Transmission of Staphylococcus aureus between healthcare workers, the environment and patients in an intensive care unit: a whole-genome sequencing based longitudinal cohort study

BACKGROUND: Health-care workers have been implicated in nosocomial outbreaks of Staphylococcus aureus, but the dearth of evidence from non-outbreak situations means that routine health-care worker screening and S aureus eradication are controversial. We aimed to determine how often S aureus is transmitted from health-care workers or the environment to patients in an intensive care unit (ICU) and a high-dependency unit (HDU) where standard infection control measures were in place. METHODS: In this longitudinal cohort study, we systematically sampled health-care workers, the environment, and patients over 14 months at the ICU and HDU of the Royal Sussex County Hospital, Brighton, England. Nasal swabs were taken from health-care workers every 4 weeks, bed spaces were sampled monthly, and screening swabs were obtained from patients at admission to the ICU or HDU, weekly thereafter, and at discharge. Isolates were cultured and their whole genome sequenced, and we used the threshold of 40 single-nucleotide variants (SNVs) or fewer to define subtypes and infer recent transmission. FINDINGS: Between Oct 31, 2011, and Dec 23, 2012, we sampled 198 health-care workers, 40 environmental locations, and 1854 patients; 1819 isolates were sequenced. Median nasal carriage rate of S aureus in health-care workers at 4-weekly timepoints was 36·9% (IQR 35·7–37·3), and 115 (58%) health-care workers had S aureus detected at least once during the study. S aureus was identified in 8–50% of environmental samples. 605 genetically distinct subtypes were identified (median SNV difference 273, IQR 162–399) at a rate of 38 (IQR 34–42) per 4-weekly cycle. Only 25 instances of transmission to patients (seven from health-care workers, two from the environment, and 16 from other patients) were detected. INTERPRETATION: In the presence of standard infection control measures, health-care workers were infrequently sources of transmission to patients. S aureus epidemiology in the ICU and HDU is characterised by continuous ingress of distinct subtypes rather than transmission of genetically related strains