Background
Responding to persistently rising health care costs and continuing concerns about the quality of health care, many health insurance purchasers have become interested in consumer-oriented strategies to improve health plan design that hold the promise of reducing health care costs and improving quality.1–3 While a number of specific activities fall within the general umbrella of consumer-oriented strategies, most of the recent developments in this area can be classified in three categories.4 Consumer directed health plans [CDHPs] are health insurance plans that use high deductibles coupled with personal health spending accounts to increase consumer accountability for health care spending. Tiered networks are health benefit structures that group providers into tiers based on their costs or quality, and reward consumers with favorable prices if they choose providers in higher quality or lower cost tiers. These strategies, which place greater financial responsibility on consumers for health care decision-making, can also be accompanied by initiatives to provide consumers with better information about the cost and quality of health care, including, for example, information about the quality of health care providers, to enable them to make more informed decisions.
While these three strategies have been increasingly discussed, there remains little consensus about their likely impacts on the health care system or how purchasers can most effectively use these strategies, either alone or in combination, to achieve the goals of reducing costs and improving quality.
Over the past several years, several studies and reviews have been published that appear to offer some insight into how consumer-oriented strategies can affect the provision of care, costs, and outcomes. However, many of these studies have examined narrow ranges of consumer-oriented strategies and often do so in a very tightly focused setting such as a single employer. A synthesis of the existing literature may be able to provide valuable information about the potential for consumer-oriented strategies to bring about beneficial outcomes in health care. Beyond a 2005 report by the Rand Corporation,4 however, existing synthesis work is limited.
The nomination of consumer-oriented health plan strategies was submitted by the Employer Health Care Alliance Cooperative [the Alliance]. Discussion concerning the relative newness of the topic and the possibility of limited scope of literature prompted the decision to conduct an initial exploratory analysis and draft a feasibility report, with the goal of developing a foundation for further evidence review work in this area.
AHRQ further determined that this initial analysis should address three issues. First, a conceptual framework should be developed that identifies important factors determining the impacts of consumer-oriented strategies, within which questions about the impacts of consumer-oriented strategies on health care utilization, costs, and quality can be addressed. Second, a review of published, peer-reviewed literature on CDHPs should be conducted. Several guiding questions specific to CDHPs were identified:
What are prevalence rates and expected trends for CDHPs?
What is the evidence on the effect of CDHPs on quality improvement or lack of improvement?
What is the evidence that CDHPs affect utilization of health care services, including doctors visits, ER visits, medications, and diagnostic tests?
Is there evidence to determine whether effects on utilization are necessary vs. unnecessary services?
What is the evidence that CDHPs discourage access to appropriate care?
Is there evidence that effects of CDHPs on utilization vary depending on the underlying health status, income or education of individuals?
What is the evidence that CDHPs reduce health care expenditures?
Third, preliminary evidence should be gathered about the ability of the literature on tiered networks and provider quality collection and dissemination to support further productive evidence review.
This report addresses these issues. We develop a conceptual framework for the study of literature on consumer-oriented strategies. We reviewed published, publicly available studies of CDHPs identified in MEDLINE® or Econlit and summarized the relatively limited number of studies that provide evidence on their effects. We also present the results of some preliminary investigation of the literature on tiered networks. We present some observations about the literature on the collection and dissemination of data on provider quality. Finally, we develop some preliminary conclusions based on the information presented.
Consumer Directed Health Plans
Definition of CDHPs
The term “consumer directed health plan,” or CDHP, has been used by many individuals in a variety of settings, and researchers and others differ on exactly how they define the term.1, 5 In some settings, the term CDHP has been used quite broadly to refer to any of a wide range of health insurance benefit design strategies that might in one way or another encourage more responsibility for and involvement in health care decisionmaking by consumers. Other definitions are narrower, and typically focus on the financial incentives in plans. One approach is to characterize CDHPs simply as plans with more cost sharing than typical health plans. While somewhat relative, this definition would capture the spirit of many current efforts to increase incentives for consumer engagement in decision making by increasing financial incentives. Within this umbrella definition, the most specific usage of CDHP is to refer to a specific set of health insurance arrangements in which individuals have a high-deductible health plan coupled with a personal health account [PHA] that they can use to pay health care expenses not covered by insurance.1 For purposes of our discussion here, we adopt this last definition of CDHPs. While in principle many of the conceptual issues we discuss would extend to other health plans that attempt to increase cost sharing, this specific set of plans is well defined and distinct in practice, which facilitates discussion, and most current policy discussions of CDHPs take this set of plans as their starting point.
Though the precise structure of CDHPs can vary, all CDHPs share the common element of a high deductible health plan. In such a plan, the consumer is responsible for all spending up to a relatively high deductible, at least $1,000 per year and in many cases $2,000 per year or more.2 After consumers reach that level of spending, the typical plan would cover all subsequent health care spending within the year, though the specific provisions of plans can vary.
The second essential component of a CDHP is the PHA, containing funds that consumers can use to pay their health care bills and can usually carry forward from year to year if not spent. The PHA portion of a CDHP can technically be structured in one of three ways: a Medical Savings Account [MSA], a Health Savings Account [HSA], or a Health Reimbursement Arrangement [also sometimes referred to as a Health Reimbursement Account, either way an “HRA”]. MSAs were the first CDHPs mechanism to become available, established as a demonstration project by the Health Insurance Portability and Accountability Act of 1996 [HIPAA]. An MSA is a tax-exempt account that an individual can use to pay for health expenses. Eligibility for these accounts is limited to employees of firms with 50 or fewer employees and the self-employed and tax-advantaged use of an MSA is contingent upon the individual being enrolled in a high-deductible health plan. The individual [or his or her employer] may deposit funds into the account tax free that can be used to pay for medical expenses, and which roll over from year to year and accumulate interest or dividends tax free if unused. Funds in an MSA cannot be used for non-medical purposes without taxes and penalties until the beneficiary turns 65, becomes disabled, or dies. Enrollment in these plans has been limited, however, never even reaching the cap on enrollment set as part of the demonstration project.6
In recent years, MSAs have been largely displaced by two newer approaches to designing CDHPs. HSAs are the most recent. They are similar to MSAs except that they are available to any individual who is also covered by a high deductible health plan, not just those who are self employed or work for small firms. Individuals are eligible to contribute to the account when they are enrolled in certain types of high deductible health plans. Funds in the HSA can be used to pay for medical expenses, and roll over from year to year if unused. HSAs were enabled by a provision of The Medicare Prescription Drug, Improvement, and Modernization Act of 2003. HSAs were first offered by insurers in January of 2004 though employers generally waited for guidance from the U.S. Treasury and I.R.S., which was issued during 2004, before offering these plans.
HRAs were introduced in 2002 and differ somewhat from MSAs and HSAs. An HRA is an employer-funded account that reimburses employees for qualified medical expenditures. The development of these types of accounts was facilitated by IRS Revenue Ruling 2002-41 and Notice 2002-45 which provided guidance on a number of matters related to the tax treatment of these accounts.
They are often set up by employers in conjunction with high-deductible health plans, but need not be. Employers setting up HRAs have considerable discretion over the specific provisions of the HRA, including whether or not the funds are allowed to roll over from year to year if unused. A typical CDHP scenario using an HRA would have an employer provide employees with a high deductible health plan, say with a deductible of $2,000, and set up and fund an HRA for each employee containing $1,000. Employees could use the $1,000 [on a tax free basis] to pay for qualifying medical expenses that were not covered by the health plan. Unused funds might roll over from year to year as long as the individual remained enrolled in the high deductible health plan. The treatment of funds in these accounts in the event that an individual leaves the employer can vary. In most cases of which we are aware, they are forfeited when an employee changes health plans or leaves the employer.
CDHPs may also be coupled with efforts to provide consumers with information about the cost and quality of providers and treatment options, which may increase the ability of now-cost-conscious patients to make optimal choices about spending their health care dollars. CDHPs may also include provisions to ensure that the prices charged by providers comply with the normal contractual arrangements between providers and health plans. In some discussions, these features are considered central parts of any CDHP. This is in many ways sensible - it may be difficult to expect consumers to improve their decisionmaking in response to increased financial responsibility without some improved information and mechanisms for making choices - but CDHPs are not required to have these features.
CDHPs are a small, though apparently growing segment of the employer sponsored health insurance market. Data from the 2005 Kaiser/HRET employer survey, a well-regarded survey, suggest that among employers offering health benefits, 1.9 percent offer a high deductible health plan coupled with an HRA, and 2.3 percent offer a high deductible health plan that meets the HSA standards. In all, 3.9 percent of benefits-offering employers appeared to offer CDHPs with an HRA or HSA.7 Other surveys report higher rates. For example, a Hewitt survey reports that 9 percent of employers offered a health account with a high deductible health plan in 2005, and 19 percent did in 2006,8 though little information is provided about the methodology of this survey. These plans are increasingly well known. Gabel et al reported that more than 80 percent of benefits managers were familiar with CDHPs in 2003, and that more than 10 percent of employers reported that they were “very likely” to offer an HRA plan in the next two years.9 The GAO reported that in 2005 a range of insurers, including Aetna, Anthem/Wellpoint, Blue Cross and Blue Shield plans, CIGNA, Humana, and United HealthCare were offering CDHPs. CDHPs are also now available in the FEHBP.10
Financial Incentives in CDHPs
A key principle behind CDHPs is that they will increase the financial interest that patients have in their decisions about consumption of medical care.3 Patients who consume medical care and deplete their PHA would be using up a resource that they could otherwise keep and use to their advantage in the future. Compared to patients with traditional insurance arrangements, at least for spending below the deductible, this would tend to raise the effective price of care and may thus make patients less likely to consume care on which they place relatively low value. This type of incentive would be strongest for individuals with MSAs or HSAs because of the more permanent nature of the benefits that accrue to them as a result of limiting their spending on health care.
A careful look at CDHPs suggests that the extent of the price incentives could vary substantially depending on the level of spending and other factors. First, the strongest incentives will face consumers who have not reached the deductible of their insurance plan. These consumers will typically face the full price of medical care. However, after they reach the deductible, and their insurance policy begins to cover their care, they would typically face low rates of cost sharing or even no cost sharing.2, 5 Compared to some traditional insurance arrangements, CDHPs may thus have more generous cost sharing arrangements in certain ranges, and there may be many situations in which the incentives under CDHPs could be very similar to those in other health plans. For example, consider the comparison between a [hypothetical] CDHP with a deductible of $2,000 above which the plan pays all medical costs, and a PPO with a deductible of $500, patient cost sharing of 20 percent until an out-of-pocket maximum is reached at $1,000, and full coverage thereafter [Figure 1]. For the first $500 in spending, consumers in the two plans will face the same incentives. Between $500 and $2,000 in annual health spending, the PPO will have more generous coverage since it will cover 80 percent of medical costs for the consumer while the consumer will be responsible for 100 percent of the costs in the CDHP. But, between $2,000 and $2,500 in annual spending, the CDHP will be more generous since at that level the consumer will be 100 percent covered by the high deductible plan but the PPO consumer will still not have met the out of pocket maximum and continue to be responsible for 20 percent of costs. After $2,500 in spending, the two plans are again equal, with consumers fully covered.
Figure
Figure 1. Healthcare spending.
Thus, while there would typically be regions in which CDHPs would strengthen the financial incentives facing consumers, CDHPs need not always do so uniformly and may even have weaker incentives than some plans in some spending regions. Specific comparisons would depend on the characteristics of the plans in question. For example, in our hypothetical example, the CDHP did not have any cost sharing provisions above the PHA amount, and some CDHPs may have this. Comparisons between CDHPs and HMOs would also tend to produce different results, since many HMOs have no deductibles and relatively low copayments.
The financial incentives associated with CDHPs will also be a function of the prices that providers charge for services. At spending levels below the deductible, patients in many CDHPs are functionally responsible for paying providers bills themselves [possibly using their PHA]. They may thus be subject to variations in prices that providers charge. Virtually all health plans negotiate prices for services with providers and monitor provider bills, but this may not happen as effectively for CDHP spending under the deductible. If patients are systematically subject to higher prices when they are in CDHPs than they would be if they were in other health plans, the effective financial incentives may also vary. In current practice, this does not appear to be a significant issue since most CDHPs appear to contain provisions that regulate the prices charged to consumers in ways that lead them to resemble prices charged in other types of health plans.
Tiered Provider Networks
The introduction of tiered benefit designs into health plans has emerged in several health plan contexts in recent years. The central concept of tiered benefits is variation in the prices consumers have to pay depending on the provider or specific type of service they use. One common form of tiered benefits uses different copayments for different pharmaceutical products. A tiered pharmacy plan could, for example, impose a copayment of $5 for a generic drug and a copayment of $15 for a branded drug with similar characteristics. Many pharmacy plans go even further to place pharmaceuticals into multiple tiers based on their prices and other characteristics, charging the highest copayments for very expensive drugs with less expensive substitutes and the lowest copayment for the least expensive generics.
In this study, we focus on a variant of this strategy that develops tiers for health care providers. In this type of plan, hospitals or physicians that meet criteria imposed by the health plan are identified and favorable financial terms are provided to patients who seek care at those providers. For example, a plan may divide hospitals in its networks into two tiers, one for those hospitals that provide the plan with favorable prices, and a second tier for hospitals that are less favored. The plan may then establish different copayments for hospitals in the two tiers, so that patients who use hospitals in the more-preferred tier face lower cost sharing. In practice, the tiers have typically been defined based on cost criteria, but in principle other criteria such as quality scores could be used as well.
Most tiered networks to date have also focused on hospitals, though some insurers have been developing tiered networks involving physicians. Aetna and UnitedHealth, for example, have worked to develop tiered networks for specialist physicians.
The fundamental strategy behind the development of tiered provider networks is to reward patients for using providers that are designated as preferred by plans. The imposition of price variability would provide incentives for patients to choose the less expensive providers favored by the plans.4
Tiered network plans have thus far been a small segment of the market. A Hewitt survey reports that 5 percent of employers offered a multi-tiered network plan in 2005, and that 7 percent did so in 2006.8 The Kaiser/HRET survey data suggested that 2 percent of employers were “very likely” and 16 percent “somewhat likely” to adopt tiered network plans in 2005.7
Quality Information
A central tenet of strategies that place greater responsibility on consumers for managing their health care is that they will have adequate information to make those choices. Thus, a key feature of consumer directed strategies is the extent to which they provide consumers with the types of information they need to make effective choices. The information needs of consumers, however, will likely vary depending on the strategy.
For example, people enrolled in consumer driven health plans may benefit from information that would help them determine whether or not they need to see a doctor [e.g., information about self-management of conditions], information that might help them determine which doctors to see [e.g., information about the appropriate treatments for given symptoms and information about the comparative cost and quality of providers], and information about the types of treatments they might choose from once they have consulted a physician [e.g., information about the cost and outcomes associated with different treatments5 or other decision support tools]. Along the way, the availability of information about their covered benefits and personal health accounts would be valuable. Individuals with CDHPs might also benefit from information about prevention if they wish to minimize their need to seek care. Correspondingly, researchers report that a variety of tools are either under consideration or are being make available to individuals enrolled in these plans.4
Provider tiering, in contrast, promotes consumer responsibility only at the point of choosing among providers for a particular service. As a result, this strategy theoretically suggests a need for providing consumers a more limited set of information including comparative provider quality and price information for a defined set of potential providers. While the lessons from tiering may be relevant for analyzing the potential effects of consumer driven health plans, they represent only a small subset of the potential issues.
Beyond the types of information collected or provided, research suggests that the ways information is presented and the ease with which consumers can grasp the information are crucial determinants of the extent to which it will influence their decisions. Individual characteristics also likely play a role, since some individuals may be more skilled at interpreting or using available information than others.
Evaluating the effectiveness of information in this setting requires addressing a range of issues including the types of information consumers need, the extent to which health plan provide this information, and whether consumers use the available information effectively. Relatively little evidence is available from recent consumer directed initiatives.4 Earlier literature raises questions regarding the extent to which consumers use the available information,12, 13 whether the information available has the desired effects,14 and whether existing information resources are developed in ways in which they will be most effectively used by consumers.15 It should be noted, however, that earlier evidence was primarily developed in settings where the incentives for consumers to use the information were not as strong as they might be in CDHPs.
1
The specific aspects of CDHP design can be quite involved. This section outlines some key features. Further information may be found in: [1] EBRI Issue Brief No. 273, September 2004, //www.ebri.org/pdf/briefspdf/0904ib1.pdf; [2] CMS Legislative Summary //www.cms.hhs.gov/mmu/hr1/PL108-173summary.asp; [3] IRS Document at //www.treas.gov/offices/public-affairs/has/pdf/notice2004-2.pdf; [4] IRS document, “Health Savings Accounts and other Tax Favored Health Plans”, IRS publication number 969, //www.irs.gov/pub/irs-pdf/p969.pdf; [5] US Department of Labor, Bureau of Labor Statistics Web site: //www.bls.gov/opub/cwc/cm20031022ar01pl.htm.
2Specific deductible levels are set by legislation or regulation, and may vary with the specific types of plan as wellas vary over time [e.g., be indexed for inflation].
3This section focuses on financial incentives facing individuals at the point of deciding what health care to purchase. There are other aspects of financial incentives that may affect things like whether or not an individual joins a CDHP or not, including premiums and expected health care costs. Some work also emphasizes the importance of tax incentives,11 which can convey a substantial subsidy to funds placed in PHAs that can be advantageous relative to using after-tax dollars to pay out-of-pocket costs in other types of plans.
4For further information about tiered networks see [1] Mays GP, Claxton, G, and BC Strunk, “Tiered Provider Networks: Patients Face Cost-Choice Trade-Offs” Center for Studying Health Systems Change, Issue Brief #71, November 2003 www.hschange.org/CONTENT/627/627.pdf; [2] Robinson JC, Hospital Tiers in health Insurance: Balancing Consumer Choice with Financial Motives" Health Affairs Web Exclusive 2003; [3] Yegian, JM, Tiered Hospital Networks, Helth Affairs Web Exclusive March 19, 2003. [4] Sweeney, K, “Health Plans Embrace Tiered-provider Networks,” Employee Benefit News. EBRI, October 1, 2003.; [5] Rosenthal and Milstein, HSR 2004;
5Ideally, information about costs would include both information about the costs of specific procedures and information about the likely longer-term costs of treatment strategies, including things like the potential need for and cost of follow-up procedures or the probability of hospitalizations.