The introduction of atypical, or "second-generation" (1), antipsychotic medications that began with clozapine in 1989 and continued with risperidone in 1994, olanzapine in 1996, quetiapine in 1997, ziprasidone in 2001, and aripiprazole in 2002 has expanded therapeutic options for persons with schizophrenia. The new medications—especially clozapine—have provided significant benefit to persons who have treatment-resistant forms of illness (2). However, these new medications have substantially increased pharmaceutical costs. For subgroups with severe illness, the increased drug costs may be offset by reduced hospital expenditures (3). And, apart from cost offsets, if atypical antipsychotics are superior at reducing suffering and improving function compared with the less costly first-generation medications, the increased spending may be seen as warranted.
This column, the third in a series describing strategies for managing psychiatric drug use that promote access, high-quality care, and more acceptable cost trends (4,5), focuses on the Department of Veterans Affairs (VA) approach to atypical antipsychotic medication use. The premise of the series is that devoting the right kind of attention to cost is an ethical requirement, not an ethical abomination (6,7).
Developing coverage policy for antipsychotic medication is not a task for the fainthearted. Policy decisions affect human well-being and even life itself. Costs are high. Psychiatrists have strong preferences and resent being told how to prescribe. Patients, influenced by their peers, by advertising, and by the Internet, also have strong preferences. Given the long history of stigma and discrimination, patient and family advocacy groups are—and should be—vigilant and vocal. A great deal of money is at stake. Pharmaceutical companies market their products vigorously through advertising, sponsored research, subsidized professional education, and lobbying. And, as is so often the case in medical decision making, the evidence base for policy making is less than ideal.
The VA system cares for approximately 200,000 veterans with psychosis. In fiscal year 2001, the VA filled more than 1.5 million 30-day prescriptions for antipsychotic medication, at a cost of $158 million. Eighty percent of these prescriptions were for atypical antipsychotics. The average cost of atypicals was 17 times that of first-generation antipsychotics. Among the various atypicals, there was an almost threefold variation in average daily cost per patient—$2.94 for quetiapine, $3.15 for risperidone, $5.01 for ziprasidone, $6.28 for olanzapine, and $8.07 for clozapine (8).
The VA serves a defined population within the budget provided to it by Congress. Funds that are not spent on medications can be used for other health services, such as the VA's compensated work therapy program (9). Managing the use of atypicals is part of seeking maximum benefit for veterans within the available budget. The issue is not cost versus care but, rather, seeking better health outcomes per dollar spent.
In February 2001, in accord with an approach to pharmacy benefit management that had been praised by the Institute of Medicine (10), the VA formed a task force of two VA psychiatrists and two VA pharmacists to develop a guideline for the use of atypical antipsychotics. After analysis of the available evidence and review of a draft by leaders within the VA system, a final guideline was approved, posted on the VA Web site in July 2001, and distributed to the regional offices.
The introduction to the guideline reads as follows: "Selection of therapy for individual patients is ultimately based on physicians' assessment of clinical circumstances and patient needs. At the same time, prudent policy requires appropriate husbanding of resources to VA to meet the needs of all our veteran patients. These guidelines are not intended to interfere with clinical judgment. Rather, they are intended to assist practitioners in providing cost effective, consistent, and high quality care. The following recommendations are dynamic and will be revised as new clinical data become available" (11).
The guideline goes on to state "there is no consensus in the literature to support one [atypical antipsychotic] being globally superior to another." Therefore, in the absence of patient-specific factors, the guideline recommends that therapy be started with a less expensive atypical agent, which "at the present time…would lead to the preference of quetiapine and risperidone over olanzapine." The guideline does not recommend that effective regimens be changed—it applies to first use of atypical antipsychotics or for patients whose current regimen requires change.
As background for this column, the first author spoke informally with dozens of psychiatrist colleagues about their "gut" beliefs about antipsychotic medications. Their answers were remarkably varied. Some expressed the belief that except for clozapine, which was widely referred to as "the gold standard," first-generation medications are as effective as atypicals. Some psychiatrists believed that second-generation medications should be used for all first starts. Among those who believed that atypicals are the standard of care, some had favorite medications, whereas others were skeptical about drug companies' claims of superiority of one agent over others. A majority of the psychiatrists appeared to agree with the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) study—sponsored by the National Institute of Mental Health—that "although a variety of claims of efficacy and safety have been made, they are often based on insufficient evidence" (12).
The VA had three broad choices with regard to making policy about the use of atypical antipsychotics. First, it could leave the choice of agents entirely to the prescriber, a policy that maximizes physician and patient autonomy but ducks responsibility for ensuring that VA expenditures produce maximum benefit for the entire population. Second, it could mandate a "fail first" policy through which nonpreferred medications could be used only if the preferred medication proved ineffective, an approach that reduces variation but risks substantial backlash from providers, patients, and the pharmaceutical industry. Finally, it could do what might be called "managed care light"—recommend a utilization strategy and depend on education as the primary tool of influence. Our reading of the public record indicates that the VA chose the third approach with a tentative tilt, at a few sites, toward the second (8,11,13).
Some patient advocates, psychiatrists, and drug companies objected to the VA policy and complained to the VA itself and to legislators, which led to a request from the House of Representative Committee on Veterans Affairs to the General Accounting Office (GAO) as to "whether this guideline may result in restricted access to the more costly atypical antipsychotic drugs and, in turn, adversely affect the quality of care for veterans" (8). The GAO report, issued in April 2002, provides guidance about clinically informed, ethically justifiable approaches to managing drug benefits that are relevant for every pharmacy program:
• As of the writing of the GAO report, and, by implication, until the NIMH CATIE trial (11) yields findings on comparative efficacy, it is sound clinical management policy to regard all atypical antipsychotics as clinically equivalent choices for first use.
• Given the fact of "equipoise" of informed opinion (14), recommending that physicians prescribing an atypical antipsychotic choose a less costly agent unless specific individual patient factors support use of a costlier medication, is sound policy.
• Excessive pressure to contain cost is not acceptable. Although the GAO report did not define "excessive," it implied that prior authorization methods that go beyond collegial monitoring, education, and discussion and impose bureaucratic barriers or perceived threat to the prescriber overemphasize the role of cost in decision making.
Arnold Relman, a respected observer of the U.S. health care system and former editor of the New England Journal of Medicine, recently commented that "doctors are taught about drugs by agents of the pharmaceutical industry, which works hard to persuade them to select the newest and most expensive medications—even in the absence of scientific evidence that they are any better than older, less costly ones" (15). The VA approach to the use of atypical antipsychotic medications provides a socially responsible counter to the power of advertising and other forms of industry influence on doctors, patients, and the public.
We have postulated that systems that make themselves accountable for the reasonableness of their limit-setting policies by actively involving stakeholders in deliberation focused on the needs of both individuals and the total population, educating constituents about the policy rationale, carefully monitoring results, and revising policy in the light of experience and new evidence, are most able to elicit cooperation in making and managing limits (7). Had the VA met the conditions of accountability for reasonableness even more fully than it did it might have achieved savings greater than the 6.5 percent that actually occurred (13).
Given that the VA could use savings on drug costs to enhance other services for veterans who have psychotic illnesses, we believe that the VA would be justified in promoting its guideline even more vigorously than it does. Nevertheless, its combination of evidence-based policy making, vigorous quality monitoring (16), and openness to policy revision on the basis of scrutiny from external bodies such as the GAO and the Institute of Medicine reflect exemplary management of care. The VA approach to managing the use of atypical antipsychotic medications should be applied in public and private settings—such as Medicaid, Medicare, and commercial insurance—that involve use of pooled funds for the benefit of a population.
The authors thank Ethan S. Rofman, M.D., and Robert A. Rosenheck, M.D.
Dr. Sabin, who is editor of this column, is clinical professor of psychiatry at Harvard Medical School and director of the ethics program at Harvard Pilgrim Health Care in Boston. Dr. Daniels is professor of ethics and population health in the department of population and international health at the Harvard School of Public Health in Boston. Send correspondence to Dr. Sabin at the Department of Ambulatory Care and Prevention, 133 Brookline Avenue, Sixth Floor, Boston, Massachusetts 02215 (e-mail, firstname.lastname@example.org).