Does stratifying surgical site infection rates by the National Nosocomial Infection Surveillance risk index influence the rank order of the hospitals in a surveillance system?

We performed a study to investigate whether stratification of surgical site infection (SSI) rates according to the National Nosocomial Infection Surveillance (NNIS) risk index could lead to a better basis for comparison of surgical units compared with simpler methods. A retrospective analysis of surveillance data of the German national nosocomial infection surveillance system (KISS, Krankenhaus Infektions Surveillance System) was completed with data from 234 volunteer surgical departments. In all, 4275 SSIs of 223 367 operations from 12 surgical procedure categories were surveyed over a 66 month period from January 2001 to June 2006. Active SSI surveillance was performed according to the NNIS method and Centers for Disease Control and Prevention definitions. For each department, two SSI rates were calculated per procedure: the crude infection rate (CIR) and the risk-adjusted standardised infection ratio (SIR) based on the NNIS risk index. Ranking was performed for the departments using both rates. The correlation between the two ranking positions was investigated by Spearman's correlation coefficient (P). For all 12 operative procedure categories, there was a strong correlation between the CIR and the SIR (P>0.95). A department's rank position does not change remarkably when the CIR, which is easier to understand and simpler to record, is taken into account for comparison instead of the SIR.