Duke 2008 | ASHE | Equity and Efficiency in Health and Healthcare

Depression is highly prevalent among older adults and is associated with substantial morbidity, higher overall health care expenditures, limited functioning and increased risk of suicide. Antidepressant medication is by far the most common treatment for depression among older adults. Studies of tiered formularies in employment-based health insurance indicate that demand for antidepressants is sensitive to out of pocket price. Few studies, however, have examined the impact of pharmacy benefit design on demand for antidepressants among older adults. Moreover, there is little empirical evidence on the impact of gaps in drug coverage such as the "donut hole" in Medicare Part D on use of antidepressants among older adults. In this study, we estimate the impact of quarterly pharmacy benefit limits on the likelihood of medication treatment for depression and spending on antidepressants, as well as total pharmacy and medical spending among elderly Medicare enrollees with depression.

We used 2005 enrollment, pharmacy and medical claims data from a large Medicare Advantage plan in Pennsylvania. We compared adults age 65 and older with three different levels of pharmacy benefit: 2 groups who enrolled in the Medicare Advantage plan as individual members who had quarterly pharmacy benefit limits of $350 or $150 depending on their county of residence, and a third group without caps on pharmacy benefits who enrolled in the Medicare Advantage plan as part of an employer group. Thus, in our study an individual's drug benefit level was determined by availability through former employers or county of residence, not by individual choices that may be influenced by expected drug spending. This limits selection biases. We first used a time series model to estimate the effect of the quarterly cap on monthly trends in pharmacy and medical spending. We also used a generalized linear model to estimate the effect of pharmacy benefit level on our outcome variables, controlling for individual characteristics. We used a logit model to estimate the effect on likelihood of receiving depression treatment.

Of the 202,251 members enrolled in the plan, 13.5% received a diagnosis of depression that was recorded on an inpatient or outpatient medical claim. We found that individuals with $150 and $350 quarterly caps were 17% and 9.3% less likely to receive antidepressant medication, respectively, than individuals with no cap. Conditional on any use, antidepressant spending in the $150 and $350 quarterly cap groups was 70% and 48% lower, respectively, than in the group with no cap. Overall pharmacy utilization and spending were also much lower in the capped groups than among individuals with no cap. There were, however, no statistically significant differences in non-drug medical spending among the 3 pharmacy benefit groups.

We conclude that benefit designs with caps on prescription drug expenditures such as those under Medicare Part D may have unintended consequences in terms of reducing rates of treatment for chronic conditions such as depression. Limited pharmacy benefits may be an important economic barrier to improving quality of depression care for older adults.