BACKGROUND: Protocols for treating recurrent infection (rCDI) through faecal microbiota transplantation (FMT) are still not standardised. Our aim was to evaluate the efficacy of different FMT protocols for rCDI according to routes, number of infusions and infused material. METHODS: MEDLINE, Embase, SCOPUS, Web of Science and the Cochrane Library were searched through 31 May 2017. Studies offering multiple infusions if a single infusion failed to cure rCDI were included. Data were combined through a random effects meta-analysis. RESULTS: Fifteen studies (1150 subjects) were analysed. Multiple infusions increased efficacy rates overall (76% versus 93%) and in each route of delivery (duodenal delivery: 73% with single infusion versus 81% with multiple infusions; capsule: 80% versus 92%; colonoscopy: 78% versus 98% and enema: 56% versus 92%). Duodenal delivery and colonoscopy were associated, respectively, with lower efficacy rates ( = 0.039) and higher efficacy rates ( = 0.006) overall. Faecal amount ≤ 50 g ( = 0.006) and enema ( = 0.019) were associated with lower efficacy rates after a single infusion. The use of fresh or frozen faeces did not influence outcomes. CONCLUSIONS: Routes, number of infusions and faecal dosage may influence efficacy rates of FMT for rCDI. These findings could help to optimise FMT protocols in clinical practice.
