Severe ulcerative colitis: prospective study of parameters determining outcome.

BACKGROUND: One-third of patients with exacerbations of ulcerative colitis (UC) require colectomy and 1% die. Although several studies have reported parameters associated with failure of medical treatment, studies using multivariate analysis of various predictors are scanty.
METHODS: Demographic, clinical and laboratory parameters within 48 h of admission of 50 consecutive patients (55 episodes) were prospectively recorded. Each episode was treated with intravenous hydrocortisone, antibiotics and fluid administration. The data of patients who failed to respond, deteriorated, developed complications requiring emergency surgery or died within 7 days of hospitalization, were compared with the data of those who responded during this period using Mann-Whitney U-test or chi-squared tests. Variables found to be significant on univariate analysis were entered into a multivariate model.
RESULTS: Medical treatment failed in 10 of 55 episodes; eight had surgery (one died), another died with bowel perforation and peritonitis and one left hospital in a moribund state. Younger age, pedal edema, transverse colonic diameter >5 cm on abdominal radiograph, low hemoglobin, low serum albumin, high C-reactive protein, presence of fibrin degradation product (FDP) in serum, low serum fibrinogen, prolonged prothrombin time, Truelove-Witt's fulminant disease, and high clinical activity index were associated with failure of medical treatment on univariate analysis. All the above variables were significant by multivariate analysis except age, serum albumin level and presence of FDP.
CONCLUSION: It may be possible to predict the failure of medical treatment in exacerbations of UC using a scoring system based on clinical parameters and simple laboratory tests.