A further analysis of the "variations in practice style" phenomenon.

To examine Wennberg's "variations in practice style" hypothesis, I examined a comprehensive data base that reflected the use of physician services in Michigan. In addition to wide variations in per capita rates of utilization across physician services markets, I found wide variations within markets in the intensity with which physicians treat their patients. I also found that within high-use markets, physicians don't treat their patients more intensively than do physicians in other markets; rather, patients are treated by a larger number of providers. Significantly, there was no relationship between variations in practice intensity within markets and per capita levels of use. I discuss the relevance of my findings for Wennberg's proposals to reduce utilization and for other health care policy issues.