Search in the Abstract Database

Abstracts Search 2019

P567 Could an escalation of therapy or intervention (ETI) calculator be used to triage appointments for patients with ulcerative colitis?

A. Walsh*1, L. Matini1, R. Kantschuster1, M. Lepetyukh1, R. Nedescu1, J. Wilson1, O. Brain1, R. Palmer1, S. Keshav1, S. Travis1

1John Radcliffe Hospital, Translational Gastroenterology Unit, Oxford, UK


Conventional follow-up for ulcerative colitis (UC) places demands on health services. Demand might be better managed by targeting appointments at those patients who need therapeutic decisions. The aim was to further validate the TrueColours UC (TCUC) Escalation of Therapy or Intervention (ETI) calculator in an IBD outpatient clinic setting.


TCUC is a comprehensive, web-based program that works through email prompts, allowing patients to enter disease-specific data. In previous work, a logistic regression model using the SCCAI (Simple Clinical Colitis Activity Index) and IBD Control-8 (quality of life), collected through TCUC, was used to create an ETI calculator. This calculator produces a probability of escalation of therapy or other intervention at an outpatient appointment (OPA). See Figure 1.

Figure 1. UC Escalation of Therapy Calculator for SCCAI and IBD Control-8

From June to November 2018, a sample of 207 patients with UC under standard hospital follow-up, actively using TCUC was examined. For each OPA, the probability of escalation was calculated using their most recent SCCAI and IBD-Control reports. Clinic letters were assessed for the outcome of escalation of therapy or intervention.


Of the 207 patients, 48 had a total of 53 OPAs over the 6-month period. Most, 33/53 (62%), OPAs resulted in no treatment escalation; 16/53 (30%) had escalation and 4/53 (8%) had de-escalation of therapy. De-escalation included stopping 5-ASA suppositories (n = 1), prednisolone enemas (n = 1), mycophenolate (n = 1), or methotrexate (n = 1). By setting the threshold for a timely OPD at a 5% estimated probability of treatment escalation, 13/16 (81%) escalation events would have been correctly identified. Of the 3 patients that would have been missed, the escalations involved increasing the dose of oral 5-ASA (n = 2) or flexible sigmoidoscopy (n = 1). By setting the estimated probability of escalation at 25%, only 9/16 (56%) would have been correctly identified. Of the 4 further patients that would have been missed, the escalation events involved starting topical therapy (n = 2), increasing the dose of azathioprine (n = 1), or increasing the frequency of vedolizumab (n = 1).


Models that predict the likelihood of the need for escalation of therapy or intervention during an outpatient appointment, based on remotely collected PROMs, have the potential to improve outpatient clinic resource utilisation. Using the ETI calculator, up to 62% of planned outpatient appointments could have been deferred if the agreed threshold for an appointment was a 5% chance of treatment escalation or intervention.