Control groups received matching placebo in all trials but 2; those 2 used usual care of no probiotics as the control. Common patient exclusions were pregnancy, immune system compromise, intensive care, a prosthetic heart valve, and pre-existing gastrointestinal disorders.
The risk for CDI was lower in the probiotic group (range 0%-11%) than in the control group (0%-40%) with no heterogeneity (I2 = 0.0%; P = .56) when the data were pooled from all 19 studies (relative risk [RR] = 0.42; 95% confidence interval [CI], 0.30-0.57). The median incidence of CDI in the control groups from all studies was 4%, which yielded a number needed to treat (NNT) of 43 (95% CI, 36-58).
The researchers examined the NNT at varying incidence rates. If the incidence of CDI was 1.2%, the NNT to prevent 1 case of CDI was 144, and if the incidence was 7.4%, the NNT was 23. Compared with control groups, there was a significant reduction in CDI if probiotics were started within 1 to 2 days of antibiotic initiation (RR = 0.32; 95% CI, 0.22-0.48), but not if they were started at 3 to 7 days (RR = 0.70; 95% CI, 0.40-1.2). There was no significant difference in adverse events (ie, cramping, nausea, fever, soft stools, flatulence, taste disturbance) between probiotic and control groups (14% vs 16%; P = .35).
WHAT’S NEW
Probiotics provide added benefit if taken sooner rather than later
This high-quality meta-analysis shows that administration of probiotics to hospitalized patients—particularly when started within 1 to 2 days of initiating antibiotic therapy—can prevent CDI.
CAVEATS
Findings do not apply to all patients; specific recommendations are lacking
Findings from this meta-analysis do not apply to patients who have an immunocompromising condition, are pregnant, have a prosthetic heart valve, have a pre-existing gastrointestinal disorder (eg, irritable bowel disease, pancreatitis), or require intensive care. In addition, specific recommendations as to the optimal probiotic species, dose, formulation, and duration of use cannot be made based on this meta-analysis. Lastly, findings from this study do not apply to patients treated with antibiotics in the ambulatory care setting.
Continue to: CHALLENGES TO IMPLEMENTATION