Perforated gastric and duodenal ulcer: an analysis of prognostic factors.

A review of patients treated operatively for perforated ulcer was undertaken to determine factors that independently predict mortality and morbidity. From 1980 to 1989 84 patients were treated with operative mortality and morbidity rates of 18 and 37 per cent, respectively. Patients with gastric ulcer were significantly older and were more likely to have concomitant medical problems, to use steroids, and be hospitalized at time of their perforation. They also had a higher perioperative mortality rate when compared to patients with gastroduodenal ulcers (P < 0.05 for each). Patients treated with definitive operation (n = 51) had a significantly lower recurrent ulcer rate when compared to an equivalent group treated with simple closure (n = 33) (mean follow-up, 47 months). Multivariate analysis confirmed a patient's likelihood of complication or death could be predicted using three variables (P < 0.03); age greater than 42 years, hospitalization at time of perforation, concomitant medical illness. While the relative distribution of these three factors in patients with gastric versus peptic ulcers appears to account for the difference in outcome between these two groups, the distribution of these factors between patients treated with patch closure and definitive operation was not significantly different. Patient outcome from operation for perforated ulcer is dependent on preoperative conditions and appears to be independent of surgical procedure performed. There was no benefit to simple closure of a perforated ulcer. Definitive procedures of perforated ulcers are associated with lower ulcer recurrence and therefore are advocated to treat perforation.