Sheikh Responds to Provider Volume???Patient Outcome Association and Policy by Luft

Abstract

Luft has presented some specific comments on my evaluation of an association between healthcare provider volume of therapeutic procedures and patient outcome (in this issue of Medical Care, pages 1118-1122), and his understanding of how to assess previous research and how to design and conduct medical care studies. These points and counterpoints need a response. Some of Luft's comments are inappropriate or unnecessary. For example, he complains that I start from a position that there is no relationship and I argue that its existence is unsubstantiated. Evidently, Luft has overlooked title of my evaluation paper and its first 2 paragraphs, which acknowledge that an association between provider volume and outcome has been found in many studies. Besides these acknowledgments, why would anyone question reliability of a nonexistent association? Luft's use of criminal law and justice system as a model for making healthcare policy decisions is erroneous. Medical care policies involve balancing therapeutic benefits against harm. These data are collected from scientific observation, experimentation, and complex evaluation. The evidence used in making judgments in courts of law lacks scientific rigor because judicial process is not a science. Luft's advocacy of separating science from the value questions and policy tradeoffs to enhance credibility of research for policy uses is very disturbing. Medical research that is not based on science or is