Predictive value of Rockall score for rebleeding and mortality in patients with variceal bleeding.

OBJECTIVE: To validate Rockall scoring system for in-hospital rebleeding and mortality in cirrhotic patients with variceal bleed. STUDY DESIGN: Cohort type of case series. PLACE AND DURATION OF STUDY: It was carried out at the Department of Gastroenterology and Hepatology, Shaikh Zayed Hospital, Lahore, from March 2005 to March 2006. PATIENTS AND METHODS: All cirrhotic patients presenting with upper GI bleeding and later found to have variceal source of bleeding on endoscopy were included. Clinical and endoscopic features were noted to calculate Rockall score. After giving appropriate pharmacological and endoscopic therapy, patients were followed for rebleeding or death till discharge from hospital. Linear regression analysis was used to determine predictive value of score and discrimination was evaluated by calculating the area under the receiver operating characteristic (ROC) curve. RESULTS: A total of 402 patients were included. Mean age was 52.57 (+/-11.39) and male to female ratio was 2:1(269/133). Esophageal varices were source of bleeding in 340 (84.5%), gastric fundal varix in 44 (11%) and ectopic duodenal varix in 3 (0.9%) patients. Both esophageal and gastric varices were present in 15 (3.6%) patients. In-hospital mortality was 6.7% while 22 (5.5%) patients had rebleeding. Rockall score was found to have good predictive value for mortality (p-value<0.001 and area under curve AUC 0.834) and in-hospital rebleeding (p-value<0.001 and AUC 0.798). CONCLUSION: Rockall scoring system has good predictive and discriminative value for in-hospital rebleeding and mortality in patients with variceal bleeding due to cirrhosis.