Medicare Part D became effective on January 1,2006, and instantly provided 43 million beneficiaries with the opportunity to receive outpatient prescription drug coverage through Medicare for the first time.1 In 2011, about two-thirds of the 29 million beneficiaries in a Medicare drug plan were enrolled in a stand-alone prescription drug plan (PDP) which provides prescription drug coverage (Part D) to beneficiaries who have Medicare Part A and/or Part B benefits—and 8 million (27.5%) received the low-income subsidy.2 Alternatively, beneficiaries can receive prescription drug coverage through a Medicare Advantage Prescription Drug Plan (MA-PD or Part C); these plans may also provide services that are not covered by original Medicare (Parts A & B). The Medicare prescription drug benefit has resulted in significant drug cost savings for beneficiaries since its inception.3

By removing cost barriers, Medicare Part D has had a significantly positive impact on the use of essential medications.3 It is estimated that 63% of beneficiaries 65 years and older who were previously without drug coverage enrolled in a Part D plan upon the program’s establishment.4 Studies examining out-of-pocket prescription drug spending between 2006 and 2009 found a 13% to 18% decrease in such expenditures.5-7 More recently, Basu et al found similar results reporting that Part D resulted in a 17% to 18% annual reduction in out-of-pocket beneficiary costs.8 In fact, the odds of spending more than $100 per month on medications were significantly reduced after the introduction of Part D.4 Another reported advantage was the reduction in out-of-pocket costs despite an increase in medication utilization by previously uninsured beneficiaries.4 However, despite all of the program successes, certain challenges remain.

Although beneficiary spending on prescription medica-tions has decreased since 2006, other cost-saving opportunities may be underused by, or unknown to, beneficiaries. Each year, beneficiaries are faced with a myriad of Part D plans from which they can receive their prescription drug coverage. Since 2006, the number of PDPs offered in California, for instance, has varied from a peak of 56 in 2008 to a nadir of 32 in 2013 (Table 1).2 While the number of PDP offerings has decreased significantly, beneficiaries were still faced with difficult choices. Aside from changes to plan offerings, each plan may also change its formulary and cost-sharing structure on an annual basis. Collectively, this results in a confusing and potentially costly situation for beneficiaries who are attempting to select an optimal Part D plan for the upcoming year.9

Most beneficiaries flock to one of just a few avail-able Part D plans. In 2011, nearly 69% of all PDP enrollees were in 7 available PDPs.2 This trend may be due to plan name recognition, effective marketing strategies, plan loyalty, lower monthly plan premiums, or any of several other reasons. As such, many beneficia-ries often fail to properly compare plan offerings before choosing their Part D plan.10 These findings suggest that beneficiaries are facing significant avoidable out-of-pocket costs. One study suggested that 75% of beneficiaries were not in the lowest cost PDP, and thus paid 30% more than necessary for their medications.11 Out of 55,000 surveyed beneficiaries enrolled in a PDP in 2006, only 6% to 9% chose the lowest cost plan, which equated to between $360 and $520 in additional out-of-pocket spending per beneficiary per year.10 Another study found that 89.7% of beneficiaries who remained in their current plan would not have been enrolled in the lowest cost PDP in the upcoming year. The same study found that beneficiaries could have saved an estimated $431 per year in out-of-pocket expenses by optimizing their PDP.12

Beneficiaries with limited income and assets who do not qualify for Medicaid may meet the less-stringent Low-Income Subsidy (LIS) criteria. The LIS provides “extra help” to further defray out-of-pocket Part D costs. LIS recipients are automatically enrolled into a basic plan (“benchmark plan”), which incorporates a $0 monthly plan premium.13 Subsidy recipients also have lower co-payments for formulary-covered medications and are not subject to the dreaded “donut hole.” However, there is a significant lack of awareness of the LIS program, which can further compound the unnecessary financial burden faced by many beneficiaries.10,14

In general, beneficiaries with limited financial resources have greatly benefited from Part D in terms of increased medication access and decreased annual costs.12,14-16 But the question remains, “Can more be done?”

The variability in beneficiary costs based on plan selec-tion can be substantial, and it is further complicated by the yearly fluctuation in the number of benchmark plans in each of the 34 Medicare regions. In 2006, 10 benchmark plans were available in California; in contrast, only 6 benchmark plans were available in California in 2013 (Table 1).17 Benchmark plan availability and changes can result in increased confusion for subsidy recipients, many of whom are non-English speaking.13

Beneficiaries who fail to fully understand the benefit may end up paying more than necessary for their prescription medications.8 Numerous studies have highlighted the financial impact of Part D on beneficiary costs in a single year.6,7,10 The present study was conducted to examine the relationship between PDP costs, cost savings from plan switching, and subsidy status in California from 2008 to 2013.



Study Design

We sought to examine the potential opportunity costs faced by beneficiaries in California from 2008 to 2013 if they failed to optimize their PDPs. The potential cost savings in each study year were evaluated in aggregate and also as a function of the beneficiaries’ subsidy status in each year.

Outreach events were conducted in cities across central and northern California during the annual open enrollment window, during which beneficiaries were assisted with their Part D plans. Beneficiaries were notified of these free events through various media outreach programs and self-selected to attend; beneficiaries in attendance were allowed to opt out of the research component while still receiving the same services. During each intervention, the estimated annual cost (EAC) of the beneficiaries’ current PDP in the upcoming year, along with the lowest EAC plan for the upcoming year, were recorded. The difference between the two represented the potential opportunity cost (OC) a beneficiary would face by remaining in their current plan in the upcoming year. The OC was calculated for each beneficiary during each study year.

Outreach Design

Beneficiaries were assisted at 59 different community outreach events from 2007 to 2012 during the fall annual open enrollment period. These outreach events were held at senior centers, retirement communities, Housing and Urban Development subsidized housing units, places of worship, and various other settings. In total, 2262 beneficiaries were assisted during the study period. Beneficiaries who were enrolled in an MA-PD, had coverage from an employer or union group that was at least as good as what Medicare could provide (creditable coverage), and those with noncreditable or no coverage were excluded from the study. Forty-one beneficiaries with a PDP opted out of the research study, and thus the final study sample size was composed of 1435 beneficiaries.

Outreach Intervention

Each beneficiary was assisted with their Part D plan by a pair of trained pharmacy students who delivered the interventional assistance in the beneficiary’s preferred language. The Medicare Plan Finder Tool (available at was utilized to assist with the Part D intervention. A personalized plan search was conducted to recognize the current plan enrollment and subsidy status of each beneficiary. The prescription medication profile (eg, drug name, strength, refill quantity) and per-sonal preferences (eg, preferred pharmacy) of each assisted beneficiary were entered into the Plan Finder Tool. Prescriptions were included only if they were regularly taken or if a general frequency of use could be defined. For example, medications taken daily for chronic dis-eases were included, as were “as needed” medications that were taken with some frequency (eg, albuterol in-haler used about once a week or sublingual nitroglycerin used at most once a month for angina). Once all relevant information was entered, the Plan Results page would reveal the name and the EAC of each available PDP. The EAC revealed the estimated out-of-pocket expenditures the beneficiary would face under that plan during the upcoming year. As such, EAC took into account expected beneficiary spending on the plan premium, deductible, copayment or coinsurance amounts, spending during the coverage gap and catastrophic coverage, and the cost of nonformulary medications.

The intent of each beneficiary intervention was to iden-tify the optimal PDP (lowest EAC) for the upcoming year. The potential OC of a beneficiary remaining in their cur-rent PDP was calculated as follows:

OC = [EAC ($) of current plan in the upcoming year – EAC ($) of least expensive PDP in the upcoming year]

Each beneficiary was presented with all available information so that they could make a well-informed decision about their Part D plan. In addition, each was offered help enrolling in a new plan onsite at the outreach event students were supervised by pharmacist faculty members who taught a Medicare course.


Data Collection

Trained pharmacy students who provided the interven-tion collected all demographic and interventional data. At the onset of each intervention, beneficiaries were informed about the intent of the session and were asked about their willingness to participate in a survey. Once permission via informed verbal consent was granted, all beneficiary-specific demographic information, data retrieved from the Plan Finder Tool, and interventional characteristics were recorded via a standardized survey form. The research was approved by the University of the Pacific’s Institutional Review Board.

Data Analysis

Descriptive statistics were reported for beneficiaries attending outreach events in each study year (Table 2). The minimum, maximum, mean, and quartile values of the potential OC ($) of remaining in the same PDP in the upcoming year for each study year are reported in Table 3. Potential OC as a function of subsidy status was also recorded for each year (Tables 4 and 5). Statistics were performed via IBM SPSS Statistics 19 (Armonk, New York).


Beneficiary demographic characteristics for those as-sisted each year from 2007 to 2012 can be found in Table 2. Additionally, Table 2 provides the aggregate total of those served during the entire study period. Although our final study sample comprised only those with a PDP, Table 2 reveals that each year during the study period between 27% and 38% of beneficiaries had something other than a PDP. Data from 1476 beneficiaries enrolled in PDPs were gathered, and all subsequent reported analyses in Table 2 focused on this sample. The majority (95%) of study beneficiaries were 65 years or older, with a mean age of 75.4 years. However, 68 study beneficiaries were aged less than 65 years and therefore qualified for Medicare under provisions unrelated to age (eg, disability, end-stage renal disease). In addition, 626 (44%) study beneficiaries were receiving either Medicaid or the LIS to further help lower their out-of-pocket prescription drug costs.

The average annual potential OC of remaining in the same PDP in the upcoming year was $728 for the 1435 beneficiaries for whom such data were recorded (Table 3). Nearly 4 in 5 beneficiaries were in a PDP that was estimated to not be the lowest cost plan in the upcoming year, and they therefore would realize an OC if they remained in their current PDP. Onsite enrollment varied from 41% of beneficiaries seen in 2008, to 69% of beneficiaries in 2009, with an aggregate average of 62% enrolling in a new PDP onsite during an event (Table 3).

Table 4 reveals that 87% of nonsubsidy recipients could have saved $792 on average by switching to a different PDP in the upcoming year during the study period. In contrast, Table 5 reveals that approximately 72% of subsidy recipients could have saved $658 on average by switching to a different PDP in the upcoming year.



When we evaluated the plan costs of PDPs, the Medicare beneficiaries enrolled in them, and the subsidy status of those beneficiaries in California between the 2008 and 2013 benefit years, the results showed that most beneficiaries face significant avoidable out-of-pocket costs if they fail to reexamine their plan annually. As identified earlier, the potential beneficiary OC defined as the difference between the estimated annual cost of remaining in the current plan and switching to the least expensive plan in the upcoming plan year averaged approximately $728 across all study beneficiaries. To our knowledge, this is the first study to examine the potential OC of failing to optimize the Medicare Part D plan of actual Medicare beneficiaries over multiple years. Additionally, this study reported costs that were obtained by the same method as beneficiaries themselves would have obtained, rather than via a retrospective database-style analysis. While the number of beneficiaries served represents only a fraction of the total Medicare population in California, the similarity in findings to larger studies, and the unique real-world data collection methods, speak to its relevance.

Results showed that 285 beneficiaries during the study period or 20% of those assisted had PDPs whose current plan would have been the least expensive in the upcoming year. The remaining 80% of beneficiaries could have realized potential cost savings by switching to a different PDP. Moreover, despite the heavily subsidized benefit and benchmark plan offerings for beneficiaries receiving the LIS or Medicaid, more than 72% of them could have potentially saved money by switching to another Part D plan in the upcoming year. Currently, subsidy recipients are auto-enrolled into a plan that offers a basic benefit and a monthly premium below the regional benchmark thresh-old; such plans are called “benchmark plans,” and the process of beneficiary assignment to these plans is random.

This random assignment does not take into account the medications regularly taken by the beneficiary. It also does not account for annual formulary changes that may result in loss of formulary coverage for beneficiaries auto-enrolled in a plan. The present findings suggest that we should move away from random assignment, and instead use an approach that considers the medication regimen of the beneficiary (beneficiary-centered assignment). This approach seems even more essential considering that subsidy recipients already have limited economic means. Beneficiaries seeking to lower out-of-pocket costs can explore getting drug samples from their prescriber, switching to cheaper therapeutic alternatives, or applying to pharmaceutical assistance programs. Otherwise, any avoidable out-of-pocket costs will likely have one or both of the following outcomes: 1) an increase in cost-related medication nonadherence, and 2) sacrifices made by beneficiaries on purchases of items of basic need (eg, food, clothing).

In contrast to subsidy recipients, it was found that those beneficiaries whose income and assets were beyond the threshold to qualify for extra help with their medication costs potentially had even more to lose by failing to evaluate their Part D plan on an annual basis; it was found that nearly 9 out of 10 nonsubsidy recipients could save money by switching to a new PDP in the upcoming year. The average potential annual savings for nonsubsidy recipients during the study window was $792. This is a significant amount when considering that most beneficiaries are seniors with fixed incomes.

Suboptimal plan selection may occur for a variety of rea-sons, including the beneficiary’s selection of an enhanced plan, which may be unnecessary based on their unique list of prescribed medications; purchasing a plan from a company only because of name recognition; fear of change and therefore finding comfort in remaining in their existing plan; or simply having a lack of resources or knowledge as to how to objectively evaluate plan offerings. Medicare Part D is fraught with many complexities, thereby making plan selection difficult for someone untrained or not well versed in the benefit,18 but beneficiaries must be educated about the importance of an annual Medicare Part D plan checkup. Governmental agencies (eg, State Health Insurance Assistance Programs, CMS) and other objective ben-eficiary advocacy groups, including supervised pharmacy students who are trained in the benefit, are sources for providing beneficiaries with the necessary assistance to minimize their out-of-pocket Part D prescription drug costs.

As the Part D program continues to evolve and the number of beneficiaries continues to increase, personalized plan assistance based on a beneficiary’s medication utilization patterns and personal preferences will become increasingly important. Moreover, other cost-saving opportunities, including the LIS and the utilization of less expensive therapeutic alternatives, may be underutilized. Pharmacists are ideally positioned, and are the most accessible healthcare professionals, to help meet the needs of Medicare beneficiaries regarding plan optimization and medication evaluation. Previous research showed that healthcare professionals assisting beneficiaries with plan selection can save them between $382 and $522 annu-ally.11,12 However, there are barriers to pharmacists providing such services, including a lack reimbursement for the service, time constrictions, and staff and logistical constraints. We remain optimistic that with the current tenor of provider status for pharmacists and the ability to bill for services, the ability to bill for providing Part D plan selection assistance in the future is a distinct possibility. If nothing else, pharmacists should educate beneficiaries with a Part D plan on the necessity of plan re-evaluation every year during the open enrollment period to avoid un-necessary out-of-pocket expenditures.

In addition, changes at multiple levels of the Part D benefit would reduce the potential costs of coverage for beneficiaries. Part D plan providers could be required to send the beneficiary the projected cost of their medications in the subsequent year based on their utilization over the past year. Currently, beneficiaries are only sent general information on premiums and co-payments, but converting these to actual out-of-pocket costs is difficult even for trained providers who do not use the Medicare Plan Finder Tool. This would take away any “sticker shock” for ben-eficiaries who do not reevaluate their plans during open enrollment and are unpleasantly surprised to see their increased co-pays in January when their altered policies have taken effect. Additionally, plans could be required to indicate if other plans exist that would be less expensive in the upcoming year. While the logistics and time involved in this effort would be substantial, a collaborative effort with CMS and pharmacists might be warranted. This could lead to further collaboration between pharmacists and Part D plans whereby pharmacists would receive compensation for plan review just as they do for performing a comprehensive medication review for eligible beneficiaries, also known as Medication Therapy Management (MTM). years to determine new versus repeat attendees. No patient tracking data were recorded at any event (except for a postcard patients could fill out to request information for the subsequent year’s events). Data from this study were collected during outreach events in 1 area (northern/cen-tral) of 1 Medicare region (California), and the participants constitute only a small fraction of all Medicare beneficiaries in this region. Confirmation of our findings in other areas and/or regions is recommended. In addition, many of the organized outreach events were targeted to those from under-represented and underserved areas, including those living at or below the federal poverty level, non-English speakers, and those from racial and ethnic minor-ity groups. Taken together, these factors may explain the findings reported in the present study.

Our overarching goal of this program was to serve seniors and other Medicare beneficiaries by providing in-person assistance with Part D plan optimization. As part of the outreach intervention, we wanted to measure the opportunity for, and potential cost savings of, Part D plan optimization. The sample size clearly does not compare to most retrospective database studies. However, this study reflects data that were obtained through the same method that beneficiaries and/or their advocates would rely upon if trying to compare PDP costs in the real world. As such, the limitations and assumptions of the present study are the same as those that would be endured by beneficiaries.



Despite fluctuations in Medicare Part D plan offerings, including benchmark plans, annual plan reexamination remains critically important for beneficiaries to maximize their Part D benefit and minimize their out-of-pocket costs. Despite many of its advantages, the Part D benefit remains challenging for many beneficiaries to fully understand and optimize. Trained advocates can assist beneficiaries to save money on their Part D prescription drug costs and thereby reduce cost-related medication nonadherence.