Statement of Michael E. Chernew, Ph.D., Professor of Health Care Policy, Harvard Medical School, Boston, Massachusetts

Testimony Before the Subcommittee on Health
of the House Committee on Ways and Means

May 14, 2008

Thank you, Chairman Stark, Ranking Member Camp, and members of the Subcommittee for inviting me to testify on the impact of cost sharing on outcomes in health care markets.  Rising health care costs represent perhaps the most important long-run challenge facing the American health care system and even the economy overall.  At the same time, we worry that too often the quality of care delivered by the health care system is below the level we would desire.  I believe many of you share my goal of finding ways to reform the health care system to control costs and improve quality. 

Today I am going to talk about the role that patient cost sharing at the point of service may play in achieving those goals. Requiring patients to pay more when they receive care is among the solutions purchasers have adopted to address the fiscal pressure represented by rising costs.  Relatively new health insurance packages, such as high deductible plans that may be accompanied by Health Savings Accounts, exemplify this trend. Yet the movement towards greater cost sharing by patients at the point of services extends much more broadly.  Patient copayment rates and deductibles have been rising even in more conventional plans. 

I will make two basic points.  First, patient cost sharing is neither good nor bad.  Its merits depend on the context.  In some cases cost sharing can promote efficiency and quality.  In other cases it can lead to inefficiency and poor health outcomes.  Second, a greater investment in clinical and behavioral research is needed to help us design systems that can use cost sharing and other tools to achieve our health care goals.

I would like to preface my remarks by noting that, as an economist, I believe that market mechanisms are, in general, the best way to achieve efficient allocations of resources.  In most settings consumers should determine what goods and services they desire and act accordingly in the market place. Over the course of our history reliance on free markets and consumer sovereignty has served us very well.  However, my general appreciation of markets does not imply a belief that they always work well.  There is a growing body of evidence in economics documenting deviations between consumer behavior and standard economic theory.  For example, contrary to standard economic models, evidence suggests that consumers are much more likely to participate in retirement savings programs if they are automatically enrolled, with the option of opting out, than if they must actively choose to participate.^[1]   Similarly, in contrast with standard economic models, evidence suggests that when consumers are given a wide choice of products (e.g. different varieties of jelly) they are less likely to purchase any product than when they are given only a few choices.^[2]  These paradoxes do not negate the merits of markets, but they do enrich our understanding of individual behavior and can suggest that policy interventions may improve welfare.     

Having studied health care markets for about two decades I believe that, despite my general faith in markets, health care markets are an instance in which we should be cautious about blindly relying on market principles.  There are a number of reasons health care markets are unique.  Perhaps most importantly, the outcomes associated with poor consumer decision making can be more serious, including death, than in other markets.  Furthermore, institutional details of decision making, including the complexity of information, increase the potential for undesired outcomes.  It seems unreasonable to expect a patient to choose between bare metal and drug eluding stents when the medical evidence is conflicting.  Even the choice of hospital or physician may be difficult because of the many attributes of different providers and because of complex provider-plan relationships.  For example, physician privileges may be limited to certain hospitals, plan provider networks often limit access to certain doctors and hospitals, and physician practices may be closed to new patients. These institutional details will limit the ability of consumers to respond to price signals. These decisions are even more difficult when patients are cognitively impaired, very emotional or stressed, or when they need to make decisions quickly.  One would not expect, for example, a patient suffering chest pain will be able to weigh tradeoffs between institutions prior to seeking care.  The role of physicians complicates the ability of patients to weigh options.  Certainly patients have a say in their care, but in many situations they are heavily influenced by physicians and it may be unlikely (perhaps appropriately in some cases) that they would respond to market signals if those signals conflicted with their physicians advice.  Finally, consumers desire protection against the financial risk of illness.  In situations in which cost sharing does not alter patient behavior, greater cost sharing does nothing to change overall spending and has no beneficial incentive effects.  It simply represents a tax on sick patients. In these instances, greater cost sharing has no beneficial incentive effects and just represents a tax on patients.  For these reasons, policy makers and private purchasers must consider the potential for unintended outcomes when promoting interventions such as greater consumer cost sharing. 

Cost sharing reduces utilization and expenditures

As with any good, the demand for health care services is responsive to price.  When patients are charged more for care, they will consume fewer health care services. Estimates from a randomized trial suggest that when patients were required to pay 95% of their care (up to an out-of pocket maximum that was based on their income) they reduced spending by over 30%.^[3] The responsiveness may be even higher as cost sharing requirements grow as a share of income.  To proponents of high cost sharing, this response is desirable.  They could rightly note that considerable evidence suggests that greater use of health care services is not consistently related to better outcomes and that it is likely we could reduce utilization and spending without adversely affecting the health of Americans.   In this view of the world, consumers, when faced with the correct incentives, would drive the system to efficiency as we believe they do in most other markets. 

When facing higher cost sharing in health care consumers forgo important services.

As much as it pains me to admit it, important aspects of standard economic models appear to be often violated in health care markets.  Specifically, economists often assume that if prices charged to consumers rise, individuals will forgo less valuable services and continue to consume services of high value.  Extensive evidence suggests that in health care markets this assumption may be incorrect in many instances. For example, the RAND health insurance experiment, which documented patient response to cost sharing, found that patients reduced utilization of services deemed clinically appropriate by the same amount as they reduced the use of services deemed clinically inappropriate.^[4]    

Similarly, more recent research has documented that relatively modest increases in cost sharing reduces utilization of important medications for managing chronic disease.^[5-11] For example, Goldman and colleagues report that a doubling of copayments reduced use of anti diabetes medications by patients with diabetes by 23% and reduced use of anti-hypertension medications by patients with hypertension by 10%.^[7]  Huskamp and colleagues report that when an employer increased cost sharing requirements by about $10 to $20 per prescription (depending on the exact medication), that 21% of patients stopped taking their medication for high cholesterol (compared to 11% in a control group).^[8]  Reducing copayment rates seems to have the opposite effect.  Research that my colleagues and I published in January found that reduction in copayments of about $10 per prescription increased patient adherence to treatment regimes for chronic disease.^[12]  Recent reviews of the literature confirm these conclusions.^{[6, 13]}

Interestingly, while a lot of attention has been devoted to measuring the quality of care in this country, we seldom appreciate the impact that greater cost sharing may have on quality of care measures.   For example, the Healthcare Effectiveness Data and Information Set (HEDIS) is a list of quality indicators maintained by the National Committee of Quality Assurance (NCQA).   Forthcoming work that my colleagues and I have done, examining a subset of those measures, suggests that higher cost sharing will reduce quality.^[14]  Other research that Trivedi and colleagues published supports that finding.^[15]  If we care about quality, and a truly hope we do, we must be concerned about the impact of cost sharing.

Not surprisingly, cost sharing may affect disparities in health care related to income.  We have recently published a study suggesting that the impact of higher cost sharing is greater among lower income individuals.^[16]  Specifically, we found that individuals living in low income areas were much more sensitive to price than individuals in high income areas.  For examples, patients with diabetes in low income areas were over three times more sensitive to costs when using blood pressure medication, a very important component of diabetes care, than patients in high income areas.  This is consistent with results from the Rand Health Insurance experiment that found the adverse health effects related to cost sharing were limited to patients with specific chronic diseases (hypertension and vision) and greater among low income individuals.^[17]    

It is important to assess how these results relate to health outcomes.  In theory we should expect to see adverse consequences of reduced use of high value services.  Evidence on this point is still developing, and conflicting evidence can be found, but I believe the best evidence suggests adverse consequences of higher cost sharing.  Hsu et al. report that higher cost sharing for prescription drugs had worse physiological outcomes (e.g. blood pressure), more visits to the emergency room, and even greater mortality.^[18]   The savings associated with reduced drug spending were almost completely offset by the higher non-drug spending.  Chandra et al. report offsets of lesser magnitudes, but the basic message, that high cost sharing can lead to worse compliance with important health care services and, in turn, result in worse health outcomes, is supported.^[19]

Proponents of cost sharing might argue that this evidence underscores the importance of patient education.  Certainly patient education is important (though I might add not costless).  While I believe education interventions can improve compliance with important services, I am skeptical that it can substantially reduce the price sensitivity of patients to higher cost sharing.  Our study of copay reductions that I referred to earlier, which demonstrated how patients responded to lower copayment rates, was conducted in a setting that already had a sophisticated care management intervention in which patients and physicians were contacted about their care and the results suggested the responsiveness to cost sharing was similar to that in other studies.^[12]    

It is important to recognize that the fact that consumers market poor decisions in health care markets does not mean that there are not settings where markets in health care work well, particularly in situations that are relatively straightforward and consumers have time to decide. Moreover, some consumers are undoubtedly more capable of successfully navigating markets than others.  Certainly when the stakes are high, some consumers can do a better job of making decisions than others.  Identification of patients or situations in which markets work well does not imply that market mechanisms should be used without modification in health care any more than identification of patients or situations in which markets cannot work implies markets should never be used.

Towards smarter cost sharing

The fundamental question is how we can design our system to recognize the failures of markets and heterogeneity of patients and treatments.  Purchasers and policy makers must strive to design benefit packages that recognize the variation in value that health care services offer and attempt to avoid creating financial barriers for access to high value services.  The paradigm of Value Based Insurance Design (VBID) reflects this approach, arguing that copays should be kept low for high value services.^[20]  

Several employers, insurers and benefit consulting firms have begun to adopt VBID style benefit packages.  For example, Pitney Bowes reduced cost sharing requirements for important chronic disease medications and reported very favorable results.  The University of Michigan designed a benefit package for employees and dependents with diabetes that focused on minimizing financial barriers to access for important services.  Insurers such as Aetna have developed a range of initiatives related to VBID, with ActiveHealth Management (a subsidiary of Aetna) using its sophisticated care manage information system as a platform to support VBID.  Hewitt Associates, a large employee benefit consulting firm has begun consulting with clients for such programs.  These are only a few examples, but they demonstrate the feasibility of such a clinically sensitive approach to cost sharing.

VBID programs are just in their infancy and are no panacea for all of the challenges facing the health care system.  Yet to the extent that consumerism, and more specifically cost sharing, is a part of the solution, VBID can help mitigate adverse effects. Moreover, VBID programs can support quality improvement initiatives by removing barriers to the services being promoted.

The potential of sophisticated cost sharing programs such as VBID depends on our ability to support the health services research upon which these programs rely.  Not only do we need the type of research that is encompassed by comparative effectiveness research, but we also need greater investment in the social science research that helps us understand patient behavior. Funding of such work will enable us to provide answers to the central questions concerning how we can design a better health care system capable of controlling costs, maintaining (or even improving) the quality of care, and providing patients with the autonomy to make decisions central to their wellbeing.

Summary

Greater patient responsibility for the costs of their care will undoubtedly be an important part of the healthcare system in the future.  However, details of the health care market suggest that cost sharing may have both beneficial and detrimental effects.  Proponents of cost sharing focus on situations in which there is over-consumption of care or consumers can be expected (but fail in practice) to shop for the provider offering the best price/quality combination.  In these cases, cost sharing can encourage efficient consumption of care.  However, in other cases, when care is appropriate or when consumers cannot respond to market signals, cost sharing can lead to worse outcomes.  Evidence suggests that in many situations cost sharing will reduce the likelihood that patients will consume appropriate services.  This could lead to hospitalization, emergency room visits, and even death.  Even if the cost sharing does not alter patient behavior, the associated cost shifting reduces wellbeing.  Specifically, consumers' desire to mitigate some of the financial risk associated with illness suggests that it is difficult to rely on the price mechanism to allocate resources in the same manner as in other markets. If we charge patients the full cost when they need health care services, we would be transferring a substantial risk to patients that is generally not considered by economists to be optimal.  More sophisticated cost sharing programs, supported by rigorous clinical and health services research are needed to balance our need to control spending with our desire to get the most from our health care system.

Thank you very much for the opportunity to speak with you today and I welcome your questions.

 References

1.       Choi JJ, Laibson D, Madrian B, and Metrick A, Optimal Defaults and Active Decisions. 2005. NBER working Paper #11074.

2.       Iyengar SS and Lepper MR, When Choice Is Demotivating: Can One Desire Too Much of a Good Thing? Journal of Personality and Social Psychology, 2000. 79(6): p. 995-1006.

3.       Manning WG, Newhouse JP, Duan N, Keeler EB, and Liebowitz A, Health Insurance and the Demand for Medical Care: Evidence from a Randomized Experiment. The American Economic Review, 1987. 77(3): p. 251-257.

4.       Siu AL, Sonnenberg FA, Manning WG, Goldberg GA, Bloomfield Es, Newhouse JP, et al., Inappropriate use of hospitals in a randomized trial of health insurance plans. New England Journal of Medicine, 1986. 315(20): p. 1259-66.

5.       Gibson TB, McLaughlin CG, and Smith DG, A Copayment Increase for Prescription Drugs: the Long-term and Short-term Effects on Use and Expenditures. Inquiry, 2005. 42(3): p. 293-310.

6.       Gibson TB, Ozminkowski RJ, and Goetzel RZ, The effects of prescription Drug Cost Sharing: a Review of the Evidence. American Journal of Managed Care, 2005. 11(11): p. 730-740.

7.       Goldman DP, Joyce GF, Escarce JJ, Pace JE, Soloman MD, and Laouri M, Pharmacy Benefits and the use of drugs by the chronically ill. The Journal of the American Medical Association, 2004. 291(19): p. 2344-50.

8.       Huskamp HP, Deverka PA, Epstein AM, Esptein RS, McGuigan KA, and Frank RG, The Effect of Incentive-Based Formularies on Prescription-Drug Utilization and Spending. New England Journal of Medicine, 2003. 349(23): p. 2224-232.

9.       Landsman PB, Yu W, Liu X, Teutsch SM, and Berger ML, Impact of 3-tier Pharmacy Benefit Design and Increased Consumer Cost-Sharing on Drug Utilization. The American Journal of Managed Care, 2005. 11(10): p. 621-8.

10.     Soumerai SB, McLaughlin TJ, Ross-Degnan D, Casteris CS, and Bollini P, Effects of a limit on Medicaid drug-reimbursement benefits on the use of psychotropic agents and acute mental health services by patients with schizophrenia. New England Journal of Medicine, 1994. 331(10): p. 650-5.

11.     Tamblyn R, Laprise R, Hanley JA, Abrahamowicz M, Scott S, Mayo N, et al., Adverse events associated with prescription drug cost-sharing among poor and elderly persons. The Journal of the American Medical Association, 2001. 285(4): p. 421-9.

12.     Chernew ME, Shah MR, Wegh A, Rosenberg SN, Juster IA, Rosen AB, et al., Impact Of Decreasing Copayments On Medication Adherence Within A Disease Management Environment. Health Affairs, 2008. 27(1): p. 103-112.

13.     Goldman DP, Joyce GF, and Zheng Y, Prescription Drug Cost Sharing. Associations with Medical Utilization and Spending and Health. The Journal of the American Medical Association, 2007. 298(1): p. 61-27.

14.     Chernew ME and Gibson TB, Cost Sharing and HEDIS Performance. Medical Care Research and Review, Fourthcoming 2008.

15.     Trivedi AN, Rakowski W, and Ayanian JZ, Effect of cost-sharing on screening mammography in Medicare Health Plans. New England Journal of Medicine, 2008. 358(4): p. 375-383.

16.     Chernew ME, Gibson TB, Yu-Isenberg K, Sokol MC, Rosen AB, and Fendrick AM, Effects of Increased Patient Cost Sharing on Socioeconomic Disparities in Health Care. Journal of General Internal Medicine, 2008.

17.     Newhouse JP, Free for All. Lessons from the RAND Health Insurance Experiment. 1996, Cambridge, MA.: Harvard University Press.

18.     Hsu J, Price M, Huang J, Brand R, Fung V, Hui R, et al., Unintended consequences of caps on Medicare drug benefits. New England Journal of Medicine, 2006. 354(22): p. 2349-59.

19.     Chandra A, Gruber J, and McKnight R, Patient Cost-Sharing, Hospitalization Offsets, and the Design of Optimal Health Insurance for the Elderly. 2007. NBER Working Paper No. W12972.

20.     Chernew ME, Rosen AB, and Fendrick AM, Value-Based Insurance Design Health Affairs, 2007. 26(2): p. w195-w203