The authors concluded that antimicrobial stewardship, glove use, hand hygiene and disposable thermometers should be used routinely for prevention of Clostridium difficile infection. The authors' conclusions appear appropriate, but a lack of validity assessment and results that were based on a small number of studies make the reliability of the conclusions uncertain.

To evaluate the efficacy of interventions for the prevention of health-care associated Clostridium difficile infection (CDI).

PubMed, CINAHL, Cochrane Database of Systematic Reviews, ACP Journal Club, DARE, Cochrane Central Register of Controlled Trials (CENTRAL) and ClinicalTrials.gov were searched up to May 2009 without language restrictions. Search terms were reported. Reference lists of relevant articles and four conference abstract proceedings were scanned for additional studies.

Controlled trials that evaluated the efficacy of interventions aimed at preventing CDI in hospital settings were eligible for inclusion. The primary outcome of interest was incidence of endemic CDI. Only studies that focused on endemic CDI (prevalence in non-outbreak settings) were eligible for inclusion. Studies had to provide a risk or rate ratio or provide data to calculate measures. For studies of probiotics, only those with a randomised controlled trial (RCT) design were included.

Nearly half of the included studies evaluated antibiotic stewardship (restriction of selective antimicrobial or antibiotic use). Other studies evaluated probiotics, disinfectants, hand hygiene, glove use, thermometers and carriage of Clostridium difficile toxin (treatment of asymptomatic carriers). Most of the included studies diagnosed CDI using a positive toxin assay. Definitions differed between studies. It appeared that several studies used multiple interventions concurrently.

The authors did not state how many reviewers selected studies for inclusion.

No formal validity assessment was conducted. The authors based recommendations on guidelines of the Infectious Disease Society of America's United States Public Health Service grading system.

Data on incidence of Clostridium difficile were extracted by three reviewers. Disagreements were resolved by discussion with a fourth reviewer.

Data were grouped by type of intervention and reported in a narrative synthesis.

Thirty-five studies were included in the review (number of participants not reported): 10 randomised controlled trials (RCTs), 22 non-randomised controlled trials (non-RCTs) and three uncontrolled studies.

Probiotics (eight RCTs): One RCT reported a statistically significant decrease in CDI with use of a combination probiotic (Lactobacillus casei, Lactobacillus bulgaricus and Streptococcus thermophilus) milkshake (p=0.001). Seven RCTs found no significant differences between groups for incidence of CDI after use of probiotics (Lactobacillus acidophilus, Bifidobacterium bifidum, Lactobacillus GG and Saccharomyces boulardii).

Antimicrobial stewardship (17 studies): Two non-RCTs (before-and-after studies) found decreased rates of CDI after clindamycin restriction. Eight non-RCTs reported that cephalosporin restriction reduced incidence of CDI. Four of the trials noted that increased usage of alternative agents did not result in any adverse consequences such as increase in infections caused by other pathogens. Seven non-randomised trials reported deceased rates of CDI after restriction of more than one antibiotic class.

Hand hygiene (three studies): Mixed evidence was identified for alcohol-based hand rubs. One non-RCT reported a reduction in incidence of CDI after implementation of the intervention compared to an historical control group (p=0.02) and one non-RCT found no differences between groups. One non-randomised controlled trial found no significant differences between a vinyl glove use group and control group for CDI incidence.

Medical apparatus (two studies): Two studies reported positive effects for use of disposable rectal thermometers. An uncontrolled study reporting a significant decrease in incidence of CDI six months after the intervention (p<0.01). A randomised crossover trial reporting a significant decrease of CDI for the intervention group compared to control group (p=0.026).

Environmental disinfection (four studies): One non-RCT reported a reduction in CDI for hypochlorite solution for bone marrow transplantation units, but not general medicine service and neurosurgical intensive care unit. One non-RCT reported a significant decrease in incidence of CDI for hydrochloride disinfectant compared with control group (p<0.05). One study reported no significant differences between groups for education compared to education and bleach. One uncontrolled study reported a significant decrease for CDI for hydrogen peroxide vapour (p=0.047).

Treatment of asymptomatic carriers (two studies): Two trials (one RCT) reported mixed effects for the efficacy of treating patients with asymptomatic carriage of Clostridium difficile for prevention of endemic CDI.

One uncontrolled study reported that use of disposable rather than electronic rectal thermometers would cost approximately an additional US$7,000 per year. Another randomised cross-over trial estimated an additional cost of $5,926 per CDI prevented for single-use disposable thermometers compared with electronic thermometers. Three trials estimated cost savings as a result of antimicrobial stewardship that ranged from $114,000 to £212,000.

Antimicrobial stewardship, glove use, hand hygiene and disposable thermometers should be used routinely for prevention of Clostridium difficile infection.

The review question was clear with broadly defined inclusion criteria. Several relevant sources were searched. Appropriate efforts were made to reduce language and publication biases. Appropriate methods to reduce reviewer error and bias were reported for data extraction of data; it was unclear whether similar methods were used for study selection. The authors noted that multiple interventions were used concurrently for many studies and this made the independent role of each preventive strategy difficult to determine. Only a few of the included studies were RCTs and the authors stated that confounding and selection bias was a concern. Some results for individual studies were reported in tables. There was no indication of the number of participants. There appeared to be a discrepancy in study numbers between text, flow chart and results tables. Some of the authors' recommendations were based on a small number of included studies. There appeared to be a lack of evidence and conflicting evidence relating to hand hygiene/glove use. No formal validity assessment was conducted and so the results of the studies and any synthesis may not have been reliable.

The authors' conclusions appear appropriate, but a lack of validity assessment and results that were based on a small number of studies make the reliability of the conclusions uncertain.

Practice: The authors stated that antimicrobial stewardship, glove use, hand hygiene and disposable thermometers should be used routinely for prevention of Clostridium difficile infection. Treatment of asymptomatic carriage of Clostridium difficile was not recommended. There was insufficient evidence to recommend for or against use of probiotics.

Research: The authors stated that further research (particularly RCTs) of interventions of environmental disinfection and probiotics were required to evaluate their roles in prevention of CDI.

No direct funding reported.

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.