Background: Approximately half of the patients with ulcerative colitis (UC) who undergo restorative proctocolectomy develop pouchitis within 10 years of surgery. Currently, there are no approved pouchitis treatments. It is important to quantify, and ultimately minimize, placebo rates to design and conduct efficient pouchitis trials.
Aims: To quantify the placebo rate observed in pouchitis randomized controlled trials (RCTs) in meta-analysis.
Methods: Embase, MEDLINE, and the Cochrane Library were searched from inception to November 3, 2017, for placebo-controlled RCTs enrolling adult UC patients with, or at risk for developing, pouchitis. A fixed-effect binomial-normal model was used to pool placebo rates on the log-odds (logit) scale. Proportions and 95% confidence intervals were reported. Outcomes of interest included development of pouchitis, induction of remission/response, and maintenance of remission/response. The Cochrane risk of bias tool was used to evaluate study quality.
Results: Twelve trials (five prevention, five induction, and two maintenance) enrolling a total of 229 placebo patients were eligible for inclusion. The pooled placebo rates for development of pouchitis and induction of response were 47% (95% CI 39-56%) and 24% (95% CI 14-37%), respectively. An insufficient number of trials prevented additional data pooling and meta-regression analysis and no consistent definitions of outcome were identified.
Conclusions: No consistent methods for measuring pouchitis disease activity or defining response and remission were identified, highlighting the need for standardized definitions of outcomes for use in pouchitis trials. Additional high-quality trials are required to evaluate existing and novel therapies in this area.
Keywords: Placebo effect; Pouchitis; Trial design; Ulcerative colitis.