GAO Report on Medicare Use of Physician Profiling

For at least two decades private health plans have examined how physicians in their provider networks practice and have given feedback on those practice patterns to the physicians.  CMS has a pilot underway to do similar profiling and reporting in regard to Medicare patients.  The Government Accounting Office analyzes a specific method of doing this profiling and provides an overview of the entire topic in a new report.   (GAO Report)

While there has been much focus on geographic variation in health care spending, ultimately the great majority of health spending is driven by individual physicians and understanding differences in how physicians treat similar patients can be very useful.  Two approaches to profiling have to been to look at patterns of treating the same episodes of care and looking at resource use on a per capita or per patient basis.  The GAO report analyzes the per capita method, finding that it is a valid way to conduct profiling.

This report follows an earlier one by GAO more generally reviewing the physician practice profiling topic and making recommendation to Congress and CMS.  The report gives a very good overview of profiling issues, including report content and form, delivery of feedback, reporting on individual physicians versus groups and  what benchmarks to compare individual doctors to.

Although the report was not intended to comment substantively on practice variation, it does have some interesting tidbits.  Physicians generally account directly for only about 2% of their patients’ total Medicare spending.  High resource use physicians tend to have much more use of institutional services.  Higher resource using patients tended to see many more physicians.  It should be noted that all these analyses were done on a patient risk-adjusted basis, so it is a fairly good apples-to-apples comparison.  Individual physician practice variation profiling seems to have more promise than bulk geographic variation studies in identifying and addressing inappropriate care delivery.  More work needs to be done to understand how to encourage physicians to adjust their practice patterns when appropriate and to ensure that the variation in practice is not justified by individual patient need.