Significant reduction of nosocomial infectious: stratified analysis of prevalence national studies performed in 1996 and 2001 in French north interegion

Background. - In France, repeated prevalence studies of nosocomial infections (NI) are part of governmental plan against NI built in 1995 by the Ministry of Health. To evaluate strand of NI prevalence, we performed a comparative analysis of two successive national point-prevalence surveys occurring in 1996 and 2001 for the Northern France. Method. - Comparison concerned the hospitals, which participated in the two studies of 1996 and 2001 in Northern France. The studies were designed, as a point-prevalence survey on voluntary basis. For each patient, risk factors and presence of active NI at the day of the study were recorded on standardised form. Criteria of NI used were these of "100 recommendations" of CTIN and of CCLIN North guideline. Prevalence rate (PR) and frequency of risk factors were compared. The risk factors significantly linked to NI by logistic regression were used to build a score of five risk levels of NI (PUVARISK) allowing an adjusted comparison of the 2 years. Results. - Total of 161 hospitals participated at the two studies, including respectively 61422 and 58 749 patients. Between 1996 and 2001, crude PR of infected patients and of NI decreased respectively from 7.8% to 7.3% and 9.0% to 8.0% (P < 10(-4)), so then relative decreases were of 6.4% and 11.1%. In contrast, the frequency of risk factors, except surgery in the past 30 days, significantly increased. Risk factors included in PREVARISK were: age >65 years, immunosuppression, surgery in the past 30 days, urinary tract and central catheter. In patients with a low risk level (PREVARISK = 0), the relative decrease of infected patients and NI PR were of 17% and 19%. The decrease was not significant for patients with high risk level (PREVARISK greater than or equal to 3). Conclusion. - Our analyses show a decrease of PR adjusted on risk factors, especially in patients with a low risk level. These result suggest an efficacy of program against NI in studied hospitals especially for patients for whom NI would be potentially avoidable. (C) 2003 Editions scientifiques et medicales Elsevier SAS. Tous droits reserves.