Comparison of hospitals with regard to outcomes of surgery.

Two overlapping studies of variation in postsurgical death and morbidity among short-term hospitals are reported. In one study, Professional Activity Study data were used to compute indirectly standardized mortality ratios for 314,000 patients in 14 surgical categories from 1,224 hospitals; in the second study, patient mix data were collected and ratios computed for 8,593 patients in 15 surgical categories from 17 hospitals. Comparison of results from the two methods in the 17 hospitals increases confidence in the methods and results. The results indicate that, for at least some types of surgery, substantial variation in outcome among hospitals does exist, independently of differences in patient mix. It is suggested that moderate expansion of routinely collected abstract data could provide a basis for continuing assessment of outcome to supplement local quality assurance programs.