Antibiotic usage in intensive care units: a pharmaco-epidemiological multicentre study.

OBJECTIVES: To prospectively assess the use of antibiotics in ICUs. PATIENTS AND METHODS: A total of 979 critically ill patients over 14 years of age were recruited in 43 Italian ICUs. For each patient, admission and discharge characteristics, information on the drugs administered, use of antibiotic susceptibility tests (ASTs), presence and severity of sepsis were collected daily until discharge or for a maximum of 21 days.
RESULTS: Most patients with sepsis (99%) received antibiotics, and in almost all (93%) the treatment was started empirically, with broad-spectrum antibiotics. ASTs followed the onset of empirical treatment in 93% of cases. De-escalation was carried out in 16 patients, while in 37.6% of cases an antibiotic had to be changed or added. Antibiotic prophylaxis in surgical patients involved widespread use of drug combinations (31% of cases) and lasted 3 days on average. In non-surgical patients antibiotic prophylaxis lasted 4.6 days and in 42% a third-generation cephalosporin was used.
CONCLUSIONS: We found an appropriate approach to the therapeutic use of antibiotics: early empirical onset with broad-spectrum antimicrobial, followed by ASTs in order to target the therapy. However, in more than one-third of the cases the first-line choice was inappropriate. As regards prophylaxis, both surgical and non-surgical patients tended to have excessive duration of treatment, with widespread use of antibiotic combinations, too often involving a third-generation cephalosporin or carbapenem. This indicates a wide gap between clinical guidelines and clinical practice that calls for close assessment.