Risk adjustment performance between NNIS index and NHSN model for postoperative colorectal surgical site infection: A retrospective cohort study

BACKGROUND: Risk stratifications to predict development of surgical site infections (SSI) are crucial methods before surgery. Hence, we aimed to compare the performance of risk adjustment between the former NNIS risk index and the new NHSN procedure-specific risk model for postoperative colorectal SSI. MATERIALS AND METHODS: A retrospective cohort study was conducted. Data of post-colorectal SSI, indicating the use of the NNIS risk index for SSI adjustment, were retrieved from the medical records. Data were taken from patients who underwent colorectal surgery procedures between January 2005 and December 2016. Additional information regarding emergency colorectal surgery was retrieved to fulfill the requirements for calculation of the risks for SSI; via the new model. The predictive performance between the two models was compared using the means of the area under the receiver operating characteristic curve. RESULTS: In total 1989 patients were included. Fifteen patients were excluded; thus, the remaining number of procedures was 1974. Surgical site infections occurred in 85 (4.3%) procedures. In colectomy surgery, the means of area under the curve (AUC) yielded 0.6196 and 0.5976 for the NNIS risk index model and the new NHSN risk model, respectively; differences in the AUC were not statistically significant (p = 0.39). In rectal surgery, the means of the AUC yielded 0.516 and 0.49 for the NNIS risk index model and the new NHSN procedure-specific risk model, respectively; differences in the AUC were not statistically significant (p = 0.56). CONCLUSION: The new NHSN procedure-specific risk model was not superior to the former NNIS risk index.