In-hospital mortality of surgical patients: is there an empiric basis for standard setting?

Several public and private groups have set minimum procedure-specific volume standards. Such standards reflect concerns about hospital quality and cost. In-hospital mortality rates are often taken as one measure of quality. To learn about variations in in-hospital mortality rates, we analyzed data on patients who underwent any of seven surgical procedures from a national cohort of 521 hospitals observed continuously between 1972 and 1981. On the average, mortality rates fell as the number of procedures performed annually at the hospital rose. Volumes at which mortality rates reached minimum levels were far higher than actual volumes achieved by the vast majority of hospitals. However, knowledge of hospital volumes left the major part of variation among hospitals' procedure-specific mortality rates unexplained. Many hospitals with low volumes of certain procedures had no associated deaths. Hospitals experienced appreciable year-to-year variation in mortality even though mortality rates fell with the number of years the procedure was performed at the hospital. Correlations among mortality rates for the procedures were low, suggesting that variation in rates is procedure rather than hospital specific. State rate-setting programs had no effect on mortality rates associated with the procedures analyzed. For several reasons, we conclude that an adequate statistical basis for setting minimum volume standards does not presently exist.