The U.S. Department of Health and Human Services (HHS) asked the committee to assess how insurers determine medical necessity. In this session, Dr. Alan Garber, staff physician at the VA Palo Alto Health Care System and Director of the Center for Health Policy at Stanford University, delineated the differences between the application of medical necessity and the development of a benefit package. He reviewed some precedents related to medical necessity, specifically federal court rulings and a definition developed by a consensus committee convened by Stanford University with its associated criteria including consideration of cost. Additionally, Dr. Barbara Warren, the Director of the Hunter College Institute for LGBT Social Science and Public Policy spoke on behalf of Consumers United for Evidence-Based Healthcare (CUE).1 She addressed consumer support for the use of evidence in determining essential health benefits (EHB), explored criteria and definitions for medical necessity from the consumer’s standpoint, and expressed ways to incorporate consumers in the process for updating the EHB.
Dr. Garber began by clarifying the difference between a coverage decision (i.e., what an insurer covers as a general benefit category) and a medical necessity determination (i.e., what intervention is deemed appropriate for a particular person). A coverage decision is a policy decision based on the general needs of the broad population group insured under a benefit plan; in general, this coverage policy refers to the broad categories of services (e.g., hospitalization) for which the insurer will pay, as well as whether the insurer will pay for a specific intervention (such as a new surgical procedure). A medical necessity determination, on the other hand, determines whether the insurer will pay for an intervention in a very specific instance—for example, when an individual patient requires hospitalization and a specific surgical intervention. To complicate these distinctions, not every treatment that is potentially medically necessary is covered by a plan. Plans usually specify categorical exclusions that are never covered. For instance, some commercial plans categorically exclude maternity care. Conversely, some plans,
1 The 29 member organizations represent a full range of health and behavioral health issues and concerns from cancer to geriatrics; minority health; addiction recovery; environmental health; lesbian, gay, bisexual, and transgender health; mental health; women’s health; disabilities; and other issues. More information can be found on their website: http://us.cochrane.org/consumers-united-evidence-based-healthcare-cue (accessed May 10, 2011).
including Medicaid, may cover certain services that are not considered medical services per se, and therefore, would not be medically necessary.
When committee member Dr. Guzick asked Dr. Garber to explain the role of medical necessity in defining the EHB, Dr. Garber replied that “once one gets away from decision making about an individual case, one is no longer speaking about medical necessity; instead, we are speaking about coverage policy.” Over the past 15 or 20 years, he said, almost “everything proven effective” has been included in the defined set of benefits. This approach, he argued, ignores the cost implications of covering a particular service, resulting in a “very costly bundle of services.” Dr. Garber advised that if value is to be considered in determining what is essential, there should be a discussion about what criteria are used to define value and how those criteria should be applied. Dr. Garber suggested that cost-effectiveness analysis or other tools implemented around the world could be applied.
Court Direction on the Meaning of Medical Necessity
The Second Circuit Court of Appeals has decided numerous cases related to medical necessity and in Mario v. P & C Food Markets, Inc.,2 specifically ruled that in the absence of a medical necessity definition in the plan document, the term medical necessity refers to what is medically necessary for a particular patient and consequently is not a blanket determination of whether coverage is appropriate:
Unless the contrary is specified, the term “medical necessity” must refer to what is medically necessary for a particular patient, and hence entails an individual assessment rather than a general determination of what works in the ordinary case.2
Furthermore, a class action lawsuit filed in 2000 and consolidated in the U.S. District Court for the Southern District of Florida resulted in a generally agreed upon definition of medical necessity.3 Under the terms of the settlement agreements, the defendants (including Aetna, CIGNA, Anthem/WellPoint, Humana, and other insurers, together with state and county medical societies and 900,000 physicians, agreed to accept the following definition:
“Medically Necessary” or “Medical Necessity” shall mean health care services that a physician, exercising prudent clinical judgment, would provide to a patient for the purpose of preventing, evaluating, diagnosing or treating an illness, injury, disease or its symptoms, and that are: (a) in accordance with generally accepted standards of medical practice; (b) clinically appropriate, in terms of type, frequency, extent, site and duration, and considered effective for the patient’s illness, injury or disease; and (c) not primarily for the convenience of the patient, physician or other health care provider, and not more costly than an alternative service or sequence of services at least as likely to produce equivalent therapeutic or diagnostic results as to the diagnosis or treatment of that patient’s illness, injury or disease. For these purposes, “generally accepted standards of medical practice” means standards that are based on credible scientific evidence published in peer-reviewed medical literature generally recognized by the relevant medical community or otherwise consistent with the standards set forth in policy issues involving clinical judgment. (Kaminiski, 2007)4
The definition that emerged from this class action settlement was only required to apply to those insurers participating in the settlement.
In determining medical necessity, one test is whether the intervention meets generally accepted standards of medical practice. Dr. Garber stated that a “sea change has occurred over the decades in what is meant by generally accepted standards of medical practice.” In the past, he said, the phrase referred to “any kind of care that a physician deemed necessary and appropriate for a patient.” The above definitions, however, state that the generally accepted standards of medical practice have evolved to mean standards that are based on credible scientific evidence. This evolution in terminology corresponds to what Dr. Garber maintained has been an increased reliance in the provider community on the published clinical and scientific literature.
2Mario v. P & C Food Markets, Inc., 313 F. 3d 758 (2nd Cir. 2002).
3 In re Managed Care Litigation. S.D. Fla. MDL No. 1334. Settlement approved December 31, 2005.
4 Emphasis by using italic font was added to excerpts from Kaminski, 2007, by Dr. Garber.
Furthermore, Dr. Garber highlighted that the court’s agreed upon medical necessity definition explicitly recognizes that relative costs can and should play a role in medical necessity determinations by stating “not more costly than an alternative service or sequence of services.” This, he said, is often a controversial aspect of medical decision making.
Stanford Definition of Medical Necessity
Before the above-discussed court decisions, a Stanford University project brought together various stakeholders to generate a model definition of medical necessity intended for use both by private health plans and state Medicaid programs (Singer et al., 1999) (see Appendix C).
After “a great deal of discussion and debate,” the definition that emerged from a 1999 workshop achieved what Dr. Garber described as “broad consensus among participants” although not necessarily agreement about each particular provision. The preamble to what is now commonly called “the Stanford definition” broadly states that for an intervention to meet the test of medical necessity and be covered by an insurer, the intervention first had to be included in a category of service not specifically excluded in the benefit contract.
The Stanford definition includes five criteria: who has decision-making authority, the purpose of the intervention, the scope of the intervention, the standards of evidence, and the value (i.e., cost-effectiveness) of the intervention.
Authority means the intervention needs to be recommended by the treating physician and determined appropriate by the plan medical director or his designee. This criterion relates to both Medicaid and private plans.
Purpose means the service in question is a health intervention that is intended to treat a medical condition. The boundaries of what is and is not considered medical care engenders a great deal of debate, such as current controversies about coverage for some treatments for common conditions like autism.5 Therefore, the purpose criterion, Dr. Garber said, is meant to “set the boundaries more or less for what is the health intervention and what is a medical condition.” In this definition, treatment encompasses diagnostics and other aspects of care, and can be thought of as the various interventions used to manage a condition from the physician’s point of view.
Scope refers to the appropriate supply or level of service. It considers potential benefits and harms to the patient, such as where the treatment or the intervention might be delivered (inpatient or outpatient, or only in specialized centers of excellence with demonstrated competence in specific types of cases).
Evidence means the intervention is effective, can reasonably be expected to produce the intended results, and that the expected benefits outweigh the potential harms. Dr. Garber noted that “evidence is infrequently cut and dried” and therefore, the following questions need definition: what is meant by evidence, what standard of evidence is appropriate, and what evidence is inadequate? Further, he said, the Stanford definition proposes a distinction between new interventions and existing interventions in assessing evidence for effectiveness. For new interventions, effectiveness is based on scientific evidence consisting of various kinds of data of varying degrees of rigor, including randomized controlled trials and properly designed observational studies. For existing interventions, effectiveness is determined first by scientific evidence, then by professional standards and expert opinion. This distinction reflects the practical difficulty of applying high standards of evidence to treatments that clinicians have long accepted as effective. If scientific evidence is unavailable, then professional standards of care are acceptable. And if professional standards of care do not exist or are contradictory, then expert opinion may be relied on to make decisions.
Value explicitly refers to the cost-effectiveness of the intervention compared to alternative interventions, including no intervention. An intervention is deemed cost-effective if the benefits relative to the costs represent an efficient use of resources for patients. Dr. Garber stated that cost-effectiveness has been “an area of contentiousness in both private and public insurance, but more openness to considering value has been developing over time.”
5 Parents seeking care for their children with autism wish to explore all potentially helpful treatments. To ensure affordability and prevent adverse selection for dependent coverage, insurers have traditionally distinguished between non-medical benefits (e.g., education) and medical benefits.
Cost Considerations in Insurers’ Decision Making
Reiterating a point made by Dr. Jonathan Gruber (see Chapter 4), Dr. Garber noted that “the more expansive your definition of medical necessity, the greater the cost of a benefit package.” That is not a surprise, Dr. Garber said, but “what may be a surprise is how small changes in expansiveness can give rise to very large changes in cost.”
When committee member Dr. Ho asked about the differences between the settlement language and Stanford definition of medical necessity with respect to cost considerations, Dr. Garber noted that the settlement came after the completion of the Stanford project, and that while he did not have a “head-to-head comparison” to utilize, there are some differences in the way that cost is considered. The settlement language confines medically necessary treatments to those that are “not more costly than an alternative service or sequence of services at least as likely to produce equivalent therapeutic or diagnostic results as to the diagnosis or treatment of that patient’s illness, injury or disease.” The Stanford definition, by comparison, applies a cost-effectiveness criterion, which assigns a much greater role to cost considerations.
The aforementioned court settlement definition of medical necessity was agreed upon by providers and large health insurers, all of whom agreed on cost consideration as a valid parameter. An earlier anonymous survey of insurers conducted in 2000 and 2001 by the Stanford Center for Health Policy (and before the settlement agreement), looked at the use of cost in plan decision making. At the time of the survey, only 40 percent of plans conducted formal cost-effectiveness analysis, and about half selectively applied cost considerations in pre- authorizations for some interventions. Fifty-four percent of surveyed plans had explicit coverage policies that included some notion of cost, and 58 percent required the use of less costly interventions before more costly ones (e.g., a generic pharmaceutical product before a branded drug) (Garber, 2004).
The same survey examined how plans evaluated costs when considering whether to cover a new health intervention. If the new intervention would be equally effective for the same costs as existing interventions, almost all plans (94 percent) said that they would cover the intervention. If the intervention was equally effective but cost more, the majority (84 percent) would not cover it. If it was less effective for the same cost, only three percent of plans said they would cover the intervention, and eight percent would cover less effective interventions for less costs (Garber, 2004). In sum, Dr. Garber said, a plan tended not to cover new interventions that were less effective, even if the intervention was substantially less costly. If the new intervention was more effective than interventions already available, plans were likely to cover it regardless of cost. The survey suggests that for plans, “effectiveness trumps costs.”
Conversely, Medicare does not consider whether a new intervention would result in equivalent or lower total costs for the program. In the 1990s, Medicare attempted to make cost a component of its coverage decision making process. The Health Care Financing Administration (HCFA) (Centers for Medicare & Medicaid Services’ [CMS’] predecessor agency) indicated in a draft rule that it would cover a service depending on four categories, including cost. The cost category assessed “whether the item or service resulted in equivalent or lower total costs for the Medicare population than the currently covered alternative.” After extensive public outcry during the public comment period, the proposed rule was withdrawn (HCFA, 2000). Subsequently, CMS has been reluctant to explicitly consider cost when making coverage decisions.
When committee member Dr. James Sabin asked for possible meanings of “fair process” in the context of the EHB program, Dr. Garber replied that fair processes could apply to both benefit coverage and medical necessity decisions. There are precedents, Dr. Garber said, for incorporating public deliberations into benefit coverage decision making. He cited the Medicare Evidence Development and Coverage Advisory Committee (MEDCAC) as “a good example of a very public process with a great deal of opportunity for public input.”
While individual medical necessity cases cannot be subject to a public process as they involve individual patients and confidential information, there can be, Dr. Garber pointed out, a public process for “vetting the rules that are used to make medical necessity decisions” and establishing an appropriate appeal process. This exchange prompted committee member Mr. Koller to ask what standards for transparency could increase consumer confidence
in the medical necessity decision-making process. Dr. Garber stated that consumers could best participate in the benefit design process if they were better trained to effectively represent the public. Other countries, he said, have embedded consumers in the benefit design processes (Sabik and Lie, 2008). While these consumers are not experts on health care or these kinds of decision-making processes, they receive training to prepare them to examine the intervention under consideration and understand complex statistical arguments. After a week of training, Dr. Garber said, the consumers can participate “at a much higher level and give more meaningful input.”
Medical Necessity Determination in the Future
Dr. Garber stated that numerous reforms introduced in the Patient Protection and Affordable Care Act (ACA) raise the question of whether medical necessity determinations will continue to be an important issue as the law’s provisions go into effect. For example, he said, if payment changes put more financial risk on the shoulders of providers, then providers “will have more of a stake in ensuring that only effective care and necessary care is delivered, so medical necessity decision making may turn out to play a lesser role.” The nation is, however, “years off from the time when medical necessity decisions will be unnecessary or much less prominent in determining which care is delivered.”
Dr. Garber concluded by noting that one of the greatest challenges for coverage policy and medical necessity decision making is how to account for individual variation in ability to benefit from an intervention. “If there is an Achilles heel in our reliance on evidence,” he said, “it is that we tend to have much better evidence about averages for populations than we have about how subgroups might derive greater or lesser benefits from any intervention. The process has to be flexible enough to accommodate that.”
Dr. Warren, who spoke on behalf of the CUE, began by describing the organization. CUE is a national coalition of health and consumer advocacy organizations established in 2003 in response to an invitation from the U.S. Cochrane Center to create a partnership between consumer advocates and scientists. The aim of the partnership is to improve consumers’ ability to engage in and demand high quality health care by providing tools that educate constituents about evidence-based health care and the use of evidence in health care decision making. The aim, Dr. Warren stated, is to get consumers “engaged in a much more proactive way.”
CUE members, Dr. Warren said, advocate for the development of a “consistent, universal definition of medical necessity that emphasizes quality and clinical effectiveness above cost and resource utilization.” She stressed the importance of transparency in the medical necessity decision-making process, particularly as it relates to who gets to make the final determination on what interventions are medically necessary, and what the appeals process will be in the event that the patient or provider who is appealing does not concur with the decision.
Dr. Warren noted that although the Stanford definition may be a viable and useful model, it is not a definition with which most consumers and consumer advocates are familiar. Therefore, any definition needs to be widely disseminated with a process for review, discussion, and revision.
Consumer Support for Evidence-Based Medicine (EBM)
Consumers, Dr. Warren acknowledged, are often represented as opposed to limits or even to discussing what care might be essential and necessary. But without an understanding of the purpose of such terminology and without being invited to engage in the discussion of what it means, the determination of EHB is “scary because it sounds like something might be limited or denied to anyone in need.” Therefore, Dr. Warren argued, inclusion and education are critical issues for consumers. By including consumer advocacy organizations and coalitions in EHB decisions, consumers can be engaged in the process.
CUE, for instance, trains consumers as effective partners in the development and implementation of systematic reviews, clinical trials, other research studies, and clinical guidelines panels. The coalition also translates evidence
into widely disseminated “plain language summaries” that are more easily understood and used by laypersons. Consumer constituents actively seek out this information and use it to more readily engage in discussions with providers about the benefits and harms of treatment.
Research, including Cochrane systematic reviews, shows that enhanced patient-provider communication and shared decision making may increase patient adherence to treatment, improve outcomes, and reduce the need for more invasive and costly treatments (Naik et al., 2008; O’Connor et al., 2003). Educated consumers, Dr. Warren argued, are more cost-effective. The first and central principle espoused by CUE is that policy makers, providers, and insurers need to commit not only to integrating the best evidence available into benefit decisions, but also to including consumers in these benefit decision processes. The second principle recommended by Dr. Warren on behalf of the CUE is that benefit decisions recognize diversity. The EHB package, she argued, should offer a “reason able range of choices” that recognize that each consumer is an individual. “One size,” she said, “cannot fit all.”
Expanding the Definition of Best Evidence
In defining what constitutes the “best available evidence,” CUE urges that reviews of the evidence take into account clinical expertise, patient values and needs, standards of care, and clinical practice guidelines developed through a combination of research, clinical expertise, and consumer input. For some populations, Dr. Warren said, this broader definition of evidence makes a critical difference in being able to access needed medical care. For instance, although there are well-established clinical guidelines on the necessity for and appropriate administration of hormone replacement therapy for transgender patients, the lack of clinical trials and systematic reviews supporting such treatment remains a barrier for many transgender patients in accessing coverage for treatment. Additionally, while clinical trials may establish research-based evidence, for example in pharmacological treatment, they do not always adequately involve diverse racial and ethnic populations, women, children, or adolescents. Furthermore, Dr. Warren argued, “efficacy in a controlled clinical setting may not prove to be effective in the real world where patients have intervening health and environmental factors that may impact their treatment outcomes.”
Consumer Engagement in Updating Covered Benefits
“Nothing About Us Without Us!” is often cited by members of the disability movement. Dr. Warren explained that this slogan means that policies should not be defined without the full and direct participation of groups affected by that policy. These groups, she said, must involve national, ethnic, ability-based, or other groups that are often marginalized from political, social, and economic opportunities. CUE promotes full consumer inclusion in the development and implementation of policies and guidelines that determine access to care. Investing in building consumer capacity for effective participation, she argued, pays off in that consumers can be invaluable members of an interdisciplinary team by providing insights and perspectives that are often not apparent to clinicians, policy makers, and industry representatives.
CUE suggests that the process for updating the EHB includes the necessary infrastructure to engage consumers in an equitable way. A national subscription to the Cochrane Library, for instance, would enable all consumers to access the wealth of systematic reviews on the effects of interventions for prevention, treatment, and rehabilitation. Cochrane resources are currently limited to government entities, hospitals, and universities that can afford paid subscriptions. Furthermore, Dr. Warren noted, providers, policy makers, and industry representatives participating in these processes are often financially supported or compensated for their participation. Travel support and stipends for consumer participation would enable more consumers and consumer advocates to engage in an inclusive and equitable way.
Dr. Warren concluded by reiterating that full disclosure, complete transparency, and consumer education will allow the committee and HHS to engage consumers in understanding medical necessity and what is meant by essential health benefits. She also noted that consumers need sufficient information to understand the benefit choices available. A “middle ground approach” has proven helpful in Medicare supplemental policies, and is now being implemented in Medicare Part D with the aim of reducing the confusion created by having too many drug
benefit designs. Reasonable and informed choice affords consumers and providers the ability to select and then implement the package that is the “best fit for that consumer’s needs and condition.”
Garber, A. M. 2004. Cost-effectiveness and evidence evaluation as criteria for coverage policy. Health Affairs W(4):284-296.
HCFA (Health Care Financing Administration). 2000. Medicare program; criteria for making coverage decisions. Federal Register 65(95):31124-31129.
Kaminiski, J. L. 2007. Defining medical necessity. http://www.cga.ct.gov/2007/rpt/2007-r-0055.htm (accessed April 20, 2011).
Naik, A. D., M. A. Kallen, A. Walder, and R. L. Street. 2008. Improving hypertension control in diabetes mellitus. Circulation 117(11):1361-1368.
O’Connor, A. M., D. Stacey, V. Entwistle, H. Llewellyn-Thomas, D. Rovner, M. Holmes-Rovner, V. Tait, J. Tetroe, V. Fiset, M. Barry, and J. Jones. 2003. Decision aids for people facing health treatment or screening decisions. Cochrane Database of Systematic Reviews (2):CD001431.
Sabik, L., and R. Lie. 2008. Priority setting in health care: Lessons from the experiences of eight countries. International Journal for Equity in Health Research 7(4):1-13.
Singer, S., L. Bergthold, C. Vorhaus, S. Olson, I. Mutchnick, Y. Y. Goh, S. Zimmerman, and A. Enthoven. 1999. Decreasing variation in medical necessity decision making. Appendix B. Model language developed at the “Decreasing Variation in Medical Necessity Decision Making” Decision Maker Workshop in Sacramento, CA, March 11-13, 1999.