Duke Research Advantage

The following post was written by Kevin Schulman, M.D., Director of Duke’s Health Sector Management Program:

As Director of Fuqua’s HSM Program, I hereby nominate the physicians and hospital managers of McAllen, Texas, for a special joint Nobel Prize in Medicine and Economics for their brilliant experiment demonstrating, beyond a shadow of a doubt, that PHYSICIANS DO RESPOND TO ECONOMIC INCENTIVES. In recognition of this work, all physicians and hospitals in McAllen should receive a bonus payment equal to their 2009 Medicare billings, and then should be permanently excluded from the program (the public system responds to few incentives beyond those of special interests, but when you’re the subject of a must-read report at the White House, you’re out of luck).

So if you’re one of the few people who has yet to find time to read Atul Gawande’s piece in the New Yorker (and it is must-read material). Here are the major highlights:

For over 40 years, Jack Wennberg and his group at Dartmouth (now including Elliott Fisher and Jonathan Skinner) have shown there is significant variation in medical practice — more variation in “discretionary” services like imaging vs. essential services like appendectomy. This group has also created a database called the Dartmouth Atlas of Medical Practice, which reports variations in medical care according to hospital referral regions in the United States. From this database, a region that was identified as one of the highest utilizing sites was McAllen, Texas. Gawandi of the New Yorker visited to try to gain a better understanding the case and uncover reasons for the high utilization.

Less developed in Gawande’s article is the idea that the Medicare program has known about this practice for years. Annually $1 billion dollars is spent on a national program of Quality Improvement Organizations. These spends are allocated for the review of medical practices within the Medicare program. State Medicare medical directors, contracted health plan managers, and Office of the Inspector General at HHS are all involved in the review. All seemed powerless to take action in the case. (In terms of the public-private plan debate, it seems the private plans in the market had the same incentives and observed the same results in McAllen, so private plans aren’t the automatic fix to this practice pattern issue).

So what are the implications of this study? Fix the incentive structure. Sounds like a simple concept, but one that has escaped us for some time. There are suggestions on how to accomplish this within a fee-for-service structure, i.e. adjusting payments at a local/state level based on utilization patterns, bundling payments into episodes of care (or even back to capitation models as Accountable Care organizations) to alter the incentives for over-provision of services.

At the core of this issue are also questions about management and accountability . Well intentioned reforms will result in a similar set of perverse incentives if left unchecked for 40 years. We need reasonable accountability and traction to enable ongoing management of the Medicare program (under civil law, not criminal law). One could easily imagine that one of the best jobs over the next 40 years would be the ‘Manager’ of the Medicare program (if it paid a little better), but that job is still open in the current Administration.

Public reporting of quality data and utilization patterns could also help the “sheep” (in the Gawande article) question their physicians about treatment recommendations and services they are receiving. These publically available data could make these utilization patterns transparent to all whose interests are being served in health care in McAllen and in other similar communities across the country.