Bed baths to beat bacteria

1 February 2013

CLINICALLY APPRAISED TOPIC (CAT): Does bathing patients with antiseptic cloths reduce bloodstream infections?


Daily bathing with a 2 per cent chlorhexidine-impregnated washcloth reduced hospital-acquired bloodstream infections rates by 28 per cent.


Ever on the alert for opportunities to reduce hospital-acquired infection rates in your hospital, you notice a new trial of chlorhexidine bathing for intensive care units. Although already participating in the national CLAB (central line-associated bloodstream infection) initiatives, you wonder if there might now be sufficient evidence to implement routine chlorhexidine bathing.


Among adult intensive care patients, does bathing with chlorhexidine reduce hospital-acquired infections?


No search strategy – paper notified through push mechanisms (Table of Contents alert).


Climo MW, Yokoe DS, Wareen DK, et al. Effect of daily chlorhexidine bathing on hospital-acquired infection, New Eng J Med 2013; 386:533-42.


Cluster randomised crossover trial conducted between August 2007 and February 2009 in eight intensive care units (ICUs) and one bone marrow transplant unit (BMT) in six hospitals in the United States. All patients admitted to the units were approached to participate and only eight declined; 7735 patients agreed to participate. The types of intensive care unit included medical intensive care (4), surgical intensive care (3), and cardiovascular intensive care (1). The clusters were randomised to bathe participants daily using a proprietary chlorhexidine-impregnated washcloth or a proprietary non-antiseptic washcloth, and then after six months, the units crossed over to using the alternate washcloth to that which they had used in the previous six months. Before the study started nursing staff trained in use of both washcloths.

Intervention: Daily bathing with 2 per cent chlorhexidine gluconate-impregnated alcohol-free and rinse-free washcloths used in accord with manufacturer’s directions.

Control: Daily bathing with non-antiseptic rinse-free washcloths used in accord with manufacturer’s directions.

Outcomes: Outcome measures included the overall multi-drug resistant organism (MDRO) acquisition, hospital-acquired blood-stream infections (HA-BSI), acquisition of MRSA (methicillin-resistant staphylococcus aureus) and VRE (vancomycin-resistant enterococci) colonisation and infection, and adverse events.


Method of randomisation was not reported, although it was reported that the individual ICUs and BMT were the unit of randomisation. Not reported how allocation concealment was maintained to the point of randomisation. No loss to follow-up and analysis was by intention to treat analysis with Poisson regression to adjust for confounders. Trial was open label with investigators, staff, and participants aware of allocation. Patient characteristics not presented, although age and sex included as characteristics in adjusted analyses. Number of admissions, total days of care, total central catheter days, mean length of stay, and MRSA and VRE prevalence similar on the two treatment periods. No evidence that participants were not treated equally. Overall the methodological quality of the study was reasonable.


There were significantly fewer infections during the chlorhexidine-bathed periods (see table), including less MDRO acquisition, less VRE acquisition, fewer HA-BSI infections, and fewer central catheter-associated bloodstream infections (CC-BSI), and fewer adverse skin reactions (2.0 per cent versus 3.4 per cent, absolute difference 1.4 per cent, 95 per cent confidence interval 0.7 to 2.1 per cent). There was no significant difference in rates of MRSA acquisition.



Treatment with chlorhexidine was interrupted by a product recall and data on infections during this period were censored from the survival analysis. A sensitivity test was conducted to ensure this did not bias the findings – incidence of infection was still lower in the chlorhexidine group (4.78 versus 6.32 cases/1000 bed-days, p=0.02) with inclusion of data from the recall period.

Originally, 12 units were recruited into the study, but one unit withdrew from the study and data from two other units were excluded because of low compliance with the protocol.

Cost-effectiveness has not been reported (although may yet be undertaken); however, any cost-effectiveness analysis would compare rates under the two conditions, which may not be applicable if the control bathing approach is not being used in local units. Local cost-benefit analyses may be needed.

It is not apparent whether the participating units had already implemented the central line bundles and whether this study represents potential for added benefit. ✚Reviewer: Dr Andrew Jull, RN PhD, Associate Professor, University of Auckland & Nurse Advisor – Quality, Auckland District Health Board.