Chlorhexidine-impregnated dressing for prevention of catheter-related bloodstream infection: a meta-analysis*.

OBJECTIVE: To assess the efficacy of a chlorhexidine-impregnated dressing for prevention of central venous catheter-related colonization and catheter-related bloodstream infection using meta-analysis. DATA SOURCES: Multiple computerized database searches supplemented by manual searches including relevant conference proceedings. STUDY SELECTION: Randomized controlled trials evaluating the efficacy of a chlorhexidine-impregnated dressing compared with conventional dressings for prevention of catheter colonization and catheter-related bloodstream infection. DATA EXTRACTION: Data were extracted on patient and catheter characteristics and outcomes. DATA SYNTHESIS: Nine randomized controlled trials met the inclusion criteria. Use of a chlorhexidine-impregnated dressing resulted in a reduced prevalence of catheter-related bloodstream infection (random effects relative risk, 0.60; 95% CI, 0.41-0.88, p = 0.009). The prevalence of catheter colonization was also markedly reduced in the chlorhexidine-impregnated dressing group (random effects relative risk, 0.52; 95% CI, 0.43-0.64; p < 0.001). There was significant benefit for prevention of catheter colonization and catheter-related bloodstream infection, including arterial catheters used for hemodynamic monitoring. Other than in low birth weight infants, adverse effects were rare and minor. CONCLUSIONS: Our analysis shows that a chlorhexidine-impregnated dressing is beneficial in preventing catheter colonization and, more importantly, catheter-related bloodstream infection and warrants routine use in patients at high risk of catheter-related bloodstream infection and central venous catheter or arterial catheter colonization.

Candida glabrata: an emerging pathogen in Brazilian tertiary care hospitals.

Candida glabrata is an infrequent cause of candidemia in Brazilian public hospitals. We investigated putative differences in the epidemiology of candidemia in institutions with different sources of funding. Prospective laboratory-based surveillance of candidemia was conducted in seven private and two public Brazilian tertiary care hospitals. Among 4,363 episodes of bloodstream infection, 300 were caused by Candida spp. (6.9%). Incidence rates were significantly higher in public hospitals, i.e., 2.42 vs. 0.91 episodes per 1,000 admissions (P< 0.01). Patients in private hospitals were older, more likely to be in an intensive care unit and to have been exposed to fluconazole before candidemia. Candida parapsilosis was more frequently recovered as the etiologic agent in public (33% vs. 16%, P< 0.001) hospitals, whereas C. glabrata was more frequently isolated in private hospitals (13% vs. 3%, P < 0.001). Fluconazole resistance among C. glabrata isolates was more frequent in private hospitals (76.5% vs. 20%, P = 0.02). The 30-day mortality was slightly higher among patients in public hospitals (53% vs. 43%, P = 0.10). Candida glabrata is an emerging pathogen in private institutions and in this setting, fluconazole should not be considered as a safe option for primary therapy of candidemia.