Nosocomial infections in intensive care unit in a Turkish university hospital: a 2-year survey

Objective. To determine epidemiology and risk factors for nosocomial infections in intensive care unit (ICU). Design. Prospective incidence survey. Setting. An adult general ICU in a university hospital in western Turkey. Patients. All patients who stayed more than 48 h in ICU during a 2-year period (2000-2001). Measurements and results. The study included 434 patients (7394 patient-days). A total of 225 infections were identified in 113 patients (26%). The incidence and infection rates were 56.8 in 1000-patient days and 51.8%, respectively. The infections were pneumonia (40.9%), bloodstream (30.2%), urinary tract (23.6%) and surgical site infections (5.3%). Pseudomonas aeruginosa (22.6%), methicillin-resistant Staphylococcus aureus (22.2%) and Acinetobacter spp. (11.9%) were frequently isolated micro-organisms. Median length of stay with nosocomial infection and without were 13 days (Interquartile range, IQR, 20) and 2 days (IQR, 2), respectively (P<0.0001). In logistic regression analysis, mechanical ventilation [odds ratio (OR): 16.35; 95% confidence interval (CI): 8.26-32.34; P<0.0001), coma (OR: 15.04; 95% CI: 3.41-66.33; P=0.0003), trauma (OR: 10.27; 95% CI: 2.34-45.01; P=0.002), nasogastric tube (OR: 2.94; 95% CI: 1.47-5.90; P=0.002), tracheotomy (OR: 5.77; 95% CI: 1.10-30.20; P=0.04) and APACHE II scores 10-19 (OR: 10.80; 95% CI: 1.10-106.01; P=0.04) were found to be significant risk factors for nosocomial infection. Rate of nosocomial infection increased with the number of risk factors (P<0.0001). Mortality rates were higher in infected patients than in non-infected patients (60.9 vs 22.1%; P<0.0001). Conclusion. These data suggest that, in addition to underlying clinical conditions, some invasive procedures can be independent risk factors for nosocomial infection in ICU.