Increased use of alcohol-based hand sanitizers and successful eradication of methicillin-resistant Staphylococcus aureus from a neonatal intensive care unit: A multivariate time series analysis

Abstract

We analyzed time series data to investigate factors that contributed to the gradual decrease and eventual eradication of methicillin-resistant Staphylococcus aureus (MRSA) from our neonatal intensive care unit (NICU). A multivariate adjusted autoregressive integrated moving average (ARIMA) model was used for time series analyses of monthly MRSA incidence density rates and their predictors in the NICU from July 2003 to July 2009. Based on our ARIMA (0,1,1), which is a nonseasonal and nonstationary moving average model, the monthly pooled mean of the amount of alcohol-based hand sanitizer used for 1 patient per day (lag time, 0 month; P = .011) was the only factor significantly associated with the MRSA incidence density rates. MRSA colonization pressure, patient-to-nurse ratios, and bed occupancy rates were not associated with MRSA acquisition in the NICU. Active surveillance, which had been practiced since the late 1980s, alone was not sufficient to control the spread of MRSA until it was accompanied by enhanced hand hygiene. Increasing the use of alcohol-based hand sanitizers by improving accessibility and providing periodic hand hygiene training sessions to health care workers is strongly recommended for decreasing the risk of MRSA acquisition among neonates in NICU settings.