Impact of the STAR*D Trial From the Perspective of the Payer

Depression is a common condition in the United States, with 12-month and lifetime prevalence rates of approximately 5% and 13%, respectively ( 1 ). However, only 60% of persons with depression report treatment for the disorder ( 1 ), and primary care physicians detect major depression among only one-third to one-half of their patients who have the condition ( 2 ). Depression results in significant impairment in social and occupational functioning and often complicates treatment of comorbid mental and physical conditions. It also carries a 4%–5% lifetime risk of suicide ( 3 ).

Even though depression is underdiagnosed and undertreated, spending on treatments for depression has increased dramatically since 1990; national Medicaid spending on antidepressants alone exceeded $2 billion annually by 2004 ( 4 ). Although treatment of depression consumes significant financial resources for an insurer (whether public or private), lack of treatment often results in either poor work productivity or an inability to work at all, which significantly increases overall cost to either the employer or the state program providing medical or social support ( 5 , 6 ). In addition, medical costs are significantly higher for patients with untreated depression ( 7 , 8 ). These factors clearly demonstrate the importance of appropriate treatment of depression.

Although there are a number of treatments for depression, including psychotherapeutic techniques and electroconvulsive therapy, the mainstay of treatment for the past 40 years has been pharmacotherapy. Drug classes initially used included tricyclic antidepressants and monoamine oxidase inhibitors, but because of the risk of significant adverse effects of those drugs, treatment has largely been supplanted by the group of drugs collectively referred to as second-generation antidepressants: serotonin and norepinephrine reuptake inhibitors and drugs with other mechanisms of action, such as mirtazapine, bupropion, and nefazodone. There are numerous drugs to choose from but no clear evidence of the superiority of one drug over another ( 9 , 10 ). In addition, efficacy of all the second-generation antidepressants has been less than optimal; studies have shown that after first-line treatment with these agents only about 60% of patients achieve a response and only 40% achieve remission ( 9 , 10 , 11 ).

Given the poor primary response rate, it is particularly important to have evidence about the comparative effectiveness of second- and third-line treatments for depression. Such evidence was sorely lacking before publication in 2006 of the results of the STAR*D trial (Sequenced Treatment Alternatives to Relieve Depression) ( 12 ). The STAR*D trial is thoroughly described in other articles in this special section ( 13 , 14 ), and STAR*D results are not presented in detail here. Instead, the implications of the STAR*D trial for payers are considered.

Implications for payers

STAR*D findings are especially important for payers because the trial focused on patients whose first-line treatment had failed and for whom there was little evidence on which to base decisions about further treatment. In addition, STAR*D was an effectiveness trial, and the applicability of findings to real-world patients is high (for example, two-thirds of patients in the trial had a concurrent medical condition). Thus STAR*D findings would be more likely than those of a clinical efficacy trial to be applicable to patients covered by an insurer, whether a private or government program. However, STAR*D participants' treatment was closely and aggressively managed by a clinician, which probably resulted in overestimation of the treatment benefits that are likely to be observed in Medicaid populations. Because of poorer access to medical care, Medicaid patients often do not have their dosages increased appropriately nor do they receive adequate counseling about the importance of adherence and expectations for improvement ( 15 ). Even in the privately insured population, it has been shown that fewer than 20% of patients receive all of the guideline-recommended follow-up for depression ( 16 ). Finally, STAR*D was publicly funded, therefore eliminating any potential funding bias.

Whereas the ARTIST trial (A Randomized Trial Investigating SSRI Treatment) ( 11 ) was the first large effectiveness trial of first-line treatments for depression, STAR*D is the first large high-quality effectiveness trial of second- and third-line pharmacologic treatments for depression, and it provided valuable information to both clinicians and payers. STAR*D findings add substantially to the body of evidence informing the treatment of depression in real-world settings. Previous research had shown no substantial differences among second-generation antidepressants used for first-line treatment. However, STAR*D showed the same to be true when these medications are used for second- and third-line treatments. STAR*D findings support the general lack of superiority of one drug over another. Also, in terms of the incidence of overall harmful effects, no drug was superior to another, although specific adverse effects were found to differ between drugs.

This information from STAR*D has allowed payers to construct formulary coverage rules with more confidence, knowing that no drug has superior effectiveness or fewer harmful effects. Because the drugs do not differ much in effectiveness but the likelihood of failure with the first medication is high, one reasonable option is to support fail-first policies in which clinicians are permitted to choose any drug in the formulary after a documented failure. For example, the Wisconsin Medicaid program has a total of ten second-generation antidepressants on its preferred drug list; payment for nonpreferred drugs is allowed by prior authorization if the patient is either established on and responding to a nonpreferred drug or has not responded to at least one preferred drug, has a medical contraindication to its use, has a potential drug interaction, or has experienced a clinically significant adverse reaction to the preferred drug.

Typically, policy makers address the issue of where resources might be best directed. In addition to policies that establish rules for preferred drug lists such as those described above, policies that promote improved access to care and promote adherence may improve the treatment of depression. In many regions, access to mental health treatment is a function of provider supply rather than limitations in coverage. Because STAR*D focused on treatment in the community delivered by primary care providers, payers might develop strategies to encourage primary care treatment of depression as a way to improve access. This could be accomplished by pay-for-performance initiatives that improve reimbursement. STAR*D demonstrated the feasibility of incorporating measurement-based care into the primary care setting, because physicians used a clinical decision support system that relied on measurement of symptoms using the QIDS-C and QIDS-SR (Quick Inventory of Depressive Symptomatology Clinician Rating and Self-Report). This suggests that pay-for-performance programs may be an effective tool for improving depression treatment. Such a project is being implemented by the Buyers Health Care Action Group in Minnesota in its Bridges to Excellence program.

Another option for improving reimbursement is to provide coverage for disease management and collaborative care interventions for depression treatment, because there is some evidence that these programs are cost-effective ( 17 ). An example of a program that provides care coordination, patient education and outreach, and the ability to measure outcomes is the DIAMOND project (Depression Improvement Across Minnesota, Offering a New Direction) ( 18 ). Another example specific to Medicaid is the Colorado Access Demonstration ( 19 ). Although increased access to care and better care management will promote adherence to treatment, payers may want to consider additional measures that encourage adherence, such as changes in benefit design, including a reduction in or elimination of copays (for prescriptions or office visits) for the treatment of depression.

Conclusions

STAR*D helped demonstrate that among the many patients who do not respond to first-line treatment with second-generation antidepressants, some respond to second- and third-line treatment options. In terms of effectiveness, there does not appear to be a clear reason to choose one medication over another for second- and third-line treatments. However, aggressive management is an important aspect of ensuring that patients receive a treatment that works. From a payer's perspective, choice of antidepressant medication may be less important than close medication management in providing effective and tolerable treatment.

Acknowledgments and disclosures

Dr. Hansen has received consulting fees from Takeda Pharmaceuticals. The other authors report no competing interests.