Twenty-five studies were included in the analysis. The number of participants was not reported. Eighteen studies reported adequate sequence generation. Fifteen studies reported adequate concealment of allocation. Twenty-one studies reported blinding. Twenty-four studies addressed incomplete outcome data and freedom from selective reporting.
Mesalamine: Compared to placebo, mesalamine was significantly more effective in preventing clinical recurrence of Crohn’s disease (WMD 8.8%, 95% CI 2.0 to 15.7%; four studies; no significant heterogeneity). There were no significant differences between mesalamine and placebo for endoscopic recurrence (five studies; evidence of statistical heterogeneity p<0.001). Surgical recurrence was not evaluated in controlled studies.
Metronidazole/ornidazole: Compared with placebo, nitroimidazolic antibiotics were more effective in preventing postoperative clinical recurrence of Crohn’s disease at one year follow-up (WMD 24.2%, 95% CI 11.6% to 36.9%; two studies) but there were no significant differences between groups at three-year follow-up. Endoscopic recurrence was significantly reduced at three-month follow-up for nitroimidazolic antibiotics compared with placebo (WMD 23.7%, 95% CI 6.4% to 41.4%; two studies). There was no evidence of statistical heterogeneity for these analyses.
Thiopurines: At one-year follow-up there were no significant differences between thiopurines and placebo or mesalamine for clinical recurrence of Crohn’s disease (four studies). At two-year follow-up thiopurines were significantly more effective than placebo or mesalamine for clinical recurrence (WMD 13.1%, 95% CI 2.3% to 23.9%; two studies). Endoscopic recurrence rates were significantly reduced at one year in the thiopurine group compared to placebo (WMD 19.7, 95% CI 8.4% to 31%; three studies). There was no evidence of statistical heterogeneity for these analyses.
Infliximab: One study with 24 patients reported that infliximab had significantly lower endoscopic recurrence rates (9.1%) than placebo (84.5%) at one year.
Compared to placebo, budenoside, probiotics, interleukin-10 or any surgical procedure did not generally demonstrate any significant difference in postoperative recurrence rates (where reported) in Crohn’s disease.
Results for individual studies were reported.