Reliability of different predictive systems in major gastrointestinal surgical patients

Abstract

APACHE II-III, SAPS II, and MPM II are commonly used predictive models. The systems designed for surgical patients include the ASA and P-POSSUM. NRS-2002 score is suggested for screening of nutritional risk. We aimed to assess the performance of prognostic models, and to compare the reliability of NRS-2002 with those, in major gastrointestinal surgical patients. APACHE II and III, SAPS II, MPM II and P-POSSUM scores, ASA grading, and NRS-2002 scores of the patients underwent major gastrointestinal surgery were collected on admission. Calculations were repeated for APACHE II and III, SAPS II, and MPM II at postoperative 24 th hour. Discrimination and calibration characteristics of the scoring systems were evaluated.APACHE II-III, and SAPS II at postoperative 24th hour, and P-POSSUM on admission, had reliable power of discrimination and calibration for mortality prediction in patients undergoing major gastrointestinal surgery. APACHE III, SAPS II and PPOSSUM had the best performance on complication prediction. ASA, MPM II, and NRS-2002 had insufficient discrimination statistics. APACHE II and III, SAPS II, and P-POSSUM are superior to other systems in predicting mortality in gastrointestinal surgical patients. APACHE III, SAPS II, and P-POSSUM are superior in predicting