Slowing the growth of health care spending is critical to improving the fiscal health of the country. A key question is whether the various approaches to patient cost-sharing, such as copayments, deductibles, and higher rates of coinsurance can be an effective means of lowering health care costs and discouraging overutilization without causing adverse health outcomes. For policymakers, employers, and health plans, a “one size fits all” approach to cost-sharing is unlikely to have the intended effects and could, in fact, generate higher costs and negative health effects. Evidence suggests that a finely targeted approach holds the greatest promise for success.
As part of its Synthesis Project, the Robert Wood Johnson Foundation recently released a comprehensive report, “Cost-Sharing: Effects on Spending and Outcomes.”1 The goal of the Synthesis Project is to gather research findings on a specific health policy question, weigh the evidence and strength of the findings, and give policymakers reliable, objective information to inform complex policy decisions. This report on cost-sharing is particularly relevant in the current policy environment. With the passage of the Patient Protection and Affordable Care Act (Affordable Care Act or ACA), there is a need for evidence-based information on the effects of cost-sharing. Rapidly rising health care expenditures have led some decision-makers to seek greater cost-sharing from consumers as a way to engage them more fully in the health care decision-making process, discourage unnecessary utilization, and reduce costs. However, there is mixed evidence on the ultimate effect of cost-sharing on consumer behavior and a concern that some populations are more negatively impacted by increased cost-sharing.
Although consumers have seen an increase over time in the dollar amount of their total out-of-pocket expenses, total health care costs have increased at a faster rate. As a result, total consumer out-of-pocket spending declined as a percentage of total private health costs from 33 percent in 1990 to 25 percent in 2002.2 Several features of the ACA will eliminate or limit patient cost-sharing for particular services, and new regulations governing insurance plans sold in health insurance exchanges will set limits on patient cost-sharing across four levels of coverage. These changes are designed to give patients who receive their coverage outside of the employer-sponsored market greater clarity and standardization in their benefits, along with a cap on out-of-pocket costs ($5,950 per year for an individual and $11,900 for families) to provide protection against catastrophic expenses.3 The ACA will eliminate copayments or cost-sharing for certain preventative care services that receive an “A” or “B” rating from the United States Preventative Services Task Force (e.g. services that have a high or moderate certainty of showing a moderate or substantial net benefit). The law also defines essential benefits for insurance packages sold in exchanges and in the individual and small-group market, and offers cost-sharing assistance for low-income patients for certain health care services.4 The expectation is that by removing the cost-sharing burden for preventative care, higher costs (such as those associated with preventable hospitalizations) will be avoided.
As policymakers continue to refine cost-sharing requirements associated with ACA provisions, information from the Synthesis Project report can help inform discussions about the effects of placing a greater share of health care costs on the consumer.
Cost-Sharing and Total Health Spending
The foundation for the report’s analysis is the RAND Health Insurance Experiment (HIE). Although the experiment was designed and conducted in the 1970s, it remains the seminal study examining variations in health insurance design. The HIE showed that higher rates of coinsurance led to declines in medical care use. However, the decline was the result in failure to initiate care; once patients sought care, the intensity of services and attendant cost was largely unaffected. The HIE findings also showed that increased cost sharing reduced both necessary and unnecessary services.
The health care landscape has changed since the HIE, but its findings have important implications for total health expenditures in the United States. The distribution of health costs is highly skewed—a large amount of spending is allocated to a small proportion of the population, largely the sickest individuals. While greater cost-sharing may lead this population to limit visits to doctors’ offices, once they seek care the amount and intensity of services they receive reflects physician preferences and norms of care. Cost-sharing does not significantly lower overall expenditures because utilization and expenses for this population are largely in the provider’s control. The HIE showed that reductions in patient-initiated care also predominantly come from the healthiest segment of the population who only account for three percent of the nation’s total health expenditures. A reduction in spending among this group will not significantly lower national health costs.
[back to top]
Cost-Sharing and Low-Income Populations
The Synthesis report discusses the strong evidence that low-income populations are disproportionately affected by increased cost-sharing, which can have adverse financial and health effects for this group.
Featured in the Synthesis report is HCFO-funded work led by Jeanene Smith, M.D., which examined how benefit reductions and increased cost-sharing impacted the Oregon Medicaid program. Her study found that while the introduction of cost-sharing decreased the utilization of health care services, the mix of services became more expensive and there were no budgetary savings.5 These findings, revealing the unintended consequences of shifting the health care cost burden to vulnerable individuals, are consistent with other work on low-income populations and cost-sharing. Another study looked at health care utilization after the implementation of increased prescription drug cost-sharing for low-income populations in Quebec. The study found that the population experienced a higher rate of emergency department visits after the implementation of greater drug cost-sharing, thus increasing total spending.6
[back to top]
Cost-Sharing and Prescription Drugs
In an effort to control costs, health plans and pharmacy benefit managers have spent the last several years implementing tiered systems of drug coverage. Commonly under this structured system, the lowest tier (usually generic drugs) will have little or perhaps no copayment, while more expensive brand-name drugs will require higher cost-sharing by the consumer. A number of studies have examined these payment structures and their effect of drug utilization and adherence. The studies have established that there is an association between increased cost-sharing for prescription drugs and a decrease in their use.7
Policymakers often support the use of generic drugs as a way to lower overall health care expenditures, as well as lower drug expenses for individuals. However, the evidence is mixed on whether patients are more likely to switch to generic drugs when faced with higher copayments for brand-name drugs. In a HCFO-funded study led by Richard Frank, Ph.D., patients were more likely to select different brand-name drugs with lower copayments (rather than generics) when confronted with increased numbers of tiers or higher copayments.8 In another study in which payments were increased for brand-name but not generic drugs, patients responded by lowering their use of both classes of drugs, suggesting a lack of understanding of the chemical equivalence of generic drugs.9
Current HCFO research is examining the effects of benefit design and cost-sharing in the Medicare Part D prescription drug program. Jack Hoadley, Ph.D. is examining generic substitution within Part D and if plan-level benefit design and formulary characteristics are associated with the use of generics. His project will provide policymakers with a clearer picture of the factors that influence generic utilization and inform efforts to create cost savings within Part D.10 Under another HCFO grant, Cindy Thomas, Ph.D. is studying the impact of state Medicaid policies to support Part D for dually eligible beneficiaries. She is examining the impact of copayment assistance for this low-income population on outcomes and overall health care utilization.11 Like the Hoadley study, the Thomas research will add to the evidence base on this important topic and help provide a clearer picture of the effects of cost-sharing on health outcomes and costs.
[back to top]
Future Directions: Value-Based Insurance Design
While the studies to date fail to provide one clear path for the use of cost-sharing as a tool to reign in costs, the evidence does suggest that carefully targeting cost-sharing may be effective. Two trends are currently being tested―increased offering of consumer-directed health plans (CDHPs) and the use of value-based insurance design (VBID).
Through its offering of CDHPs, purchasers are moving toward higher deductibles and copayments in both preferred provider organization and health maintenance organization insurance products. These initiatives are part of an effort to incentivize consumer cost-consciousness and efficient delivery of care.12
A second innovation, which the Synthesis report cites as an understudied area of promise for reducing costs and improving outcomes, is VBID. The core premise of VBID involves removing or eliminating financial barriers to high-value drugs and services. Conversely, cost-sharing can be increased for low-value services. The overall goal is to improve compliance and health, and avoid the potentially costly complications of non-adherence. Ultimately, plans that utilize features of VBID could tailor benefits for patients with certain medical conditions (particularly chronic conditions that involve a high degree of maintenance), or, more broadly, eliminate barriers to high-value services for all participants.
Two HCFO-funded studies will add to the evidence base on VBID. In the first study, Niteesh Choudhry, M.D., Ph.D., has found that while VBID has potential to improve the overall value of health care, there needs to be a greater effort to target copayment reductions effectively. Additionally, VBID faces hurdles due to the fragmented system of care in the United States. Payers make an initial outlay to implement the program but are not always able to recoup savings from its positive effects.13 In the second study, Matthew Maciejewski, Ph.D., is examining “Medication Dedication,” a program of BlueCross Blue Shield of North Carolina that eliminated copayments for generic medications and reduced copayments for certain brand-name drugs for several chronic conditions. His preliminary results have shown that the program improved medication adherence among the study population, as compared with patients whose employer did not offer a similar program; but more evidence is needed on longer-term adherence trends and spending.14
[back to top]
The Synthesis Project report concludes that there are still many questions unanswered on the effects of patient cost-sharing. In particular, there are gaps in knowledge around the long-term effects of cost-sharing on health, the specific types of services that are reduced when patients face greater cost-sharing, and the types of interactions that occur between cost-sharing and different types of health insurance. While the studies analyzed in the report confirm the broad conclusion first put forth by the HIE that demand for most health care services is price-sensitive, much is still unknown on how best to structure cost-sharing without creating unintended adverse cost and health consequences.
As payers and patients struggle with the rising cost of health care, cost-sharing initiatives will remain in the spotlight as a potential tool to control costs and encourage responsible utilization. Policymakers implementing ACA initiatives will need to balance the evidence on the advantages and drawbacks of patient cost-sharing as they work to develop effective insurance coverage.
Details on studies led by HCFO grantees are available at http://www.hcfo.org. The report on cost-sharing, as well as other reports in the Robert Wood Johnson Foundation’s Synthesis Project can be found at http://www.rwjf.org/pr/synthesis.jsp.
[back to top]
Related HCFO Grants
Title: Surviving the Perfect Storm: Impacts of Benefit Reductions and Increased Cost-Sharing in a Medicaid Program
Grantee Institution: Office of Oregon Health Policy and Research
Principal Investigator: Jeanene Smith, M.D.
Grant Period: June 1, 2004-August 31, 2006
How have benefit reductions and increased cost sharing impacted the Oregon Health Plan (OHP)? The researchers examined: (1) impacts on economic viability, including whether cost savings accrue to Medicaid or whether additional costs will be incurred as beneficiaries shift from one benefit to another; (2) impacts on access, including whether access and continuity of care will be compromised as a result of cost sharing and benefit reduction strategies; and (3) impacts on coverage, including the degree to which Medicaid beneficiaries leave the program due to these changes. The objective of this study was to inform state decision makers who continue to seek efficient cost-saving strategies and consider competing approaches for maintaining and rebuilding benefits following reductions in Medicaid and reshaping publicly financed health care.
Title: The Impact of Pharmaceutical Formularies on Prescription Drug and Health Care Costs and Utilization
Grantee Institution: Harvard University
Principal Investigator: Richard Frank, Ph.D.
Grant Period: May 1, 2001-April 30, 2004
What are the effects of a health plan instituting a three-tiered co-payment (TTCP) financing mechanism on prescription drug spending, total health care spending, and patients' compliance with treatment protocols and quality of care? The study, conducted by researchers at Harvard University in conjunction with Merck-Medco, involved an analysis of Merck-Medco administrative, medical and pharmaceutical claims, and encounter data. The researchers investigated the effects of the three-tier co-payments on drug use and costs for both drugs and other health care services as well as the effects of the three-tier formulary on patterns of care for patients diagnosed with depression, congestive heart failure, and hypercholesterolemia. This study informs public and private policymakers - particularly those involved in designing proposals for adding a prescription drug benefit to Medicare - on the range of implications a three-tier copay strategy for prescription drug cost containment may have for patients, plans, and the market.
Title: Generic Substitution within a Class of Drugs for Medicare Part D Plans
Grantee Institution: Georgetown University
Principal Investigator: Jack F. Hoadley, Ph.D.
Grant Period: April 1, 2010-March 31, 2011
The researchers will examine the extent of generic drug use in Medicare’s stand alone drug plans for three classes of drugs: cholesterol drugs, antidepressants, and hypertension drugs. They will determine whether plan-level benefit design and formulary design characteristics and the program’s overall design are associated with the use of generics. Specifically, they will examine four research questions: 1) are shifts to generic alternatives within a drug class influenced by differences in benefit and formulary design; 2) does the impact of benefit and formulary design features differ between beneficiaries eligible for the Part D program’s low-income subsidy (LIS) and other, unsubsidized beneficiaries; 3) do effects vary by drug and drug class; and 4) does the coverage gap design feature of the Part D program affect the use of generics? The objective of this project is to provide policymakers with a better understanding of the forces that influence generic utilization and the potential opportunities to create health care cost savings within the Part D program.
Title: Impact of State Policies Supporting Medicare Part D for the Dually Eligible
Grantee Institution: Brandeis University
Principal Investigator: Cindy Parks Thomas, Ph.D.
Grant Period: May 1, 2010-April 20, 2012
The researchers will assess two state Medicaid policies - co-payment assistance to reduce cost sharing and beneficiary centered assignment. They will examine the impact of these policies for the dually eligible population on health outcomes (health care utilization and sentinel events), beneficiary switching among plans, continuity of drug treatment, and Medicare program costs. The researchers will compare beneficiaries in six states that provide full co-payment assistance to those states without such assistance. They will also compare beneficiaries in Maine, the only state with CMS-approved beneficiary centered assignment for dually eligible beneficiaries, to similar beneficiaries in other states. The objective of this project is to inform the Medicare program and state policymakers on the impact of the Part D benefit on dually eligible beneficiaries.
Title: Factors Influencing the Success of Value-Based Insurance Design Programs
Grantee Institution: Brigham and Women’s Hospital
Principal Investigator: Niteesh K. Choudhry, M.D., Ph.D.
Grant Period: May 1, 2010-October 31, 2011
The researchers will conduct a series of natural experiments examining value-based insurance design plans (VBID) implemented by CVS Caremark, a large pharmacy benefit manager, on behalf of numerous clients. They will evaluate the impact of different VBID characteristics on patterns of adherence, discontinuation and use of statins by (1) developing a system to classify differences in VBID plan characteristics that may influence the ability of these plans to stimulate the use of evidence-based medications; (2) surveying plan administrators and plan sponsors to identify additional plan features that might confound the relationship between VBID plan characteristics and medication use; and (3) using pharmacy claims to determine which VBID design features (and combinations thereof) most effectively stimulate appropriate medication use. The objective of this project is to determine the factors that influence the success of value-based insurance design plans (VBID) and to develop “best practices” for future implementation.
Title: The Patient and System Benefits of Value-Based Insurance Design
Grantee Institution: Duke University
Principal Investigator: Matthew L. Maciejewski, Ph.D.
Grant Period: April 1, 2010-March 31, 2011
The researchers will examine “Medication Dedication,” a BlueCross BlueShield of North Carolina (BCBSNC) program that eliminated copayments for generic medications and reduced copayments for brand-name medications to treat hypertension, congestive heart failure, hyperlipidemia and diabetes. The researchers will determine whether VBID improved medication utilization behavior among BCBSNC enrollees with hypertension during the first two years of the program and address the following research questions: (1) did the initiation of Medication Dedication impact medication initiation, switching and adherence among program participants with hypertension; and 2) did Medication Dedication impact inpatient, outpatient or emergency room health services utilization and expenditures among program participants with hypertension. The objective of this project is to explore the business case for value-based insurance design (VBID) and to inform benefit design changes and cost-containment strategies being considered by insurers and Medicare.
[back to top]
1. Swartz, K. “Cost-Sharing: Effects on Spending and Outcomes,” Synthesis Project, The Robert Wood Johnson Foundation, December 2010. Also see http://www.rwjf.org/pr/product.jsp?id=71583
2. Goff, V. “Consumer Cost Sharing in Private Health Insurance,” Issue Brief, No. 798, National Health Policy Forum, May 14, 2004.
3. Rabin, R.C. “In Health Law, a Clearer View of Coverage,” The New York Times, May 17, 2010.
4. Rosenbaum, S. “Value-Based Health Care Purchasing: Essential Health Benefits and State Health Insurance Exchanges,” HealthReformGPS, January 25, 2011. Also see http://www.healthreformgps.org
5. Wallace, N.T. et al. “How Effective are Copayments in Reducing Expenditures for Low-Income Adult Medicaid Beneficiaries? Experience from the Oregon Health Plan,” Health Services Research, Vol. 43, No. 2, 2008, pp. 515-530. Also see http://www.hcfo.org/files/hcfo/HCFOfindings0708.pdf
6. Tambryn, R. et al. “Adverse Events Associated with Prescription Drug Cost-Sharing Among Poor and Elderly Persons,” JAMA, Vol. 285, No. 4, 2001, pp. 421-429.
7. Goldman, D.P. et al. “Prescription Drug Cost Sharing: Associations with Medicarion and Medical Utilization and Spending and Health,” JAMA, Vol. 298, No. 1, 2007, pp. 61-69.
8. Huskamp, H.A. et al. “The Effect of Incentive-Based Formularies on Prescription Drug Utilization and Spending,” New England Journal of Medicine, Vol. 349, No. 23, 2003, pp. 2224-2232.
9. Gibson,T.B. et al. “A copayment Increase for Prescription Drugs: the Long-Term and Short-Term effects on Use and Expenditures,” Inquiry, Vol. 42, No. 3, 2005, pp. 293-310.
10. http://www.hcfo.org/grants/generic-substitution-within-class-drugs-medicare-part-d-plans
11. http://www.hcfo.org/grants/impact-state-policies-supporting-medicare-part-d-dually-eligible
12. Robinson, J.C. “Renewed Emphasis on Consumer Cost Sharing in Health Insurance Benefit Design,” Health Affairs, Web, w139-w154, March 20, 2002.
13. Choudhry, M.K., et al. “Assessing the Evidence for Value-Based Insurance Design,” Health Affairs, Vol. 29, No. 11, November 2010, pp. 1988-1994.
14. Maciejewski, M.L. et al. “Copayment Reductions Generate Greater Medication Adherence in Targeted Patients,” Health Affairs, Vol. 29, No. 11, November 2010, pp. 2002-2008.