Seven studies were included in the review: two RCTs and five interrupted time series (ITS) (20,384 participants, range 256 to 5,043). There was no perceived risk of selection bias or attrition bias in any study. One study was double-blinded and another was partial-blinded. Detection bias was considered to be present in three studies. Compliance with chlorhexidine body washing was measured in one study.
Incidence of acquisition of MRSA carriage was significantly reduced with chlorhexidine body washing in the three ITS studies in which this was the primary outcome.
Five studies (four ITS, one RCT) assessed MRSA infection rates (three studies as the primary endpoint) and a significant reduction in MRSA infection rate was found in one ITS study.
Carriage and bacteraemia rates due to vancomycin-resistant enterococcus were assessed as the primary endpoints in one ITS study that found a significant reduction in both outcomes.
Three studies (two RCTs, one ITS) considered all-cause infection rates as the primary endpoint and one RCT found a significant reduction in incidence of all-cause primary blood stream infections with chlorhexidine body washing.
None of the studies had incidence of colonisation and/or infections with antibiotic-resistant gram-negative bacteria as primary endpoints. One RCT found a higher incidence of infection with antibiotic-resistant gram-negative bacteria after introduction of chlorhexidine body washing. Two studies (one ITS, one RCT) found a reduction in incidence of infection with chlorhexidine body washing. Rates were very low (1% or less) in both study periods in one ITS study.