Surgical site infection: the principles and practice of surveillance. Part 2: analysing and interpreting data

In Part 1 of this two-part series on surveillance of surgical site infection (SSI) the principles of surveillance methodology and the role of surveillance in reducing the risk of infection were discussed. This second part focuses on the analysis and interpretation SSI surveillance data, the challenges this presents and of some of the solutions. The risk of SSI is conventionally expressed as the percentage of operations that develop SSI. However, this metric is strongly dependant on the length of post-operative stay, since infections take several days to become apparent and are difficult to identify after discharge. Comparisons based on more severe infections detected in inpatients or those readmitted with SSI are more likely to provide reliable data for inter-hospital comparisons. The precision of the estimated rates and adjustment for intrinsic risk factors are important considerations, although ultimately mechanisms for discriminating significantly higher rates merely indicate a problem requiring further investigation rather than definitive evidence of poor practice