention of ventilator-associated pneumonia (VAP) in a meta-analysis.
Computerized PubMed and MEDLINE search supplemented by manual searches for relevant articles.
Randomized controlled trials evaluating efficacy of topical chlorhexidine applied to the oropharynx vs. placebo or standard care for prevention of VAP.
Data were extracted on patient population, inclusion and exclusion criteria, diagnostic criteria for VAP, form and concentration of topical chlorhexidine used, incidence of VAP, and overall mortality.
Data on incidence of VAP and mortality were abstracted as dichotomous variables. Pooled estimates of the relative risk and 95% confidence intervals were obtained using the DerSimonian and Laird random effects model and the Mantel-Haenszel fixed effects model. Heterogeneity was assessed using the Cochran Q statistic and I2. Subgroup analyses were used to explore heterogeneity.
Seven randomized controlled trials met the inclusion criteria. Topical chlorhexidine resulted in a reduced incidence of VAP (relative risk, 0.74; 95% confidence interval, 0.56–0.96; p = .02) using a fixed effects model. Using the more conservative random effects model, the point estimate was similar (relative risk, 0.70; 95% confidence interval, 0.47–1.04; p = .07), but the results failed to achieve statistical significance. The I2 test showed moderate heterogeneity. Subgroup analysis showed that the benefit of chlorhexidine was most marked in cardiac surgery patients (relative risk, 0.41; 95% confidence interval, 0.17–0.98; p = .04). There was no mortality benefit with chlorhexidine although the sample size was small.
Our analysis showed that topical chlorhexidine is beneficial in preventing VAP; the benefit is most marked in cardiac surgery patients. A large randomized trial is needed to determine the impact of topical chlorhexidine on mortality.
From the Section of Infectious Diseases, Department of Medicine, University of Wisconsin Medical School, Madison, WI.
Supported, in part, by a 2006 Vision Grant from the Society of Critical Care Medicine (NS).
The authors have not disclosed any potential conflicts of interest.
Address requests for reprints to: Nasia Safdar, MD, MS, University of Wisconsin Madison, H4/572 Section of Infectious Diseases, 600 Highland Avenue, Madison, WI 53792. E-mail: email@example.com