Introduction Between 2008 and 2014, the UK national audit of adult ulcerative colitis (UC) admissions revealed a fall in mortality from 1.54 to 0.75%, a rise in anti-TNF therapy in steroid non-responders from 12 to 43%, and a fall in emergency surgery from 12 to 11%.

Methods Using Hospital Episode Statistics, patients aged between 18 and 60 years coded with a first emergency admission with UC were identified. The influence of demographic factors, comorbidity and infused anti-TNF therapy on mortality, surgery and emergency readmissions was examined using multivariate logistic regression.

Results Between 2004 and 2014, 17,344 patients (47.5% female and mean age of 36 (IQR 26–45)) were identified. Mortality was 0.13% at 30 days, 0.17% in hospital and 0.55% within 1 year. During admission, 11.5% of patients had surgery (median time to surgery 6 days (IQR 1–17)) and 1.93% had infused anti-TNF therapy. Surgery during admission fell non-significantly from 12.4 to 11.7% between 2004 and 2014, but the fall in surgery within a year between 2004 and 2013 was significant (OR 0.65 (95% CI 0.52–0.83) p < 0.001). Anti-TNF therapy rose from 0.9 to 4.6% between 2006 and 2014. In-hospital and 1 year mortality fell from 0.25 and 0.69% in 2004 to 0.14 and 0.56% in 2014 but this was not statistically significant. Patients aged 35–60 had a higher in-hospital (3.69 (1.37–9.94) p 0.010) and 1 year mortality (2.68 (1.66–4.33) p < 0.001) than those aged 18–34. Increased comorbidity was associated with 30 day mortality (29.73 (9.89–89.41) p < 0.001) and non-white patients had a lower 1 year mortality (0.59 (0.38–0.92) p = 0.010). Females were less likely to have surgery during admission (0.67 (0.61–0.74) p < 0.001) or within 1 year (0.87 (0.79–0.96) p < 0.001), but gender was not associated with mortality. Patients aged 35–60 (1.17 (1.06–1.29) p 0.001) and those of non-white ethnicity (1.27 (1.12–1.42) p < 0.001) were more likely to have surgery during admission. Patients given anti-TNF therapy during admission were more likely to need surgery at the time (1.40 (1.03–1.89) p = 0.031) and within 1 year (1.44 (1.04–2.00) p = 0.030). Emergency readmissions within 30 days were associated with younger age (35–60 years 0.89 (0.81–0.97) p = 0.011) and increased comorbidity (1.78 (1.22–2.62) p = 0.003).

Conclusion For patients with a first emergency admission for UC, there was no change mortality between 2004 and 2014. Rates of anti-TNF therapy during emergency admission have increased and surgery decreased over time. Older men and non-white ethnicity were associated with surgery during admission and the use of anti-TNF agents was associated with an increased risk of surgery, likely reflecting severe colitis.

Disclosure of Interest J. Rees Grant/research support from: £10,000 from Merck Sharp and Dome (MSD) to contribute to costs of HES access, F. Evison: None Declared, J. Mytton: None Declared, P. Patel: None Declared, N. Trudgill: None Declared