Validation of surgical site infection surveillance in Perth, Western Australia

There is marked variation in the rate of surgical site infections (SSIs) following total hip and knee replacements reported by hospitals submitting data to the Healthcare Infection Surveillance Western Australia. Ablinded retrospective case review was performed in order to determine if the differences were due to varying surveillance and reporting effectiveness. To further assess the quality of surveillance methodology the infection control professional at each site completed a paper-based, structured interview. A total of 444 patient charts were reviewed. Overall, the sensitivity of routine surveillance was 83%, the specificity 99%, the positive predictive value 94% and the negative predictive value 97%. The sensitivity varied between hospitals, ranging from 20 to 100%. The results of the methodology survey revealed fairly comparable surveillance practices at nine of the 10 hospitals. The hospitals with the lowest SSI rates were also those with the highest sensitivity to detect an SSI, suggesting that the difference in the quality of surveillance and reporting by each hospital is not responsible for the variation in SSI rates. The results of this study support the notion that those hospitals reporting the lowest SSI rates demonstrate high performance in both detection and prevention of SSI following hip and knee replacement surgery.