The authors concluded that there was a lack of rigorous evidence linking specific hand hygiene interventions with the prevention of health care-associated infections. This was a reasonably well-conducted review and the authors' cautious conclusion reflected the evidence, but the poor quality of included studies and potential for language bias mean that some caution is required when interpreting its reliability.

To assess the relationship between hand hygiene interventions and the incidence of health care-associated infections in acute care and long-term care settings.

PubMed, CINAHL, EMBASE and the Cochrane Library were searched for English-language studies from 1996 to July 2006; search terms were reported. Additional studies were sought through consultation with experts.

Experimental studies (of any design) that assessed the relationship between hand hygiene practices and health care-associated infections were eligible for inclusion. Eligible studies has to be carried out in acute or long-term care facilities in developed countries. Studies that examined hand hygiene compliance, efficacy of alcohol hand gels, plain soap, and antimicrobial soap in reducing bacteria count recovered from hands were excluded.

Outcomes comprised rates of health care-associated infections, methicillin-resistant Staphylococcus aureus (MRSA), vancomycin-resistant Enterococcus, bloodstream infections, rotavirus, surgical infections and Clostridium difficile.

In the included studies, interventions included: multimodal hand hygiene initiatives; introduction of a hand hygiene product; implementation or modifications of infection control practices; implementation of infection control policies; administration intervention; electronic monitoring; and survey of current practices. At least two health care-associated infections were examined in three-quarters of the studies. Over half of the studies were conducted in the USA.

Two reviewers selected studies for inclusion in the review, with discrepancies resolved through discussion.

Two reviewers independently assessed study quality using Larson's 2005 scoring tool, where fatal methodological flaws included: inadequate sample size; uncontrolled bias or confounding; unclear operational definitions or description of intervention; inadequate statistical analysis; or lack of evidence that the intervention was implemented. Studies with a score of 1 had at least one fatal flaw in the study, studies with a score of 2 had had at least one other (nonfatal) flaw, but no fatal flaws; studies that scored 3 had no fatal or other flaws; and studies that scored 4 were randomised controlled trials (RCTs) with no fatal or other flaws.

Discrepancies were resolved through discussion between three reviewers.

Differences in rates of health care-associated infections were extracted.

The authors did not state how many reviewers performed the data extraction.

The studies were combined in a narrative synthesis, organised by type of hand hygiene intervention.

Thirty-one studies were included in the review (n unknown): two were randomised trials, four were cohort studies with a control group, three were cohort studies with no control group, four were before-and-after studies with a control group, and 18 were before-and-after studies without control groups. Eleven studies had a quality score of 1, 10 studies scored 2, nine studies scored 3, and one scored 4. The most frequent flaws were: failure to identify the presence of confounding factors (eight studies); inadequate statistical analysis (six studies); and inadequate sample size (five studies).

Multimodal hand hygiene initiatives (seven studies): Two studies reported a significant reduction in health care-associated infections rates and two studies did not report a significant reduction in these rates. Two studies reported a significant reduction in methicillin-resistant Staphylococcus aureus (MRSA) rates. One study found a reduction in rotavirus rate.

Hand hygiene products (11 studies): A decrease in infection rates following the introduction of a hand sanitiser was reported in three studies, but two studies found no significant reduction in health care-associated infections rates between handwashing and the introduction of a hand sanitiser. Other studies reported a decrease or a significant reduction in: MRSA with or without reductions in vancomycin resistant Enterococcus (three studies); and infection rates in both active intervention groups (one study). One study reported no a significant difference between two active interventions.

Infection control practices (eight studies): A reduction in health care-associated infections rates were reported in two studies, but it was not possible to assess the role of hand hygiene in these studies. Other studies reported a decrease or a significant reduction in: bloodstream infections (one study); bacteraemia (one study); MRSA (two studies); vancomycin resistant Enterococcus (one study); and Clostridium difficile (one study).

Infection control policies (two studies): Significant reductions were noted for Clostridium difficile and health care-associated infections (one study), and the risk of hospital-associated gastrointestinal infections (one study).

Other interventions (three studies): Three studies assessed organisational interventions (administration intervention, electronic monitoring and a survey of current practices), but none reported significant reductions in infections.

There was a lack of rigorous evidence linking specific hand hygiene interventions with the prevention of health care-associated infections. The varied nature of the interventions used, and the diverse factors affecting the acquisition of health care-associated infections, made it difficult to show the specific effect of hand hygiene alone.

The review question was clear. Detailed inclusion criteria were stated for all aspects, although criteria for outcomes were broad. The search strategy included some relevant sources. No attempts were made to minimise language bias. There was no indication that unpublished material was sought. Study selection and quality assessment were carried in out duplicate, which minimised the potential for reviewer error and bias, but it was unclear whether this extended to data extraction.

Limited study details were provided. The chosen method of synthesis appeared to be appropriate. The restriction to studies conducted in developed countries mean that the findings may not be generalisable to settings elsewhere.

The authors' cautious conclusion reflected the evidence, but the poor quality of included studies and potential for language bias mean that some caution is required when interpreting its reliability.

Practice: The authors did not state any implications for practice.

Research: The authors stated that a modified version of Larson's recommendations should form the basis for the design and dissemination of future studies. Studies should consist of multi-layered hand hygiene interventions, with comparable control groups, to assess which combination of interventions are cost-effective for reducing health care-associated infections.

Not stated.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.