Objective: To compare the predictive performance of the new CDC NHSN logistic risk model and the former National Nosocomial Infections Surveillance (NNIS) risk index for post-cholecystectomy surgical site infection (SSI) classification.

Materials and Methods: Surveillance of data for post-cholecystectomy SSIs were retrieved from the Infection Control Unit of Songklanagarind Hospital. The surveillance system uses the former NNIS risk index for SSI risk adjustment. The study included 2,422 patients that underwent gall bladder surgery in the hospital between January 2005 and December 2016. Medical records were reviewed for additional information including emergency operation, and gastrointestinal cancer. The predictive performance of the former NNIS risk index was then compared to the new NHSN model by mean of area under receiver operating curve (AUC).

Results: The study identified 71 operations with post-cholecystectomy SSI while the former and the new model predicted 21.6 and 14.0 SSIs, respectively. AUC of the new model (65.7%; 95% CI 59.1 to 72.3) was not significantly (p=0.5) different from the former one (64.5%; 95% CI 58.0 to 71.0).

Conclusion: The predictive performance of the new CDC NHSN and the former NNIS risk index model was about the same for classification risk of post-operative cholecystectomy SSI.

Keywords: Cholecystectomy, Surgical site infection (SSI), Risk stratification, Surveillance

Received 20 Mar 2019 | Revised 23 May 2019 | Accepted 24 May 2019