[Perforated gastroduodenal ulcer. II. Risk analysis in the prognostic evaluation of the postoperative course].

The data of 245 patients of our clinic, which perforation was treated with simple suture, were analysed by uni- and multivariate statistics to estimate risk factors. The aim was to predict the postoperative outcome. In a first step univariate test methods gave some orientation of the relevance of the factors. There was a highly significant correlation of age, preoperative shock, medical illnesses, time interval perforation-operation and clinical diagnosis to postoperative mortality. In a second step a multivariate decision model was used to calculate a prognostic index of the outcome after operation of perforated peptic ulcer. The index was tested in a cohort study. The results of validation were good for all quality parameters, so that the index allows a relatively safer statement about the so-called risk patients.--The prognostic problems of perforated gastroduodenal ulcer, including diagnosis, are also represented with graphic figures in the form of decision trees.