In Reply: Our study examined whether the data used for publicly reported performance could be used by patients to choose higher-quality hospitals and developed a composite measure that may help find a statistically significant association. Dr Hyder asks whether statistically significant findings in our composite measure (S-INF) and lack of statistical significance in the individual measures may have occurred because of differences in the populations meeting inclusion criteria for any SCIP measure vs those available for the S-INF composite measure. In fact, the individual measure analysis is a building block for the composite measure analysis, as pointed out by Dr Gold in his letter. A hospital that does not perform well on the individual measures cannot perform well on the composite measure. Furthermore, if performance on the individual measures is not aligned with improved quality of care, neither will be the composite measure.
Stulberg JJ, Koroukian SM, Delaney CP. Surgical Care Improvement Project Adherence and Postoperative Infections—Reply. JAMA. 2010;304(15):1669-1672. doi:10.1001/jama.2010.1463