1. Introduction
Medicare Part D is the primary source of prescription drug coverage for Americans with Alzheimer's disease and related dementias (ADRD). Cognitive and functional limitations may hinder beneficiaries' pharmacy access. Mail‐order pharmacies can offer convenience [1], although their use among beneficiaries with ADRD and implications on drug utilization and spending remain unknown. This study examined utilization and spending patterns among ADRD Medicare Part D beneficiaries using mail‐order and retail pharmacies.
2. Methods
Pharmaceutical claims data were obtained for beneficiaries with ADRD using 2021 Medicare Prescription Drug Events (PDE) files from a 20% national random sample of Medicare beneficiaries enrolled in traditional fee‐for‐service Medicare (see Supporting Information) [2]. We identified beneficiaries with ADRD using Medicare‐defined diagnosis codes from Medicare's 2020 Master Beneficiary Summary File (MBSF) Chronic Conditions segment [3] and 2021 MBSF Base segment [4]. We restricted the sample to beneficiaries continuously enrolled in Medicare, living at home, and alive through 2021. We excluded beneficiaries without any Part D prescription fills and those with claims from pharmacy types other than retail or mail‐order pharmacies in 2021. Beneficiary characteristics were compared between primarily mail‐order users (> 50% of prescriptions dispensed in mail‐order) and primarily retail pharmacy users (≥ 50% of prescriptions from retail pharmacies).
Drugs were defined as the combination of active ingredient, dosage form, and strength [5]. Acute‐use drugs were excluded. For each beneficiary‐drug pair, the proportion of days covered (PDC), the beneficiary out‐of‐pocket cost, plan spending, and total spending (plan plus beneficiary plus any third‐party payments) were calculated per 30‐day supply. Comparisons were implemented at the beneficiary‐drug level between beneficiary‐drug pairs with 100% of the claims in mail‐order versus beneficiary‐drug pairs with 100% in retail pharmacies. Because low‐income subsidies can lower beneficiary out‐of‐pocket costs, spending metrics were also examined separately for beneficiaries not receiving subsidies. Statistical significance was set at two‐tailed p < 0.05. Two‐tailed independent‐samples T tests and Chi‐square tests were used for the comparisons; all analyses used STATA Version 16 (Stata Corp, College Station, TX). The study was approved by the Internal Review Board and followed STROBE reporting guidelines for observational studies.
3. Results
The sample totaled 121,754 Medicare Part D beneficiaries with ADRD, with 13,393 (11%) primarily mail‐order and 108,361 (89%) primarily retail pharmacy users (Table 1). Mail‐order users were slightly older (mean (SD) age: 81.5 (7.8) vs. 80.3 (9.6) years, p < 0.001); less often female (56% vs. 62%, p < 0.001); and more frequently white (87% vs. 70%, p < 0.001). Fewer MO users received low‐income subsidies (7% vs. 36%, p < 0.001) and fewer were dually eligible for Medicaid (6% vs. 34%, p < 0.001). As measured by PDC, drug possession was greater among mail‐order beneficiary‐drug pairs (mean (SD) PDC: 1.0 (0.1) vs. 0.9 (0.2), p < 0.001).
TABLE 1.
Beneficiary characteristics by primary mail‐order versus retail pharmacy use.
| Beneficiary characteristics | All | Mail‐order a (> 50%) | Retail a (≥ 50%) | p |
|---|---|---|---|---|
| N (% of total) | 121,754 (100%) | 12,988 (10.7%) | 108,766 (89.3%) | N/A |
| Demographic characteristics | ||||
| Age—mean (SD) b | 80.4 (9.4) | 81.5 (7.8) | 80.3 (9.6) | p < 0.001 |
| Female, N (%) b | 75,211 (62%) | 7234 (56%) | 67,977 (62%) | p < 0.001 |
| Race, N (%) b | p < 0.001 | |||
| White | 87,724 (72%) | 11,354 (87%) | 76,370 (70%) | |
| Black | 13,034 (11%) | 778 (6%) | 12,256 (11%) | |
| Other/unknown | 20,996 (17%) | 856 (7%) | 20,140 (19%) | |
| Place of residence, N (%) b , c | ||||
| Metropolitan area | 98,980 (81%) | 10,485 (81%) | 88,495 (81%) | p = 0.090 |
| Rural area | 3925 (3%) | 454 (4%) | 3471 (3%) | |
| Suburban area or small town | 18,763 (15%) | 2043 (16%) | 16,720 (15%) | |
| Clinical characteristics | ||||
| Original reason for Medicare enrollment, N (%) b | ||||
| Old age and survivor's insurance | 101,077 (83%) | 11,425 (88%) | 89,652 (82%) | p < 0.001 |
| Disability insurance benefits | 20,186 (17%) | 1547 (12%) | 18,639 (17%) | |
| ESRD d | 491 (0.4%) | 16 (0.1%) | 475 (0.4%) | |
| N chronic conditions, mean (SD) e , f | 5.5 (2.8) | 5.6 (2.6) | 5.5 (2.8) | p = 0.008 |
| Medicare enrollment characteristics | ||||
| Low‐income subsidy recipient b | 40,511 (33%) | 850 (7%) | 39,661 (36%) | p < 0.001 |
| Eligible for Medicaid b | 37,944 (31%) | 735 (6%) | 37,209 (34%) | p < 0.001 |
| Proportion of days covered (PDC) | ||||
| PDC (first fill to end of year), mean (SD) g | — | 0.9 (0.3) | 0.8 (0.3) | p < 0.001 |
| PDC (first fill to supply end date), mean (SD) g | — | 1.0 (0.1) | 0.9 (0.2) | p < 0.001 |
Obtained from 2021 Medicare Part D Event (PDE) file.
Obtained from 2021 Medicare A/B/C/D file.
Geographic designation as metro, rural, or suburban and small town was based on zip code. N = 121,668 beneficiaries with information.
ESRD: end‐stage renal disease.
Obtained from 2020 Medicare Chronic Conditions file.
Count of 25 chronic conditions included in the A/B/C/D file.
Proportion of days covered (PDC) was calculated at the beneficiary–drug level limited to pairs with 100% of claims in mail‐order or 100% in retail pharmacies. The overall sample included 587,494 pairs, of which 45,095 (7.7%) were mail‐order and 542,399 (92.3%) were retail. It was derived from the date of the first fill in 2021 through both December 31, 2021, and the final date of supply, as the total days' supply of any ADRD medication divided by the number of days remaining in the calendar year.
Source: Authors' analysis of the Medicare Chronic Conditions file, A/B/C/D file, and Part D Event (PDE) file for enrollment year 2021.
Overall, mail‐order beneficiary‐drug pairs had lower plan cost per 30‐day supply (mean (SD): $17.6 ($56.0) vs. $19.1 ($72.2), p < 0.001); lower total cost (mean (SD): $34.8 ($102) vs. $37.8 ($116), p < 0.001); and lower beneficiary out‐of‐pocket cost as a percentage of total cost (mean (SD): 31.6% (37.5%) vs. 39.4% (39.3%), p < 0.001), as compared to retail pharmacy pairs (Figure 1). Among beneficiaries not receiving low‐income subsidies, mail‐order beneficiary‐drug pairs had lower out‐of‐pocket costs (mean (SD): $6.3 ($17.5) vs. $9.6 ($22.9), p < 0.001), and lower beneficiary out‐of‐pocket cost as a percentage of total cost (mean (SD): 33.1% (38.0%) vs. 58.2% (37.5%), p < 0.001) than retail pharmacy pairs.
FIGURE 1.
Prescription drug spending for 30‐day supply among Medicare Part D beneficiaries with Alzheimer's Disease and Related Dementias in U.S. dollars, 2021. (A) Beneficiary payments: Average beneficiary out‐of‐pocket spending. (B) Plan payments: Average Medicare Part D plan spending. (C) Beneficiary share of total cost (%): Average claims‐level percentage contribution of beneficiaries' out‐of‐pocket spending to the total drug cost. (D) Total drug cost: Average plan plus beneficiary plus any third‐party spending. All measures are reported at the beneficiary–drug level for the overall sample and for beneficiaries not receiving low‐income subsidies, stratified by pharmacy type (mail‐order vs. retail). p values: ᵃ0.06; ᵇ< 0.001; ᶜ0.007.
4. Discussion
The study findings suggest that mail‐order pharmacies may offer benefits for Medicare Part D beneficiaries with ADRD through higher medication possession rates and reduced out‐of‐pocket costs, particularly for those not receiving low‐income subsidy. However, mail‐order pharmacies were significantly less likely than retail pharmacies to serve low‐income beneficiaries. The Medicare Part D program should require that plans offer mail‐order pharmacy information for all beneficiaries, especially low‐income subsidy recipients and those dually eligible for Medicaid.
Study limitations include not controlling for confounders such as the degree of cognitive impairment, caregiver support, pharmacy preferences, health status, and medication regimen complexity. PDC may not actually reflect medication intake, and while the differences in drug possession were statistically significant, their clinical significance remains uncertain. The study findings may not be generalizable to all Medicare beneficiaries with ADRD, including newly diagnosed cases and Medicare Advantage beneficiaries.
Future research should explore how patient and caregiver preferences, medication characteristics, and pharmacy type interact to affect drug utilization and clinical and financial outcomes. To inform policy interventions, studies should assess specific strategies that could help Part D plans increase medication access and utilization in the Medicare population with ADRD.
Author Contributions
Mariana P. Socal: conceptualization, methodology, writing – reviewing and editing, supervision, project administration, funding acquisition. Jenny Markell: methodology, data curation, formal analysis, investigation, visualization, writing – original draft preparation, writing – reviewing and editing. Pineal I. Bareamichael: data curation, formal analysis, investigation, visualization, writing – original draft preparation, writing – reviewing and editing. Ge Bai: conceptualization, methodology, writing – reviewing and editing, supervision. Jeromie Ballreich: conceptualization, methodology, writing – reviewing and editing, supervision.
Disclosure
The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the article; and decision to submit the article for publication.
Conflicts of Interest
Mariana P. Socal and Jeromie Ballreich had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. The other authors declare no conflicts of interest.
Supporting information
Text S1: Data sources and identification of the study sample.
Text S2: Definition of comparison groups for the beneficiary‐level and claims‐level analysis.
Figure S1: Sample selection flow diagram: beneficiary level.
Figure S2: Sample selection flow diagram: claims level.
Text S3: Definition of outcome variables.
Acknowledgments
The authors are grateful for feedback received from faculty from the Hopkins Economics of Alzheimer's Disease & Services (HEADS) Center at the planning phases of this project, especially Prof. Cynthia Boyd, M.D. MPH. The data examined in this study was obtained from Medicare Part D Prescription Drug Event data (“PDE”) files; Medicare Chronic Conditions files; and Medicare beneficiary enrollment information (“A/B/C/D”) files. The data is property of the Centers for Medicare and Medicaid Services (CMS) and can be obtained at: https://resdac.org/.
Socal M. P., Markell J., Bareamichael P. I., Bai G., and Ballreich J., “Mail‐Order and Retail Pharmacy Use Among Medicare Beneficiaries With Alzheimer's Disease and Related Dementias,” Journal of the American Geriatrics Society 74, no. 2 (2026): 585–588, 10.1111/jgs.70188.
Funding: This work was funded by the Hopkins Economics of Alzheimer's Disease & Services Center and Arnold Ventures (P30AG066587).
References
- 1. Fernandez E. V., McDaniel J. A., and Carroll N. V., “Examination of the Link Between Medication Adherence and Use of Mail‐Order Pharmacies in Chronic Disease States,” Journal of Managed Care & Specialty Pharmacy 22, no. 11 (2016): 1247–1259, 10.18553/jmcp.2016.22.11.1247. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Centers for Medicare & Medicaid Services , Medicare Part D Event (PDE) File [Data Set] (Chronic Conditions Data Warehouse, 2021), https://resdac.org/cms‐data/files/pde. [Google Scholar]
- 3. Centers for Medicare & Medicaid Services , Master Beneficiary Summary File (MBSF): Chronic Conditions Segment [Data Set] (Chronic Conditions Data Warehouse, 2020), https://resdac.org/cms‐data/files/mbsf‐27‐cc. [Google Scholar]
- 4. Centers for Medicare & Medicaid Services , Master Beneficiary Summary File (MBSF): Base Segment [Data Set] (Chronic Conditions Data Warehouse, 2021), https://resdac.org/cms‐data/files/mbsf‐base. [Google Scholar]
- 5. IBM Micromedex , Red Book Online [Database] (IBM Corporation, 2021), https://www.micromedexsolutions.com/home/dispatch. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Text S1: Data sources and identification of the study sample.
Text S2: Definition of comparison groups for the beneficiary‐level and claims‐level analysis.
Figure S1: Sample selection flow diagram: beneficiary level.
Figure S2: Sample selection flow diagram: claims level.
Text S3: Definition of outcome variables.