[New nutritional index for preoperative evaluation of malnutrition as a risk factor in surgery]

In a retrospective trial preoperative anthropometric and biochemical dates of 72 cancer patients were measured. All patients underwent a major surgical resection. Discriminant analysis and a computer-based stepwise regression procedure were used to identify the most important variables. It was able to compute a linear predictive model of 7 variables, called Nutritional Index, relating risk of operative mortality to preoperative nutritional status. In a prospective control trial it was possible to confirm the predictive value of the Nutritional Index by nutritional assessment of 204 surgical patients undergoing a major surgical procedure. This Nutritional Index provides a quantitative classification in surgical patients at high, low and without increased risk of nutritionally based complications.