Waldo’s Optimal Fragmentation
Thank you Frank, Glenn, and Concurring Opinions for inviting me to participate in the online Fragmentation book review symposium. As serendipity would have it, your timing could not have been better, for two reasons.
First, just last week in my Health Care Finance class here at the University of Georgia, we began the chapter on “The Structure of the Health Care Enterprise,” which opens, as those familiar with the Furrow, Greaney, et al. Health Law casebook know, with a classic fragmentation scenario: Our 35-year-old patient, Waldo, a recurring character throughout the casebook, must navigate the late 1970s health care system and all of its poorly coordinated care, minimal attention to quality, self-interested referrals, fee-for-service payment, and balance-billing flaws. Students are urged to observe the deficiencies and incentives driving the U.S. health care “system” — or lack thereof — and prepare, in future chapters, to compare Waldo’s treatment under, first, 1980s managed care and, later, contemporary consumer-driven health care. Preparing for this Symposium primed me for discussing the Waldo example with my students.
Second, over the weekend I presented at the Midwestern Law and Economics Association conference, where I had the good fortune to see my friend and mentor, David Hyman. Thanks again to the upcoming Symposium, I had just read David’s chapter on my flight to Denver and could engage his thoughts on the Fragmentation book. As I commented to David, what I find so refreshing and valuable about this book is that it does not start from the well-worn and, to me, uninteresting, suggestion made whenever one is asked to talk about what’s wrong with the United States health care system: “Well, really, don’t we just need universal health care, a single-payor system? If Canada can do it, why can’t we?” (I mean no disparagement to Vickie Williams’s excellent post earlier today on this point, for those who are less jaded than I.) Instead, the book accepts the system we have — an admittedly disjointed amalgam of government programs and regulations, along with competitive markets and incentives — and asks how it might be improved.
David’s chapter focuses on payment systems as a source of fragmentation and provides some nice examples even within particular health care programs or payors. For example, Medicare A (hospitalization) providers have no incentive to consider Part B (outpatient and physician services) providers’ costs because the Parts operates as “payment silos,” under distinct reimbursement methodologies, with no need to coordinate care or service delivery. With dually eligible beneficiaries, the fragmentation is even worse: Medicare faces higher costs if poor quality care in Medicaid-covered long-term care settings precipitates an acute hospitalization, while Medicaid faces higher long-term care costs if Medicare prescription drug coverage fails to treat beneficiaries’ chronic conditions. The chapter suggests some payment reform strategies, such as medical homes, pay-for-performance, and bundled payments, which have been debated and adopted to varying degrees by government and private payors, including in the recent Patient Protection and Affordable Care Act.
The end of the chapter was the most provocative. In David’s characteristically contrarian, and, to me, always entertaining way, he concludes by begging the question: Before we get all worked up about fragmentation and how to fix it, we should step back and consider that, at least in some cases, we seem to (and, he would suggest, at least sometimes, should) prefer fragmented health care delivery. He points to the popularity of Wal-Mart retail clinics, medical tourism, and specialty hospitals. And I would add the decline of managed care, rise of concierge medicine, and popularity of $4 prescription drugs. Of course, as David acknowledges, those trends reveal other, deeper flaws in our health care delivery “system.” But before jumping on the de-fragmentation bandwagon, Waldos should consider carefully what they may have to give up.