Laboratory confirmed bloodstream infection aetiology in an intensive care unit: eight years study.

To evaluate laboratory confirmed bloodstream infection (LC-BSI) aetiology we carried out a prospective study in the general 13 bed ICU of the teaching hospital Policlinico Umberto I in Rome. According to CDC case definitions for LC-BSI, all patients admitted >48h between 2000-2007 to ICU were included. Risk factors (i.e. age, sex, SAPS II), invasive procedures (i.e. endotracheal intubation, vascular and urinary catheterisation), microbiological isolates and their antibiotic susceptibility were screened. Overall 1741 patients (64.8% males, 35.2% females) were included, mean age was 58.1 +/- 19.8, SAPS II score 45.1 +/- 17 and ICU stay 14.0 +/- 21.1 days. Finally, 167 (9.6%) patients developed 203 (11.7%) ICU-acquired LC-BSI and sources of infection were CVC (39.8%), unknown (39.3%), respiratory tract (12.4%), surgical wound (6.5%) and urinary tract (2.0%). Between 2000 and 2007 the incidence of LC-BSI/1000 patient days (14.8 per thousands vs. 7.8 per thousands: p85%). P. aeruginosa and A. baumannii showed respectively 25% and 68.7% multidrug-resistance. Over 1/3 of Eneterobacteriaceae isolates were extended spectrum beta-lactamase (ESBL), but non resulted resistant to carbapenems. Surveillance showed a high incidence of LC-BSI associated to invasive procedures and the presence of multiresistant bacteria