Comparative effectiveness of different therapies for infection in adults: a systematic review and network meta-analysis of randomized controlled trials.

Background: infection (CDI) is a leading cause of healthcare-associated diarrhea, with substantial morbidity and mortality. CDI is a severe and growing problem with numerous treatment options. We evaluated the effectiveness of all therapies in recurrent and non-recurrent infections and their prevention.

Methods: This network meta-analysis and systematic review of randomized controlled trials (RCTs) compared all CDI therapies and preventions. We included RCTs published until 19 August 2024 and focused on adult population. We performed a systematic search in MEDLINE, EMBASE, and Cochrane Central Register of Controlled Trials. Inclusion criteria were patients: adults (>16) treated against CDI; study type: randomized controlled trial; outcome: cure rate, recurrence or effectiveness of prevention. Any publication not meeting all criteria was considered to be ineligible and excluded. We applied random-effects meta-analysis using frequentist methods. We reported our main results as odds ratios (as a symmetric effect size measure, OR) with 95% confidence interval (95% CI). We used the Cochrane risk-of-bias tool to assess the risk of bias. Our study protocol was preregistered in PROSPERO (CRD42022371210).

Findings: We assessed 73 RCTs with 28 interventions, involving 27,959 patients (49.2% female) in five networks. Fecal microbiota transplantation (FMT) was the most effective treatment in terms of the cure rate overall (P-score: 0.9952) and in recurrent cases (P-score: 0.9836). In recurrent cases, fidaxomicin (P-score: 0.6734) showed significantly greater effectiveness than vancomycin (P-score: 0.3677) and tolevamer (P-score: 0.0365). For non-recurrent CDI treatments ridinilazole, fidaxomicin, FMT and nitazoxanide were equally effective. Ridinilazole (P-score: 0.7671) and fidaxomicin (P-score: 0.7627) emerged as the most effective in preventing recurrence. Probiotics were not effective in preventing CDI, since network meta-analyses did not show significant differences between probiotics and placebo. In probiotics' subgroups pairwise meta-analyses proved to be significantly more effective in prevention than placebo. Oral and colonoscopic FMT administration methods were equally effective. The study-level aggregated risk of bias of the publications included ranged from low to high. We observed relevant heterogeneity among studies in therapeutic doses, treatment durations, and follow-up times.

Interpretation: The superiority of FMT in the treatment of CDI highlights the potential for increased use of FMT in clinical settings. Further research on optimizing FMT protocols and exploring its long-term safety and efficacy in larger samples is needed. Our findings suggest that the preventive use of probiotics might be questioned.

Funding: None.