P191 Systematic review and meta-analysis: Patient-reported outcomes and endoscopic appearance in ulcerative colitis

To evaluate the association of the patient-reported outcomes rectal bleeding and stool frequency with mucosal healing in ulcerative colitis (UC). A systematic review of studies reporting the association of patient-reported outcomes and mucosal healing was conducted. Diagnostic accuracy meta-analysis was performed using the hierarchical bivariate method. Five studies were included with a total of 2132 participants. For rectal bleeding subscore of 0, the pooled sensitivity was 81% (95% confidence interval (CI): 73–86%), specificity 68% (95% CI: 61–75%), positive likelihood ratio (LR) 2.5 (95% CI: 2.2–3.0), and negative LR 0.28 (95% CI 0.22–0.37). For stool frequency subscore of 0, the pooled sensitivity was 40% (95% CI: 25–58%), specificity 93% (95% CI: 86–97%), positive LR 6.0 (95% CI: 3.7–9.7), and negative LR 0.64 (95% CI 0.50–0.82). For combined rectal bleeding and stool frequency subscores of 0, the pooled sensitivity was 36% (95% CI: 22–54%), specificity 96% (95% CI: 91–98%), positive LR 8.4 (95% CI: 5.5–12.8), and negative LR 0.66 (95% CI 0.53–0.84). Forest plots of coupled sensitivity and specificity of combined rectal bleeding and stool frequency subscores = 0 for mucosal healing UC patients with normal rectal bleeding and stool frequency subscores likely have attained mucosal healing. Rectal bleeding is often absent in those with mucosal healing. Normal stool frequency predicts mucosal healing, but often remains abnormal in patients despite mucosal healing.