Determinants of Implementation of Isolation Precautions Against Infections by Multidrug-Resistant Microorganisms: A Hospital-Based, Multicenter, Observational Study.

OBJECTIVES We aimed to ascertain the factors associated with lack of isolation precautions (IP) in patients infected or colonized by third-generation cephalosporin-resistant Enterobacteriaceae (3GCR-E) and methicillin-resistant Staphylococcus aureus (MRSA) in hospital settings. DESIGN Prospective surveillance and audit of practices. SETTING The study included 4 university hospitals in Lyon, France. PARTICIPANTS All patients hospitalized between April and June in 2013 and 2015 were included. Case patients had ≥1 clinical sample positive for MRSA and/or 3GCR-E. METHODS Factors associated with the lack of IP implementation were identified using multivariate logistic regression. The incidence of MDRO infections was expressed per 10,000 patient days. RESULTS Overall, 57,222 patients accounting for 192,234 patient days of hospitalization were included, and 635 (1.1%) MDRO cases were identified. MRSA incidence was 2.5 per 10,000 patient days (95% confidence interval [95% CI], 2.1-3.0) and 3GCR-E incidence was 10.1 per 10,000 patient days (95% CI, 9.2-11.0), with no crude difference between 2013 and 2015 (P=.15 and P=.11, respectively). Among 3GCR-E, the main species were Escherichia coli (43.8%) and Klebsiella pneumoniae (31.0%). Isolation precautions were implemented in 78.5% of cases. Lack of IP implementation was independently associated with patient age, year, specialty, hospital, colonization compared with infection, and lack of medical prescription for IPs (adjusted odds ratio, 17.4; 95% CI, 8.5-35.8; P<.001). CONCLUSIONS MRSA and 3GCR-E infections and/or colonizations are frequent in healthcare settings, and IPs are implemented in most cases. When IPs are lacking, the main factor is the absence of medical prescription for IPs, underscoring the need for alerts to physicians by the microbiological laboratory and/or the infection control team. Infect Control Hosp Epidemiol 2017;38:1188-1195.