Benefit Design Innovations: Implications for Consumer-Directed Health Care

Issue Brief No. 109
February 2007
Ha T. Tu, Paul B. Ginsburg

Current health insurance benefit designs that simply rely on higher, one-size-fits-all patient cost sharing have limited potential to curb rapidly rising costs, but innovations in benefit design can potentially make cost sharing a more effective tool, according to a new study by the Center for Studying Health System Change (HSC). Innovative benefit designs include incentives to encourage healthy behaviors; incentives that vary by service type, patient condition or enrollee income; and incentives to use efficient providers. But most applications of these innovative designs are not widespread, suggesting that any significant cost impact is many years off. Moreover, regulations governing high-deductible, consumer-directed health plans eligible for health savings accounts (HSAs) preclude some promising benefit design innovations and dilute the incentives in others. A movement away from a one-size-fits-all HSA benefit structure toward a more flexible design might broaden the appeal of HSA plans and enable them to incorporate features that promote cost-effective care.

Increasing the Potential of Cost Sharing to Control Spending

aced with rapidly rising health care costs but eager to moderate premium increases, employers have increased patient cost sharing in recent years, primarily through higher deductibles, copayments and coinsurance. The federal government also has fostered the trend toward increased patient cost sharing by creating tax-favored health savings accounts (HSAs) linked to high-deductible health plans. HSA-eligible plans must have minimum deductibles of $1,100 for self-only coverage and $2,200 for family coverage in 2007.

While higher patient cost sharing has become a major strategy to slow rising premium trends, the cost-containment potential of current benefit designs built on greater cost sharing is constrained by several factors. First, many consumers’ ability to handle the additional financial responsibility is limited. When cost sharing is too large in relation to a consumer’s resources, the result is either serious financial strain or reduced access to care.

The potential of higher cost sharing as a cost-control strategy is also limited because a relatively small proportion of patients accounts for a large portion of medical spending. Studies indicate that about 10 percent of patients account for 70 percent of spending in a given year.1 As a result, a large portion of spending is beyond the reach of patient financial incentives, even in a high-deductible plan. Spending that exceeds the deductible is subject to relatively weak financial incentives, and spending that exceeds a plan’s annual out-of-pocket maximum is not subject to any financial incentives at all. As a result, high-deductible plans are unlikely to have much impact on the utilization of the sickest patients who account for a large part of health spending.

Another limitation in current benefit designs is that there are few incentives to choose more efficient providers. In physician services, for example, copayments typically are the same for all network providers and even coinsurance—where the patient pays a percentage of the total cost—sharply dilutes price differences among network providers. Moreover, benefit structures do not distinguish between services that are considered extremely important, such as testing, insulin and physician visits to manage diabetes, and services that are more elective, such as knee surgery to play recreational sports. Finally, benefit structures tend to be designed uniformly without regard for variation in consumers’ financial resources. This means that cost-sharing requirements that would provide meaningful incentives for a higher-income family likely are impractical because the burden on lower-income families would be too great.

As patient cost sharing continues to increase, it is important to examine innovations in benefit design and explore whether there are alternative, more flexible and potentially more effective designs to help raise patients’ cost consciousness without impeding access to important care. To that end, HSC researchers interviewed experts, including benefits consultants and representatives from health insurance companies and employers, about benefit design innovations (see Data Source). In seeking employers to interview, the focus was on large employers known for innovative benefit designs. Experts identified several types of innovations recently introduced or planned:

Back to Top

Encouraging Healthy Behaviors and Self-Management

n the past few years, large employers seeking to curb health costs and increase worker productivity have begun providing incentives aimed at engaging consumers in maintaining their own health and reducing health risk factors. For a wide range of programs designed to increase wellness—including nutrition and physical activity programs, health risk appraisals, disease management, lifestyle management, and personal health coach programs—financial incentives can boost participation.

Until five to ten years ago, employers rarely offered meaningful incentives for taking part in such programs, and participation consequently tended to be quite low. Johnson & Johnson was a pioneer in using financial incentives to encourage participation: After the company began offering a $500 health insurance premium discount for participating in a wellness program (including completion of a health risk assessment), enrollment shot up to 90 percent. Some employers followed suit in introducing premium differentials; a few employers make a health risk appraisal a prerequisite for insurance eligibility.

With the introduction first of health reimbursement accounts (HRAs) and then HSAs, many large employers began contributing to these spending accounts in place of premium differentials or cash bonuses as incentives for self-management activities. Benefits consultants and employers both say that HRAs are more popular funding vehicles for these incentives than HSAs, because HRAs offer greater flexibility (see Table 1).2

Among large employers offering self-management incentives, many design the incentives incrementally: first, an employee is offered a certain amount—typically $50 to $200—to complete a health-risk appraisal; next, the employee is given an additional amount to enroll in a personal health coaching program; and finally, the employee is given another incentive upon graduation from the health coaching program. In total, an aggressive employer might offer $400 to $500 for completion of all three steps. A benefits consultant specializing in this area noted: “A health assessment alone has no return on investment. You need to get people engaged in a follow-up program.”

One employer that has been particularly aggressive in integrating self-management incentives into benefit design is a public employer, King County government in Washington. The county uses a three-tier benefit structure based on enrollees’ participation in healthy and risk-reduction behaviors. All plan enrollees are automatically subject to the highest (bronze) level of cost sharing. Completion of a wellness assessment qualifies the enrollee for the middle (silver) level of cost sharing, and complying with an individual action plan, designed by the county’s wellness vendor, enables the enrollee to achieve the lowest (gold) level of cost sharing. The initial success of King County’s new benefit structure—with more than 90 percent participation in wellness assessments—has stirred interest among other employers, but most have been slow to follow suit, instead waiting for solid evidence on cost impact.

There appears to be no broad consensus among experts on the extent of health benefits and cost savings achievable from self-management incentives, or which programs are most effective, within the broad array of programs being offered. Some respondents noted that capturing the full extent of both direct savings (reduced health care costs) and indirect savings (increased workplace productivity) from self-management programs can be a challenge for employers. Indeed, many employers do not have the ability to measure productivity gains from programs that seek to encourage healthy behaviors. Several experts also noted that some of the savings from improved health may be realized only in the long run, and, therefore, may not accrue to the employer currently paying for self-management programs.

One point that experts do appear to agree on is that just providing financial incentives is not enough to guarantee widespread employee participation and tangible benefits; strong communication also is key. Even with sizeable financial incentives, experts expect only low to medium participation in health risk appraisals and other health-promotion activities, unless accompanied by effective communication from management.

Back to Top

Table 1
Selected Characteristics of Personal Health Accounts

Health Reimbursement Arrangement (HRA)
Health Savings Account (HSA)
Employer-funded account to reimburse employees’ qualified medical expenses
Tax-free account to pay for qualified medical expenses and serve as a retirement savings account
Account Owner
Account Funders
Employer only
Employee, employer or both
Annual Contribution Limits
No federal limits
Federal limit of $2,850 (self-only) or $5,650 (family) in 2007; annual deductible limitation no longer applies
Tax Treatment of Contributions
Employer contributions are excluded from gross income and not subject to taxes
Employer contributions are excluded from gross income and not subject to taxes; employee contributions are tax-deductible
Rollover Provisions
Yes; unused funds revert to employer when employee leaves or retires from company
Non-Medical Use
Not allowed
Taxed as income plus 10% additional tax
Required Companion Plan
None required
Must be paired with high-deductible plan (minimum $1,100 self-self only, $2,200 family, in 2007); preventive care expenses exempt from deductible
Enabling Legislation
Authorized by Treasury Department Revenue Ruling 2002-41 in 2002, and IRS Guidance 2002-45
Medicare Prescription Drug, Improvement, and Modernization Act of 2003, amended by Tax Relief and Health Care Act of 2006

Back to Top

Varying Incentives by Service or Patient Condition

xperts who question the extent of cost savings possible from consumer incentives to engage in healthy behaviors and self-management point to the potential of “evidence-based benefit design” in achieving more cost-effective care. This includes two types of incentives: those designed to decrease the use of treatment options that are more expensive than alternatives, especially those without greater proven effectiveness; and those designed to encourage—or at least avoid discouraging—the use of services known to be clinically effective.

Regarding incentives to decrease the use of expensive, overused services, little concrete innovation was reported by experts, who noted that insurers and employers tend to rely on administrative controls, rather than benefit design, to curb overuse. For a few services, such as bariatric surgery, some large employers apply differential coinsurance rates based on clinical risk indicators, but the overall list of services subject to higher rates of coinsurance tends to be short. Some experts suggested using a reference-pricing system for certain services; for example, in hip and knee replacement, the least expensive implant would provide the reference price, and the price differential between the price of the implant used and the reference price would be subject to a higher rate of cost sharing (perhaps even completely out of pocket). However, such approaches appear not to have moved beyond the discussion stages.

Among incentives to avoid discouraging the use of particularly effective services, the major approach reported was a reduction in cost sharing applicable to treatment regimens—most commonly prescription drugs—for certain chronic conditions. One prominent example among employers is Pitney Bowes, which reduced coinsurance for all drugs used in treating three chronic conditions—diabetes, asthma and hypertension—to the lowest pharmacy-benefit tier (10% coinsurance vs. 20% or 50%, resulting in a 50-85% reduction in the average cost of a 30-day prescription refill). For diabetes, reduced cost sharing also applies to all testing supplies. Since introducing the program in 2001, Pitney Bowes has observed reductions in both direct medical costs (emergency department visits and hospitalizations) and indirect costs (substantially reduced sick-leave and disability rates). While more than 20 large employers nationwide have introduced similar cost-sharing reductions for prevalent chronic conditions, many employers appear to be waiting for more definitive evidence of cost savings before adopting similar strategies.

Recent moves by insurers to offer HSA-eligible high-deductible plans with an expanded preventive care safe harbor (not subject to the minimum deductible) represent another effort to prevent underuse of effective services. While Internal Revenue Service (IRS) guidance for HSA plans’ preventive care safe harbor stipulated that maintenance drugs for existing chronic conditions could not be exempted from the deductible, insurers now have broadened the list of prescription drugs eligible for first-dollar coverage to include medications for several chronic conditions, contending the drugs help to prevent acute complications arising from the chronic conditions. Aetna was the first insurer to offer an expanded preventive care safe harbor, including drugs for arthritis, diabetes, asthma and high cholesterol. Nearly all other major insurers have followed suit, citing demand from large employers that wanted to offer HSA plans but were concerned about the lack of first-dollar prescription drug coverage for common chronic conditions.

Some experts noted that, while reduced cost sharing for certain chronic-condition treatments represents a step forward, it is still a blunt tool (limited to two categories—highly effective and other) in need of further refinements. One such refinement would be a “value-based benefit design” that identifies subgroups of patients who would benefit most from a given treatment and reduces cost sharing—perhaps even to zero—for that group..3 An example is the use of statins, which provide the greatest benefit to patients with previous heart attacks—a group that with the suggested refinement would be eligible for lower cost sharing than those whose only cardiac issue is high cholesterol.

Incorporating refinements into benefit design poses major challenges: Insurers and employers resist retooling information systems and rewriting contracts to incorporate such nuances; communication to enrollees about varying benefit levels for specific services can be difficult; and the difficulty is compounded when different patients have different cost sharing for the same service. For fully insured products, differential benefits also must pass muster with state insurance regulators. Developing incentives to limit overuse of inappropriate services and encourage use of clinically effective services also is hindered by the limited knowledge base available to guide decisions about whether care is appropriate. It has been estimated that only 15 to 25 percent of medical services are supported by credible evidence on clinical effectiveness.4 Because of this limitation, as well as the other noted barriers, incentives to comply with evidence-based care generally have been limited to the most prevalent chronic conditions.

Back to Top

Encouraging Use of Efficient Physicians and Hospitals

everal experts noted that a complicated benefit structure that requires consumers to process a great deal of complex information may not result in optimal consumer decision making. As one respondent commented, “The fewest, simplest rules have the best chance of success.” These experts suggested that a better alternative is to identify network providers distinguished by both high quality and high efficiency (low cost) and then give consumers incentives to use those providers. This approach simplifies options for consumers, who can rely on the insurer to gather and analyze data on quality and efficiency. Recently, health plans—in response to requests from large national employers—have introduced, in selected markets, networks of physicians designated as “high performers” on the basis of efficiency scores and quality measures, both of which vary significantly by insurer.5

High-performance networks generally are built around major physician specialties, in particular those that account for a large share of medical expenditures. Primary care physicians typically are not included in high-performance networks, in part to avoid disrupting existing relationships between patients and their regular care providers. Hospitals are not targeted directly by high-performance networks, but since hospital costs are included in assessments of physician efficiency, physicians treating patients in relatively high-cost hospitals will have more difficulty meeting the efficiency standards.

Although insurers have developed high-performance networks, it is employers who decide whether to incorporate them into their plans and whether to alter benefit structures to provide financial incentives to use particular providers. Some large employers have used these networks solely as information tools for enrollees and have yet to introduce financial incentives for using higher-performing physicians. Other large employers have introduced differential cost sharing: for example, 10 percent coinsurance for high-performing physicians vs. 20 percent for another network physician. A plan executive noted that strong financial incentives are uncommon because most employers do not yet have a “high comfort level” with these networks.

One reason for the low comfort level and slow adoption of high-performance networks appears to be the lack of consensus on methods for classifying high performers, especially on quality measures, according to respondents. Even on the cost side, a physician may be rated “efficient” by one insurer and not another, leading to questions about validity of the rankings. Differences could result from separate, small samples of cases in different insurers’ claims files or from using different software to group claims into episodes of care. Allowing insurers access to physician-specific Medicare Part B claims data—something employer groups have pushed—would help produce more accurate assessments of physician efficiency and quality.

Back to Top

Varying Benefits by Income

lthough the practice does not appear to be widespread, some employers vary health plan premium contributions and cost-sharing levels by employee earnings. Many of the employers who take this approach report that it is a long-standing practice that reflects an organizational emphasis on equity and recognition that employees’ ability to bear risk varies depending on their financial means. Some employers that vary benefits by earnings choose to vary only premium contributions, while other employers also differentiate deductible and out-of-pocket maximum levels. (Copayment and coinsurance levels generally are not subject to income tiering, because of complexity and confidentiality issues.) One large employer that varies premium contributions, deductibles and out-of-pocket maximums based on salaries reported that the company aims for the out-of-pocket maximum as a proportion of gross salary to be roughly constant across all wage categories.

Information technology advances are making possible greater refinement of income-based benefit design variations and online determination of premiums, deductibles and out-of-pocket maximums. In spite of such advances, however, employers that practice income tiering still face an important limitation: They have access only to information about their own employees’ earnings, rather than total family income, which would be a more accurate indicator of ability to assume risk.

Back to Top

Innovations and HSA Regulations

ith HSAs a central part of policies to promote consumer-directed health care, it is important to analyze the extent to which promising innovations in benefit design are compatible with HSA rules. Incentives to encourage healthy behaviors and patient self-management generally are permitted under HSA rules, but with two important caveats: Incentives cannot be structured to reduce the deductible below the statutory minimum, and, until recently, the sum of employer and employee contributions could not exceed the plan deductible. Because of these and other restrictions, many large employers making incentive-based contributions into a spending account preferred to do so into an HRA, which has no such restrictions. However, the Tax Relief and Health Care Act, enacted in December 2006, contains several HSA changes, including eliminating the requirement that the annual contribution can be no higher than the HSA plan deductible. This change will make it less likely that employers’ incentive-based HSA contributions will exceed the contribution limit.

The new law, however, leaves statutory minimum deductibles in place for HSA plans, posing barriers to improved patient compliance with chronic-condition treatment regimens through reduced cost sharing. For example, deductible waivers for many expenses required for effective disease management—diabetes testing supplies, for instance—still are not permitted. The existence of the high minimum deductible also dilutes enrollees’ incentives to use high-performance network physicians, because cost-sharing reductions for using these physicians would not apply until the deductible has been met.

Until the HSA legislation was passed in December 2006, HSA regulations did not allow employers to differentiate HSA contributions or plan cost-sharing provisions by employee income. The new law allows employers to make higher HSA contributions for workers not classified as highly compensated—workers earning less than $100,000 a year. The law, however, does not allow further income-based refinements beyond the distinction of highly compensated vs. not highly compensated workers.

Back to Top

Policy Implications

lthough many promising benefit design innovations were identified in interviews for this project, the limited extent of innovative activity was surprising. It would have been difficult to identify enough innovative employers to interview without the advice of benefits consultants, who develop some of these innovations, and insurers, who also create innovations and administer most of them for large employers. Indeed, consultants pointed out that employers known for innovation in some areas often are not on the leading edge in other areas. So if innovative benefit designs are to eventually affect large numbers of people, it likely will be many years into the future.

To the extent that government wants to encourage innovative benefit designs, policy makers might consider examining current laws and regulations to identify and reduce factors that discourage innovation. Despite the loosening of some HSA restrictions by the recent legislation, the benefit structure for HSA-eligible plans remains quite inflexible—so these plans either cannot accommodate promising benefit design innovations or they dilute the incentives.

The administration’s fiscal year 2008 budget includes a proposal to allow plans with at least a 50 percent coinsurance requirement and a minimum out-of pocket exposure resulting in the same premium under existing HSA requirements to be excluded from the minimum deductible requirement. This would increase flexibility somewhat but still would rule out many of the innovative benefit structures now under consideration by health plans and employers.

Alternative HSA benefit structures that allow more flexibility could appeal to a broader base of purchasers and consumers while maintaining the substantial cost-sharing requirements. One approach, proposed by several experts, is to allow HSA plan deductible and out-of-pocket maximum amounts to vary by income, to reflect the fact that consumers’ ability to assume risk varies based on their financial resources.

Another option would remove the minimum-deductible structure altogether, replacing it with the stipulation of a maximum actuarial value—or the total proportion of allowed amounts for covered medical expenses paid by insurance. For example, the actuarial value could be limited to 75 percent, meaning that for enrollees in the aggregate, the plan would pay no more than 75 percent of allowed amounts for medical care. This approach would permit health plans to offer extensive variations in benefit structure to better reflect the goals and evolving tools of consumer-directed health care. For instance, insurers would be able to offer HSA plans with first-dollar prescription-drug coverage, usually with incentives to choose generic or preferred brand drugs, balanced by higher cost sharing for other services, such as hospital stays.

Such a flexible benefit design has a precedent in Medicare Part D, where prescription drug plans can vary benefit structures as long as they assure the government that the actuarial value is at least as high as the structure for the standard prescription-drug plan as defined by statute. If this approach were to be adopted for HSA-eligible plans, insurers would have to certify that the actuarial value is no higher than the amount mandated by law. The standard high-deductible plan still could be retained as one of the options. With a more flexible HSA structure, there is the potential to allow variation in actuarial values by income levels—for example, by allowing a higher actuarial value for lower-income enrollees.

Some argue that innovation in the use of patient cost sharing is limited by the current tax treatment of employer-based health insurance. With employer premium contributions—and many employees’ contributions—exempt from income and payroll taxes, use of patient cost sharing is discouraged, notwithstanding the recent trend of increased patient cost sharing. The administration recently proposed replacing this unlimited tax exclusion with a standard deduction for those with health insurance. Such a step would motivate changes in benefit structures toward more extensive cost sharing, increasing the need for benefit design innovations.

Back to Top


1. Berk, Marc L., and Alan C. Monheit, “The Concentration of Health Care Expenditures, Revisited,” Health Affairs, Vol. 20, No. 2 (March/April 2001).
2. There is no required companion plan for HRAs as there is for HSAs (see Table 1). Incentive-based employer HSA contributions can only be made if the HSA is organized as part of a Section 125 cafeteria plan; other Section 125 nondiscrimination rules still apply.
3. Chernew, Michael E., Allison B. Rosen, and A. Mark Fendrick, “Value-Based Insurance Design,” Health Affairs, Web Exclusive (Jan. 30, 2007).
4. Eddy, David M. “Evidence-Based Medicine: A Unified Approach,” Health Affairs, Vol. 24, No. 1 (January/February 2005); “Medical Guesswork,” Business Week (May 29, 2006).
5. Draper, Debra A., Allison Liebhaber and Paul B. Ginsburg, “High-Performance Health Plan Networks: Early Experiences,” Center for Studying Health System Change, forthcoming.
Back to Top

Data Source

The information in this Issue Brief is derived primarily from interviews HSC researchers conducted with approximately 25 thought leaders, benefits consultants, and representatives of large employers and national and regional health insurers. Information also was obtained from literature reviews on benefit design. Interviews were conducted by telephone, primarily in May-August 2006, using semi-structured interview protocols and two-person interview teams.

Back to Top

ISSUE BRIEFS are published by the Center for Studying Health System Change.

600 Maryland Avenue, SW, Suite 550
Washington, DC 20024-2512
Tel: (202) 484-5261
Fax: (202) 484-9258

President: Paul B. Ginsburg