Evaluation of the contribution of isolation precautions in prevention and control of multi-resistant bacteria in a teaching hospital.

From February 1999 to January 2000, a control programme to prevent the spread multi-resistant bacteria (MRB) was implemented in a French teaching hospital. This programme focused on methicillin-resistant Staphylococcus aureus (MRSA) and Enterobacteriaceae producing extended-spectrum beta-lactamases (ESBL), and was based on the application of barrier precautions (washing hands with antiseptic soaps, wearing disposable gloves and gowns, identifying MRB carriers). No changes in antibiotic policy occurred during the year. Our aim was to conduct an evaluation of this programme by measuring incidence rates. Concurrently, the effect of barrier precautions was estimated in an indirect way, by documenting the availability of barrier precautions in MRB carriers' rooms and by analysing the monthly correlation between the supply of such material and the theoretical cumulated length of MRB carriers' isolation in six randomized wards. All MRB isolated in hospitalized patients were recorded, and differentiated between acquisition in our hospital or from elsewhere. For the analysis of trends, the year was divided in three periods of four months. Over the year, the global MRB incidence was 1.26 per 1000 patient-days (PD) [95% confidence interval (95%CI)=1.16-1.36]. The MRSA incidence was 0.89 per 1000 PD (95%CI=0.81- 0.97) and the ESBL incidence was 0.38 per 1000 PD (95% CI=0.33-0.43). The MRB incidence decreased significantly in all types of specialties except for surgical wards. The incidence decreased by 17.9% for MRSA, 54.9% for ESBL and 34.8% for both MRB. Concurrently, the proportion of strains acquired in our hospital decreased for MRSA (P for trend > or = 0.05) and ESBL (P for trend > or = 0.01), whereas the incidence of imported strains increased slightly. The proportion of multiresistant strains in S. aureus (36.8%) and Enterobacter aerogenes (37.0%) remained similar throughout the year. Thus, the decrease of the incidence concerned both resistant and susceptible strains. The availability of antiseptic soaps increased significantly (P for trend > or = 0.01). The amount of antiseptic soap ordered and the theoretical lengths of isolation were correlated on a monthly basis (Spearman coefficient = 0.72; P > or = 0.02). These results shows the efficacy of such a programme of MRB containment in a large hospital, provided barrier nursing is instigated, together with the availability of such material as antiseptic soap, to allow implementation.