A nomogram to predict postoperative surgical site infection of adult patients who received orthopaedic surgery: a retrospective study

Abstract

Surgical site infection is a common postoperative complication with serious consequences. This study developed a nomogram to estimate the probability of postoperative surgical site infection for orthopaedic patients. Adult patients following orthopaedic surgery during hospitalization were included in this study. We used univariate and multivariate logistic regression analyses to establish the predictive model, which was also visualized by nomogram. To evaluate the model performance, we applied the receiver operating characteristic curve, calibration curve, and decision curve analysis, which were utilized in external validation and internal validation. From January 2021 to June 2022, a total of 787 patients were enrolled in this study. After statistical analysis, five variables were enrolled in the predictive model, including age, operation time, diabetes, WBC, and HGB. The mathematical formula has been established as follows: Logit (SSI) = − 6.301 + 1.104 * (Age) + 0.669 * (Operation time) + 2.009 * (Diabetes) + 1.520 * (WBC) − 1.119 * (HGB). The receiver Operating Characteristic curve, calibration curve, and decision curve analysis presented a good performance of this predictive model. Our nomogram showed great discriminative ability, calibration, and clinical practicability in the training set, external validation, and internal validation.