Another Geographic Variation in Health Care Spending Study

Multiple pieces of research over at least the last two decades have demonstrated extensive geographic variation in health spending, primarily in Medicare, but in employer-sponsored health plans as well.  If you hold prices constant, much of the variation is revealed to be in utilization.  Utilization is largely driven by physician choices, what are referred to as physician practice patterns.  Physicians order most health care, so they control utilization.  They learn practice patterns first in medical school and then tend to pick up the patterns of physicians in the community where they begin to work.  Some research has found that when a physician moves, he or she will adopt the practice patterns in their new community.  While there is room for some judgment about optimal care, most research leads to evidence-based guidelines that should not vary from place to place.  Any care above the optimal contributes to our excessive health spending problem.

In this paper the authors followed military members who were moved from one base to another to see if their care changed as they changed locations and physicians.  Military physicians are salaried so they don’t have an incentive to provide more care to make money.  Military health benefits are provided without significant deductibles and copays so consumer cost isn’t likely a factor in care.  The researchers found that care patterns did change after a move and that those geographic differences in care patterns accounted for about 80% of total variation in spending.

Over one million people who moved were compared with over two million in the control cohort.  In the geographic regions with higher utilization and spending, there was no difference in quality measures.  Utilization differed by as much as twice across regions.  That is a huge variation if everyone is practicing according to the same standards of care.  It is highly unlikely that patient preference plays a signficant role in the variation, given that the patient moves are essentially random and involuntary in the military system.   (NBER Paper)