In an ideal world, Somnath Saha said, benefit packages would be designed with only patient care, safety, and effectiveness in mind. In reality, though, states, health insurance plans, and the federal government must also consider cost-effectiveness and affordability. In this panel, Dr. Saha, a practicing physician and volunteer chair of Oregon’s Health Services Commission (HSC), and Jeanene Smith, the administrator of the Office for Oregon Health Policy and Research (OHPR), described Oregon’s experience in setting priorities when funds are limited. Additionally, they discuss developing value-based insurance design across the public and private sector in Oregon. Jean Fraser, the current chief of the San Mateo County Health System and the former CEO of the San Francisco Health Plan, built on these points by discussing her experience “making hard choices” when developing benefit plans for two California coverage programs.
Dr. Saha explained the function of the Oregon HSC and his role as chair. The commission develops the state’s health care priorities in the form of a rank-ordered list of health care services. They then “hand” this list to the state legislature, which uses it to develop Medicaid benefits packages. The HSC first began its work in 1989, when the governor and legislature recognized the problem of rising health care costs and determined that in order to maintain the state’s Medicaid expansion, it would need to “trim covered services, not trim people” (The Economist, 1998). The HSC is now comprised of 12 members, principally generalist physicians—including Dr. Saha, who is a primary care physician at the Portland Veterans Affairs Medical Center—and consumer advocates. Dr. Saha described its role as that of a gardener: the commission oversees a list of prioritized services and “tends the garden” by correcting errors, incorporating new services, and improving prioritization processes.
Developing a Prioritized List of Covered Services
The HSC developed and now maintains a prioritized list of covered services that are “rank-ordered” according to impact on health, treatment effectiveness in improving and promoting health, and public values and priorities. Since the implementation of the rank-ordered list in 1994, the HSC has regularly updated the list and biennially submits it to the legislature, which “draws a line” at which the state stops covering benefits based on the amount the
state can afford for the covered population. This latter decision of where to draw the line is primarily made using actuarial analysis of the cumulative costs of services. Currently, Dr. Saha said, approximately 75 percent of the over 600 lines of condition-treatment pairs are covered (Table 13-1) (Oregon Health Services Commission, 2011). The HSC’s prioritization methodology was “reinvented” in 2006. The HSC first ranks nine categories of care based on healthy life years and impact on suffering, among other criteria. Each category of care is given a weight, ranging from maternity care with a weight of 100 to inconsequential care with a weight of 1 (see Table 13-2). The category list does not mean that all maternity care is prioritized over preventive care; rather, these category rankings are one component in an overall scoring formula. Dr. Saha pointed out that these categories of care are based on Oregon’s priorities and not on organ systems or type of provider. As a result of this distinction, dental and mental health care are integrated in the category list. Second, the HSC goes “line by line” and scores each condition and treatment within these categories of care based on eight impact measures: impact on health life years, impact on suffering, population effects beyond the affected patient (e.g., contagious diseases), vulnerability of the population affected, prevention of downstream complications, treatment effectiveness, the need for medical services, and net cost. The comprehensive list of conditions and treatments is built using diagnostic codes and procedure codes (i.e., ICD-9, CPT, and HCPCS [Healthcare Common Procedure Coding System] codes). Thus, a procedure without a CPT code is not included on the list, and prescription drugs and DME apply to many different lines on the list. To address this latter shortcoming, these products are called “ancillary services” and are “blanketly covered for conditions that fall within the covered range of the list.” Additionally, the list is used to determine the coverage of treatments only after the necessary diagnostic services establish the condition.
|Line Number||Examples of Services||Coverage|
|101||Medical treatment of acute lymphocytic leukemia|
|201||Surgical treatment of brain hemorrhage|
|301||Treatment for rheumatic heart disease|
|401||Laser therapy to prevent retinal tear|
|501||Treatment for noninflammatory vaginal disorders|
|551||Treatment for back pain without neurologic impairment|
|651||Treatment for calcium deposits|
SOURCE: Oregon Health Services Commission. 2011.
|Ranking||Category of Care||Weight|
|2||Primary and secondary prevention||95|
|3||Chronic disease management||75|
|6||Fatal conditions-acute care||40|
|7||Nonfatal conditions-acute care||20|
|SOURCE: Oregon Health Services Commission, 2007.|
The HSC uses a mathematical formula to rank health services within these categories. The formula considers the category weight (shown in Table 13-2), the service’s total impact score (derived from the set of scaled impact measures), the effectiveness of the service, and the need for the service. An example of its application for Type II diabetes (with a weighting of 75 for chronic disease management) is illustrated in Box 13-1.
Effectiveness, Dr. Saha said, is an important multiplier because you may have a “very high-impact illness,” but if a treatment for that illness is ineffective, this multiplier will ensure that the condition-treatment pairing gets a score of 0. Additionally, the “need for service” score ensures the exclusion of services that do not require medical care. Conservative management of a sprained ankle, for example, might be effective but in most cases does not require medical care. The requirement for a service multiplier reduces the priority score for this type of treatment. The HSC only uses net cost, Dr. Saha said, as a “tiebreaker.” Although actuaries provide information about what it costs to deliver the service, the HSC has little reliable information about the cost of not delivering the service. Because the commission often found itself “guessing,” it excluded net cost from the main formula.
Consumer and Provider Pushback
Dr. Saha acknowledged that consumers and providers have, on occasion, expressed discontent with the priority rankings of specific conditions and treatments. He said, though, that HSC’s “straightforward” formula, which was revised from a more complex formula developed in 1991, contributes to the commission’s transparency. Because the HSC’s process does not include a Delphi panel to review each impact score, he recognized that individuals can and do “quibble with every single line” in terms of whether the commission “got the scoring right.” The HSC believes that what it “sacrifices in rigor, it counters with transparency.” The scores are available on its website, and an individual or group disagreeing with a score or ranking can present evidence to the HSC as to where and how the scoring for a condition or treatment should change. While the HSC does not change rankings without legislative approval, it does have an open forum to respond to providers, patients, and others who have concerns. The HSC meets several times a year, and the legislature biennially reviews HSC’s proposed changes.
|Impact on Healthy Life Years:||7|
|Impact on Suffering:||2|
|Effects on Population:||0|
|Vulnerability of Population Affected:||2|
|Need for Service:||1|
|Category 3 Weight||75|
Total Score: 3300 -» Line: 33
SOURCE: Smith and Saha, 2011.
|Service Tier||Examples of Service||Cost Sharing|
|Value-Based||Routine vaccinations, prenatal care, chronic illness management, smoking cessation treatment||0-5%|
|Tier 1 (Lines 1-112)||Highly effective care for severe chronic disease and life-threatening illness and injury (e.g., rheumatoid arthritis, heart attack, and low birth weight)||Low|
|Tier 2 (Lines 113-311)||Effective care of other chronic disease and life-threatening illness and injury (e.g., glaucoma, breast cancer, ADHD)||Moderate|
|Tier 3 (Lines 312-502)||Effective care for non-life-threatening injury and illness (e.g., ear/sinus infections, herniated disk, reflux, dentures)||High|
|Tier 4 (Lines 503-679)||Less effective care and care for self-limited illness and minor injury and illness (e.g., chronic back pain, viral sore throat, seasonal allergies, and acne).||Highest|
|SOURCE: Smith and Saha, 2011.|
Using the Prioritized List to Develop EHB
In 2007, during a push for a universal coverage plan in Oregon, the state legislature directed the development of “recommendations for defining a set of essential health services that would be available to all Oregonians under a comprehensive reform plan.”1 The legislature, Dr. Saha said, recommended using Oregon’s prioritized list of health services as the basis for developing the state’s essential benefits package. To undertake this task, the state formed the Oregon Health Fund Board Benefits Committee that was comprised of health professionals (including Dr. Saha), consumers, and public and private insurers. These diverse parties faced difficulty determining which benefits should be considered “essential,” and found it easier to agree on what Dr. Saha called “levels of essentiality.” These levels resulted in a tiered benefits package in which higher-priority services have lower cost sharing, one approach to value-based insurance design (VBID). The committee used the prioritized list to create the service tiers. Dr. Saha described the top “value-based tier” as including tests and treatments that are highly effective, low cost, and that the committee wanted to encourage in the population (Table 13-3).
Since the 2007 development of the value-based benefits package described by Dr. Saha, the state of Oregon has explored how it might implement this package in a state-based health insurance exchange. Dr. Smith described her office’s role in developing this option. The OHPR is the policy office within the Oregon Health Authority (OHA), which includes the state’s Medicaid and state employee programs, the Oregon Educators’ Benefit Board, and the state’s high-risk pool (a premium subsidy program). Thus, the OHA has influence over 850,000 covered lives (approximately 30 percent of the state’s population), and the cost of covering these individuals plays a large role in state budget discussions. To address rising costs, the OHA aims to use value-based benefit design for these state-covered programs, said Dr. Smith, as well as promote its use in the private sector.
A Hypothetical Example of Oregon’s Proposed Insurance Exchange
To illustrate how the program could work, Dr. Smith cited the fictional example of Robert, a single male earning $20,000 per year (Box 13-2). His income is just above Medicaid eligibility, but when an insurance exchange exists in the state, he could purchase insurance through the exchange and get tax credits to assist with his premium. If he chose a Patient Protection and Affordable Care Act (ACA) “silver-level plan” that was based on the valuebased
1 Healthy Oregon Act (2007).
Robert is single, earns $20,000 per year.
- He purchases insurance through an insurance exchange.
- He will get tax credits to assist with his premium.
- He chooses value-based insurance design (VBID) with 10%/30%/50%/70% tiered co-insurance.
- His deductible is $300; out of pocket max is $1,600—amounts limited due to income level.
- Plan uses evidence-based formulary for medications:
- $10 for generic,
- $30 for preferred, and
- 50% for nonpreferred.
SOURCE: Smith and Saha, 2011.
insurance design (VBID) model, he would have certain benefits with no cost sharing and variable cost sharing for others. Based on his income level, his deductible would be $300 and his out-of-pocket (OOP) maximum would be $1,600. If Robert had Type 2 diabetes, under his silver-level plan, his insulin, eye exams, and supplies would be covered with little or no cost sharing. If his doctor found a diabetic foot ulcer and referred Robert to a surgeon for an antibiotic and surgical treatment, these services would be covered under tier 1. Robert would have a $10 co-pay for the antibiotic and a 10 percent co-insurance for the surgical procedure. Robert’s total OOP cost for his diabetic ulcer would be $470, half of what he would pay had he been insured by a typical employer plan in the state. Under a typical commercial plan, his OOP costs would be $810 plus the cost of exams, insulin, and supplies. Thus, the VBID in the insurance exchange would “drive incentives for the patient to get the care they need,” while creating barriers that help patients and payers “avoid marginally effective care,” she said. For now, though, the exchange is not yet operational, but the VBID could be used by the OHA’s current lines of coverage or by other purchasers of benefits.
Actuarial Estimates for Bending the Cost Curve
Actuaries used information from approximately 100,000 covered lives under Oregon Medicaid and the Oregon Educators’ Board plan to price the silver-level plan described by Dr. Smith. The actuaries used “judgment, rules of thumb, and many assumptions,” she said, to “tease out” the first estimates of the cost implications of tiered, evidence-based benefit design; initial analysis suggests that a 3 to 5 percent premium reduction would be possible compared to a traditional commercial plan. Dr. Saha explained that this reduction is conservative as it did not incorporate estimates of utilization changes based on extra cost sharing for low-priority services. Dr. Santa, a committee member, asked whether the OHPR had explored how many additional people could be covered if costs were reduced by 3 to 5 percent. Dr. Smith noted that because the state is experiencing a 10 to 15 percent cost growth, the magnitude of this reduction is just one way to “bend the trend” and help minimize the need to “cut populations and benefits.”
The OHPR is currently developing a more “robust” unit-cost model to allow for modeling reimbursement by tier. This model will allow Dr. Smith and her colleagues to better determine how they could further reduce costs by coupling the tiered benefits with physician payments. For example, could the plans pay providers more for tier 1 services than for marginally effective services in the bottom tier? If the answer were yes, the exchange would be a “two-way street”: cost-sharing arrangements would incentivize patients to use value-based services while payment systems would incentivize providers to deliver evidence-based care. When committee member
|Service Tier||Change in Utilization Due to Cost Sharing|
|Tier 3||Modest decrease|
|Tier 4||Moderate decrease|
|Prescription Drugs||Moderate decrease|
|SOURCE: Smith and Saha, 2011.|
Mr. Schaeffer asked for further details about how the payment portion of this design would change physician behavior, Dr. Smith acknowledged that her department has not yet determined how “exactly that would happen.” She said, though, that the plan could pay cardiologists less for those procedures in tiers 3 and 4 and pay them more for procedures in tier 1. Similarly, she said, obstetricians could be paid more for a vaginal delivery than for a full-term elective caesarean-section.
In addition to the cost implications of tiered benefit design, the OHPR has explored the expected utilization offset by changes in cost sharing. As shown in Table 13-4, use of value-based services (those that are highly effective, low cost, and have strong evidence supporting their use) are estimated to increase by 10 to 20 percent. Dr. Smith noted that the analysis assumed ambulance care would be similarly utilized under regular plans and value-based plans, but explained that because the value-based plans would incentivize primary care and care coordination, ambulance and emergency department use would likely decrease. The analysis, though, did not account for these secondary effects.
Focus Group Findings
In late 2010 and early 2011, the OHPR conducted focus groups with insurers, providers, hospitals, employers, and consumer advocates to gauge their perspectives on a value-based benefit package for non-Medicaid participants. Dr. Smith summarized relevant findings:
- Value-based benefits with low or no cost sharing are appealing.
- Tiered benefit plans are complicated and would require a lot of provider and consumer education.
- The value-based plan should have a greater emphasis on wellness.
- One-size-fits-all plans will not satisfy consumers.
While acknowledging that there are challenges associated with evidence-based benefit design, Dr. Smith closed by expressing that a value-based benefit design could be a vision for balancing access with cost and quality. Using such a plan design in the ACA insurance exchanges would be one way to realize that vision.
Before assuming her current role as chief of San Mateo County Health System, Ms. Fraser served as the CEO of the San Francisco Health Plan, where, in conjunction with the San Francisco Department of Public Health, she designed the Healthy San Francisco universal coverage program. Ms. Fraser’s experiences with benefit design provide her with ample “on-the-ground” perspectives about making, what she called, “tough decisions.” In both her current and former roles she faces the same dilemma: how do we provide as much benefit as possible with
limited funds? She began her presentation by suggesting that the committee faces a similar dilemma as it designs the essential health benefits (EHB), and posed a question for the committee’s consideration: “Fundamentally, what you’re going to have to decide is what is the goal of the ACA? Is it to cover the most people for most conditions, or is to cover all care for some people?" The committee’s answer to this question, she said, would affect San Mateo County Health System and other public indigent care providers across the country because the people who will be “priced out of coverage will end up” getting care from public hospitals and clinics.
Creating Healthy San Francisco
To address the tradeoffs between cost and coverage, the developers of Healthy San Francisco made some “really, really, difficult decisions.” First, Ms. Fraser said, they developed a list of exclusions ranging from acupuncture to long-term care and organ transplants. “Yes, these are some very serious things,” she said, but these exclusions allow Healthy San Francisco to “cover most people for most things.” Furthermore, the plan has a very narrow provider network. When initially implemented, beneficiaries could only get hospital care at San Francisco General Hospital. Even “perfectly legitimate ED [emergency department] visits” were not covered if treatment was received at a different hospital. While the network has expanded since its inception, the provider group is still “extraordinarily narrow,” said Ms. Fraser.
While the exclusions and limited provider network impose significant limits, they also allow the program to cover a “limited set of core services,” including prevention and treatment for “most medical conditions for tens of thousands of people who did not have coverage before.” Thus, the “choice” the developers made in benefit design “was not between the perfect and the good. It was nothing or something,” she said.
Consumer Response to Coverage Limits
Ms. Fraser believes the committee’s task is “one of the most important decisions regarding whether the ACA is going to be successful or not,” and that an important component of success will be getting stakeholder buy-in on the design of the EHB. She acknowledged that at the outset of implementing Healthy San Francisco, she was concerned that the limits she previously described would make people believe the program did not provide value and would undermine confidence in the plan’s mission. Once the plan was unveiled, though, she was surprised that the plan’s limits were “accepted almost without objection.” While some individuals were unhappy and wanted more benefits, fundamentally, she said, people were happy that coverage was being extended to a previously uninsured population.
She noted that one of the keys to public acceptance of these limitations was that the program was transparent about them by providing a straightforward explanation of the coverage. “There really isn’t any fine print.” She advised the committee to consider the importance of simplicity over detail because such simplicity and transparency will foster public acceptance.
Making Tough Decisions
In Ms. Fraser’s current role, she continues to face complex coverage decisions. She cited a recent coverage determination to highlight the tough decisions faced by public providers. San Mateo County will provide joint replacements when this treatment is the only way to keep a patient out of a wheelchair. The plan does not, however, cover skilled nursing care. Confronted with the issue of what to do with their first patient who was scheduled for hip replacement, Ms. Fraser upheld the decision not to pay for the skilled nursing care. However, she said, the county offered to provide the nursing care at a discounted rate and with a payment plan for the individual, or to teach the family to care for the patient at home. The family elected the latter. While this decision was “tough,” it was financially necessary for the county, she said, and ultimately, the medical outcome “was fine,” although that might not always be the case.
Ms. Fraser closed by sharing some lessons. First, she said, “under-promising so that we can exceed expectations” is crucial. Adding a benefit is much easier than removing one, so if the committee begins with an expansive list of EHB, it will be “virtually impossible to cut it back” should the benefits be found to be unaffordable.
Second, Ms. Fraser expressed strong support for the use of federally funded comparative effectiveness research in benefit decisions. Currently, she said, each provider and each plan is left to figure out these decisions on their own, which results in inconsistent decisions.
To conclude the panel discussion, committee member Mr. Koller asked the panelists whether the committee’s time “would best be spent telling the Feds what to do based on your experiences, doing it for the Feds, or telling the Feds what to tell the states to do?" Dr. Saha replied that the latter option would be the most fruitful; whereas Ms. Fraser noted that regardless of whether states do it or the federal government tells the states what to, the committee should develop mechanisms to ensure some level of consistency. She suggested that if the federal government delegates this work to states, states ought to be given the option of following federal rules and/or joining regional consortia to take advantage of economies of scale and data. Also, if states make these decisions, she said, the states should be required to report them to a publicly available central database. The federal government should track results and publish guidance when the evidence becomes clear that certain treatments are more appropriate than others.
The Economist. 1998. John Kitzhaber’s prescription. http://www.economist.com/node/160924 (accessed May 10, 2011).
Oregon Health Services Commission. 2007. Prioritization of health services: A report to the Governor and the 74th Oregon Legislative Assembly. Salem, OR: Oregon Health Services Commission.
____. 2011. Prioritized list of health services: April 1, 2011. http://www.oregon.gov/OHA/OHPR/HSC/docs/L/Apr11List.pdf (accessed May 10, 2011).
Smith, J., and S. Saha. 2011. Oregon’s value based benefits package. PowerPoint Presentation to the IOM Committee on the Determination of Essential Health Benefits by Jeanene Smith, Administrator, Office for Oregon Health Policy and Research and Somnath Saha, Staff Physician, Portland VA Medical Center and Chair, Oregon Health Services Commission, Costa Mesa, CA, March 2.