In a prospective study of children with severe ulcerative colitis (UC), we aimed to assess outcomes and to identify predictors of nonresponse to intravenous corticosteroids.

A total of 128 children (47% males; 12.9 ± 3.9 y) hospitalized for severe UC were enrolled from 10 pediatric centers. Clinical and laboratory data and the Pediatric UC Activity Index (PUCAI) were recorded throughout the admission. Patients were followed up for 1 year postdischarge.

Thirty-seven (29%; 95% confidence interval [CI], 22%–37%) children failed intravenous corticosteroids and received, within 10.5 ± 6.4 days, cyclosporine (n = 1; 3%), colectomy (n = 3; 8%), or infliximab (n = 33; 89%). Several predictors were associated with intravenous corticosteroids failure, but the best model included number of stools, amount of blood, age, and new-onset disease (odds ratio [OR], 1.9; 95% CI, 1.1–3.5; OR, 2.5; 95% CI, 1.3–4.6; OR, 1.2; 95% CI, 1.04–1.36; and OR, 0.27; 95% CI, 0.1–0.7, respectively). The PUCAI, followed closely by the Travis rule, strongly predicted response when compared with other measures (Seo and Lindgren indices, C-reactive protein level, and fecal calprotectin level) (P < .001). Aiming for sensitivity on day 3, a PUCAI greater than 45 screened for patients likely to fail intravenous corticosteroids (negative predictive value, 94%; positive predictive value, 43%; P < .001). Aiming for specificity on day 5, a PUCAI score greater than 70 optimally guided implementation of salvage therapy (positive predictive value, 100%; negative predictive value, 79%; P < .001). Twenty-five of 33 children treated with infliximab responded. The overall cumulative colectomy rate was 9% and 19% by discharge and 1-year, respectively. The day 3 PUCAI score predicted response up to 1 year postdischarge (P < .001; time to salvage therapy).

The PUCAI, calculated on days 3 and 5 of steroid therapy, can identify patients requiring salvage therapy. Infliximab is an effective therapy in steroid-refractory pediatric UC.