What question did this study set out to answer?

To validate and compare the performance of Day 1 scores (ADMIT-ASC, ACE, ROR) for predicting intravenous corticosteroid failure in acute severe ulcerative colitis.

January 23, 2026

P0191Head-to-head comparison and validation of current day 1 scores for predicting intravenous corticosteroid failure in acute severe ulcerative colitis

Key Points

To validate and compare the performance of Day 1 scores (ADMIT-ASC, ACE, ROR) for predicting intravenous corticosteroid failure in acute severe ulcerative colitis.
Retrospective analysis of adult patients with acute severe ulcerative colitis.
Data collected from three tertiary IBD centres in Australia from January 2018 to December 2023.
Primary outcome was intravenous corticosteroid failure defined by escalation to medical rescue therapy or colectomy.
Included 238 patients, 44.5% were female, and 56.3% required medical rescue therapy.
ROR score had the highest AUROC (0.72), followed by ADMIT-ASC (0.68) and ACE (0.66), but none were statistically significant.
ADMIT-ASC and ACE showed high specificity but low sensitivity, while ROR had high sensitivity with lower specificity.

Abstract

Abstract Background Early identification of patients with acute severe ulcerative colitis (ASUC) likely to fail intravenous corticosteroids (IVCS) can expedite the need for medical rescue therapy (MRT) and improve patient outcomes. We aimed to validate and directly compare the performance of three available Day 1 scores, ADMIT-ASC, ACE, and ROR, in a contemporary cohort. Methods We retrospectively analysed consecutive adult patients admitted with ASUC, as defined by the Truelove Witts criteria, to three tertiary IBD centres in Australia between January 2018 to December 2023. Clinical, biochemical and endoscopic data were collected. The primary outcome was IVCS failure, defined as escalation to MRT or colectomy during the index admission. The secondary outcomes aimed to validate and compare the ADMIT-ASC, ACE and ROR scores. Dichotomised performances were evaluated using published cut-off points of the risk scores using the AUROC, sensitivity, specificity, positive and negative predictive values. Results We included 238 patients, 106 (44.5%) were female and 134 (56.3%) required MRT. The AUROC was highest for ROR (0.72, 95% CI 0.66-0.79), followed by ADMIT-ASC (0.68, 95% CI 0.62-0.74) and ACE (0.66, 95% CI 0.60-0.73), however there was no statistical significance (p = 0.17). At published cut-off points, ADMIT-ASC ≥3 and ACE =3 had high specificity (90% and 89% respectively) but low sensitivity (22% and 23% respectively), whereas ROR ≥50 had high sensitivity (82%) but lower specificity (49%). Conclusion All three scores showed moderate discrimination for IVCS failure. ROR provides higher net benefit across typical decision thresholds, while ADMIT-ASC and ACE offered higher specificity when ruling in high-risk patients. Conflict of interest: Dr. Subhaharan, Deloshaan: No conflict of interest Jones, Mark: No conflict of interest Mohsen, Waled: No conflict of interest Kakkadasam Ramaswamy, Pradeep: None

Mark Helpful

Bookmark

Relay