Does practice make perfect? Part II: The relation between volume and outcomes and other hospital characteristics

The effect of a greater volume of patients with similar conditions being treated at a hospital on the outcomes achieved is investigated for almost 500,000 selected surgical and medical patients treated in over 1,200 nonfederal United States hospitals. In Part I the authors found strong and consistent evidence for surgical patients that high volume is associated with better outcome; evidence for medical patients was mixed. In this paper the authors include other hospital variables related both to volume and outcome--hospital size, teaching status, and expenditures--to determine whether they mask the true relationship; still, strong and consistent evidence that greater volume produces better outcome was found for both surgical and medical patients. This relation was significant for low-, medium-, and high-risk patients. Among the hospital variations added, only size was consistently and strongly related to outcome; greater size was associated with poorer outcome after accounting for volume. The potential importance of the findings for reducing deaths and days in hospital on a national level is discussed. The evidence is strongly supportive of the need for policies that would promote greater regionalization of a given service, and not greater size, to obtain better quality outcome for patients treated.