Background: ICU-acquired infections constitute an important world-wide health problem. Our aim was to determine the incidence, predictive factors and impact of ICU-AIs in ICU patients in Tunisia.
Material/Methods: We conducted a prospective observational cohort study over a 3 month period in the medical surgical intensive care unit of Habib Bourguiba University Hospital (Sfax-Tunisia).
Results: During the study period 261 patients were surveyed; 44 of them (16.9%) developed 55 episodes of ICU-AI (34.7 ICU-AI/1000 days of hospitalization). The most frequently identified infections were ventilator-associated pneumoniae (58.2%), and primary bloodstream infection (18.2%). The most frequently isolated organisms were multidrug-resistant P. aeruginosa (44.7%), and A. baumannii (21.3%). The initial antibiotic prescription for ICU-AI was inadequate in 9 cases (16.4% of episodes of ICU-AI). At ICU discharge, overall mortality was 29.9%. Independent risk factors for acquiring infection in ICU were the use of central venous catheter (p=0.014) and antibiotic prescription on admission for more than 24 hours (p=0.025), those of mortality in ICU were SAPS II of more than 35 points (p<0.001) and ICU-AI (p=0.002), and those of mortality at 28 days after an episode of ICU-AI were septic shock (p=0.004) and inadequate initial antimicrobial treatment (p=0.011).
Conclusions: We conclude that the occurrence of ICU-AI is significantly related to increased mortality, and that focusing interventions on better use of antibiotics would have a benefit in terms of prevention and consequences of ICU-AI.

Keywords: Intensive Care Units, Cross Infection - therapy, Child, Preschool, Child, Bacteria - isolation & purification, Aged, 80 and over, Adolescent, Risk Factors, Tunisia - epidemiology, young adult