NNIS risk score or COLA score: which predicts better the risk of surgical site infection among patients undergoing colorectal resection in a tertiary care hospital setting?

Abstract

Background: By using predictive scores to identify and quantify risk of Surgical site infections (SSI), preventive measures can be directed effectively to improve patient outcome post operatively. This study was undertaken to compare the National Nosocomial Infections Surveillance (NNIS) risk index and COLA (contamination, obesity, laparotomy and ASA grade) scores to predict risk of SSI among patients undergoing colorectal resection.Methods: A total of 77 patients who underwent colorectal resection at a tertiary care hospital were enrolled in this study. Wound surveillance was performed in all patients during the hospital stay and follow-up information for 30 days postoperatively was collected. Data analysis was done in SPSS version 20.0 for windows. Descriptive analysis was used to exhibit the clinical parameters. Differences between groups were tested by Pearson chi-square and fisher’s exact test. ROC curves were utilized. All the statistical tests were examined with 5% (p≤0.05) level of significance.Results: Incidence of overall SSI was 28.6%. Superficial SSI was more common (14.3%). Serious organ/deep space infection occurred only in 9.1%. Area under the curve (AUC) for NNIS score was 0.645 (95% CI: 0.510-0.781, p<0.05) and 0.611 for the COLA score. (95% CI: 0.472-0.750, p>0.05). Therefore, both these scores have less than acceptable accuracy in this data set to predict SSI.Conclusions: The risk prediction models (NNIS, COLA) although simple, did not accurately predict the risk of SSI in the given study population.