Factors contributing to releak after surgical closure of perforated duodenal ulcer by Graham's Patch.

BACKGROUND: Perforated duodenal ulcer is one of the common surgical emergencies. Releak after duodenal ulcer perforation closure is an important cause of mortality. This study was planned to analyse risk factors if any, which could predict releak following duodenal ulcer perforation closure and to ascertain the contribution of releak towards ultimate outcome.

METHODS: A prospective study was undertaken between September 1997 and August 1999 including all patients undergoing surgery for perforated duodenal ulcer. All patients (119) underwent a Graham's patch closure and were put on parenteral H2 antagonists and antibiotics postoperatively. Patients with releak were included in case group (9), and those without releak were included in control group (110). Factors considered for comparison among the two groups were age, pulse rate, systolic blood pressure at presentation, anthropometeric parameters, haemoglobin, serum total protein/albumin, total lymphocyte count and operative findings including size of perforation, evidence of chronicity of ulcer, quantity and nature of peritoneal fluid.

RESULTS: Age greater than 60 years (p-0.0470, CI-0.76-31.54), pulse rate greater than 110/minute (p-0.0217, CI-1.04-34.48), systolic blood pressure less than 90 mm Hg (p-0.0016, CI-2.04-71.9), haemoglobin level less than 10 g/dl (p-0.0009, CI-2.25-135.02), serum albumin less than 2.5 grams/dl (p-0.0145, CI-1.21-38.31), total lymphocyte count less than 1800 cells/mm-3 (p-0.0003, CI-8.9-42.2), size of perforation greater than 5 mm (p-0.0011, CI-1.09-36.13) were identified as risk factors for releak. Serum albumin, hemoglobin and size of perforation were independent risk factors for prediction of releak on multivariate analysis. The anthropometric parameters namely mean triceps skin fold thickness, mean mid arm circumference and mean body mass index were all significantly less in cases as compared to controls. Releak was found to be a significant cause of death in patients with perforated duodenal ulcer. A total of 8 patients died in both the groups. The mortality rate in the releak group was 55.6% (5 out of 9 patients) compared to 2.7% (3 out of 110 patients) in the control group [p-0.0001].

CONCLUSION: Releak was a significant factor influencing mortality rate after omental patch closure of perforated duodenal ulcer.