Approaches to Improving Value— Consumer Incentives
Previous discussions at this workshop series, such as the stakeholder perspectives described in Chapter 2, highlight the importance of consumers in reorienting health care in the United States toward a value-driven system. It was heard that consumers play a critical role in the medical decision-making process and make multiple decisions in the path of care that ultimately impact the value of care delivered on both individual and societal levels. The presentations in this session focused on specific examples of consumer-focused approaches to achieve greater value, exploring the research to date and the evidence of impact.
A. Mark Fendrick emphasizes the continued underutilization of highvalue health services, with research indicating that U.S. adults receive only about 50 percent of recommended care (McGlynn et al., 2003). He discusses the potential for value-based insurance design—which focuses on consumer choices, adjusting patients’ out-of-pocket costs for specific services based on an assessment of the clinical benefit achieved (with the more clinically beneficial interventions associated with lower out-of-pocket costs)—to be utilized as a tool for increasing value in health care.
Building on the concept of value-based insurance design, Melinda Beeuwkes-Buntin discusses consumer-directed, high-deductible health plans (CDHPs). This presentation elaborated on the mechanisms though which CDHPs attempt to provide patients with financial incentives to make wiser healthcare choices while spurring them to take greater responsibility for their care. The impact of evolving “consumer-directed” plan designs on
expenditures, access to care, and clinical outcomes is also reviewed, with gaps in knowledge and future areas of needed research identified.
Approaches such as pharmaceutical or hospital tiering programs have attempted to increase the transparency of value of different medical interventions and providers. Dennis Scanlon describes in further detail how tiering classifies healthcare providers, pharmaceuticals, or treatments on the basis of objective or subjective criteria such as cost, quality, and value, and engages patients and consumers in making informed decisions. One example discussed in detail is a hospital tiering program and the impact of the program on consumer choices and quality of care.
Concluding the session, Ronald Goetzel details the value of worksite health promotion and chronic disease prevention programs, indicating that they can yield significant health and economic benefits for employers and employees. In addition to discussing how workplace wellness programs can serve as vehicles for health behavior change, he outlines recommendations to increase employer engagement in providing evidence-based health promotion programs to their employees.
VALUE-BASED INSURANCE DESIGN: RESTORING HEALTH TO THE HEALTHCARE COST DEBATE
A. Mark Fendrick, M.D., University of Michigan Medical Center, and Michael E. Chernew, Ph.D., Harvard Medical School
As healthcare premiums escalate, both private and public purchasers are forced to decide how best to address this unsustainable economic burden. Unfortunately, value—the clinical benefit achieved for the money spent—is frequently excluded from the dialogue on how to manage the growth of healthcare spending.
If the desirable clinical effects of health insurance are ignored, constraining healthcare cost growth can be achieved simply by providing less generous coverage or no coverage at all. In fact, the numbers of Americans who are uninsured or underinsured is at an all-time high, reflecting the trade-off between the high cost of health benefits and remaining viable in today’s economy (Kaiser Family Foundation, 2005). Although rising healthcare costs are the main impetus behind the redesign of health benefits, concerns regarding the quality of care share the limelight. This clear and unresolved tension between cost containment and suboptimal quality of care has led to two prevailing trends in benefit design:
Cost containment strategies that use financial incentives to alter patient and provider behavior: This approach includes increases in cost sharing (e.g., deductibles, copays, coinsurance rates) in exist-
ing plan designs and the introduction of high-deductible health plans that allow employees to set aside tax-free money for health expenses. A recent Kaiser Foundation Employer Benefit Survey showed relative moderation in the growth in healthcare premiums, largely attributable to increasing cost shifting from employer to beneficiary (Kaiser Family Foundation, 2005).
Improving the quality of care and keeping individuals healthier longer: Employers and insurers are implementing wellness and disease management (DM) initiatives to help individuals manage their health in an effort to avoid more costly care. Pay-for-performance (P4P) programs, which pay providers more for adhering to evidence-based clinical practices and delivering specific health outcomes, are disseminated widely. While many proponents of these initiatives contend that better health will lead to lower spending, fiscal savings from quality-oriented interventions have not materialized.
Since higher patient cost sharing discourages use of high-value medical services, these two trends inherently conflict. The main challenge is to devise benefit packages that openly address the problem of spending growth, yet explicitly aim to optimize the health of beneficiaries through the incorporation of features that complement each other in the effective and efficient delivery of care.
Role of Cost Sharing
From the patient perspective, increased cost sharing is the principal instrument of change. There is little debate over the economic theory that an increase in out-of-pocket expenses will lead to less consumption of healthcare services. Many studies demonstrate that when confronted with higher costs, individuals will purchase less care (Gibson et al., 2005). Ideally, higher patient copayments would discourage only the utilization of low-value care. For this important assumption to be achieved, patients must be able to distinguish between high-value and low-value interventions. However, when this ability to differentiate among services does not exist, increased cost sharing has the potential to cause negative clinical outcomes. A large and growing body of evidence demonstrates that in response to increased cost sharing, patients decrease the use of both highvalue (e.g., immunizations, cancer screening, appropriate prescription drug use) and low-value services, and may have worse health outcomes as a result (Figure 4-1).
Value-Based Insurance Design
In response to the adverse clinical effects of “one-size-fits-all” cost shifting, we propose “value-based insurance design” (VBID), a system that bases patients’ copayments on the relative value—not the cost of the clinical intervention (Chernew et al., 2007; Fendrick and Chernew, 2006). In this setting, cost sharing is still utilized, but a “clinically sensitive” approach is explicitly employed to mitigate the adverse health consequences of high out-of-pocket expenditures. Originally referred to as the “benefit-based copay” for prescription drugs, VBID has broadened to all sectors of healthcare delivery. The principle tenets of a VBID program are that (1) medical services differ in the clinical benefit achieved and (2) the value of a specific intervention likely varies across patient groups. We believe that more efficient resource allocation can be achieved when the amount of patient cost sharing is a function of the value of the specific healthcare service to a targeted patient group.
Although cost sharing may be ill advised in certain clinical circumstances, it would be absurd to completely ignore the need for interventions to reign in spending. Increased cost sharing seems inevitable given the lack of demonstrated savings from, or unwillingness to adopt, other approaches. In the VBID paradigm, patients’ out-of-pocket costs are determined by the costs and benefit of care—zero or low copayments for interventions of highest value (e.g., mammogram for women with a first-degree relative
with breast cancer, lipid-lowering therapy for an individual with a history of myocardial infarction) and higher cost sharing for interventions with little or no proven healthcare benefit (e.g., total body computed tomographic scanning). This more sophisticated benefit design is made possible by advances in health information technology and comparative effectiveness research. While some believe that such benefit packages are too complex to be accepted by consumers or too difficult to create in certain clinical conditions, the inability to construct the perfect program should not lead to abandonment of key VBID principles. The cost of maintaining the status quo, in terms of higher spending and worse health outcomes, is staggering.
Barriers to VBID implementation certainly exist and create several challenges (Chernew et al., 2007). From experience in the field, VBID programs are feasible, are acceptable to all stakeholders, and have been very well received by beneficiaries. VBID can address several important inconsistencies in the current system and work synergistically with other initiatives such as high-deductible health plans, disease management, patient-centered medical home, and P4P programs. By allowing different cost-sharing provisions for different services, value can be enhanced without removing the role of cost sharing in the system overall.
Types of VBID Programs
In practice, there are two general approaches to VBID programs. The first simply targets services known to be of high value (e.g., ACE [angiotensin converting enzyme] inhibitors). While some users of the services have the target high-value condition(s) (e.g., congestive heart failure, myocardial infarction), others do not (e.g., essential hypertension), and the system does not attempt to differentiate between these patient groups.
The second approach targets patients with select clinical diagnoses (e.g., coronary artery disease) and lowers copays for specific high-value services (e.g., statins, beta-blockers) only for those patient groups. This diagnosis-driven strategy, which requires more sophisticated data systems to implement, creates a differential copay based on patients’ health conditions.
A controlled evaluation of a VBID program that lowered copayments for all users of five high-value pharmaceutical classes, demonstrated significant increases in patient compliance (Chernew et al., 2008) (Table 4-1).
The financial impact of VBID programs on healthcare spending is under investigation. Economic effects depend on the level and precision of targeting and the extent or direction of the changes in copayments. Since many clinical services provide higher value for a select subset of patients, the better the system is at identifying those patients, the greater is the likelihood of achieving a high financial return. Employers with more targeted
TABLE 4-1 Copay Reductions Increase Adherence to High-Value Drug Classesa
% MPR Increase
% Reduction in Nonadherence
2.59 (p < .001)
3.02 (p < .001)
4.02 (p < .001)
3.39 (p < .001)
1.86 (p = .134)
NOTE: ACE = angiotensin coverting enzyme; ARB = angiotensin receptor blocker; MPR = medication possession ratio.
aWhen a large services industry employer reduced copays for certain classes of drugs, nonadherence rates decreased by 7-14%. Copayment rates for generic medications were reduced from $5 to $0; copayments for brand-name drugs were cut in half for five classes of drugs. A similar employer with identical disease management offerings and similar, but stable, copayments served as a control group.
SOURCE: Copyrighted and published by Project HOP/Health Affairs as Chernew, M. E., M. Shah, A. Wegh, S. Rosenberg, I. Juster, A. Rosen, M. Sokol, K. Yu-Isenberg, and M. Fendrick. 2008. Impact of decreasing copayments on medication adherence within a disease management environment. Health Affairs 27(1):103-112. The published article is archived and available online at www.healthaffairs.com.
programs incur lower treatment costs, because fewer individuals are eligible for copay reductions and the targeted patients who receive copay relief are most likely to benefit from increased utilization.
Offsetting these direct costs of copay reduction are the savings incurred by reductions in future services avoided due to better clinical outcomes. For example, savings due to fewer emergency room visits for acute asthma exacerbations would offset the direct costs of lower copays for asthma controller medications, at least partially. The net financial benefit improves if the underlying risk of an adverse outcome is high, if the cost of that adverse outcome is high, if consumers are responsive to lower copays, and if the service is effective at preventing the adverse outcome. Additional return on investment accrues if the nonmedical benefits of improved health (e.g., reduced disability and absenteeism, enhanced productivity) are included.
The following financial scenarios are likely to occur, depending on the goals of the VBID program and the willingness to raise copayments on low-value services:
Targeted copay reductions will result in higher value for each market basket of services only if there are incentives to use services that produce high levels of health benefit. There will be an uncertain effect on total healthcare cost trends.
Copay reductions with global or targeted copayment increases to offset short-term costs of increased utilization of targeted services (actuarial equivalence) will result in higher value for each market basket of services due to incentives to use services that produce high levels of health benefit. Total healthcare costs will be equal or lower, depending on the extent of savings yielded due to offsets from improved health and lower utilization of low-value services as a result of higher copays.
Efforts to control costs should not produce preventable reductions in quality of care. Multiple private and public sector employers, health plans, and pharmacy benefit managers have implemented VBID programs encouraging the use of high-quality services. In 2001, Fortune 500 employer Pitney Bowes lowered copayments for asthma and diabetes medications, reporting to the Wall Street Journal a $1 million savings from reduced complications (Furhrman, 2007). The city of Asheville, North Carolina, Marriott Corporation, Mohawk Carpets, Wal-Mart, CIGNA, the State of Maine, and the University of Michigan are among those who have implemented VBID. Leading health plans and health benefit consultants are working to make these packages accessible nationwide.
Payers desiring to optimize health gains per dollar spent should avoid “across-the-board” cost sharing and instead implement a “value-based” design that removes barriers and provides incentives to encourage desired behaviors for patients and providers. Targeted efforts to reduce utilization of low-value services are more likely to contain cost growth while maintaining quality of care.
We do not expect VBID to solve the nation’s healthcare crisis. Technological advances will continue to generate upward pressure on costs, and the ability of individuals and their employers to afford such coverage will increasingly be strained. That said, the alignment of financial incentives— for patients and providers—would encourage the use of high-value care, while discouraging the use of low-value or unproven services, and ultimately produce more health at any level of healthcare expenditure.
CONSUMER-DIRECTED HEALTH PLANS: WHAT ARE THEY, WHAT DO WE KNOW ABOUT THEIR EFFECTS, AND CAN THEY ENHANCE VALUE?
Melinda J. Beeuwkes-Buntin, Ph.D., RAND Health
The health insurance options available to Americans have changed dramatically in the last five years. Higher deductibles and personal health savings accounts—either health reimbursement accounts (HRAs) or health savings accounts (HSAs)—are increasingly popular features that are intended to make patients more cost conscious in their healthcare choices. Plans with these features, often referred to as consumer-directed health plans, are intended to engage consumers in understanding the costs of care and in making healthcare choices.1 Critics charge, however, that they are currently little more than a device to shift costs to enrollees. This paper sketches the scope of consumer-directed care, discusses what is known about the effects on CDHPs to date, and concludes with some thoughts about how consumer-directed care might be used to increase the value of the healthcare services we receive.
Scope of Consumer-Directed Care
Consumer-directed health plans emphasize the role consumers can play in making decisions about their healthcare choices. These plans usually provide patients with upfront financial incentives to choose care wisely in the form of deductibles that are typically higher than those of traditional plans—the typical consumer-directed plan has a single deductible of $1,000 or more. In addition, as mentioned above, they are often coupled with personal savings accounts that roll over if funds in them are not spent; this gives patients additional incentives to save for future expenses rather than consume care in the current period. Proponents of consumer-directed plans hope these measures will spur patients to take greater responsibility for their own care and seek information about care options. They also point out that more informed and motivated consumers can in turn spur healthcare providers to compete for their business on the basis of higher quality and/or lower costs.
Since the advent of CDHPs early in the decade, an increasing number of employers have offered them either exclusively or as a choice among
other types of insurance. Large employers are more likely than midsize and small businesses to provide CDHPs as an option. In 2008, almost a quarter of firms with 1,000 or more workers offered CDHPs, compared to 15 percent of midsize and 13 percent of small firms (Figure 4-2). Half of very large firms with more than 5,000 enrollees offer a CDHP option (Watson Wyatt/RAND, 2007). Across the board, the rate of firms offering CDHPs has increased substantially since 2005 when U.S. Treasury regulations laid out the criteria for HSAs. However, despite employer enthusiasm for CDHPs, consumer take-up remains relatively low. In 2007, 11 percent of Americans were enrolled in a high-deductible health plan (HDHP). Only 2 percent of Americans were enrolled in a CDHP featuring an HDHP plus an HRA or an HSA (EBRI/Commonwealth, 2008).
Economic forces, employer choices, and evolving plan designs all have the potential to change this picture. Even before the extent of the economic downturn was known, a survey of human resources executives in early 2008 indicated that companies are still warming to the idea of CDHPs and see them becoming more important in the future. Nearly nine-tenths of respondents were optimistic about the future of CDHPs at their firms: 34 percent anticipated more employees in a CDHP, 26 percent anticipated
offering a “full replacement” CDHP (i.e., offering only a CDHP to their employees), and 29 percent anticipated a “majority” of employees in a CDHP (Figure 4-3). All of these rates represented an increase from 2007, and now that companies are more financially stressed, they are reportedly even more interested in CDHPs (Watson Wyatt/RAND, 2007). Finally, given that CDHPs are still relatively new products in the market, it may not be surprising that interest in them is growing as offerings are refined. For example, most CDHP plans now exempt some or all preventive care services from the deductible and vendors are offering more sophisticated incentive programs designed to help people stay healthy and manage chronic illnesses. In short, although CDHP enrollment has been lower than many in the industry had initially expected, it is still growing and forces may now be aligned to help it expand more quickly.
Research to Date on CDHPs
Given the growing enrollment in CDHPs and the increased focus on improving the value of health services delivered, what is known about the effects of CDHPs on healthcare costs and quality? The most authoritative estimates of their likely effects come from the decades-old RAND Health Insurance Experiment (HIE). The HIE found that higher deductibles could reduce spending by about 5-10 percent (Newhouse, 1992). The reductions in care use implied by these savings came through reductions in patient-initiated visits, however, not through the choice of more cost-effective treatments. Indeed, the reductions occurred in all types of services—preventive care, routine chronic care, and care deemed necessary by physician experts as well as care deemed unnecessary. Yet these reductions in care use had little effect on health outcomes during the evaluation period. The exception was for patients who were both sick and poor; their health declined more in the higher cost-sharing plans.
There are reasons, however, why the effects of today’s CDHPs might differ from the high-deductible plans studied in the RAND experiment. First, medical technology has advanced dramatically since the experiment began in the early 1970s and, thus, the consequences of going without care could be more severe. Second, the Internet has made medical information much more widely available, potentially increasing the ability of patients to make informed decisions about care. Third, new consumer-driven designs often emphasize prevention and are coupled with other programs such as disease management programs and health risk assessments. Finally, personal savings accounts could provide patients with the liquidity they need to initiate care. Still, despite these changes, it is important to know that the overall financial risk borne by CDHP enrollees is likely greater than that borne by the original HIE participants, because the latter were compensated for enrolling in
the experiment and had their financial exposure limited to a percentage of their income. Thus, the question of whether modern CDHP designs can spur consumers to seek higher-value care—rather than indiscriminately cut back on care when faced with higher costs—remains an open one.
So what does the recent literature suggest about CDHPs? Unfortunately, evidence is still largely limited to early adopters of these plan types and to studies of the experience of a single employer or a single insurance carrier. A review in 2006 of the evidence to date about CDHPs concluded that moving consumers from traditional plans to high-deductible plans would result in a one-time reduction in service utilization of about 4-15 percent (Buntin et al., 2006), but that coupling these plans with funded savings accounts could reduce this effect by half. Some evidence supported lower average spending, smaller premium increases, and lower use of care across a range of services, but mixed results were found for changes in individual spending over time and for quality of care.
Recent work has largely confirmed these findings—mixed as they are. A follow-on study by Feldman and colleagues (Feldman et al., 2007) suggested that after the initial cost savings, CDHP enrollees might actually spend more in subsequent years (Figure 4-4). The authors also found that expenditures for hospitalizations were higher for CDHP members than for members of preferred provider organizations (PPOs) and that the savings within the CDHP group came only through reduced pharmacy costs. However, Feldman and colleagues concluded that the CDHP they studied did not have high enough cost sharing to limit care use—in particular, it had an HRA account to which the employer made a substantial contribution.
Other studies have found mixed results for different types of care use. Wharam and colleagues (Wharam et al., 2007) showed that emergency room (ER) visits, hospitalizations, and hospital days all decreased among CDHP patients and increased among members of traditional plans. Interestingly, they also found that CDHP enrollees were much less likely to have a second ER visit, indicating that they learned something about the costs of ER use during their first visit. However, the average cost for a CDHP member’s hospital stay compared to the costs for members in traditional plans was higher, suggesting that CDHP consumers may delay care until reaching a critical point. Rowe and colleagues (Rowe et al., 2008) found that preventive care visits decreased among both Aetna CDHP and PPO enrollees, but the decrease among the PPO enrollees was actually slightly larger (Figure 4-5).
Preventive service use and screening rates among CDHP participants are comparable to those of members of traditional plans if their CDHP offers first-dollar coverage for preventive services. Rowe and colleagues (2008) found that preventive service use for diabetes and preventive screen-
ings rates showed no consistent pattern of association with CDHP enrollment. Busch and colleagues (2006) also found either no effect or a positive effect of CDHP enrollment on the rate of cancer screenings and well-child visits (Figure 4-6). It is worth noting, however, that rates of use of preventive care for both the CDHP and the comparison plan show substantial room for improvement.
One area in which CDHPs do compare unfavorably to traditional plans is appropriate prescription drug use. Building on prior work reporting that CDHP members are more likely to forgo filling a prescription because of cost than those in traditional plans, Greene and colleagues (2008) established that CDHP enrollees are more likely to discontinue using medication for chronic diseases than consumers in a basic three-tier copayment plan (Figure 4-7). However, these authors also found that CDHP consumers have the same rates of medical adherence if they do continue their medication and are no more likely to substitute to generic drugs.
While the section above summarizes the recent published literature, it is important to note a few limitations in what has been learned to date. First, all of the studies mentioned are observational studies of a single insurer or employer, most with a pre/post design and an attempt to control statistically for selection into different plans. Given that favorable selection into CDHPs is now fairly well documented (e.g., Busch et al., 2006; Dixon et
al., 2008; Government Accountability Office, 2006), the results should be interpreted in that light. Second, CDHP vendors have conducted their own internal studies that paint a brighter picture of CDHP cost savings and health outcomes than the results discussed here. These industry studies have the benefits of larger sample sizes and more recent data. However, most of these publications are not peer-reviewed so their methods and assumptions have not been subject to outside scrutiny. Third, there is considerable variation in benefit design among CDHPs. Plans have different constellations of coinsurance rates, deductibles, and personal accounts—all of which can affect care use. Most of the CDHPs studied in the published literature have moderately high deductibles and are coupled with employer-funded HRA accounts. Newer plans may be coupled with HSA accounts and may employ a broader array of incentives and information tools that can facilitate consumer decision making and potentially increase value.
How Could CDHPs Be Shaped to Increase Value?
Consumer-directed health plans are predicated on the idea that informed and incentivized consumers can make decisions about their healthcare use that yield better outcomes at lower cost: in other words, their goal is to improve value. Putting aside whether or not CDHPs are currently doing this, the goal is a good one. CDHPs should be shaped to increase their ability to reach that goal. One way to do this, which is entirely in keeping
with the philosophy of consumer-directed care, is promoting the collection and dissemination of information about the cost and quality of care. Such information can change behavior and increase value. For example, when provided with information and faced with differential cost sharing, people do switch from brand-name medications to generic drugs. However there is relatively little information available about the cost and quality of hospital procedures and services, and even less on the outpatient side. Human resources executives at firms offering CDHPs to their employees nearly universally agreed that the information tools available to their employees were of fair or poor quality (Watson Wyatt/RAND, 2007). Worse, they did not cite any improvements in them between 2007 and 2008 (Figure 4-8). Without such information, it is hard for consumers to make meaningful choices among providers that take value into account.
A second way to increase the value of CDHPs is to disseminate and deploy “best practices” in CDHP design. For example, many CDHPs cover preventive services before the deductible is met, but not all do. This should be encouraged, especially for high-value preventive services (Masciosik et al., 2006). In addition, it would be beneficial to define some drugs and services that support secondary prevention—such as drugs for the prevention of the sequelae of diabetes and foot exams for those with diabetes—as preventive care exempt from the deductible. Many CDHPs currently offer financial incentives to participate in health improvement programs (Figure 4-9). It could be valuable to expand the combination CDHPs with these and other
value-oriented approaches that are discussed in this volume, including value-based benefit designs, tiering, and workplace-based wellness initiatives.
Finally, value could be increased with more research about the effects of different CDHP designs on the use and outcomes of care. This includes research on the specific aspects of CDHP design such as deductible levels, deductible exemptions, and personal savings accounts. It also includes research on the value of health improvement programs that are being promoted under the broad rubric of “consumerism.” The effectiveness of these programs alone, in conjunction with CDHPs and in combination with other value-based programs, is not yet known, but the need for strategies that can yield improvements in value is clear.
THE ROLE OF TIERED BENEFIT DESIGNS FOR IMPROVING VALUE IN HEALTH CARE
Dennis P. Scanlon, Ph.D., and Grant R. Martsolf, M.P.H., R.N., Ph.D. Candidate, The Pennsylvania State University
Definition and Motivation for Tiered Benefit Programs
As discussed in this report and many other sources, the healthcare system in the United States suffers from substantial deficits in quality, safety,
efficiency, and value (IOM, 2000, 2001; McGlynn, 1997). Policy makers, academics, and clinicians from across the ideological spectrum have proposed a wide array of strategies to improve value in the healthcare system, many of which are discussed in other chapters of this volume. For example, improvement efforts may come through the regulatory system, utilizing mechanisms such as accreditation and professional licensure (Brennan, 1998). Other strategies include providing “supply-side” incentives to hospitals, physicians, and health plans to improve the quality of their care. These incentives may include both financial (e.g., pay-for-performance programs; Rosenthal et al., 2005) and nonfinancial or “reputational” incentives (e.g., public reporting systems; Marshall et al., 2000).
There has also been a growing interest in “demand-side” interventions to address quality, safety, and value deficiencies. For example, increasing attention has been paid to the notion of “consumer engagement,” which describes a variety of activities designed to help patients become more active in their care. These activities may include helping consumers utilize public reports and providing self-management education (Hurley et al., 2009). Another “demand-side” strategy that is garnering significant attention is the concept of “tiering.” Tiering can be defined as the classification of healthcare providers (e.g., hospitals, physicians) or treatments (e.g., pharmaceuticals, durable medical equipment, physical therapy) into different groups—or tiers—based on objective or subjective criteria such as measures of cost, quality, safety, or value.
Prevalence of Tiering Programs
Tiering of prescription drugs is the most prominent form of tiering program and is emerging as a nearly universal benefit characteristic. In 2008, the Kaiser Family Foundation found that 92 percent of beneficiaries face some type of differential cost sharing. This is equal to a 21 percent increase from 2000 (76 percent) (Kaiser Family Foundation and the Health Research Educational Trust, 2008a). Furthermore, in 2000, only 27 percent of beneficiaries were in plans with three tiers and no plans reported four tiers (Kaiser Family Foundation and the Health Research Educational Trust, 2000). By 2008, 77 percent of all beneficiaries were in plans with at least three tiers, an increase of 185 percent. Four-tier structures are also growing in popularity as 7 percent of all beneficiaries in 2008 were enrolled in such plans (Kaiser Family Foundation and the Health Research Educational Trust, 2008a).
Although far less prevalent than tiered prescription drug benefits, health plans are also beginning to institute tiered benefit designs for hospital care and physician services. There are few estimates of the prevalence of tiered benefit structures for providers, but in 2005, the Kaiser Family Foundation estimated that 13 percent of beneficiaries were enrolled in plans with
tiered provider benefits (Kaiser Family Foundation and the Health Research Educational Trust, 2005). Historically, provider tiering has been advanced by large employers or purchaser coalitions. Therefore, markets with firms that are active in health benefit reform are more likely to see these design changes (Draper et al., 2007).
Although provider tiering programs are not prevalent, employers continue to be somewhat interested in instituting these benefit changes. In Kaiser’s 2008 Employer Health Benefits Survey, 18 percent of firms stated that they were at least somewhat likely to adopt tiered copayments for office visits or hospital stays (Kaiser Family Foundation and the Health Research Educational Trust, 2008a). Multistakeholder quality improvement collaboratives are also beginning to investigate the possibility of developing provider tiering programs in local communities (Anthem Blue Cross Blue Shield, 2007; Maine Health Management Coalition, 2008).
There is no published evidence that tiering has been used for other types of therapies or services such as durable medical equipment, physical therapy, or long-term care, but it is conceivable that tiering could be utilized in these sectors of health care as well.
Goals of Tiering Programs
By assigning providers or products to different tiers and offering information about the tiering method and the financial implications of choosing a tier, healthcare purchasers and payers seek to accomplish several goals. Commonly cited goals of tiering programs include (1) engaging patients and consumers in making informed decisions about providers or therapies; (2) steering patients toward better-quality or safer hospitals and doctors; (3) cost savings through the use of lower-cost providers or therapies; and (4) helping patients to better self-manage their health conditions and to receive appropriate care. Thus, the “first-order” goal of a tiering program is to influence a decision such as the choice of a prescription drug or the use of the best doctors and hospitals. This choice is important only in so far as it leads to the “second-order” goal(s), which include improved quality, value, or safety.
Although many different methods can be used to place providers into tiers, these decisions are generally based on cost, quality, or safety measures or a combination of these. As part of their tiering programs, employers or health insurance plans generally provide consumers with information about how the tiers are constructed and which options are in each tier. Consumers often, but not always, face financial incentives in tiering programs, such as reduced premiums or copayments for choosing the tiers with preferred providers or treatments (Draper et al., 2007).
Although the popular press might suggest otherwise, tiering is not a
new concept in health care, but instead originates from methods used previously. For example, a strict pharmaceutical formulary essentially establishes a “two-tiered” prescription drug benefit since if a patient chooses a drug that is not on the formulary, the health plan simply does not pay for it. Similarly, closed-model health maintenance organizations (HMOs) essentially offer a “two-tiered” provider benefit. The first tier would include in-network providers, whereas out-of-network providers would be in the second tier and would not be covered. One key distinction however is that modern tiering systems, having learned from the “managed care backlash,” generally do not place hard limits on consumer choice. Instead, beneficiaries are allowed to choose any provider but face differential cost sharing based on the tier in which the provider is placed. Another distinction from historical managed care programs is that today’s tiering programs increasingly incorporate quality and safety as factors used to define tiers. In short, tiering programs are attempting to utilize the cost-sharing benefits of managed care to drive improvements in value, safety, and quality without restricting consumer choice or focusing exclusively on costs.
Stakeholders’ Perspectives on Tiering Programs
While tiering programs are increasing in popularity, stakeholders vary in their perspective on these programs. Purchasers (e.g., self-insured employers, government healthcare programs) and payers (e.g., health insurance plans) view tiering as a potentially attractive way to reward high-value providers and give other providers incentives to improve their outcomes. They also view tiering as a way to increase the utilization of more cost-effective treatments or therapies, such as the substitution of generic drugs for brand-name drugs. Tiering is also attractive to purchasers and payers because it allows these entities to continue to offer broad provider networks, thus avoiding the negativity that comes with restricting consumer choice (Draper et al., 2007). The perspective of other stakeholders is likely mixed. For example, physicians or hospitals may differ in their assessment of tiering on either philosophical or empirical grounds. Philosophically, some providers are opposed to differential consumer incentives, such as lower copayments or coinsurance, for the same types of providers. On the other hand, some providers view tiering as an opportunity to demonstrate value and to be rewarded with increased market share and reimbursement. For this latter group, the primary concern is making sure that the methods used to develop tiering programs are scientifically valid and robust.
Patients and manufacturers of drugs and other medical devices are also likely to have mixed views on tiering. On the one hand, tiering programs may provide incentives for all parties in the healthcare system to demonstrate better value, making the primary concern the validity and fairness
of the tiering criteria. On the other hand, since tiering is a departure from historical norms, many patients, providers, and producers view it as an attempt to resurrect the old-style managed care programs.
Requirements and Technical Considerations for Tiering Programs
Several factors and considerations are important when designing and implementing tiered benefits programs. These are discussed in turn.
Measurement and Data Availability
Because tiering programs distinguish providers, drugs, or therapies based on cost, quality, or safety, it is important that these dimensions can be measured accurately and reliably from readily available data. The risk of faulty measurement is significant for all stakeholders since, for example, inaccurate measurement can lead to the false classification of providers, resulting in unintended consequences for payers, purchasers, and patients, not to mention reputational, revenue, and market share consequences for providers.
Measuring and comparing the quality of brand-name drugs and generics within the same class is relatively straightforward as is measuring drug costs, but even so there are still many important technical details to consider when establishing a pharmaceutical tiering program (e.g., how to handle discounts and rebates when computing pharmacy costs). Healthcare costs are also notoriously difficult to measure and as such can potentially lead to erroneous tiering classifications (Fishman et al., 2004). When it comes to hospital care and physician services, measurement is more difficult. For quality and safety, national consensus measures are emerging through the efforts of groups such as the National Quality Forum (NQF, 2009) and should eventually serve to alleviate concerns about the measurement aspect of tiering programs. For now, however, there remain substantial debate regarding the ability of these consensus measures to accurately capture the quality and safety of care and concerns about the appropriateness of the available measures for use in tiered hospital and physician benefits programs.
In some cases the measurement science is ahead of purchasers’ and payers’ practical abilities to collect the necessary data to measure and classify providers into tiers. In these cases the technical issues often relate to issues such as data availability, data source (claims vs. electronic record vs. provider self-report), ability to attribute patients to providers, and necessary minimal sample size required for accurate measurement (Fishman et al., 2004; Iezzoni, 1997; Krein et al., 2002). Since most general acute care and tertiary care hospitals provide a broad range of treatment and services,
including medical and surgical care, there is also the decision about whether to tier hospitals across an entire range of services or to tier separately based on specific clinical services (e.g., cardiac or cancer care) within a hospital.
Market-Level Provider Capacity
Although tiering programs are appealing conceptually, they will not be successful without a sufficient number of providers in the preferred tiers or unless providers in the preferred tiers are operating at full capacity. The first issue, while potentially mutable over time, can be a limiting factor if few options are available in the preferred tier. When this is the case, it is a matter of judgment about when to proceed with a tiering program. In some cases, sponsors may proceed immediately to provide an incentive for providers to qualify for the preferred tier (i.e., presumably through improving quality or value), while others may wait until there is a viable set of options in the preferred tier that can be selected at the launch of the program. The second situation is more challenging and not easily mutable in the short term. In other words, if physicians or hospitals in the preferred tiers are operating at capacity, such that they cannot take on new patients, then the effect of the tiering program is inherently limited. This is an important consideration because there are locations in the United States where both hospitals and physicians are operating at full capacity or are facing significant physician shortages in primary care and other specialties (Bazzoli et al., 2003; Cooper et al., 2002; Trude, 2003).
Tiering programs, particularly those that include financial incentives, have budgetary implications for the program sponsor since changes in the distribution of providers, drugs, or products used have real actuarial implications. It is impossible to make firm general statements about the direction and magnitude of actuarial implications of tiering programs. Instead, each program must be assessed individually. For example, tiered pharmaceutical benefit programs that encourage generic substitution would have to estimate the impact on overall cost to the plan sponsor of varying copayment amounts by tier. This estimate would depend on the baseline utilization and cost of drugs in each tier as well as the rate by which patients substituted drugs in different tiers. Although this projection is still subject to some uncertainty regarding the substitution rate, plan sponsors should have reasonable information about the costs of drugs in the various tiers. The same is often not true for physician or hospital tiering, where the link between quality and safety measures and costs is not as well established and often depends on probabilistic expectations regarding complication
rates and length of stay. In this case, plan sponsors should make budgetary projections based on the best available evidence and in consultation with actuaries, while accounting for the uncertainty and also considering an appropriate time horizon for achieving a return.
The Evidence Base for Tiered Benefit Programs
Tiered benefit designs for prescription drugs have emerged in response to growing prescription drug costs and represent the most common type of tiering program, although drug tiering is clearly different from hospital and physician tiering. While the details of the benefit structures differ between plans, there are common characteristics. Drugs tend to be divided into either two or three tiers (and sometimes four) based on cost and clinical efficacy criteria. In two-tier plans, generic drugs are placed in Tier 1 while brand-name drugs are placed in Tier 2. In three-tier plans, the brand-name drugs are further differentiated into preferred and nonpreferred drugs, which are brand-name drugs with similar clinical indications and effectiveness but different prices. Patients can choose from any of the drug types but face increasing copayments as they move up the tiering ladder. The copayments for each of the tiers vary by health plan but average $11, $25, and $43, respectively, for Tier 1, Tier 2, and Tier 3 (Kaiser Family Foundation, 2008).
Because pharmaceutical tiering strategies have grown in popularity, there is a substantial collection of literature investigating their effect. The literature focuses on both “first-order” goals, such as changes in drug choice, and “second-order” goals, such as total expenditures and clinical outcomes (e.g., drug discontinuation rates). The literature does suggest that tiered drug benefits affect drug choice. The effect is most apparent for preferred brand-name drugs. Specifically, introducing a third tier tends to decrease spending for nonpreferred drugs and increase spending for preferred drugs (Gibson et al., 2005). For example, one study shows that the use of preferred drugs increased for ACE inhibitors (13 percent), protein pump inhibitors (8.9 percent), and statins (6 percent) (Rector et al., 2003). However, Gibson and colleagues (2005) suggest that tiering has been less effective in encouraging switching to generic medications, perhaps because generic brand price differentials have been too small to induce substitution of generics for brand-name drugs.
The literature also suggests that the prescription drug switching induced by tiering schemes tends to reduce total expenditures on prescription drugs (Fairman et al., 2003; Gilman and Kautter, 2007; Motheral and Fairman, 2001). However, these reductions in expenditure tend to be captured by
health plans, while beneficiary costs may actually increase (Gibson et al., 2005; Hodgkin et al., 2008; Huskamp et al., 2003, 2005). For example, Landon and colleagues (2007) show that as overall drug spending decreased by 5-15 percent, health plan spending decreased around 20 percent, while out-of-pocket payments for beneficiaries increased by at least 20 percent to more than 100 percent.
Literature investigating the effect of tiered benefit designs on clinical outcomes, such as morbidity, is scant. However, some literature does focus on intermediate outcomes such as drug discontinuation. There is some evidence that tiering may lead to drug discontinuation in certain drug classes, among certain groups of patients, and within certain benefit designs, but these results have not been consistent across a diversity of settings (Gibson et al., 2005; Huskamp et al., 2003, 2005, 2007; Landsman et al., 2005).
Although there are no estimates on the exact prevalence of physician tiering, it is believed that the prevalence of these programs is small. Health plans and employers have introduced physician tiering schemes in only a limited number of markets (Draper et al., 2007). Despite its low prevalence, physician tiering has garnered quite a bit of critical media attention.
Existing physician tiering strategies vary but commonly employ a two-or three-tiered system, which groups physicians based on either cost or quality or a combination of both. Tiering strategies have been used for both primary and specialist services. There is not only substantial variation in the criteria (cost, quality, or both) used to classify physicians but also in the methodology used to determine the tier placement. Depending on the methodology, plans place anywhere from 25 to 80 percent of physicians in the highest tiers. In some plans, beneficiaries are required to pay different copayments depending on the tier in which their physician is placed, whereas other plans simply use the tiers as informational tools for beneficiaries (Draper et al., 2007).
One employer that has been out in front of physician tiering has been the Commonwealth of Massachusetts, thus providing a useful case study of how physician tiering programs have been executed in the real world. Health benefits for state employees in Massachusetts are overseen by a quasi-governmental organization called the Group Insurance Commission (GIC). GIC contracts with a variety of health plans including Unicare, Harvard Pilgrim, and Tufts Health Plan to administer the benefits. In June 2006, GIC introduced a physician tiering program called the Clinical Performance Improvement (CPI) Initiative, which was designed to both reduce costs and increase quality while maintaining provider choice for
beneficiaries (Commonwealth of Massachussetts Group Insurance Commission, 2007).
Each of the plans provided claims data to GIC for all of its members. Based on clinical guidelines, GIC established process performance standards for each of the specialties. From the claims dataset, GIC used a novel algorithm to assess the quality of care for each physician in Massachusetts. Quality is essentially based on the percentage of eligible patients for which a physician complies with the clinical guidelines. Similarly, GIC was able to use the claims dataset to determine the cost of care provided by each physician for similar conditions, adjusted for geographic and market characteristics (Wellpoint, Inc., 2008).
Each of the GIC plans has the freedom to execute CPI in its own way. Some of the plans’ programs tiered all Massachusetts physicians, while others only tiered certain specialties. One example of how a plan has executed CPI comes from Unicare, which placed physicians in one of three tiers using standard benchmarks for cost and quality. Unicare classified physicians as “excellent” and placed them in Tier 1 if they have performed at or above the benchmark for both cost and quality. Physicians are considered “good” and placed in Tier 2 if they are below (but not more than 25 percent below) the benchmark for either cost or quality. Physicians are considered “standard” and placed in Tier 3 if they are at least 25 percent more expensive or at least 25 percent lower in quality than comparable physicians (Wellpoint, Inc., 2008).
Based on the tier placement, beneficiaries are rewarded for choosing high-quality, efficient providers by being subject to lower copayments for physicians in higher tiers. Basic members in the Unicare plan face copayments of $10, $20, and $25 for primary care physicians respectively in Tier 1, 2, or 3. For specialty care, beneficiaries face copayments of $10, $20, and $30 (Wellpoint, Inc., 2008). Although the GIC program has received substantial publicity, other plans, such as Regence Blue Cross in Washington as well as UnitedHealthcare and Cigna in Connecticut, have also experimented with physician tiering (King, 2008).
To our knowledge there is no peer-reviewed literature examining the effect of this or similar physician tiering strategies on any kind of outcome, including physician choice, quality improvements, clinical outcomes, costs, or expenditures. However, there is a single study investigating the potential effects of physician tiering on care inequities for minorities, which the authors conclude is likely minimal (Brennan et al., 2008).
Physician tiering systems have been extremely controversial, leading to a series of lawsuits across the nation. In November 2006, the Washington State Medical Association filed suit against Regence Blue Shield, claiming that the tiering methodology did not adequately measure the quality of physician care. The suit was ultimately settled in August 2007 and Regence
agreed to postpone the tiering program until it could receive sufficient input from physicians. In 2007, the Fairfield County Medical Society in Connecticut filed a similar suit against UnitedHealthcare and Cigna (King, 2008). GIC has not been immune from legal action. In May 2008, the Massachusetts Medical Society filed suit against GIC, Unicare, and Tufts Health plan claiming both defamation and fraud (Krasner, 2008).
These lawsuits illustrate the importance of measurement and data in physician tiering programs. In each of the suits, the plaintiffs claim that the measurement of care quality was arbitrary or did not represent the actual quality of care provided by physicians. Bruce Auerbach, the president of the Massachusetts Medical Society stated that “the GIC has refused to correct the CPI’s most glaring problem, which is its ranking of individual physicians using inaccurate, unreliable, and invalid tools and data” (Krasner, 2008).
Even as lawsuits continue, there is some indication that progress is being made toward consensus on physician tiering methodologies. In July 2007, Attorney General of New York Andrew Cuomo sent letters to CIGNA, Aetna, and UnitedHealthcare expressing concern that physician tiering methodologies were based on inaccurate data and were disproportionately based on cost rather than quality (King, 2008). The health plans quickly agreed to adopt Cuomo’s “Doctor Ranking Model Code.” Among other provisions, the code required plans to disclose tiering methodology, to use risk adjustments, and to use national consensus measures (New York State Office of the Attorney General, 2007). Similar codes have been adopted in other states including Colorado (Berry, 2008).
A coalition of consumer, labor, and employer organizations called the Consumer-Purchaser Disclosure Project, which includes AARP, the Leapfrog Group, and the National Business Coalition on Health, has been working on a similar initiative. This coalition has agreed on a set of principles that will guide future efforts in performance measurement, reporting, and tiering. Called the “Patient Charter,” this set of principles includes reliance on national consensus measures and disclosure of measurement methods (Robert Wood Johnson Foundation, 2008). The Robert Wood Johnson Foundation recently funded a study by George Washington University investigating the legality of tiering programs. That study has affirmed that these types of transparency provisions should ensure the legality of tiering programs (Robert Wood Johnson Foundation, 2007). These, and similar, agreements may facilitate the expansion of physician tiering.
There are also a variety of methods used to place hospitals in specific tiers, but placement is generally based on cost and/or quality measurements. Blue Cross and Blue Shield of California (BCBSCA) introduced a tiered hos-
pital network in 2002 based purely on costs (cost per inpatient episode and cost per outpatient procedure) (Robinson, 2003). Boeing, in its self-funded plan, placed hospitals in tiers based on compliance with the safety goals outlined by the Leapfrog Group (Rosenthal et al., 2007). However, other health plans have attempted to create composite indices. For example, the Tufts Health Plan uses quality measures that have been required by Medicare and the Joint Commission (JCAHO) for regulatory purposes as well as cost data to place hospitals into tiers (Rosenthal et al., 2007). Hospitals are then rated as “good” if they meet no cost or quality standards, “better” if they meet one of the standards, or “best” if they meet all standards for cost and quality (Rosenthal et al., 2007).
These measurement details prove to be very important in hospital tiering programs. Rosenthal and colleagues (Rosenthal et al., 2007) tested whether or not rating methodology had an effect on hospital tier placement. The authors tested two “extreme” strategies based on only cost or quality as well as two hybrid strategies— one that used cost data with minimal quality data and another that weighted cost and quality equally. This study found that there was little agreement between the four strategies, even the hybrid strategies, suggesting that measurement methodology is extremely important.
Clearly, the grouping methodology can have a great effect on health plans’ tiering programs. However, if done correctly, there is some early evidence that if the financial incentives lead patients to higher-quality hospitals, the tiering structures may actually lead to improved quality of care for patients. In a study released in November 2008, the Blue Cross/Blue Shield Association (BCBSA) showed that hospitals that were awarded a “Blues Distinction” had lower readmission rates for cardiac patients. “Blues Distinction” is a designation developed by the BCBSA that is used by local Blues plans to structure quality-based tiering schemes (Nylen, 2008). However, wider evidence of the effect of quality designation on outcomes has yet to emerge and there are no similar studies on cost-based designations.
Much of the empirical evidence on the effect of tiering programs on hospital choice has emerged from an evaluation of a single firm’s hospital tiering initiative. In July 2004, this firm changed the standard hospital benefit for union employees from 100 percent to 95 percent coverage. However, if the beneficiaries used a “safe” hospital, defined as one that complied with the Leapfrog Group’s three safety “leaps,” the benefit would return to 100 percent (Scanlon et al., 2008). The evaluation has tested the effect of the tiering program on awareness and attitudes regarding patient safety as well as its actual impact on hospital choice (Scanlon et al., 2008).
The results of this intervention were mixed. There was no evidence to suggest that the education associated with the tiering program had any effect on beneficiaries’ awareness and attitudes. The intervention group
was no more likely than the control group to seek information on health benefits and quality of care, acknowledge variation in medical errors among hospitals, or express willingness to go to a different hospital (Scanlon and Christianson, 2008). Despite no significant changes in awareness and attitudes, there were some effects on hospital choice among members of the engineers’ union admitted for medical conditions. Specifically, this group was 2.92 times more likely to visit a so-called “safe” hospital after the intervention than before. However, there was no significant effect for surgical admissions or within the machinist union’s employees, who were less well educated and earned less than the engineers (Scanlon et al., 2008).
The effects of this study may have been mediated by the limitations of the program. Only 17 percent of all hospitals were placed in the top tier. Perhaps, there would have been a greater effect if the patients had more hospitals to choose from in the preferred tier. Also, because the analysis was limited to a single payer, it is reasonable to assume that the results cannot be generalized to other plans or firms. Regardless of these limitations, the results do draw into question the effectiveness of “demand-side” interventions executed by employers and plans for the purpose of influencing provider choice, particularly when considered in relationship to alternative approaches such as changing supply-side behavior or providing supply-side incentives (Scanlon et al., 2008).
Conclusions and Policy Implications
Tiered benefit designs are now commonplace for pharmaceuticals, and while not nearly as prevalent for other healthcare products or services, this type of benefit design is becoming increasingly popular for physician and hospital care. Tiering is theoretically appealing because it provides an incentive for patients to select high-value providers, which in turn is expected to stimulate supply-side improvement for fear of losing revenue and market share. Yet with the exception of tiered drug benefits, there is unfortunately little published evidence about the impact of tiering, making it difficult for purchasers and payers to easily predict the outcome of adopting tiering strategies. Like many advances in healthcare finance and insurance design, while there is a dearth of evidence in the published literature, it is likely that some of the more innovative private sector payers and purchasers who have implemented tiering programs have amassed unpublished evidence on the topic. Despite this dearth of evidence, our analysis has highlighted some key considerations when thinking about the potential benefit and impact of tiered benefit design programs.
Consider alternative strategies to achieve goals As discussed above, tiered benefit designs are considered a means to achieving the ultimate end of
improved value, quality, and safety. As such, it is important to consider alternative approaches to accomplish these goals, including regulation, pay-for-performance, professional development, and continuing education in the areas of quality of care and efficiency. While the evidence base for improving value in these areas may not be strong either, those considering tiering programs should also understand the cost-benefits of these different options.
Evidence from pilot studies is needed The currently scant evidence base will be improved only if purchasers and payers implementing tiered benefit programs study the financial and clinical impact of these programs, including the return on investment to program sponsors. Thus, program sponsors may want to implement programs gradually, allowing first for a pilot phase to gauge the likely outcomes of full-scale program implementation.
Healthcare market characteristics matter Providing incentives for consumers presumes a set of viable options. Therefore, the impact of tiered benefit programs, particularly on physicians and hospitals, will depend on local market characteristics, including the supply and capacity of healthcare providers. This is particularly important in markets where there are shortages of physicians or hospital beds.
Targeting the incentive at the decision maker Tiered benefit designs provide incentives to consumers for decisions that are made primarily (or at least heavily influenced) by physicians or specialists. For example, patients are referred and admitted to a hospital by a physician and often in a weakened state where time is of the essence. In this case, a tiered benefit design may have limited impact if aimed at the patient rather than the admitting physician. Likewise, while physicians heavily influence hospital admission decisions and referrals to specialists, the degree to which they are able to consider alternative hospitals or specialists is increasingly dictated by admitting privileges, contractual affiliations, distance and geographic location, and ownership interest in healthcare facilities (Scanlon et al., 2006). All of these factors can be extremely important in the overall success of a tiering program.
Methods matter As discussed above, tiering programs are only as good as the methods and data used to define the tiers. Related to this point, program developers need to consider the “appropriate unit” for tiering programs. For example, should hospitals be tiered across their entire range of services or for a specific subset of services? Likewise, should individual physicians be subject to tiered benefit programs or only physician practices comprised of multiple physicians? The answers to these questions need to be made
after considering both the practical goals and objectives of the program and the data requirements needed to construct reliable and valid performance measures.
Communication matters Many consumers and patients have insufficient knowledge of health insurance plan benefits let alone the various plan options that might be available to them (Hibbard et al., 1998). Because tiered benefit plans often require information at the point of service, when either filling a prescription or making a decision about doctors and hospitals, it can be challenging to communicate this information to beneficiaries. While it is not clear that a perfect communication strategy exists, plan sponsors should develop a robust strategy for communicating information about tiered benefit designs to program beneficiaries.
A critical mass is important One potential problem with existing tiered benefit designs and other health finance innovations is that they have not reached a critical enough mass in the marketplace to be taken seriously by healthcare providers. In other words, the threat of losing market share may not be that great when a single, self-insured employer implements a tiered hospital benefit program in a community. However, if many other employers, health plans, and even the state and federal governments’ health programs implement similar tiering strategies, the threats and incentives become very real. Thus, those developing tiered benefit design programs should consider the need for a critical mass and seek leverage by partnering with others in the market.
POLICY PERSPECTIVES: HEALTH PROMOTION AND DISEASE PREVENTION (AKA WELLNESS)
Ronald Z. Goetzel, Ph.D., Emory University and Thomson Reuters
We are at a pivotal point in our nation’s history. Not only are we ushering in a new administration, headed by a charismatic and visionary leader, we are also at a juncture in American history where dramatic and significant changes in our healthcare system are not only possible, but probable. The manner in which health system reform unfolds during the coming months and years can and should be influenced by all Americans. As a collective, Americans are presented with a wondrous opportunity to transform a clunky, inefficient, and at times harmful healthcare delivery system to one that provides quality and cost-effective care, with an increased emphasis on prevention and health promotion.
Employers can play a significant role in improving peoples’ lives and their health. Their part has not been fully vetted in discussions of healthcare
reform, although the promise and potential for achieving large-scale health and economic impacts among working-age adults is undeniable.
Why employers? About 160 million Americans go to work every day, spending a significant portion of their waking hours in the work environment. In fact, more than ever before, work spills over past the traditional office hours in the form of e-mails, voicemails, and Blackberries. Work influences health, and in turn, workers’ health influences work performance. Astute employers wishing to improve the health and well-being of their workers can reach large segments of the population who would not normally be exposed to and engaged in organized health improvement efforts. Thus, an opportunity presents itself to positively influence population health and, at the same time, mitigate the rise in healthcare costs through workplace health promotion programs.
In many ways, the workplace represents a microcosm of society and an ideal setting for introducing and maintaining health promotion programs. Employers establish work rules and procedures that reflect the distinct norms and culture of the organization. Six key employer attributes supporting the potential of using workplaces as a venue for improving workers’ health include the following: (1) workplaces often contain a concentrated group of people who share a common purpose; (2) communication with workers is well established and straightforward; (3) social and organizational support for behavior changes is available; (4) data on program impacts can be tracked using existing organizational health surveillance programs; (5) certain policies, procedures, and practices can be introduced and organizational norms can be shaped; and (6) financial or other types of incentives can be offered to gain participation in programs.
Certain employers have recognized the benefits of worksite health promotion and are already far along in providing effective health improvement programs. Over the past 30 to 40 years, there has been a noticeable rise in the number of employers engaged in health promotion and, more broadly, health and productivity management programs at their workplaces. However, many of these initiatives have design and implementation flaws that reduce their potential for effective and positive change. A recent federally funded study published by Laura Linnan (Linnan et al., 2008), a professor at the University of North Carolina, found that although about 90 percent of employers say they have health promotion programs in place, only 6.9 percent actually offer effective programs containing these essential ingredients to make them successful: (1) health education, (2) links to related employee services, (3) supportive physical and social environments for health improvement, (4) integration of health promotion into the organization’s culture, and (5) employee screenings with adequate treatment and follow-up. In other words, most employers do not offer evidence-based programs that achieve clear health and financial objectives. Furthermore,
even those who do implement well-structured and evidence-based programs may not always be sure these programs work because they have not put effective measurement systems in place.
Making the Business Case for Workplace Health Promotion
Many employers are familiar with a growing body of literature showing that theory-based and empirically sound programs can improve workers’ health, lower their risk for disease, save businesses money, and improve an organization’s competitiveness (Goetzel and Ozminkowski, 2008). However, others lack the hard evidence necessary to convince program sponsors and company management that investing in workers’ health is “worth it”—not just from a humanitarian point of view but also because it is good for their business.
How might a health promotion program champion respond when confronted by the boss who says, “Convince me—why should a business invest in the health and well-being of its workers?” The response may take the form of a series of statements supported by a growing body of empirical evidence.
The logic flow for worksite health promotion can be articulated as follows. A large proportion of the diseases and disorders affecting workers is preventable (Partnership to Fight Chronic Disease, 2008). Modifiable health risk factors are precursors to a large number of diseases and disorders and, at the extreme, premature death (Amler and Dull, 1987). Many modifiable health risks are associated with increased healthcare costs and reduced worker productivity within a relatively short time window (Goetzel et al., 1998a). By utilizing a workplace-sponsored health promotion and disease prevention program, employers can target modifiable health risk factors and achieve improvements in the health risk profile of their population that can lead to reductions in healthcare costs and improvements in productivity (Goetzel et al., 2002; Heaney and Goetzel, 1998; Ozminkowski et al., 1999). Finally, worksite health promotion and disease prevention programs thus save companies money and produce a positive return on investment (ROI) (Aldana, 2001).
Where is the evidence supporting each of the above statements? What follows is a synopsis of the research linked to each of the major points.
There is little argument that poor health costs employers significant amounts of money and that many chronic health conditions, such as heart disease, Type 2 diabetes, and certain cancers, are largely caused by behavior and lifestyle (Mokdad et al., 2000). Excess spending has its source in increased and avoidable healthcare services, employee absenteeism, short- and long-term disability payments, higher accident rates, and diminished worker productivity (Goetzel et al., 2004). There is also growing evidence that workers
at higher risk for modifiable conditions such as obesity, inadequate physical activity, smoking, poor diet, unmanaged stress, and high biometric values for cholesterol, blood pressure, and glucose also cost more than those lacking these risks (Anderson et al., 2000) (Figure 4-10). Further, employees exhibiting several risk factors cost significantly more than employees with fewer risk factors. These higher costs affect not only employers but also employees, since the dollars spent on health care and other employee benefits are subtracted from employee salaries and total compensation. Thus, improving the health risk profile of workers can benefit employers and employees alike.
Workplace Programs as a Vehicle for Behavior Change
What is the evidence supporting the positive effects of workplace health promotion on health risks and behaviors? A systematic literature review commissioned by the U.S. Centers for Disease Control and Prevention (CDC) in 1995, and more recently in 2007, concluded that well-designed, evidence-based programs built on behavioral theory can achieve long-term health and productivity improvements in employee populations (Soler et al., 2009). In an earlier review, Catherine Heaney and I examined 47 peer-
reviewed studies over a 20-year period and found that workplace programs, in spite of their variability in terms of comprehensiveness, intensity, and duration, achieved long-term behavior change and risk reduction among workers (Heaney and Goetzel, 1998). The most effective programs were those that offered individualized risk reduction counseling to the highest-risk employees within a “healthy-company” workplace environment in which broader health awareness initiatives were already under way.
The review released in 2007 by the CDC Community Guide Task Force examined data from more than 50 studies that reported workplace program participation outcomes based on a range of health behaviors, physiological measures, and productivity indicators. The review was largely positive with sufficient and strong evidence supporting the view that workplace programs exerted a positive effect on poor behaviors and biometric values. When measured at an individual level, many of the changes in these outcomes were small, but at the population level they were considered substantial (Centers for Disease Control and Prevention, 2007) (Table 4-2).
Workplace Programs’ Financial Outcomes
What, then, is the evidence of cost savings? Here too several literature reviews that weigh the evidence from experimental and quasi-experimental study designs suggest that workplace programs using tailored communications and individualized counseling for high-risk individuals are likely to produce a positive ROI; that is, for every dollar invested over a three-year period, the ROI ranges from about $1.40 to $4.70 (Chapman, 2005; Goetzel et al., 1999). Studies often cited for the strongest research designs and the largest numbers of subjects include those performed at Johnson and Johnson, Citibank, Dupont, Bank of America, Tenneco, Duke University, the California Public Retirees System, Procter & Gamble, Highmark, and Chevron Corporation (Figure 4-11). Even when taking into consideration the inconsistencies in design and results across these studies, most of these workplace studies have shown positive financial outcomes.
TABLE 4-2 Evaluation of Worksite Health Promotion Programs— February 2007 Analysis Summary Results
Body of Evidence
Magnitude of Effect
SOURCE: Soler et al., 2007.
With the above discussion as background, following are personal suggestions for employer-directed initiatives as part of healthcare reform. The overall aim of these recommendations is to increase employer engagement in providing state-of-the-art and science-based health promotion programs to their employees.
Improve employer communication and education about the benefits of effective health promotion programs We need to do a better job communicating the human and economic costs associated with poor health, the effects of not achieving health improvements, and the options available to reduce health risks. Federal, state, and local health agencies, in partnership with businesses, should leverage their extensive marketing and communication networks to share information about exemplary programs to employers. This means a broader dissemination of results from scientific studies and translation of findings into understandable business language;
convening business group meetings on workplace topics; and speaking before legislators and policy makers with testimonial evidence regarding program successes.
Increase funding for applied research in “real-world” settings There are not enough federally funded, applied research studies that examine real-world applications of health promotion programs in the workplace. Until recently, much of the workplace research came from the private sector and was paid for by private sources. This research is improving but is still relatively primitive and limited. To enhance knowledge of best practices, more government support is needed for studying the science underlying workplace-based programs and the relative effectiveness of various component parts in improving health, lowering costs, and increasing worker productivity.
Develop tools and resources to support employer efforts in health promotion Several tools and resources have already been developed and disseminated with the support of government funding. However, additional ones are needed to help employers design, implement, and evaluate their workplace programs.
Pilot innovative health promotion programs at federal, state, and local health departments and agencies It is ironic that most government agencies do not have state-of-the-art programs for their own employees and dependents. Some noteworthy exceptions can be found in King County, Washington, and the State of Delaware, where experimental health promotion programs are now being implemented and evaluated. Public agencies should serve as laboratories for testing innovative approaches to improving workers’ health as well as role models that other employers can emulate.
Honor and reward America’s healthiest organizations We need to recognize and reward employers who are the champions of health improvement. Innovative organizations that have successfully implemented extraordinary programs deserve recognition. There are good award programs already in place, including those at the National Business Group on Health, the Health and Human Services Secretary’s Innovation in Prevention Award, and the Health Project’s C. Everett Koop National Health Award. These efforts should be further supported and expanded.
Create an employers’ health promotion resource center A government-supported resource center would collect, develop, and disseminate objective, easy-to-use, and accessible information and act as a clearinghouse for resources, tools, and expertise to support employer efforts.
Establish a public-private technical advisory council The council would be set up just like other government advisory panels, such as the U.S. Preventive Services Task Force and the Community Guide Task Force.
Establish collective purchasing consortia for small employers These consortia would define common health and business objectives for employers in a given community, achieve consensus on health promotion program designs, issue requests for proposals to vendors and health plans, and establish performance guarantees related to the success of these programs.
Support establishment of health promotion program certification and accreditation programs Several established review and accreditation organizations, such as the National Committee for Quality Assurance (NCQA), the Utilization Review Accreditation Committee (URAC), and the Health Enhancement Research Organization (HERO), have recently introduced review processes focused on workplace vendors and health plans. These organizations should be encouraged because they will improve the quality of health promotion programs and introduce a level playing field of competition across programs and vendors.
Provide financial incentives for employers to adopt evidence-based programs An immediate and effective way to capture the attention of businesses is to provide them tax credits for implementing bona fide health promotion programs. These tax credits would partially offset the costs of providing a qualified program. (See Senator Tom Harkin’s Healthy Workforce Act—S. 1753; Library of Congress, 2009.)
In sum, we need to make sure that there is a clear focus on workplaces as a venue for health system reform. The current U.S. healthcare system has major flaws. We are spending more than $2 trillion per year on health care, with three-fourths of this amount being directed toward the treatment of chronic diseases. Almost two-thirds of the growth in spending is attributable to Americans’ worsening health habits, particularly the epidemic rise in obesity. Our healthcare delivery system favors paying for treatment rather than prevention. For the United States to continue to be an economic leader worldwide, supported by a healthy and productive workforce, we need to direct our national attention and energy toward concerted health promotion and disease prevention. We can start by focusing on improving the health and well-being of employees who work within our organizations. Prevention and health promotion are essential to comprehensive health
system reform because improving the health of Americans will reduce the social and financial burdens imposed by preventable illnesses.
Aldana, S. G. 2001. Financial impact of health promotion programs: A comprehensive review of the literature. Am J Health Promot 15(5):296-320.
Amler, R. W., and H. B. Dull, eds. 1987. Closing the gap: The burden of unnecessary illness, New York: Oxford University Press.
Anderson, D. R., R. W. Whitmer, R. Z. Goetzel, R. J. Ozminkowski, R. L. Dunn, J. Wasserman, and S. Serxner. 2000. The relationship between modifiable health risks and group-level health care expenditures. Health Enhancement Research Organization (HERO) Research Committee. Am J Health Promot 15(1):45-52.
Anthem Blue Cross Blue Shield. 2007. http://www.anthem.com/member/me/f1/s0/t0/pw_ad087947.pdf (accessed January 9, 2009).
Bazzoli, G. J., L. R. Brewster, G. Liu, and S. Kuo. 2003. Does U.S. hospital capacity need to be expanded? Health Aff (Millwood) 22(6):40-54.
Berry, E. 2008. Colorado moves to regulate tiered networks. http://www.ama-assn.org/amednews/2008/05/05/bisb0505.htm (accessed January 5, 2009).
Brennan, T. A. 1998. The role of regulation in quality improvement. Milbank Q 76(4):745-747.
Brennan, T. A., C. M. Spettell, J. Fernandes, R. L. Downey, and L. M. Carrara. 2008. Do managed care plans’ tiered networks lead to inequities in care for minority patients? Health Aff (Millwood) 27(4):1160-1166.
Buntin, M. B., C. Damberg, A. Haviland, K. Kapur, N. Lurie, R. McDevitt, and M. S. Marquis. 2006. Consumer-directed health care: Early evidence about effects on cost and quality. Health Aff (Millwood) 25(6):w516-w530.
Busch, S. H., C. L. Barry, S. J. Vegso, J. L. Sindelar, and M. R. Cullen. 2006. Effects of a cost-sharing exemption on use of preventive services at one large employer. Health Aff (Millwood) 25(6):1529-1536.
Centers for Disease Control and Prevention. 2007. Task force community preventive services. Proceedings of the task force meeting: Worksite reviews. Atlanta, GA.
Chapman, L. S. 2005. Meta-evaluation of worksite health promotion economic return studies: 2005 update. Am J Health Promot 19(6):1-11.
Chernew, M. E., A. B. Rosen, and A. M. Fendrick. 2007. Value-based insurance design. Health Aff (Millwood) 26(2):w195-w203.
Chernew, M. E., M. Shah, A. Wegh, S. Rosenberg, I. Juster, A. Rosen, M. Sokol, K. YuIsenberg, and M. Fendrick. 2008. Impact of decreasing copayments on medication adherence within a disease management environment. Health Affairs 27(1):103-112.
Commonwealth of Massachusetts Group Insurance Commission. 2007. Raising the bar in health care: Fiscal year 2007 annual report. Boston.
Cooper, R. A., T. E. Getzen, H. J. McKee, and P. Laud. 2002. Economic and demographic trends signal an impending physician shortage. Health Aff (Millwood) 21(1):140-154.
Dixon, A., J. Greene, and J. Hibbard. 2008. Do consumer-directed health plans drive change in enrollees’ health care behavior? Health Aff (Millwood) 27(4):1120-1131.
Draper, D., A. Liebhaber, and P. B. Ginsburg. 2007. High-performance health plan networks: Early experiences. Washington, DC: Center for Studying Health System Change.
EBRI/Commonwealth. 2008. Findings from the 2008 EBRI consumer engagement in health care survey. Washington, DC: Employee Benefit Research Institute.
Fairman, K. A., B. R. Motheral, and R. R. Henderson. 2003. Retrospective, long-term follow-up study of the effect of a three-tier prescription drug copayment system on pharmaceutical and other medical utilization and costs. Clin Ther 25(12):3147-3161; discussion 3144-3146.
Feldman, R., S. T. Parente, and J. B. Christianson. 2007. Consumer-directed health plans: New evidence on spending and utilization. Inquiry 44(1):26-40.
Fendrick, A. M., and M. E. Chernew. 2006. Value-based insurance design: Aligning incentives to bridge the divide between quality improvement and cost containment. Am J Manag Care 12 (Spec No.):SP5-SP10.
Fishman, P. A., M. C. Hornbrook, R. T. Meenan, and M. J. Goodman. 2004. Opportunities and challenges for measuring cost, quality, and clinical effectiveness in health care. Med Care Res Rev 61(3 Suppl):124S-143S.
Furhrman, V. 2007. New tack on copays: Cutting them. Wall Street Journal, May 8, D1.
Gibson, T. B., R. J. Ozminkowski, and R. Z. Goetzel. 2005. The effects of prescription drug cost sharing: A review of the evidence. Am J Manag Care 11(11):730-740.
Gilman, B. H., and J. Kautter. 2007. Consumer response to dual incentives under multitiered prescription drug formularies. Am J Manag Care 13(6 Pt 2):353-359.
Goetzel, R. Z., and R. J. Ozminkowski. 2008. The health and cost benefits of work site health-promotion programs. Annu Rev Public Health 29:303-323.
Goetzel, R. Z., D. R. Anderson, R. W. Whitmer, R. J. Ozminkowski, R. L. Dunn, and J. Wasserman. 1998a. The relationship between modifiable health risks and health care expenditures. An analysis of the multi-employer HERO health risk and cost database. J Occup Environ Med 40(10):843-854.
Goetzel, R. Z., B. H. Jacobson, S. G. Aldana, K. Vardell, and L. Yee. 1998b. Health care costs of worksite health promotion participants and non-participants. J Occup Environ Med 40(4):341-346.
Goetzel, R. Z., T. R. Juday, and R. J. Ozminkowski. 1999. What’s the ROI? A systematic review of return on investment (ROI) studies of corporate health and productivity management initiatives. AWHP’s Worksite Health. Washington, DC.
Goetzel, R. Z., R. J. Ozminkowski, J. A. Bruno, K. R. Rutter, F. Isaac, and S. Wang. 2002. The long-term impact of Johnson & Johnson’s health & wellness program on employee health risks. J Occup Environ Med 44(5):417-424.
Goetzel, R. Z., S. R. Long, R. J. Ozminkowski, K. Hawkins, S. Wang, and W. Lynch. 2004. Health, absence, disability, and presenteeism cost estimates of certain physical and mental health conditions affecting U.S. employers. J Occup Environ Med 46(4):398-412.
Goldman, D. P., G. F. Joyce, J. J. Escarce, J. E. Pace, M. D. Solomon, M. Laouri, P. B. Landsman, and S. M. Teutsch. 2004. Pharmacy benefits and the use of drugs by the chronically ill. JAMA 291(19):2344-2350.
Government Accountability Office. 2006. Consumer directed health plans: Small but growing enrollment fueled by rising cost of health care coverage. Washington, DC.
Greene, J., J. Hibbard, J. F. Murray, S. M. Teutsch, and M. L. Berger. 2008. The impact of consumer-directed health plans on prescription drug use. Health Aff (Millwood) 27(4):1111-1119.
Heaney, C., and R. Z. Goetzel. 1998. A review of health-related outcomes of multi-component worksite health promotion programs. American Journal of Health Promotion (11):290-307.
Hibbard, J. H., J. J. Jewett, S. Engelmann, and M. Tusler. 1998. Can Medicare beneficiaries make informed choices? Health Aff (Millwood) 17(6):181-193.
Hodgkin, D., C. Parks Thomas, L. Simoni-Wastila, G. A. Ritter, and S. Lee. 2008. The effect of a three-tier formulary on antidepressant utilization and expenditures. J Ment Health Policy Econ 11(2):67-77.
Hurley, R. E., P. S. Keenan, G. R. Martsolf, D. D. Maeng, and D. P. Scanlon. 2009. Early experiences with consumer engagement initiatives to improve chronic care. Health Aff (Millwood) 28(1):277-283.
Huskamp, H. A., P. A. Deverka, A. M. Epstein, R. S. Epstein, K. A. McGuigan, and R. G. Frank. 2003. The effect of incentive-based formularies on prescription-drug utilization and spending. N Engl J Med 349(23):2224-2232.
Huskamp, H. A., P. A. Deverka, A. M. Epstein, R. S. Epstein, K. A. McGuigan, A. C. Muriel, and R. G. Frank. 2005. Impact of 3-tier formularies on drug treatment of attention-deficit/hyperactivity disorder in children. Arch Gen Psychiatry 62(4):435-441.
Huskamp, H. A., P. A. Deverka, M. B. Landrum, R. S. Epstein, and K. A. McGuigan. 2007. The effect of three-tier formulary adoption on medication continuation and spending among elderly retirees. Health Serv Res 42(5):1926-1942.
Iezzoni, L. I. 1997. Assessing quality using administrative data. Ann Intern Med 127(8 Pt 2):666-674.
IOM (Institute of Medicine). 2000. To err is human: Building a safer health system. Washington, DC: National Academy Press.
———. 2001. Crossing the quality chasm: A new health system for the 21st century. Washington, DC: National Academy Press.
Kaiser Family Foundation. 2005. E-health and the elderly: How seniors use the Internet for health. http://www.kff.org/entmedia/entmedia011205nr.cfm (accesses January 2009).
———. 2008. Prescription drug trends fact sheet—September 2008 update. Menlo Park, CA: The Henry J. Kaiser Family Foundation.
Kaiser Family Foundation and the Health Research Educational Trust. 2000. Employer health benefits annual survey. Menlo Park, CA: The Henry J. Kaiser Family Foundation.
———. 2005. Employer health benefits 2005 annual survey. Menlo Park, CA: The Henry J. Kaiser Family Foundation.
———. 2008a. Employer health benefits 2008 annual survey. Menlo Park, CA: The Henry J. Kaiser Family Foundation.
———. 2008b. Employer health benefits 2008 annual survey. Menlo Park, CA: The Henry J. Kaiser Family Foundation.
King, P. 2008. Physician ranking: Health care transparency or scam? http://www.netdoc.com/Physician-Practice-Articles/General-Medical-Practice/Physician-Ranking:-health-care-transparency-or-scam?/ (accessed January 6, 2009).
Krasner, J. 2008. Physicians group files suit over rankings. Boston Globe, May 22.
Krein, S. L., T. P. Hofer, E. A. Kerr, and R. A. Hayward. 2002. Whom should we profile? Examining diabetes care practice variation among primary care providers, provider groups, and health care facilities. Health Serv Res 37(5):1159-1180.
Landon, B. E., M. B. Rosenthal, S. L. Normand, C. Spettell, A. Lessler, H. R. Underwood, and J. P. Newhouse. 2007. Incentive formularies and changes in prescription drug spending. Am J Manag Care 13(6 Pt 2):360-369.
Landsman, P. B., W. Yu, X. Liu, S. M. Teutsch, and M. L. Berger. 2005. Impact of 3-tier pharmacy benefit design and increased consumer cost-sharing on drug utilization. Am J Manag Care 11(10):621-628.
Library of Congess. 2009. THOMAS. http://www.thomas.gov/cgi-bin/bdquery/z?d110:s.01753 (accessed March 27, 2009).
Linnan, L., M. Bowling, J. Childress, G. Lindsay, C. Blakey, S. Pronk, S. Wieker, and P.Royall. 2008. Results of the 2004 national worksite health promotion survey. Am J Public Health 98(8):1503-1509.
Maine Health Management Coalition. 2008. http://www.mhic.org/PTE/index.html (accessed January 9, 2009).
Marshall, M. N., P. G. Shekelle, S. Leatherman, and R. H. Brook. 2000. The public release of performance data: What do we expect to gain? A review of the evidence. JAMA 283(14):1866-1874.
Masciosik, M., A. Coffield, N. Edwards, T. Flottemesch, M. Goodman, and L. Solberg. 2006. Priorities among effective clinical preventive services: Results of a systematic review and analysis. American Journal of Preventive Medicine 31(1):90-96.
McGlynn, E. A. 1997. Six challenges in measuring the quality of health care. Health Aff (Millwood) 16(3):7-21.
McGlynn, E. A., S. M. Asch, J. Adams, J. Keesey, J. Hicks, A. DeCristofaro, and E. A. Kerr. 2003. The quality of health care delivered to adults in the United States. N Engl J Med 348(26):2635-2645.
Mokdad, A. H., M. K. Serdula, W. H. Dietz, B. A. Bowman, J. S. Marks, and J. P. Koplan. 2000. The continuing epidemic of obesity in the United States. JAMA 284(13):1650-1651.
Motheral, B., and K. A. Fairman. 2001. Effect of a three-tier prescription copay on pharmaceutical and other medical. Health Services Research Journal 1293-1304.
New York State Office of the Attorney General. 2007. Attorney General Cuomo announced doctor ranking agreement with independence health. http://www.oag.state.ny.us/media_center/2007/dec/dec12a_07.html (accessed January 5, 2009).
Newhouse, J. P. 1992. Medical care costs: How much welfare loss? Journal of Economic Perspectives 6:3-21.
NQF (National Quality Forum). 2009. http://www.qualityforum.org/ (accessed January 9, 2009).
Nylen, L. 2008. Study: Hospitals designated for quality care had lower readmission rates for cardiac patients. CQ HealthBeat. Congressional Quarterly, Inc. November 18, 2008.
Ozminkowski, R. J., R. L. Dunn, R. Z. Goetzel, R. I. Cantor, J. Murnane, and M. Harrison. 1999. A return on investment evaluation of the Citibank, N.A., health management program. Am J Health Promot 14(1):31-43.
Partnership to Fight Chronic Disease. 2008. Almanac of chronic disease—Statistics and commentary on chronic disease and prevention. The Partnership to Fight Chronic Disease. http://www.fightchronicdisease.org (accessed January 2009).
Rector, T. S., M. D. Finch, P. M. Danzon, M. V. Pauly, and B. S. Manda. 2003. Effect of tiered prescription copayments on the use of preferred brand medications. Med Care 41(3):398-406.
Robert Wood Johnson Foundation. 2007. Study underscores legality of physician ranking and public reporting systems when used to improve care quality. http://www.rwjf.org/qualityequality/product.jsp?id=22591 (accessed January 3, 2009).
———. 2008. Consumers, health care purchasers, physicians, and health insurers announce agreement on principles to guide physician performance reporting. http://www.rwjf.org/qualityequality/product.jsp?id=27932 (accessed January 6, 2009).
Robinson, J. C. 2003. Hospital tiers in health insurance: Balancing consumer choice with financial incentives. Health Aff (Millwood) Suppl Web Exclusives:W3-135-146.
Rosenthal, M. B., R. G. Frank, Z. Li, and A. M. Epstein. 2005. Early experience with pay-for-performance: From concept to practice. JAMA 294(14):1788-1793.
Rosenthal, M. B., M. B. Landrum, E. Meara, H. A. Huskamp, R. M. Conti, and N. L. Keating. 2007. Using performance data to identify preferred hospitals. Health Serv Res 42(6 Pt 1):2109-2119; discussion 2294-2323.
Rowe, J. W., T. Brown-Stevenson, R. L. Downey, and J. P. Newhouse. 2008. The effect of consumer-directed health plans on the use of preventive and chronic illness services. Health Aff (Millwood) 27(1):113-120.
Scanlon, D. P., and J. B. Christianson. 2008. Increasing awareness and changing attitudes towards patient safety: Evaluation of an employer initiative. http://ihea2009.abstractbook.org/presentation/102/ (accessed January 2009).
Scanlon, D. P., J. Beich, E. W. Ford, and J. Christianson. 2006 (unpublished). Assessing physician perceptions and opinions of variation in quality, patient safety and cost among health care facilities.
Scanlon, D. P., R. C. Lindrooth, and J. B. Christianson. 2008. Steering patients to safer hospitals? The effect of a tiered hospital network on hospital admissions. Health Serv Res 43(5 Pt 2):1849-1868.
Soler, R., D. Hopkins, S. Razi, K. Leeks, and M. Griffith. 2007. Proceedings of the task force meeting: Worksite reviews. Atlanta, GA: Centers for Disease Control and Prevention.
Soler, R. E., M. Griffith, D. P. Hopkins, and K. D. Leeks. 2009. The assessment of health risks with feedback; results of a systematic review. In American college of sports medicine worksite health handbook, edited by N. Pronk. Champaign, IL: Human Kinetics. Pp. 82-91.
Trude, S. 2003. So much to do, so little time: Physician capacity constraints, 1997-2001. Washington, DC: Center for Studying Health System Change.
Watson Wyatt/RAND. 2007. The CDHP implementation experience with large employers. http://www.watsonwyatt.com/us/research/featured/cdhp.asp (accessed January 2009).
Wellpoint, Inc. 2008. Physician tiering: For Unicare state indemnity plan basic, plus. In Unicare state indemnity plan. Andover, MA.
Wharam, J. F., B. E. Landon, A. A. Galbraith, K. P. Kleinman, S. B. Soumerai, and D. Ross-Degnan. 2007. Emergency department use and subsequent hospitalizations among members of a high-deductible health plan. JAMA 297(10):1093-1102.