Effect of procedural volume on outcome of coronary artery bypass graft surgery in Japan: Implication toward public reporting and minimal volume standards

Abstract

Since the Japanese government updated the medical practice laws, each hospital has to submit procedural volume from April 2007 and may sometime in the future have to submit some outcome indicators. It is very important to examine whether procedural volume is accurate and appropriate. We analyzed 4581 procedures from 36 centers between 2003 and 2005 by clinical database. The effect of hospital volume on each outcome was tested by a hierarchical mixed-effects logistic regression model, covering clinical risk factors, procedural year, clinical processes, and hospital volume/surgeon volume as a fixed effect and random intercepts for sites. Logistic regression model revealed a significant association between hospital bypass graft volume and 30-day mortality (P < .05) and operative mortality (P < .01). Surgeon procedural volume, however, did not have a significant effect on those outcomes. The effect of hospital procedural volume was associated with better outcomes in most patient subgroups: age younger than 65 years (P < .05), age 65 years and older (P < .01), low risk (P = .58), and high risk (P < .01). In Japan, high-volume compared with low-volume providers had better outcomes. As for public reporting in Japan, hospital-based evaluation might be more credible than surgeon-based evaluation. Although minimal volume standards might be effective to improve quality to some extent, volume has limitations as a marker of quality because of its wide range of variance.