|
PERSPECTIVEWhich Way For Competition? None Of The Above
Despite growing documentation that the conditions needed to support competition in health care do not exist, consumer-directed health care has been offered as the new market-based solution to cost inflation. Yet typical consumer-based insurance products undermine the very logic of expecting consumers to make good health care decisions by making preventive services—the category of services about which consumers are best able to make informed decisions using their own money—exempt from cost sharing. Consumer-directed care threatens important societal values—in particular, the goal of establishing relationships between patients and clinical professionals based on trust.
At some point, a coalition of business interests, consumer advocates, and providers is likely to unite in a call for more federal involvement in health insurance and health care, although they will disagree on the form of that involvement. Few local-market participants remain optimistic that market forces will be strong enough to improve the efficiency of our health system without some kind of government intervention. Have you noticed the way traveling Americans attempt to communicate to people who do not understand English? They speak much louder in English, to the evident consternation of the still noncomprehending respondent. I’ve tried this technique; it doesn’t work. Advocates of market-based reform of the U.S. health care system similarly speak a language different from that of most patients, clinicians, and others directly involved in health care. But instead of shouting to a noncomprehending public, they develop idealized marketplace scenarios replete with new jargon and repeat them over and over, in an attempt to create an air of inevitability about markets working in health care. Like English, their language is the most widely used one in the world, and it works quite well in many sectors of the economy. When applied to health care, however, it remains a foreign language. It is increasingly evident that the necessary conditions for market competition simply are not met in the local markets where health care is provided.1 Commenting on the "competition revolution" then rapidly spreading, in 1988 Victor Fuchs pointed to structural limitations in health care that in his view, unfortunately, made the ideal of true market competition in health care unlikely.2 If anything, these conditions are even worse than when Fuchs was writing nearly two decades ago. For example, one essential condition is that there be a large number of buyers and sellers, no one of whom is so big as to have a strong influence on market prices.3 In 1988 Fuchs argued that this condition was rarely met. Subsequently, facing pricing pressure from payers, hospitals, in particular, consolidated through mergers and cut excess capacity to increase their negotiating leverage to demand higher prices.4 More recently, health insurers have begun concentrating their market power as well. In highly concentrated markets, buyers and sellers exert countervailing market power, but as Fuchs points out, that possibly socially desirable behavior by buyers and sellers should not be mistaken for competition.5
So now comes a new approach to achieve competition, one that envisions consumers as the buyers of their own health care at the point of service. The objective is to attack directly "moral hazard"—people’s tendency to spend more of a third-party payer’s money than they would their own. Hence we now have a new nostrum for getting a competitive market right: "consumer-directed health care." This model focuses on reducing the influence of insurance by expecting people to spend their own money through high deductibles and co-insurance at the point of service, with insurance reserved for catastrophic expenses. The papers by James Robinson and by Mark Hall and Clark Havighurst in this volume call for a gentle form of this approach—gentle because increased patient cost sharing would be placed on a secure insurance platform, where, for example, provider and supplier prices continue to be negotiated by the insurers and not by patients in a retail market.6 Mark Pauly envisions a more ambitious role for health plans as the source of differential rationing of new technology, but in doing so, he reaffirms the crucial role of private and public insurers in our health care system.7 The Hall/Havighurst paper usefully describes many of the important functions that third-party payers provide—functions that individual consumers are not well positioned for or interested in doing themselves. The authors make good arguments for "active managed care," predicting that the advocates of purer consumer-drive care would likely object, as Greg Scandlen surely does.8 On a secure health insurance platform, what is potentially new and useful as developed in the Robinson paper, is the prospect for "smarter" cost sharing. Instead of seeing cost sharing merely as a device for shifting costs to employees, Robinson suggests that cost sharing can be titrated carefully to influence patients’ behavior in desirable ways, based on the nature of each service that makes up a comprehensive benefit package. Robinson takes a stab at benefit design, sorting the appropriateness of cost sharing by whether services are "discretionary," "sensitive to patient preference," or "supply sensitive," borrowing concepts developed by Jack Wennberg and colleagues.9 To address costs, he suggests relying on varying mixes of consumer cost sharing on the one hand and provider network interventions on the other. Robinson’s recommendation that some services should not be subject to significant cost sharing and his insight that there is inevitable interaction between cost sharing and a range of provider-oriented approaches are powerful reasons, again, for not leaving consumers out there on their own, but rather involving a plan sponsor. He emphasizes the need for "organizational management" to get all of this right, thereby also running afoul of consumer-directed health care theorists.10
Nevertheless, there are both theoretical and practical problems with Robinson’s suggested model for benefit and network design. What some label as "discretionary" and "supply-sensitive" are only so in the aggregate. Services such as intensive care unit (ICU) stays and magnetic resonance imaging (MRI) scans are surely provided in excess, and more so in some geographic areas than others.11 But in innumerable individual clinical situations, they are as essential as the services labeled "medically necessary." As a firm basis for differential cost sharing, these categories will not hold up. Although it has become fashionable to minimize the importance of Kenneth Arrow’s incisive, four-decades-old observations regarding information asymmetry and uncertainty in health care, the reality remains that for the most part, clinicians are far better equipped than patients to grapple with the complexity of clinical decision making.12 Acknowledging that patients want to more actively contribute to clinical decisions that affect their well-being and reflect their values does not imply that they should become their own physicians. Patients surely could have incentives through differential cost sharing where information asymmetry is minimal and where uncertainty is not much of an issue. Pauly and Frank Sloan have provided helpful guidance by distinguishing among different types of medical care based on the likelihood that patients will have sufficient experience to gain purchasing competence.13 The choice of biologically and functionally equivalent prescription drugs is a good example of where consumers should be expected to make choices with their own money; most people can understand the trade-offs in selecting among medications that do the same thing. However, asking a patient to decide whether to spend substantial funds to spend time in the ICU or have a recommended MRI would place an unreasonable intellectual burden on many and produce lots of bad decisions. Shared clinician/patient decision making for these services is one thing; cost sharing–induced consumer decisions are something else altogether. Further, there are plenty of other available, but underused, approaches for addressing the overproduction of discretionary services that should be targeted to the providers.14 The flawed logic of relying heavily on consumer judgment, influenced by cost sharing, to reduce health care costs can be illustrated with preventive services. For basic prevention, there is relatively little information asymmetry; most people can understand the purpose of a screening test or the role of a medication in reducing the odds of an unwanted medical event, and they can understand what is involved in the actual intervention and its price. And while there is often scientific uncertainty as new studies contradict old ones, the uncertainty usually applies equally to clinicians and patients because the underlying science is unclear. In fact, sometimes clinicians may rely too much on evanescent evidence and become stubbornly "certain," whereas aware and informed patients adopt a reasonably skeptical attitude that influences shared decisions. Nevertheless, even consumer-directed plans, relying on high deductibles and health savings accounts (HSAs), often cover important prevention services with no cost sharing. That is, they paternalistically do not trust consumers to make the correct decision to spend their own money for these services. If consumers have to be protected from themselves in situations in which they actually do have the skills and clear interest to make "value-for-money" decisions, what possibly can be the rationale for relying on their judgment, by imposing substantial cost sharing, for other clinical decisions that are much more laden with clinical complexity and uncertainty? Pauly included routine care for people with chronic conditions in the services that are purchased relatively frequently by those with the conditions, which suggests that these people could gain sufficient experience to be smart buyers for these services. Yet the documented deficiencies in chronic care provision, even where there is no financial barrier, suggest that patients have not been smart shoppers of chronic care services.15
Contrasting tiered drug pricing and proposed higher cost sharing for services such as MRIs raises another practical consideration that Robinson’s model needs to address. Simply put, a market mechanism using variable direct patient payments should be favored when there are large numbers of transactions—services—with relatively low unit prices because it is much more efficient than trying to directly regulate the provision of those services.16 Some "regulatory" approaches directed to clinicians, such as prior authorization and centers of excellence, are too intrusive and administratively costly for high-volume services. Thus, on operational feasibility grounds, tiered drug pricing might be preferred to relying primarily on provider-side interventions. In contrast, relatively infrequent, high-price services, especially those that are often discretionary, are amenable to strategies targeted to providers. Ignoring operational feasibility is one reason why many managed care companies extended prior authorization, which was successful when used for certain elective hospitalizations and overused advanced imaging procedures, such as MRIs, even for routine, office-based referrals—justifiably incurring the wrath of patients and physicians alike.17 A form of Robinson’s differential cost sharing, distinguishing among services that are medically necessary or discretionary, already exists for emergency care. After the fact, it is relatively easy to label a diagnosis or intervention as having been discretionary. But ex ante, patients and physicians may truly be uncertain about what is going on. A patient with abdominal pain and vomiting may seek emergency room (ER) care and turn out to have appendicitis, a prototype of a condition that Robinson would not apply much patient cost sharing to because surgery for acute appendicitis is always considered "medically necessary." Already, cost sharing associated with ER visits that result in hospitalizations are often waived (although a first-day hospital deductible might apply). However, if the problem turned out to result from a different cause, treatable on an ambulatory basis, the standard ER cost sharing would apply. So what Robinson suggests already exists in a simple form. Nevertheless, practically, for other situations, it is hard to see how one would administer a program of differential cost sharing associated with symptoms, rather than firmly established diagnoses.
Despite practical barriers to implementation, smarter cost sharing makes sense as a cost containment strategy when included in comprehensive insurance coverage, whether in government-administered social insurance programs or in good private plans. But the goal of many advocates of consumer-directed care, including Scandlen, is to create a retail market for health care that sees a greatly reduced role for health insurance protection. They fundamentally reject the long-settled reality that most people seek insurance to protect against the risk of having to pay large bills and the inevitability that they will want to consume more care than they would without insurance.18 Observing vibrant price competition in other sectors, such as recreation, proponents of consumer direction want to restructure health care to rely more on price competition in retail provider markets, while minimizing the role of the insurance middleman. In its idealized form, a retail market would promote the development of so-called focused factories engaging in price competition for consumers over each disease and treatment.19 Surely focused factories have a role in a modern health care system. Indeed, it is instructive that some of the most prominent examples, such as the Shouldice hernia hospital in Toronto and the Coxa Hospital in Finland (hip and knee replacements), achieve their success in government-directed, social health insurance systems. The British National Health Service—the prototypical socialized health care system—plans to open fifty facilities targeted to specialized surgical procedures.20 So to the extent that centers of excellence and focused factories have merit, where, as Robinson instructs, there are significant scale and experience efficiencies in the clinical technology, those advocating retail purchasing markets have no unique claim to the concept.
My concern, ultimately, is less with economic issues of competition than with the values that consumer-directed care embodies. An inevitable objective of this model is to alter the role of physicians and other clinicians from that of trusted professionals acting in the best interests of their clients to that of vendors selling their wares in the market. Fuchs, as usual, understands that the patient/physician relationship is very different from the one we accept in commercial marketplaces: The production function for health is a peculiar one; it usually requires patients and health professionals to work cooperatively rather than as adversarial buyers and sellers. Mutual trust and confidence contribute to the efficiency of production. Thus the model of atomistic competition...often is not the right goal for health.21 With all of the demonstrable failures of medical professionals to act professionally, it is tempting to consider the patient/physician relationship an ancient artifact of a Marcus Welby era that is long gone. However, Hall and Havighurst importantly observe that consumer-directed care theory has yet to confront the reality of how patients actually make medical decisions. That is, they still make decisions as patients do, not as would-be consumers might. Just as most Americans want to be protected from health care costs through insurance, making moral hazard inevitable, most also want to place their trust in professionals to bear the primary responsibility for medical decisions, whether or not such trust is warranted. In fundamentally challenging this unexceptional expectation, advocates of consumer-directed care do not speak the language of the public, no matter how often they invoke the ideal of consumer sovereignty. It is one thing to try to empower patients to be less passive in the face of professional authority and to share decision making with their selected care professionals. It is quite another to turn patients into consumers persistently questioning the advice and recommendations of the clinicians from whom they seek care. Role of quality information. Perhaps nothing better exemplifies the difference between a patient and a consumer than the muted, but quite live, controversy over the decision by a former U.S. president to actually follow his physician’s recommendations about where to undergo coronary artery bypass graft (CABG). Market advocates seriously assert that a person in highly stressful, life-threatening circumstances should have sought out, understood, and relied on three-year-old data showing statistically significant, but clinically questionable, differences in mortality after CABG to reject his cardiologist’s advice about where to undergo the surgery.22 Perhaps the referring cardiologist provided faulty advice, but surely there were relevant considerations other than dated data, which, again, are more useful for management review and policy making than for individual patient decisions. And if the cardiologist’s advice was faulty, the focus of attention should be on improving that advice and the performance of the particular heart center, and not on putting the onus on patients, in effect, to become their own physicians. Using focus groups, the Foundation for Accountability found that patients typically do not want to use quality information to select their physicians; rather, they want their selected physicians to do better.23 Evidence-based medicine. In many ways, the consumer-directed movement is in conflict with developing progress toward reliance on evidence-based medicine. A prime example of the difference can be found in how different parts of the health care sector dealt with the dissemination and use of COX-2 inhibitors. There was never solid evidence that these drugs provided better pain relief than traditional nonsteroidal anti-inflammatory drugs (NSAIDs), such as ibuprofen. Patients who might benefit from preferential use of Vioxx, in particular, were only those at high risk for gastrointestinal bleeding.24 Kaiser Permanente, adhering to the evidence, developed formulary guidelines that limited use of COX-2 drugs. As a result, very few Kaiser patients were affected when Vioxx was shown to be dangerous.25 In contrast, vibrant direct-to-consumer (DTC) advertising and expert marketing to doctors achieved levels of Vioxx and Celebrex use that far exceeded what any prudent, informed review of the science would have supported.26 The physician advice-and-consent function, which calls on the physician to write the prescription, surely failed in this case. Because it is unlikely that we will put the DTC advertising genie back in the bottle in such a consumer-empowered society as ours, the needed reform is on the provider side—to figure out how to introduce group practice–style adherence to evidence into currently unaccountable, fee-for-service delivery.
The brave new world we are supposed to look forward to was candidly captured by a UnitedHealth executive: "The people who develop potato chips know just how crunchy the chip has to be and how much air to put in the bag, but we’ve been at a much cruder level in health care. We’re [now] thinking more like a retailer thinks."27 Consumer-directed care would have even hospital staff and clinicians thinking like retailers, presumably adopting commercial ethics, rather than the professional ethics of acting in patients’ best interests. As such, it is a much more basic challenge than other competition models to the prevailing ethos of health care. In the right hands, market competition ideas can be made consistent with this ethos. Alain Enthoven’s managed competition approach, for example, in the words of Uwe Reinhardt, was an attempt to "fuse a price-competitive framework for health care with production processes designed to produce medical treatments efficiently and with income transfers designed to achieve a desired level of social equity."28 This sounds fine to me, because it might have been done in ways that did not threaten the foundation of trust.29 It’s just that markets haven’t followed Enthoven’s vision—for lots of reasons.30 True consumer-directed health care—the kind Scandlen talks about, not just smarter cost sharing embedded within health insurance plans—is different and more ambitious. It looks to other sectors of the economy for guidance and inspiration. We should acknowledge that food and shelter are also essential human needs, and yet their provision relies on retail markets where no one invokes a sacred grocer/customer relationship.31 Nevertheless, most citizens do not want to restructure health care so that it resembles other parts of the economy, even if that means forgoing some of the consumer-friendly and price-competitive innovations that characterize much of the economy. Nor do we establish commissions to study "recreation disparities," or food or housing disparities, for that matter. To the extent that inequitable access to food or housing is a social issue at all (and New Or-leans suggests that it isn’t), the focus is on providing a decent minimum to everyone and not worrying about reducing access disparities. In contrast, society is very concerned about racial, ethnic, sex, and income disparities in health care and devotes a lot of time and resources to their reduction. Could it be that health care really is different?
Bob Berenson (rberenso{at}ui.urban.org) is a senior fellow at the Urban Institute in Washington, D.C.
This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||
Top
Smarter Cost Sharing
Theoretical Problems
