Defining Surgical Site Infection in Colorectal Surgery: An Objective Analysis Using Serial Photographic Documentation.

Abstract

Surgical site infection is common following colorectal surgery, yet the incidence varies widely. CDC criteria include "diagnosis by attending physician," which can be subjective. Alternatively, the ASEPSIS score is an objective scoring system based on the presence of clinical findings. The aim of this study is to compare the interrater reliability of the ASEPSIS score vs CDC definitions in identifying surgical site infection. This 24-month prospective study used serial photography of the wound. Three attending surgeons independently reviewed blinded photographic/clinical data. This study was conducted at an academic institution. Patients undergoing elective colorectal surgery were selected. Surgeons assigned an ASEPSIS score and identified surgical site infection by using CDC definitions. The interrater reliability of ASEPSIS and the CDC criteria were compared by using the κ statistic. These data were also compared with the institutional National Surgical Quality Improvement Program database. One hundred seventy-one patients were included. Four surgical site infections (2.4%) were identified by the National Surgical Quality Improvement Program. Data from the surgeons demonstrated significantly higher yet discrepant rates of infection by the CDC criteria, at 6.2%, 7.4%, and 14.1% with a κ of 0.55 indicating modest interrater agreement. Alternatively, the ASEPSIS assessments demonstrated excellent interrater agreement between surgeons with 96% agreement (2.4%, 2.4%, and 3.6%) and a κ of 0.83. This was a single-institution study. This study demonstrates the relatively poor reliability of CDC definitions for surgical site infections in comparison with an objective scoring system. These findings could explain the wide variability in the literature and raise concern for the comparison of institutional surgical site infection rates as a quality indicator. Alternatively, an objective scoring system, like the ASEPSIS score, may yield more reliable measures for comparison.