P241 Early prediction of outcome in acute severe ulcerative colitis (ASC): comparison of clinical and endoscopic indices at presentation, with day 3 criteria

V. Gupta1, W. Mohsen1, A. Adams1, A. Murphy1, D. Sathananthan1, S. Cripps1, O. Brain1, R. Palmer1, T. Ambrose1, B. George2, A. Walsh1, S. Travis1, J. Satsangi1

1John Radcliffe Hospital, Translational Gastroenterology Unit, Oxford, UK, 2John Radcliffe Hospital, Department of colorectal surgery, Oxford, UK


The management of ASC needs early characterisation of factors predictive of outcome to allow appropriate patient counselling and stratification for second-line therapy or surgery. Previous data suggest that the number of additional Truelove and Witts’ (TW) criteria (fever, tachycardia, anaemia or CRP elevation) on admission may predict response to medical therapy. We compared the predictive accuracy of these clinical criteria on admission with a validated endoscopic scoring system (UCEIS), and with accepted day 3 criteria.


Cases of ASC were retrospectively evaluated. The number of TW criteria, UCEIS, inpatient medical therapy, same admission outcome and follow up were recorded. Pre-specified endpoints included rescue therapy, colectomy during the same admission and colectomy within 1 year of follow up.


Consecutive 131 admissions (117 patients) between 2015–9 were analysed. All satisfied modified TW definition of ASUC. Sixty-eight patients (58%) were female,index presentation 38 (29%),median age at presentation 40 years (16–76),median disease duration 1 year (1–43), median follow up 23 months (1–49).Seventy-one (54%) received rescue therapy (ciclosporin 35/71 and anti-TNF 36/71).Colectomy rates were 16% (19/117) during same admission and 26% (30/117) within 1 year of follow up.Outcomes were stratified according to UCEIS score, additional TW criteria on day 0,and D3 response (Oxford criteria) (Table 1).UCEIS score ³7 predicted higher need for rescue therapy (Chi square,p = 0.01) but not colectomy during same admission (p = 0.68) or within 1 year (p = 0.41).In logistic regression analysis,UCEIS predicted rescue therapy (p = 0.01) but not colectomy during same admission (p = 0.68) or within 1 year(p = 0.55); whereas day 0 TW criteria predicted need for rescue therapy (p = 0.02),colectomy during admission (p = 0.04) and within 1 year (p = 0.03). D3 response predicted colectomy during same admission (p = 0.001) and within 1 year(p = 0.0002).

Table 1.
UCEIS3(n = 5)4(n = 11)5(n = 37)6(n = 40)7(n = 30)8(n = 8)
Rescue therapy15(46%)18(49%)19(48%)21(70%)7(88%)
ColectomySame admission02(18%)7(19%)6(15%)3(10%)1(13%)
Colectomy (1 yr)03(27%)10(27%)6(15%)9(30%)2(25%)
TW criteria1(n = 43)2(n = 62)3(n = 25)4(n = 1)
Rescue therapy17(40%)37(60%)16(64%)1(100%)
Colectomy same admission3(7%)9(15%)7(28%)-
Colectomy (1yr)5(12%)16(26%)9(36%)-
D3 responseComplete responder(n = 21)Partial responder(n = 54)Non-Responder(n = 56)
Colectomy same admission03(6%)16(29%)
Colectomy(1 yr)08(15%)22(39%)


Clinical criteria assessed by D3 response are the strongest predictors of colectomy on that admission or within 1 year.TW criteria on presentation predict colectomy within 1 year of follow up.UCEIS predicts the use of rescue therapy but does not predict colectomy rates.