Objective: We carried out a large multicentre study to determine whether wearing gloves was associated with poorer hand hygiene compliance.
Methods: 232 hours of observations (7020 observations) were carried out on 15 Intensive Therapy Units (ITUs) and 41 Acute Care of the Elderly /General Medical (ACE/GM) wards in 20 hospitals in England & Wales, whilst conducting a randomised controlled trial of an intervention to improve hand hygiene compliance (FIT trial N0256159318, NRR website). Gloves use was not part of the intervention. Hand hygiene moments & behaviours were recorded using a rigorously standardised hand hygiene observation tool (the HHOT) , and noting whether the HCW was using gloves or not. Compliance with & without gloves was compared overall & for different hand hygiene moments.
Results: Of the 7020 observations, 1729 (25%) were associated with glove use in both ITUs and ACE/GM wards. Gloves were used in 78% of high risk hand hygiene moments (aseptic technique, body fluid exposure) and in 16% of low risk moments (before & after patient contact, after environmental contact). Overall hand hygiene compliance was lower when wearing gloves (42.6%) than when not wearing gloves (51.2%). The odds of HCWs cleaning their hands were less when gloves were worn (OR = 0.71; [CI 95% 0.61-0.82]). This was true for all hand hygiene moments. On ACE/GM wards there was a greater difference in overall compliance (35.3% with gloves v 49.5% without; odds 0.56 [0.45, 0.69]) than on ITUs (50.1% v 54.9%).
Conclusions: This large study found that glove use was associated with worse hand hygiene compliance, and may present a barrier to HH. Modelling studies suggest that small increments in compliance may result in large reductions in transmission of organisms. It is therefore possible that interventions to reduce unnecessary glove use & ensure appropriate compliance before & after glove use could have a significant effect on spread of infection. Further work examining the predictors of glove usage & associated hand hygiene by HCWs may be warranted to guide such interventions.