Surveillance of nosocomial infections associated with anaesthesia. A multicentre survey.

Objectives: To determine incidence rate, main characteristics and risk fact ors of nosocomial infections associated with anesthesia (NIAA). Study design: Prospective, descriptive multicentre survey. Patients: All patients aged more than 15 years and undergoing surgery (exce pt cardio-thoracic, ENT or ambulatory surgery) under general or regional an aesthesia. Methods: Volontary participation of surgical units from public or private h ospitals. Use of pre-established definitions of infections and a 72 hours p ostanesthetic follow-up. Anaesthesia and operation related risk factors col lected. End point based on occurrence, or not, of clinical infection. Recor d, control, treatment and analysis of the data by Epi Info - 5.0 software. Statistics used: Fischer's exact test, Mantel-Haenszel test, Anova method, Kruskall-Wallis test. Results: Among 7,300 patients belonging to 13 hospitals, 25 developed an in fection (nine vascular catheter related infections, 12 respiratory tract in fections, two infections of the eye and two of the mouth). Only two infecti ons have been bacteriologically documented. The overall incidence of NIAA w as 3.4 per 1,000 patients. It was significantly higher after an anaesthetic of more than 2 hours and after transfusion. Conclusions: This first prospective survey of NIAA confirmed that nosocomia l infections are a real problem in the practice of anaesthesia and the nece ssity to use preventive measures. A survey with a larger sample size would allow to specify the respective part of the various risk factors and to dev elop a risk index. (C) 2000 Editions scientifiques et medicales Elsevier SA S.