Development and validation of a nomogram predicting postoperative pneumonia after major abdominal surgery

Abstract

Purpose: Postoperative pneumonia (POP) is a common complication that can adversely affect the outcomes after surgery. This study aimed to devise and validate a model for stratifying the probability of POP in patients undergoing abdominal surgery.

Methods: We included 1050 patients who underwent major abdominal surgery between 2012 and 2013. A nomogram was devised by evaluating the predictive factors for POP.

Results: Of the 1050 patients, 56 (5.3%) developed POP. Multivariable logistic regression analysis revealed that the independent predictive factors for POP were age, male sex, history of cerebrovascular disease, Brinkman Index (BI) ≥ 900, and upper midline incision. A nomogram was devised by employing these five significant predictive factors. The prediction model showed a relatively good discrimination performance, with a concordance index of 0.77.

Conclusions: A nomogram based on age, male sex, history of cerebrovascular disease, BI ≥ 900, and upper midline incision may be useful for identifying patients with a high probability of developing POP after major abdominal surgery.

Keywords: Major abdominal surgery; Postoperative pneumonia; Predictive factor; Upper midline incision.