P577 “Treat to target” recommendations in ulcerative colitis in practice: clinician perceptions and potential barriers

Bryant R.^*1,2, Costello S.^1,3, Schoeman S.⁴, Sathananthan D.⁴, Lau S.³, Schoeman M.⁴, Mountifield R.^1,5, Tee D.⁶, Travis S.^7,8, Andrews J.^1,9

¹University of Adelaide, School of Medicine, Adelaide, Australia ²Royal Adelaide Hospital, Department of Gastroenterology and Hepatology, Adelaide, Australia ³Queen Elizabeth Hospital, Department of Gastroenterology, Adelaide, Australia ⁴Royal Adelaide Hospital, Department of Gastroenterology, Adelaide, Australia ⁵Flinders Medical Centre, Department of Gastroenterology, Adelaide, Australia ⁶Lyell McEwin Hospital, Department of Gastroenterology, Adelaide, Australia ⁷Oxford University Hospital NHS Trust - John Radcliffe Hospital, Department of Transitional Gastroenterology, Oxford, United Kingdom ⁸University of Oxford, Oxford, United Kingdom ⁹Royal Adelaide Hospital, Department of Gastroenteorlogy and Hepatology, Adelaide, Australia

Background

A “Treat to Target” (T2T) approach has been proposed for ulcerative colitis (UC), with a target of combined clinical and endoscopic remission. This has yet to be evaluated in real-world care.

Methods

A multicentre, retrospective, cross-sectional review of patients with UC attending outpatient services in South Australia between Jul 2013 and Nov 2015 was conducted. Clinician assessment of disease activity and objective assessment (endoscopy, histology, and/or biomarkers) were recorded. An on-line survey of local Gastroenterologists evaluated their perceptions of T2T in UC. Multivariate logistic regression, Fisher's exact tests and Kappa statistics were performed.

Results

Of 246 patients with UC, 61% were documented to be in clinical remission (normal bowel habit and no rectal bleeding), 35% documented to be in clinical and endoscopic remission (Mayo endoscopic sub-score ≤1), and 16% documented to be in concordant clinical, endoscopic and histological (Truelove and Richards' Index) remission.

Figure 1. UC treatment targets achieved in real-world practice. *Clinical remission: normal stool frequency and absence of rectal bleeding. Endoscopic remission: Mayo endoscopic sub-scores of ≤1. Histological remission: Truelove and Richards Index remission.

Table 1. UC treatment targets achieved in real-world practice. Overall proportions of patients in UC cohort (n=246) documented to attaining remission

Rather than disease-related factors (extent/activity), clinician-related factors dominated outcome. The hospital location at which care was delivered and choice of therapy predicted combined clinical and endoscopic remission (OR 3.6, 95% CI 1.6–8.7, p<0.001; OR 3.3, 95% CI 1.1–12.5, p=0.04, respectively)

Table 2. Clinical factors associated with documentation of combined clinical and endoscopic remission. Logistic regression analyses, n=218 (endoscopy during follow-up).*p<0.05 statistical significance; ^Any therapy analysed in a separate model

Clinicians used C-reactive protein (CRP) more often than endoscopy as a biomarker for disease activity (75% vs 47%, p<0.001), despite observed random discordance between CRP and endoscopy (kappa 0.13, 95% CI 0.0–0.27). In the survey, 45/61 Gastroenterologists responded, with significant disparity between clinician estimates of targets achieved in practice and real-world data (p<0.001 for clinical and endoscopic remission)

Figure 2. Clinician reported achievement of treatment targets in UC vs. real world data. Actual vs. expected proportions compared using a Fisher's exact test. ***p value ≤0.001.

Conclusion

Most patients with UC do not achieve composite clinical and endoscopic remission in real-world practice. Clinicians overestimate their achievements and practice behaviour is a barrier to achieving stipulated targets.

Posted in: Poster presentations: Clinical: Therapy and observation (2017)