Objectives: To evaluate the feasibility of a point prevalence survey for monitoring antibiotic use in a voluntary sample of French hospitals.
Methods: Demographic and medical data were collected for all inpatients. Additional characteristics regarding antimicrobial treatment, type of infection and microbiological results were collected only for patients receiving antimicrobials.
Results: Among 3964 patients in 38 hospitals, 343 (8.7%) received antimicrobial prophylaxis and 1276 (32.2%) antimicrobial therapy. The duration of surgical antimicrobial prophylaxis was >1 day in 41 out of 200 (21%) of the cases. Among patients with antimicrobial therapy, 959 (75.2%) received β-lactams (including 34.8% penicillins with β-lactam inhibitors, 22.1% third-generation cephalosporins and 7.8% carbapenems) and 301 (23.6%) received fluoroquinolones (50% orally). A total of 518 (40.6%) patients were treated with more than one drug and 345 (27.2%) were treated for >7 days. Patients treated for hospital-acquired infections (39.2%) were more likely to receive combinations (47.6% versus 34.4%, P < 0.01), carbapenems (14.4% versus 2.6%, P < 0.01), glycopeptides (14.4% versus 3.7%, P < 0.01) and antifungals (17% versus 5.3%, P < 0.01) for a longer duration (7.8 versus 6 days, P < 0.01). Fifty-six patients (4.4%) were treated for >7 days and did not have any microbiological sample drawn. The time allocated for the survey represented 18.3-25.0 h for 100 patients.
Conclusions: The data provide directions for further interventions, such as better use of diagnostic tools, decreasing the treatment duration and the use of combinations. In addition, the survey shows that, although cumbersome, it is feasible to improve the representativeness of national data in European surveys.