Over-utilization, Under-utilization… How would one spot an excessive health care treatment if it walked by?

Dahlia K. Remler, Ph.D.

School of Public Affairs

Baruch College

The City University of New York

and NBER

Kamiar Khajavi**, JD, MD

Formerly of Department of Public Health

Weill Cornell Medical College

and presently in the private sector

Work in Progress

Do not cite or circulate without the authors’ permission

October 20, 2008

Note: Khajavi has not seen recent drafts, which have substantial changes. Thus, Remler, and not Khajavi, bears responsibility for some portions of this current draft.

**Khajavi’s work was done primarily while he was at Cornell; he is now in the private sector.

We thank Diane Gibson, Marthe Gold and participants in seminars at the Cornell Department of Public Health, Pfizer Pharmaceutical Outcome Group and the 2005 International Health Economics Association meeting for comments on earlier drafts. We also thank Karl Kronebusch and many seminar participants for comments on a closely related paper. We thank Joe Newhouse for inspiring the title in a very different context. Khajavi gratefully acknowledges funding from Pfizer Pharmaceuticals. All responsibility for the content and opinions in the paper resides with the authors and no endorsement or responsibility for the funder or commentators is implied.

Abstract:

We frequently hear about over-utilization of health care—excessive health care—as a primary culprit of our health care cost crisis. Some advocate cost-sharing and policies based on it, like consumer driven health plans, as a solution. We also frequently hear about under-utilization of health care. Others, on the opposite side of the fence, advocate less cost-sharing and more insurance coverage, as a solution. Of course, we may have both problems simultaneously, but among different groups. However, implicitly, perhaps even unconsciously, the different sides have different definitions of excessive care. At one extreme, a health care treatment is deemed excessive if the recipient is unwilling to pay out of pocket. Nyman (2003, 2007) has challenged what he describes as the traditional analysis of health insurance, finding it accords with the extreme view above. At the other extreme, any health care treatment expected to have any positive benefit is deemed not excessive, no matter how great its cost. This view ignores economic (and political) reality. To the extent a middle view is articulated, it is usually based on some form of cost-benefit analysis (CBA). A clear conceptual definition of what constitutes excessive care would help prevent both sides talking past one another and support a constructive health care policy debate about cost containment. We extend Nyman’s work in several ways. First, we develop a simple contingent claims contract (CCC) view of the value of health insurance that is accessible to a wide policy audience. In it, if consumers are willing to pay the higher premium needed to ensure coverage of care under particular circumstances, then that care is not excessive. Second, we illustrate how the identification of the term “market” with out-of-pocket payment is misguided, because it ignores the health insurance market and the inter-relatedness of health care and health insurance. Third, we illustrate, with specific medical examples, the extent to which the value of a specific medical treatment depends on a tremendous clinical detail and the consequences for cost containment practices. Fourth, we contrast the CCC model with CBA, to which it is related. We find that CCC model fits better with an individualistic and market-based society because: (a) it does not require all benefits be expressed in dollars; (b) it is individualistic, rather than collectivist.

I. Introduction

“[T]here’s a systemic problem. Health-care costs are on the rise because the consumers are not involved in the decision-making process. Most health-care costs are covered by third parties. And therefore, the actual user of health care is not the purchaser of health care. And there’s no market forces involved with health care.”

-- President George W. Bush, Third Bush-Kerry Debate of 2004 Presidential Campaign, October 13, 2004. (

“Market forces”, individual choice, consumerism and other such terms have figured prominently in the health policy debates of the last decade. Moreover, the policies identified with those terms, such as health savings accounts (HSAs) and consumer driven health care (CDHC), are supposed to increase out-of-pocket payment by patients at the point of service and reduce payments by insurers (Atlas 2005). The policies are often described as creating “more skin in the game.”[1] Supporters envision greater competition in the health care services market, driven by empowered and informed patient/consumers.

This market rhetoric, however, blurs the distinction between consumer choice and market forces at the level of health care and consumer choice and market forces at the level of health insurance. These are not the same and are often in conflict (Remler, Brown and Glied). Enthoven’s (1980) original managed competition vision was based on market competition at the health insurance plan level, driven by ordinary individuals who might or might not end up using health care services. Enthoven’s vision has been partially realized in managed care, but the ensuing backlash against managed care fueled interest in cost-sharing based CDHC. Thus, interest in tradeoffs at the patient level, rather than at the insurer level, increased.

Underlying the out-of-pocket based proposals is the widely taught and widely repeated moral hazard rationale: With insurance, patients pay only a fraction (if any) of the cost of health care services. Therefore, they choose to consume health care treatments whose marginal benefits (to the patients) exceed the marginal costs (to society). Such care is deemed excessive. It is said to make society worse off than if only care the patient would choose to purchase out of pocket were consumed. In the tradition story, the magnitude of such excessive care is assumed to be large and a significant contributor to our health care crisis. I will term both the policies based on this perspective and their supporters “out-of-pockets.”

The traditional excessive care story, however, ignores the perspective of those purchasing insurance. The basic value of insurance comes partially from financial protection in the event of illness but also from increasing the use of beneficial medical care beyond what patients would or could purchase if they had to pay entirely out of pocket. Nyman (2003, 2007) has shown and stressed that the widely taught form of traditional model has simplifications that preclude the second value of health insurance, which he has termed the access value. In effect, the traditionally taught model compares the benefits to the patient to the costs to society. It ignores the ex ante perspective, the desirability of transferring money to unhealthy states in order to increase health care usage—a major reason that health insurance is valuable. In the standard simplified model, the optimal level of health care services is that which would be purchased with no insurance whatsoever and payment being entire out of pocket; anything else is “excessive” and an undesirable side-effect of the financial protection role of insurance.

While many commentators today, including Fuchs and Emanuel are decrying our “perfect storm of over-utilization” (2008), others are decrying our tremendous problems of under-utilization (XX cite). Of course, it is possible to have both problems simultaneously among different groups, with under-utilization being present among the uninsured. However, some of the different perspective is probably among the same people. Much of the current under-utilization is said to be due to precisely the kind of cost-sharing advocated by the “out-of-pockets.”

Some of those who decry under-utilization are what Glied (1997) has termed medicalists: they believe that there are well-defined health care needs and everyone has a right to have those needs provided. Implicitly, if not always explicitly, many medicalists hold a belief at the opposite end of the spectrum from the “out-of-pockets.” Specifically, they reject the notion of any role for costs and believe that any health care treatment expected to have any positive value should be covered by insurance. They further believe that everyone should insurance to cover all such care is a right that everyone should have. Thus, no health care treatment with an expected positive value, no matter how small, is excessive. I will term these advocates and the policies they support the “at any costs.”

While attractive, this opposite extreme ignores both economic and political reality. Funding everything effective for everyone is beyond our means, certainly on a global scale and probably on a national scale. Trying to come close would require taking funds away from other valuable sectors, such as education and infrastructure, to fund less valuable aspects of health care. Politically, trying to come close requires far more redistribution that United States citizens have been willing to do and a level of redistribution that Europeans are moving away from.

Cost-effectiveness analysis (CEA) provides the only explicit definition of excessive care which takes an intermediate position and avoids both the “out-of-pocket” and “at any cost” extremes. CEA trades off the costs to society of health care with the benefits to the society as a whole—the benefits to ordinary citizens who might or might not get a particular condition (Gold et al 1996). While specific uses of CEA have major informational and technical problems, the conceptual definition of excessive care is clear and intuitive.

CEA is widely used to make coverage decisions by the National Health Service in Britain and in some other countries (Pearson and Rawlins 2005). However, CEA has made few inroads in the US, although its rise is both advocated and predicted (e.g., Garber 2004). Some allege that the resistance to CEA in the US is based on opposition to “rationing” (any limitations on effective health care) but Gold et al 2007 have some evidence that this is not so. We contend that part of the difficulties CEA has in gaining acceptance in the USis its collectivist perspective that is at odds with American’s emphasis on individualism, markets and choice.

Managed competition implicitly contains an intermediate definition of excessive care, but it suffers from the lack of explicitness. Managed competition envisions consumers trading off the premiums of plans (their costs) with the benefits they would provide in health circumstances that might or might not occur (Enthoven 1980). The benefits encompass both the quality of the care and the breadth of coverage. This intermediate perspective, however, gives neither consumers nor plans any guidance about how to make specific tradeoffs. Indeed Eddy (1991a, 1991b) advocates that plans use CEA to make coverage decisions. While the backlash against managed care has many causes, a critical one is that the plans were trying to skimp on care to save money (Enthoven and Singer 1998). Americans did not like the idea of some big anonymous entity trying to save money, even though they also prefer plans with lower premiums. The lack of transparency and the anonymity of the tradeoff are certainly problematic. However, a significant problem is that plans lack language with which to articulate what coverage limitations might be legitimate, unless the care is shown to be utterly ineffective (Garber 2004). We contend that managed competition would be more acceptable if limits in coverage could be presented to Americans as choices that they freely and knowledgably made.[2]

Currently in the US, health care policy discussion often consists of the two extreme views, the “at any costs” and “out-of-pockets”, talking past one another without common language or paradigms. They use terms like over-utilization and under-utilization but they share no common definitions. The one conceptually clear intermediate definition, CEA, is based on a collectivist paradigm. Managed competition, another intermediate perspective, has no clear definition of excessive care, and no way to teach consumers about how to make tradeoffs.

Productive discussion of health care policy requires a conceptual definition of excessive care that all sides can accept and that fits with US political, philosophical and economic norms. To create such a definition, we extend and simplify Nyman’s (2003, 2007) work. We develop a simple contingent claims contract (CCC) model and use it to create a conceptual definition of excessive care that is accessible to a wide policy audience. Specifically, our proposed definition is that if consumers are willing to pay the higher premium needed to ensure coverage of care under particular circumstances, then that care—that coverage-- is not excessive.

We rush to stress that we do not believe that plans could articulate every coverage situation—indeed we stress the overwhelming impossibility of this. We are even more sure that consumers could not understand all the contingencies. Nonetheless, we contend that the conceptual definition helps a great deal. It helps avoid the extremes and provides intellectual clarity for a center vision. It can provide legitimacy to CEA and managed competition.

The paper is structured as follows. In section II, we develop the contingent claims contract (CCC) model of health insurance and illustrate the value of health insurance. The model inevitably makes a variety of simplifications but we develop it fully enough to make all of those simplifications explicit. In section III, we illustrate why health insurance inevitably causes genuinely excessive care, along with desirable increases in care, and illustrate the definition of excessive care. In section IV, we discuss the effect of relaxing some of the model’s simplifications and how that complicates applying our definition of explicit care. In section V, we discuss the relationship of the CCC model to CEA. In section VI, we return to the big picture of health care policy, discussing how CCC furthers managed care and fits into the big picture of health care perspectives. Our proposed definition and its value can be understood without extensive discussion of the history of thought about moral hazard in health care. However, that history has shaped policy and parts are likely to be familiar to readers. Therefore, in an appendix, we relate our modelto the history of the use of the term moral hazard in health care and to Nyman’s work.

II. Contingent Contracts: Realizing the Value of Health Insurance

The purpose of our model is to support our proposed definition for excessive care caused by insurance: are you willing to pay the higher premium needed to provide coverage under particular circumstances? If so, it’s not excessive. Obviously, such a definition neglects many philosophical and technical issues. In this section, we lay out the basic contingent claims contract (CCC) model of health insurance in some detail.[3] In order to make the assumptions clear and allow comparison to other models, it is more technical than is essential. We also provide an accessible simple numerical example to support the definition. Our model builds on and simplifies Nyman’s model (1991a, 1999b, 2001, 2003, 2004, 2007) and in the appendix, we discuss the relationship of our model to his. In section V, we address many of the issues neglected due to our assumptions.

The CCC model plays multiple roles. First, it captures the essence of what health insurance is and how it works. Second, it illustrates what an ideal health insurance contract, illustrating what we would like to have but cannot, for reasons of medical complexity. Finally, and most importantly, the model provides a conceptual definition of the first-best for health insurance coverage and medical care consumed. With such a definition, we can conceptually disentangle care we would like policies to eliminate or reduce from care that we want policies to enable or promote.

We take the traditional perspective of economists: individualistic and utilitarian. So, individual preferences are assumed to be given and we consider how to best pursue those preferences in light of limited resources. Most importantly, our perspective precludes the idea that health care is a “basic human right” and neglects idea of distributive justice. In some sense, the individualistic perspective facilitates inequality However, less restrictive forms of the model which incorporate heterogeneity in wealth allow can address redistribution and mechanisms to accomplish it. We return to the issue of equity in section V.

Formal CCC Model

The most restrictive assumptions of our model are that there are no selection in insurance markets or inequalities in wealth and income. Specifically, our model assumptions are as follows: First, individuals are assumed to be identical in income or wealth and identical in their tastes for medical care and other goods. Second, neither behavior nor prior medical care can affect an individual’s risk of illness. Third, medical care is a purely private good, with no externalities: care such as vaccines or mandatory treatment of infectious diseases, which can prevent individuals from infecting others, is not included. Fourth, all individuals are identical in their initial-- or ex ante -- risk of illness. Thus, each person is ignorant ex ante of his medical or family history. Finally, and most importantly, there is no real notion of time: uncertainty does not unfold over time but in one single shot. Thus, insurance does not consist of repeated limited-time contracts but rather consists of a single one-shot contract. We return in section IV to some of these complexities.