Validation of surgical site infection registration in colorectal surgery

Background: Surgical site infection (SSI) represents almost 40% of nosocomial infections. Risk for SSI is most prominent following colorectal surgery. There is no Swedish standard SSI classification or surveillance. We have failed to find validation of SSI surveillance programs presently in use. This study investigates SSI registration routines at Halmstad County Hospital, Sweden (HCH) as well as Swedish surgery departments' SSI classification and reporting practices. Methods: We requested information regarding SSI surveillance routines from all surgical departments performing colorectal surgery in Sweden. The reliability of SSI data collected routinely at HCH between 1999 and 2008 was approximated by comparing routine reports with data from a prospective study on infection prophylaxis. We calculate how much the reported SSI frequency must change for detection with current surveillance routines. Results: In all, 36 of 56 surgical departments answered a questionnaire; 24 have active SSI registration. Most commonly, SSI is detected with a combination of postoperative control in the outpatient unit and review of medical records after discharge. The department of surgery at HCH reported, from 1999 to 2008, an average SSI rate of 12.7% including emergency surgery. Compared to prospective registration, the routine provided a sensitivity of 0.56 and a specificity of 1.0. A reported frequency of 26% is required to detect a statistically significant increase in SSI. Conclusion: Swedish hospital SSI rates cannot be compared, there is neither uniform SSI surveillance nor classification system. The SSI monitoring routine explored in this study is uncertain. Methods must be developed further to obtain reliable data.