A one-year prospective study of nosocomial bacteraemia in ICU and non-ICU patients and its impact on patient outcome

A one-year, prospective, two-observational cohort study, was performed to evaluate the incidence and outcome in hospitalized patients (ICU and non-ICU) of nosocomial bacteraemia, and to assess its prognostic value in the ICU group. A group of 18 098 hospitalized patients and a group of 291 consecutive ICU patients were followed. Prognostic factors were determined using single and multivariable analyses. 109 (90 non-ICU and 19 ICU) patients developed I IS nosocomial bacteraemic episodes. The incidence of nosocomial bacteraemia was 6.0 per 1000 admissions (95% confidence interval (CI): 5-7%) and 65 per 1000 admissions in ICU patients (95% CI: 4.5-8.5%). Gram-positive and Gram-negative bacteria were 63/133 (47%) and 70/133 (53%) of the isolated micro-organisms respectively. Crude mortality rates were 41/109 (38%) with adverse outcome associated with mechanical ventilation (OR: 3.65 95% CI: 1.4-9.2, P=0.01), neutropenia (OR: 7.7; 95% CI: 0.8-73.1; P=0.07) while gastro-intestinal surgery was associated with an improved outcome (OR: 0.4; 95% CI: 0.16-0.96; P=0.04). Of the 291 ICU patients, 19 acquired 22 episodes of nosocomial bacteraemia. and IS were referred from the wards with documented nosocomial bacteraemia. Of these 37 bacteraemic patients, 17 (46%) died. When adjusting for predictors of death (SAPS II greater than or equal to 40, cardiac and neurological failure), nosocomial bacteraemia markedly influence the outcome in ICU patients (OR: 3.4; 95% CI: 1.3-8.7; P=0.010). This study suggests that the outcome of nosocomial bacteraemia in hospitalized patients is poor in ventilated and neutropenic patients and that nosocomial bacteraemia per se influenced outcome in ICU patients. (C) 2000 The Hospital infection Society.