Objective: Update a 2011 review of differences in accuracy of diagnostic tests and the effects of interventions to prevent and treat Clostridium difficile infection (CDI) in adults.
Data sources: Medline®, the Cochrane Clinical Trials Registry, and Embase® from 2010 through April 2015 plus reference lists of included studies and recent systematic reviews.
Methods: Two investigators screened abstracts and full texts of identified references for eligibility. Eligible studies included studies of sensitivity and specificity for diagnostic tests in patients at risk for CDI. We included randomized controlled trials or high-quality cohort studies enrolling adult patients with CDI or suspected CDI for treatment interventions. Prevention studies also included adult patients at risk for CDI and observational study designs. Two investigators extracted data, assessed individual study risk of bias, and evaluated the strength of evidence for each comparison and outcome. Pooled estimates were analyzed to assess the efficacy and comparative effectiveness of a variety of treatments.
Results: We identified 37 diagnostic studies and 56 studies evaluating prevention or treatment interventions to update the review. High-strength evidence showed that nucleic amplification tests were sensitive and specific for CDI when using culture as the reference standard. Low-strength evidence was found that some institutional prevention interventions, such as antibiotic prescribing practices and transmission interruption (terminal room cleaning with hydrogen peroxide vapor and handwashing campaigns), reduce CDI incidence. Low-strength evidence also suggested that prevention programs can be sustained over several years. For CDI treatment, vancomycin is more effective than metronidazole (high-strength evidence), and the effect does not vary by severity (moderate-strength evidence). Fidaxomicin remains noninferior to vancomycin for the initial cure of CDI (moderate-strength evidence) but is superior to vancomycin for prevention of recurrent CDI (now high-strength evidence). Although both fecal microbiota transplantation (FMT) and probiotics were the subject of a significant number of new studies, the overall high risk of bias of many of these studies necessitated ratings of low strength of evidence. Specifically, low-strength evidence suggests that FMT may have a significant effect on reducing recurrent CDI. Similarly, low-strength evidence suggests that lactobaccilus strains and multiorganism probiotics also can reduce recurrent CDI. However, Saccharomyces boulardii was no more effective than placebo in preventing recurrent CDI. Evidence for FMT for refractory CDI was insufficient. Few studies reported adverse events; when reported, few events were noted.
Conclusions: Research on diagnostic testing for and interventions to treat CDI expanded considerably in 4 years. Nucleic acid amplification tests have high sensitivity and specificity for CDI. Vancomycin is more effective than metronidazole for initial CDI, while fidaxomicin is more effective than vancomycin for the prevention of recurrent CDI. FMT and lactobacillus probiotics to restore colonic biodiversity and improve patient resistance to CDI or recurrence have low-strength but relatively consistent positive evidence for efficacy.