Emergency surgery for bleeding oesophageal varices.

PURPOSE OF STUDY--This case-series reviews patients who underwent emergency transgastric ligation or oesophageal transection for uncontrolled bleeding oesophageal varices despite initial conservative therapy with vasoactive drugs, balloon tamponade and injection sclerotherapy. The study seeks to identify factors determining outcome of surgery and the problems of transection surgery following endoscopic sclerotherapy. SELECTION OF STUDY SUBJECTS--One hundred and ninety-eight patients (median age 54.6 years) were treated for bleeding varices between 1981 and 1991. Of these, 36 (18%) required emergency surgery and they formed the cohort under study. OBSERVATIONAL METHODS AND MAIN FINDINGS--Twenty-three patients underwent transabdominal oesophageal devascularisation and oesophageal transection while 13 had transgastric ligation. Nine patients were graded Child-Pugh's A, 20 Child-Pugh's B and 7 Child-Pugh's C-mortality was 11%, 25% and 100% respectively. Of all the risk factors, the Child-Pugh's grade was the single most important factor determining outcome (Chi-square test; 2 degrees of freedom, p < 0.0005). The data also showed that patients who were transfused less than 3 litres of blood had a mortality of 18% compared to 55% for those requiring 3 or more liters (Fisher's exact test p = 0.0155). There was no difference in mortality between transgastric ligation and oesophageal transection when patients were evaluated according to the Child-Pugh's grade. The results indicate that oesophageal transection is to be recommended for better control of bleeding (0% vs 23% post procedure) but conversely is associated with higher anastomotic leakage rate (8.7% vs 0%). PRINCIPAL CONCLUSIONS--We conclude that the decision to operate on bleeding variceal patients should be made early as procrastination is detrimental to their surgical outcome. In view of the extremely poor results with Child-Pugh's C patients, emergency surgery is best offered to the Child's A and B cases.