Background: The decision to re-operate after abdominal surgery is still difficult, especially in the setting of intraabdominal sepsis. Mathematical models provide a good aid to both diagnosis and decision-making. Methods: A prospective observational study was conducted with 300 patients consecutively admitted to the intensive care unit of an academic institution affiliated to Calixto García Medical Faculty following abdominal surgery from January 2008 to January 2010. The patients were randomly separated (2:1) into estimation and validation groups. Logistic regression analysis was used in the estimation group to develop three models for decision-making related to re-operation including related factors such as age, ARPI, IAP, type of surgery (elective or emergency), and the duration of surgery. The three models developed were validated on the other group. Results: The acute re-operation predictive index-intra-abdominal pressure (ARPI-IAP) model was the best of the three models, with an excellent calibration, using the Hossmer-Lemeshow goodness-of-fit statistical test (C = 9.976, P = 0.267), as well as discrimination (AUC = 0.989; 95% CI: 0.976-1.000).
Conclusion:The combination of IAP with ARPI in a mathematical model can add accuracy to the prediction of need for re-operation related to intra-abdominal infectious complications in patients following abdominal surgery. This may be useful in all medical settings, but especially those with limited resources.