This Article Abstract Submit a response to this article Services E-mail this article to a friend Alert me to new issues of the journal

P E R S P E C T I V E S
V A R I A T I O N S
W E B  E X C L U S I V E   

7 October 2004  
  

Perspective: 
Everything New Is Old Again

After thirty years of insightful research and debate,
unexplained variations continue to persist
throughout U.S. health care.

By Bruce C. Vladeck

ABSTRACT:

For more than thirty years, John Wennberg and his colleagues have been documenting variations in patterns of health care use from one community to the next, which are not explained by illness or demographic patterns. Twenty years ago Health Affairs devoted an issue to a symposium on this work, and it is striking how little some things have changed in the intervening years. In fact, there have been enormous changes in physicians’ behavior and patterns of medical practice, but our cost problems seem as intractable as ever, perhaps because policymakers continue to focus erroneously on the relationship between use and costs.

Twenty years ago Health Affairs devoted a special issue to “Variations in Medical Practice.” The lead paper, by John Wennberg, reviewed what was then already a decade’s worth of research on small-area variations in use of medical care (which, in turn, built on earlier, more limited research on tonsillectomies extending back to the 1930s) and proposed a three-step process for reducing unjustifiable and costly variations. Those three steps were routine, systematic monitoring of performance across hospital markets; working with physician communities to develop better data on outcomes and to effect behavior change toward more effective and efficient practice; and providing financial incentives to hospitals to reduce capacity and move their clinicians toward better practice patterns.¹

Wennberg’s paper was accompanied by a number of commentaries. Almost all referred to the crisis in health care costs and the sheer unsustainability of then-current rates of cost growth. Medical leaders emphasized the profession’s responsibility to take this evidence to heart and to assert leadership in the process of moving clinical practice toward a more uniform, scientifically defensible basis. David Eddy contributed a brilliant essay on the relationship between variations and intellectual uncertainty, from the most abstract level to the realities of clinical practice.² Two officials of the Reagan administration pointed with alarm to the impending bankruptcy of the Medicare Hospital Insurance trust fund and then advocated for policy changes including higher out-of-pocket costs for beneficiaries in the short run and turning much of Medicare over to private managed care plans in the long run.³ I alone among the contributors discussed the significance of variation from the perspective of government regulation.⁴

So here we are, again—or perhaps, still. To put first things first, it’s important to emphasize the magnitude of the contribution that Wennberg and his colleagues have made during the past thirty years. Their work is part of the basic canon of health services research—a field to which they have contributed not only in substance but in the mobilization of congressional support and public interest. The Dartmouth Atlas of Health Care is a standard reference work for anyone involved in health services management or health care policy. Much of the contemporary concern with the quality of care and efforts to improve clinical quality have strong intellectual roots in Wennberg’s work, as does the impetus toward broader adoption of evidence-based medicine.

Still, a first glance at the papers in the collection of essays on which I comment here might reasonably elicit questions about how much progress we have actually made in the past two decades.⁵ Widespread, otherwise unexplained variations in patterns of clinical practice continue to exist, and they appear to be strikingly persistent: High-utilization hospitals and communities tend to stay that way, as do low-utilization hospitals and communities. Much of medical practice falls far short of professionally accepted norms of high-quality care. The rise and intervening fall of managed care doesn’t seem to have had much of an impact on all of this. And costs continue to go up, unacceptably and “unsustainably,” threatening, among other things, the future solvency of Medicare.

Changing medical conditions and practices. In fact, a closer comparison of this collection of papers with those of twenty years ago reveals some areas of demonstrable progress. In some other ways, however, we do seem to be going in circles or even backward. First the good news: Among the highest-variation causes for hospital admissions in Wennberg’s paper twenty years ago were peptic ulcers, tonsillectomies, tubal ligations, and dental extractions. None of those procedures accounts for more than a negligible number of admissions at most hospitals today. Similarly, many of the diagnostic procedures discussed in Eddy’s paper twenty years ago are now largely historical relics.

In other words, actual medical practice has changed dramatically before our eyes, something the largely cross-sectional analyses that dominate the Wennberg canon can never adequately reflect. Old problems get solved or ameliorated, and new ones spring up to take their place. Wennberg’s paper in this collection implicitly reflects that phenomenon: Congestive heart failure and chronic obstructive pulmonary disease (and, to a lesser extent, cancer) are currently foci of interest because, to some extent, fewer people are dying earlier of coronary or cerebrovascular diseases, and it’s only relatively recently that health care professionals have begun developing broader consensus on appropriate models of treatment for those more chronic diseases.⁶ There is an important general lesson here, although it runs counter to at least the tone of the Wennberg school’s work: The performance of the health care system may always look worse than it is because the last generation’s successes are taken for granted—or forgotten—as the (non-age-adjusted) mortality rate in the overall population remains at 1.0.

More generally, in fact, physicians’ behavior may be changing much more rapidly, and in much more desirable ways, than most students of health policy acknowledge. Medicare data, as reported by Stephen Jencks and colleagues, seem to demonstrate that it is, at least for some targeted diagnoses that were the subject of some targeted interventions.⁷ The problem is that we still really don’t know very much about why physicians do what they do, or how to get them to do different things. An important subsidiary problem is that most of our contemporary health services research, whether of the Wennberg epidemiological school or the more hegemonic microeconomic models, doesn’t help us very much in developing that understanding. The existing establishment may be able to demonstrate that we have a problem, but we need the methods and insights of organizational sociologists, social psychologists, and professional educators to help us solve it, and no one listens to those folks inside the Beltway—or in most of academe, either.

Persistence of geographical patterns. The other problem with thinking about behavior change and variation reduction is that even as variation falls for some procedures or some diagnoses, high-utilization areas appear to remain high-utilization areas, and low-utilization areas remain low. This is always a sensitive issue for a New Yorker, but, in fact, New York City has remained a relatively high-utilization area even as total inpatient use per capita has fallen by a third in the past two decades, and Rochester has remained a low-utilization area even as its metropolitan area has experienced enormous economic setbacks. We still don’t really know why people in Oregon use so much less hospital care than people in Louisiana, and it’s far from clear that the supply differences to which Wennberg attributes much of the difference are really a cause, rather than an effect.

What has changed in the past twenty years or so is the recognition by politicians in low-utilization areas that the presumably more efficient behavior of providers reduces the inflow of Medicare dollars to their constituencies. I used to argue that from the perspective of Medicare policy, the clearest implication of Wennberg’s data was that Medicare could save a lot of money if it could only convince all of its beneficiaries in Florida to move to Minnesota. In fact, the resentment of public officials from low-utilization states and congressional districts has produced, in the past decade, some bizarre adjustments to Medicare payment formulas, which, among other things, largely doomed the Medicare+Choice provisions of the Balanced Budget Act (BBA) of 1996 to failure. So eager were representatives of low-utilization areas to redress the perceived inequity in payments that they reduced payments to communities in which Medicare health maintenance organizations (HMOs) were thriving by enough to drive a majority of those plans out of the business, while raising payment rates in rural and other low-cost areas for hypothetical plans that didn’t exist and didn’t spontaneously spring into existence after the BBA was enacted.

Association of variations and spending. More generally, the principal reason why health care policy has continued to go around in circles even as the data on variations have become more sophisticated and more refined is because of the continuing preoccupation, which Wennberg and his colleagues continue to reinforce, with the presumed association between use and spending. As Gerard Anderson and his colleagues recently reminded us in this journal, a comparative international perspective makes clear that health care in the United States is so much more costly than it is anywhere else not because Americans use so much more health care, but because our prices are higher and because our decentralized, pluralistic system generates extraordinary overhead costs.⁸ The Germans, Canadians, and Australians all have considerable problems with small-area variations, but in radically different ways they all manage to cover their entire populations, produce better outcomes than we do, and spend a lot less money.

The American obsession with the notion that excess utilization is the principal source of excessive health care costs is partially cultural: Our Puritan ethos rebels at the thought of all those overinsured consumers frivolously using up all that valuable health care, and if the consumers are using health care because they’re sick, then that’s probably their own fault, too. It’s partially methodological: Our principal analytic approaches assume a relationship between long-range prices and long-range costs that has absolutely no congruence to the realities of actual U.S. health care markets. But it’s mostly political: If we focus our discussions on utilization, we don’t talk about prices—or incomes—and thus won’t directly threaten existing structures of power and prestige within the health care system or the political system. At the most aggregate level, focusing on utilization permits both providers and payers to change the subject away from the more embarrassing discussion of who’s getting paid, by whom, and how much.

Overuse and underuse. As the burgeoning literature on quality of care increasingly suggests that there may be as many instances of underuse in the health system as overuse, and as a separate stream of activity should continually remind us of how much underservice—especially for highly “discretionary” procedures such as cardiac catheterization and bypass surgery—there is among the growing U.S. minority populations, the seductive, simple (and simple-minded) syllogism that “better care means lower use rates means lower costs” blows up on the front end. It’s never been particularly valid on the back end: Lower utilization doesn’t always equate to lower costs. Solving the cost problem will not be that easy. For the level of spending to be less than it otherwise might be, some providers will have to have less income than they might otherwise expect.

More than thirty years ago the work of Wennberg and his colleagues began in part from the conviction that overuse of certain services, especially surgery, was bad for patients. It still is. For a variety of reasons, in the environment of U.S. health services and health care policy, however, that concern was drowned out by the preoccupation with the costs of care. But nowhere in all of the variations literature is concealed the magic key that is going to unlock all our cost problems. It just isn’t there. We’re going to have to figure out something else to do about costs. At the same time, however, whenever there’s an opportunity to protect patients by forgoing services or treatments that are as likely to harm as to benefit them, we must seize it. We’re going to need to learn a lot more about how to do that, but the Wennberg opus should serve as a constant reminder that it must be done. That’s the real lesson of these thirty-plus years of careful, thoughtful, and often creative research, and one whose implications are not so much economic as moral. When patients’ health and well-being are involved, we should do the right thing because it’s the right thing to do.

NOTES

1. J.E. Wennberg, “Dealing with Medical Practice Variations: A Proposal for Action,” Health Affairs 3, no. 2 (1984): 6–33. The entire issue is archived online at www.healthaffairs.org and is freely available to all site visitors.
2. D.M. Eddy, “Variations in Physician Practice: The Role of Uncertainty,” Health Affairs 3, no. 2 (1984): 74–89.
3. R. Rubin and G. Hackbarth, “ReViews: The Federal Government,” Health Affairs 3, no. 2 (1984): 38–45.
4. B.C. Vladeck, “Variations Data and the Regulatory Rationale,” Health Affairs 3, no. 2 (1984): 102–109.
5. A special collection of papers and commentaries on variations is available at content.healthaffairs.org/cgi/content/full/hlthaff.var.108/DC1.
6. J.E. Wennberg et al., “Use of Medicare Claims Data to Monitor Provider-Specific Performance among Patients with Severe Chronic Illness,” Health Affairs, 7 October 2004, content.healthaffairs.org/cgi/content/abstract/hlthaff.var.5.
7. S.F. Jencks, E.D. Huff, and T. Cuerdon, “Change in the Quality of Care Delivered to Medicare Beneficiaries, 1998–1999 to 2000–2001,” Journal of the American Medical Association 289, no. 3 (2003): 305–312.
8. G.F. Anderson et al., “It’s the Prices, Stupid: Why the United States Is So Different from Other Countries,” Health Affairs 22, no. 3 (2003): 89–105.

Bruce Vladeck (bruce.vladeck{at}mountsinai.org) is a professor of health policy and geriatrics at Mount Sinai School of Medicine, in New York City.

DOI: 10.1377/hlthaff.var.108
©2004 Project HOPE–The People-to-People Health Foundation, Inc.