Book Review
Purchasing Population Health:
Paying for Results
David A. Kindig. Purchasing Population Health: Paying for Results. Ann Arbor: University of Michigan Press, 1997. 216 pp. $24.95 paper.
Aimed primarily at an audience of business, political, and professional leaders, this book provides yet another analysis of the American health care system, with suggestions for reform. But it breaks new ground in two important ways.
First, Kindig's real concern is health, not health care, and his proposals are built on the rapidly growing understanding of those determinants of population health that lie outside the health care system. And second, he advocates a market-based "technical fix" not only for the provision of health care but for the promotion of health through these broader determinants.
Musing on the failure of the Clinton Health Security Act, Kindig concludes that "a more technical purchasing standard would have a much greater chance of realistic implementation than a massive centralized, national program" (p. xii). Such a technical approach, avoiding "sweeping national legislation" and "massive new bureaucracy," would be "more consistent with U.S. values and approaches" (p. 7).
This "more technical purchasing standard" is to be a single, population-based measure of health outcome, the Health-Adjusted Life Expectancy (HALE). The development of such a measure is a major focus of the book. On the principle that "what is rewarded, is done," contracts would then be let to private or public sector organizations, paying them according to their success (however achieved) in increasing the HALE measure for particular populations. In Kindig's view, maximizing the amount of health we achieve will not occur until a measure of health outcome becomes the purchasing standard for both the private and the public sectors.
This book was written during the author's sabbatical at the Universities of York and British Columbia. It offers a creative synthesis of work on the measurement of health outcomes at the York Centre for Health Economics (and elsewhere) with that of the Population Health Program, Canadian Institute for Advanced Research (CIAR). But the principal contribution is Kindig's effort to insert these intellectual perspectives into the American policy context, and to operationalize them in working programs.
The first few chapters review the stylized facts of U.S. health care. The United States spends far more per capita than any other nation. Yet basic [End Page 1238] measures of health outcomes, such as infant mortality and life expectancy, are much worse than elsewhere. None of this is new, but it is a compact and accessible summary to remind Kindig's target audience of the extraordinary mismatch between effort and outcome.
Perverse financial incentives within the health care system itself are largely responsible for the long-term expansion in the volume of services and health expenditures. These incentives have been significantly changed over the past decade, as managed care plans of various forms have come to dominate the American landscape. Costs have stabilized--though they remain extraordinarily high in world terms--and there is some evidence of improving efficiency. But public inconvenience and dissatisfaction appear to be growing rapidly, as is the number of people--now about 43 million--excluded from insurance coverage.
In any case, for Kindig the "managed care revolution" does not go nearly far enough. It focuses on health care, not health. Using cross-national data from the World Bank, he argues that in developed countries, increased spending on health care yields diminishing returns of population health gain. He concludes therefore that, for the United States, "further overall increases in health care expenditure will have limited return in terms of health outcomes" (p. 40). The opportunities for health improvement are clearly still substantial, but the high-payoff interventions lie elsewhere.
If so, it would seem to follow that the financial incentives faced by health plans should link reimbursement to measurable increases in the health of the populations served. Improvements in the volume, effectiveness, or efficiency of health services are simply one possible means to this end. The expansion and/or profit (or survival!) of health plans would then depend upon their ability to identify and influence factors, in or out of the health care system, that contribute to population health. Measured against a single common outcome standard, health plans would be more comparable and accountable, and upward pressures on health budgets should be mitigated. The key, of course, is the development of a single agreed-upon measure of population health.
In chapter 4, Kindig summarizes the state of play in the measurement of health status, emphasizing that any measure must incorporate both quantity (mortality) and quality of life. Quality, in turn, depends upon functional capacity as well as presence or absence of illness. The HALE, a mortality measure adjusted for both disability and morbidity, is offered as the best single population health measure currently available.
Chapter 5 provides a detailed discussion of the multiple determinants [End Page 1239] of health. This crucial chapter draws heavily on the work of the CIAR Population Health Program, and presents current knowledge of the range of determinants, and their complex interactions, again in a compact and accessible way. The reader will come away with a good understanding of the limits of medical care and the importance of nonmedical factors such as education, income, genetics, and culture in producing good population health outcomes.
The reader will also appreciate, however, the scope of the problem faced by a health plan under contract to supply "health"--or at least HALEs--rather than health care. The rest of the book addresses various aspects of the task of operationalizing these ideas. What organizations would purchase HALEs on behalf of populations? What would the "sellers" of HALEs look like? How would they behave, and how might the contracts be structured? And how does one get from "here"--that is, the contemporary United States--to "there"?
It is a bold and challenging vision, with a powerful internal logic and a very broad evidentiary base. But does it come off? All new ideas have flaws--so do all old ones--but are they fatal? We see problems on four levels.
Consider first the strategy of the "technical fix"--defining a common, objective standard for measuring health that can be used to compare health plans and determine their relative rewards. Not only does no such "objective" standard now exist, logically it cannot exist. A population-based measure requires aggregation across multiple dimensions of health, and across large numbers of people who are very differently placed on these dimensions. Any set of weights for aggregating the different dimensions chosen to represent health will simultaneously rank some people's problems as more significant than others, and their amelioration as more rewarding. The definition of the single standard is inherently a political, as well as a technical, process.
On the other hand, Americans have a long history of disguising political problems as technical ones, perhaps because their grasp of technology is superb and their political system is . . . what it is. Americans trust technology; they do not trust politics. If a bit of illogical self-deceit permits them to get on with tasks that their political system cannot handle, well, there's a lot of it about. But the "technical" process will turn out to be pretty controversial--ask John Kitzhaber!
More fundamentally, however, Kindig's proposals rest upon the presumption--largely unspoken--that the American people, given the appropriate mechanisms, would actually want to purchase health for the [End Page 1240] population at large. Themselves and their families, yes; "People Like Us" (PLUs), probably. But the population in general?
The persistent failure of efforts to establish a universal health insurance program in the United States has led Uwe Reinhardt (1996), for example, to conclude that Americans are different from the citizens of other developed societies, at an ideological or philosophical level. They do not accept, as others do, a collective responsibility to provide people in different socioeconomic strata with equal protection against the vicissitudes of illness or injury--even in the more limited domain of the health care system. If so, it seems a fortiori implausible that Americans would embrace Kindig's much more ambitious agenda. He wants to achieve better health for all by addressing the full range of determinants of population health. They don't care.
Others, however, argue that American exceptionalism is rooted, not in hearts and minds, but in political institutions. The latter are uniquely incapable of creating a universal, public health care system such as other countries enjoy, and Americans say they want. Kindig's bet has to be on this side. If one could find an appropriate institutional structure more or less separate from conventional politics, Americans would be no less willing than others to support the advancement of health for the whole population. Well, maybe.
The technical approach, however, still has a central technical problem, one faced by any system based on multiple, competitive, profit-driven health plans. Powerful incentives for innovation and efficiency are also powerful incentives for opportunistic behavior. A system that focuses on health, rather than health care, offers a wider range of possibilities for such behavior, and its control will be correspondingly more difficult.
The classic problem in private insurance is selection. The most profitable strategy for a private insurer, or managed care plan, is to insure healthy people. Therefore a good deal of effort is devoted to identifying unhealthy people, and finding ways not to insure them. Don't enroll them, push them out, or reduce your exposure to their losses. It is known as minimizing the loss ratio.
Advocates of competitive markets reply that selection bias would disappear if reimbursement were adjusted to risk on a person-specific basis. And this is true--if anyone knew how. Like the Holy Grail, such an adjustment process has been much sought after, never found. Meanwhile private insurers, driven by the stern law of survival in the competitive marketplace, continue to "dump the dogs."
The same law will drive the competitive producers of HALEs. Whatever [End Page 1241] the measure, it seems certain that HALE gain will be easier, less costly, to produce in some populations than in others. The economically motivated health plan will seek out the former and avoid the latter.
Coverage of the whole population will thus require a single, monopoly buyer of HALEs (a national government?) that is willing and able to pay very different prices for different subpopulations. One might then "let the market decide"--let the competitive tendering process sort out how much it really costs to produce a HALE for a given population. Those who get it right will prosper; others will leave the market. And HALE prices will adjust over time as information accumulates.
Well and good. But HALEs are produced over time, often over many years. Investigators are increasingly exploring factors in early childhood that influence health in adult life. Unless contractors--and populations--are locked in over long periods, health plans will find it unprofitable to put resources into long-term HALE production. If contracts are up for frequent renegotiation, another plan will bid away the nurtured population before it bears its HALE fruit. The information that may be generated in an active short-term market comes at the cost of long-term incentives.
Let us say, however, that we have our HALE producers locked into long-term contracts, at prices per HALE that are believed, by both parties, to reflect fairly (on current knowledge, perhaps with a risk premium) the cost of producing HALEs in the covered population. What now might we expect the contractors to do? Here we encounter a degree of ambiguity in Kindig's presentation of the determinants of population health.
Current understanding of these goes well beyond the individual-based factors, behavior and "lifestyle" that are at least in practice the principal targets of conventional health promotion. Studies of population health show quite convincingly that socioeconomic circumstances not only powerfully condition individual lifestyle "choices" and behavior, but also have a direct influence on health independent of individual-level factors.
Consider, for example, the growing evidence suggesting that the degree of economic inequality in a society has a direct impact on overall health levels, independent of the level of income (Wilkinson 1992, 1995; Kaplan et al. 1996). Inequality is simply not defined, much less controllable, at the individual level. Could a competitive health plan be expected, in the pursuit of HALEs for its charges, to take on a social/political issue of this magnitude? (Would its executives want to?)
Kindig is well aware of the evidence for the role of social context in health. His audience may, however, lose this crucial distinction when he [End Page 1242] observes that "[Wikler] reminds us that smoking, sloth, or other dangerous but enjoyable pastimes are still the decision of each individual" (p. 94), and quotes Knowles: "The solution to the problems of ill health in modern American society involves individual responsibility in the first instance, and social responsibility . . . in the second" (p. 93).
The ambiguity may be a tactical concession to the realpolitik of American health ideology, particularly that of the audience Kindig hopes to reach. But a reversion to an emphasis on individual-level (and, worse, individual-controlled) determinants of health would also provide a ready-made excuse for health plans that find the production of HALEs to be too difficult, costly, or politically threatening. (There is also the important detail that, despite their popularity, evidence for the effectiveness of individual-level interventions is meagre to nonexistent.)
The more profitable strategy may be to ignore the broader social determinants, with their heavy load of political freight and ideological discomfort. Traditional notions of individual responsibility are ready at hand. "If you are not healthy after all our well-intentioned promotional preaching, that's your fault. Get out of our plan--or pay extra--our shareholders should not be penalized for your shortcomings." As always, principle is invoked in support of interest.
Kindig's real hope seems to be, under cover of the market, to recruit more powerful allies to a broader social agenda--after all not much progress is being made in the United States through the conventional political system! If one put the issue of "what creates health?" squarely in front of powerful, profit-driven organizations, and the people who run them, they may work their way through to an understanding of the population health evidence and then to action. To maximize HALE production, health plans would have to bring pressure to bear on those institutions outside the health services sector that influence health.
This is a key issue. "It is appropriate to expect private sector health care organizations with powerful political and community standing to use their community and political influence to indirectly impact on appropriate health-enhancing investments in other sectors (environment, education, social services, public health), which will accrue to the population's health status improvement as well as to that of the community at large." Kindig admits that "currently there is no entity or market force driving the cross-sectoral outcome of population health" (p. 130). But if there were, would this book be needed?
And times change, particularly when, as Kindig begins by reminding his readers, the present situation is nothing to be proud of. His proposals [End Page 1243] are carefully couched in marketlike language that may offset any "socialistic" undertones. Confronted with technical and managerial challenges of a high order--their strong suits --the American "can do!" spirit might be engaged to find ways around the problems identified above. "You don't have to out-run the bear. . . ."
In any case Kindig has done an excellent job of adapting a population health framework, developed within the context of a Canadian--or European--perception of health as a public responsibility, for application within the very different American health care system and ideological culture. This book is a very creative attempt to take an academic model and mould it into a practical tool for use by American managers who, it would appear, are currently in some need of new ideas. At the same time it should serve the useful function of providing an influential audience with a very accessible summary reminder of the limits of medicine and the multiple social determinants of health.
Robert Evans and Aleck Ostry
University of British Columbia
Robert Evans is a professor of economics who is well known internationally for his work in health economics. He is also the past head of the Canadian Institute of Advanced Research's Population Health Program. Evans teaches at the University of British Columbia.
Aleck Ostry is a research associate in the Department of Health Care and Epidemiology at the University of British Columbia and is also affiliated with the Canadian Institute of Advanced Research.
References
Kaplan, G., E. Pamuk, J. Lynch, R. Cohen, and J. Balfour. 1996. Inequality in Income and Mortality in the United States: An Analysis of Mortality and Potential Pathways. British Medical Journal 312:999-1003.
Reinhardt, U. E. 1996. A Social Contract for 21st Century Health Care: Three-Tier Health Care with Bounty-Hunting. Health Economics 5:479-499.
Wilkinson, R. G. 1992. Income Distribution and Life Expectancy. British Medical Journal 304:165-168.
------. 1995. From Material Scarcity to Social Disadvantage. In Daedalus: Journal of the American Academy of Arts and Sciences 123(4):61-78.
Previous Article
Medicare HMOs: Making Them Work for the Chronically Ill (review)
