Effects of Medicare Part D medication therapy management on racial/ethnic disparities in adherence to antidementia medications among patients with Alzheimer’s disease and related dementias: An observational study

doi:10.21203/rs.3.rs-2070681/v1

Download PDF

Research Article

Effects of Medicare Part D medication therapy management on racial/ethnic disparities in adherence to antidementia medications among patients with Alzheimer’s disease and related dementias: An observational study

https://doi.org/10.21203/rs.3.rs-2070681/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 01 Mar, 2024

Read the published version in Exploratory Research in Clinical and Social Pharmacy →

Version 1

posted

You are reading this latest preprint version

Background

Evidence is sparse on the effects of Medicare medication therapy management on racial/ethnic disparities in medication adherence among patients with Alzheimer’s disease and related dementias. The objective of this study was to examine the Medicare medication therapy management program’s effects on racial/ethnic disparities in the adherence to antidementia medications among patients with Alzheimer’s disease and related dementias.

Methods

This is a retrospective analysis of 2016 and 2017 Medicare Parts A, B, and D data linked to Area Health Resources Files. The study outcome was nonadherence to antidementia medications, and intervention was defined as new medication therapy management enrollment in 2017. Propensity score matching was conducted to create intervention and comparison groups with comparable characteristics. A difference-in-differences model was employed with a logistic regression, including interaction terms of dummy variables for the intervention group and racial/ethnic minorities.

Results

Unadjusted comparisons revealed that Black, Hispanic, and Asian/Pacific Islander (Asian) patients were more likely to be nonadherent than non-Hispanic White (White) patients in 2016. Differences in odds of nonadherence between Black and White patients were lower in 2017 than in 2016 by 27% (odds ratios [OR]: 0.73, 95% confidence interval [CI]: 0.65–0.82). The same difference was also lower between Hispanic and White patients by 26% (OR: 0.74, 95% CI: 0.63–0.87). The MTM enrollment was associated with reduced disparities for Black-White patients of 33% (OR: 0.67, 95% CI: 0.57–0.78) and Hispanic-White patients of 19% (OR: 0.81, 95% CI: 0.67–0.99).

Conclusions

Medicare medication therapy management program was associated with lower disparities between Black and Hispanic patients with Alzheimer’s disease and related dementias than their White counterparts. Expanding the medication therapy management program may particularly benefit racial/ethnic minorities in Alzheimer’s disease and related dementias care.

health care policy

health disparities

health services use

Alzheimer’s disease and related dementias (ADRD) is an umbrella term encompassing Alzheimer’s disease (AD) and other forms of dementia, including frontotemporal degeneration, Lewy body dementia, vascular contributions to cognitive impairment and dementia, and mixed etiology dementias [1]. More than six million older Americans aged ≥ 65 years are estimated to have ADRD in 2021, accounting for 11% of the U.S. population in this age group [2]. This estimate is projected to increase to nearly 14 million by 2060, marking more than a twofold increase in less than four decades [2].

ADRD is a devastating condition whose burden on patients and society manifests in multiple aspects. Pharmacologically, no treatment cures or slows the progression of ADRD. Thus far, the U.S. Food and Drug Administration (FDA) has approved six medications, five of which (i.e., donepezil, galantamine, rivastigmine, memantine, and memantine combined with donepezil) are antidementia medications and merely intended to mitigate cognitive symptoms.² Aducanumab, a medication that received accelerated FDA approval in June 2021, is the only treatment option touted to delay the progression of AD.³ While it is still too early to fully evaluate the efficacy of Aducanumab, evidence on the other five medications is mixed [4, 5]. Nonetheless, due to their prospects of improving patients’ quality of life, the medications have been included in recommended treatment guidelines and are widely prescribed to patients with ADRD [5, 6].

The availability of these antidementia medications, however, does not equate to effective utilization of them. Medication nonadherence is common among older patients with ADRD due to their memory loss and cognitive impairment [7]. The proportion of patients nonadherent to oral antidementia medications ranged from 33% [8.9] to over 40% [10]. Nonadherence increases the risk of adverse drug events and the need for caregivers to help with medication management [7]. Aside from pharmacological and medication management challenges, ADRD care is costly for both individuals and society. The estimated cost of care in 2021 totals $355 billion, with the largest share (51%) covered by Medicare [2]. Older Medicare beneficiaries with ADRD incur out-of-pocket expenses that are more than four times higher than those without ADRD [2].

To control costs and improve therapeutic outcomes for patients with challenging diseases such as ADRD, the Centers for Medicare & Medicaid Services (CMS) has since required Medicare Part D sponsors to provide medication therapy management (MTM) services to eligible Medicare beneficiaries [11]. Mostly provided by pharmacists, such services typically include medication reviews and follow-up action plans [12]. Each Part D sponsor establishes its own MTM eligibility criteria using the CMS guidelines, which mandate sponsors to target beneficiaries who have multiple chronic conditions, take multiple Part D medications, and are likely to incur high medication expenditures [12]. In addition, the guidelines require sponsors to choose from a list of core chronic conditions, including AD, when determining the specific conditions to target in their MTM eligibility criteria [12].

Evidence on the positive effects of Medicare MTM on medication adherence has thus far been limited. A CMS-commissioned evaluation reported a positive effect on MTM enrollees with diabetes, congestive heart failure, and chronic obstructive pulmonary disease [13]. Among more recent studies, Peasah and colleagues noted that a more robust MTM intervention led to increased adherence to anticoagulants among Medicare beneficiaries [14]. Another study focusing on AD patients found that comprehensive medication review, a required component of the Medicare MTM program, had reduced nonadherence to medications for diabetes, hypertension, and hyperlipidemia [15]. These findings were consistent with results from the broader literature on non-Medicare MTM services that generally noted a positive effect of such services on adherence, although the evidence is far from conclusive due to heterogeneity in methodology and sample size across studies [16, 17].

In the MTM literature, few studies have examined the effects of MTM on medication utilization by racial/ethnic minorities. This is primarily due to the lack of these patient characteristics in publicly available data [16, 18]. Several studies reported racial/ethnic disparities in the use of antidementia medications among ADRD patients, such as finding minorities less likely to be treated with antidementia medications than non-Hispanic Whites (Whites) [19]. Minorities also used fewer antidementia medications [6, 20, 21] and discontinued the medications faster [22]. We bridge the gap in MTM and ADRD research by examining the effects of the Medicare MTM program on racial/ethnic disparities in adherence to antidementia medications among older beneficiaries with ADRD.

Data Source

We conducted a retrospective study using 100% 2016 and 2017 Medicare Parts A/B/D administrative data linked to the Master Beneficiary Summary File and Area Health Resources Files (AHRF). Medicare administrative data and Master Beneficiary Summary File provided information on diagnosis records, medication fills, MTM enrollment, and beneficiary demographic characteristics [23]. The AHRF provided county-level data on the socioeconomic and health service characteristics of the beneficiaries’ county of residence [24].

Study sample

We identified our study sample based on the following inclusion criteria for beneficiaries: (1) had an ADRD diagnosis; (2) were at least 65 years of age at the beginning of 2016 and alive at the end of 2017; (3) maintained continuous Parts A, B, and D coverage; and (4) had at least two prescription claims for antidementia medications and more than 60 days of supply in each of the two years [25]. ADRD diagnoses were identified utilizing the International Classification of Diseases, Ninth and Tenth Revisions, Clinical Modification (ICD-9-CM and ICD-10-CM) codes specified in the algorithm for ADRD in CMS Chronic Conditions Warehouse [26].

Based on the Pharmacy Quality Alliance (PQA) technical specifications, we included donepezil, galantamine, rivastigmine, and memantine as antidementia medications [25]. Established by CMS in 2006, PQA is a non-governmental nonprofit organization that develops and maintains medication performance measures, some of which have been adopted by CMS in its Star Ratings program for Parts C and D health insurance plans [27]. Race/ethnicity in our study sample was categorized into five groups: Whites, Blacks, Hispanics, Asian/Pacific Islanders (Asians), and Other. The Other category encompassed American Indians, Alaska Natives, unknown, and other races/ethnicities.

Study groups construction

To facilitate a difference-in-differences analysis, we constructed an intervention group and a comparison group from beneficiaries who met the inclusion criteria. The intervention was defined as new MTM enrollment in 2017. The intervention group consisted of new MTM enrollees in 2017. The comparison group included MTM non-enrollees who satisfied the CMS guidelines for MTM eligibility but were not enrolled in MTM during the study period.

For the years studied, the CMS guidelines required plans to target beneficiaries who met the following criteria: (1) had at least two to three chronic conditions; (2) were prescribed at least two to eight covered medications; and (3) were likely to incur at least an annual medication expense of $3,507 in 2016 and $3,919 in 2017 [28, 29]. Because most plan sponsors adopted three chronic conditions and eight covered medications in their MTM eligibility criteria [30, 31], we used these two values in conjunction with the annual medication spending requirement to identify beneficiaries in the comparison group. Given that data on the exact chronic conditions targeted by individual MTM programs were not available, a beneficiary’s number of chronic conditions was determined based on the 25 prevalent chronic diseases in MTM programs [32].

To reduce the imbalance in beneficiary baseline characteristics between intervention and comparison groups, we matched MTM non-enrollees to enrollees in a 1:1 ratio based on a propensity score that predicted the likelihood of a beneficiary’s MTM enrollment. We employed a nearest neighborhood algorithm without replacement for the matching because the algorithm created the most balanced groups compared with alternative algorithms [33].

Outcome measure

The outcome measured was nonadherence to antidementia medications. A binary outcome variable was constructed based on the proportion of days covered (PDC), calculated as the total number of days that a beneficiary was covered by at least one antidementia medication over the total number of days in the measurement period. For the comparison group in both years and the intervention group in 2016, the measurement period began from the earliest date of an antidementia medication fill through the end of the year. For the intervention group in 2017, the period started on the first prescription fill date after the MTM enrollment date. Beneficiaries were classified as nonadherent if their PDC was less than 80% and adherent otherwise (1 = PDC < 0.80; 0 = PDC ≥ 0.80).

Covariates

The conceptual framework for this study was Gelberg-Andersen’s Behavioral Model for Vulnerable Populations, and medication utilization patterns were considered as a function of predisposing, enabling, and need factors in this study [34]. Predisposing factors influence a beneficiary’s inclination to seek medications. Enabling factors facilitate or hinder the use and access of medications. Need factors signify a beneficiary’s need for medication. We used the following covariates for this study: the predisposing variables included age, sex, race/ethnicity, the proportion of married-couple families, the proportion of people with at least high school education, per capita income, and the proportion of people without health insurance. The enabling variables included dummy variables for metropolitan statistical area (MSA), Health Professional Shortage Area (HPSA), and census regions. The need variable was a CMS risk adjustment summary score. A higher score indicated greater expected health care utilization. Among the covariates, age, sex, race/ethnicity, and risk adjustment summary score were beneficiary-level, while all other covariates were measured at community-level based on a beneficiary’s county of residence.

Empirical approach and statistical analysis

We first conducted descriptive analyses concerning the baseline characteristics and study outcomes of the intervention and comparison groups. We then analyzed the effects of MTM on racial/ethnic disparities in three steps by estimating a different multivariate logistic regression model at each step.

First, we included in our model a dummy for each racial/ethnic minority, with Whites being the reference group, to determine whether racial/ethnic disparities in nonadherence to antidementia medications existed in the baseline year for both intervention and comparison groups. If the odds ratio (OR) of a minority dummy was significantly greater than one, this would indicate the minority group was more likely than Whites to be nonadherent.

Next, we incorporated interaction terms between dummy variables for the year 2017 and racial/ethnic minorities to examine changes in racial/ethnic disparities over time. If the OR of the interaction term for a minority group was significantly lower than one, this would indicate that the disparities between Whites and the minority group decreased in 2017.

Finally, we used a difference-in-differences (DID) approach by incorporating three-way interaction terms of dummy variables for MTM enrollment, the year 2017, and racial/ethnic minorities to compare the changes of racial/ethnic disparities between the intervention and comparison groups. If the OR of the three-way interaction term for a minority group was significantly lower than one, this would indicate that the difference in the odds of Nonadherence between Whites and the minority group in the intervention group was reduced by a greater magnitude over time than the comparison group. This would demonstrate that MTM enrollment was associated with reduced racial/ethnic disparities concerning medication nonadherence.

In all regression models, standard errors were clustered at the county level to account for potential within-county correlation. Statistical significance was determined at a two-sided alpha value of 0.05. All analyses were conducted with SAS Enterprise Guide 7.1 at the CMS Virtual Research Data Center. All important data analysis codes can be found on the study website (https://mtmstarvalue.uthsc.edu/codes/). The Institutional Review Board at the University of Tennessee Health Science Center US approved the study (approval number: #20-07197-XM).

Descriptive statistics for the baseline beneficiary characteristics are reported in Table 1. After propensity score matching, the final study sample consisted of 56,426 Medicare beneficiaries, with the intervention and comparison groups each accounting for 50%. There were 72.2% Whites, 11.7% Blacks, 10.8% Hispanics, 4.0% Asians, and 1.4% Other in the study sample. All characteristics were balanced and were not significantly different across the two study groups (P > .05).

Table 1. Baseline characteristics of the study population in 2016 by intervention status

Characteristics	Intervention Group (n = 28,213, 50%)		Comparison Group (n = 28,213, 50%)
Characteristics	Number	%	Number	%
Predisposing Factors
Age, mean (SD)	80.71 (7.02)		80.68 (7.24)
Male	9,262	32.83	9,262	32.83
Race/Ethnicity
Non-Hispanic Whites	20,355	72.15	20,355	72.15
Blacks	3,302	11.70	3,302	11.70
Hispanics	3,043	10.79	3,043	10.79
Asians/Pacific Islanders	1,119	3.97	1,119	3.97
Other	394	1.40	394	1.40
Proportion of Married-Couple Families, mean (SD)^a	0.72 (0.07)		0.72 (0.07)
Proportion of Education >= High School, mean (SD)^a	0.86 (0.06)		0.86 (0.06)
Per Capita Income (in $1,000), mean (SD)^a	47.53 (15.84)		47.75 (15.60)
Proportion of No Insurance, mean (SD)^a	0.11 (0.05)		0.11 (0.05)
Enabling Factors
Metropolitan Statistical Area^a	22,426	79.49	22,426	79.49
Health Professional Shortage Area^a	25,677	91.01	25,677	91.01
Census Regions^a
Northeast	5,803	20.57	5,842	20.71
Midwest	6,364	22.56	6,418	22.75
South	12,089	42.85	12,010	42.57
West	3,957	14.03	3,943	13.98
Need Factor
Risk Adjustment Summary Score, mean (SD)	2.41 (1.37)		2.42 (1.22)

All characteristics were not statistically different between the intervention and comparison groups (P > .05).

SD, standard deviation.

^aIndicates a county-level characteristic.

The percentages of nonadherent beneficiaries by race/ethnicity and study year for both study groups are illustrated in Fig 1. In the baseline year 2016, Whites had lower percentages of nonadherence to antidementia medications than Blacks, Asians, Hispanics, and Others in both the comparison group (P <.001 for all race/ethnicities) and the intervention group (P < .001, P < .001, P < .001, P =.0148 respectively). Specifically, within the intervention group, the difference was 8.6 percentage points (22.4%-13.8%) between Blacks and Whites, 7.3 (21.1%-13.8%) between Hispanics and Whites, 3.2 (17.0%-13.8%) between Asians and Whites, and 4.2 (18.0%-13.8%) between Other and Whites. Similarly for the comparison group, the difference was 4.1 percentage points (13.8%-9.7%) between Blacks and Whites, 6.8 (16.5%-9.7%) between Hispanics and Whites, 6.4 (16.1%-9.7%) between Asians and Whites, and 5.0 (14.7%-9.7%) between Other and Whites. By comparison, the intervention group experienced smaller differences between Whites and two racial/ethnic minorities in 2017 (P < .001): the difference was 3.7 percentage points (18.5% versus 14.8%) between Blacks and Whites and 2.9 (17.7% versus 14.8%) between Hispanics and Whites. The comparison group did not see any significant changes concerning racial/ethnic differences in nonadherence across time.

Table 2 first presents the multivariate regression results concerning racial/ethnic disparities in medication nonadherence for both intervention and comparison groups. In 2016, racial/ethnic disparities were found in nonadherence to antidementia medications for both study groups. After adjusting for beneficiary characteristics, racial/ethnic minorities, except for Other in the intervention group, were found to have significantly higher odds of being nonadherent than Whites. Specifically for the intervention group in 2016, the adjusted ORs for Blacks, Hispanics, and Asians were respectively 1.64 (95% confidence interval [CI]: 1.47-1.82), 1.53 (95% CI: 1.35-1.73), and 1.36 (95% CI: 1.16-1.59). Regarding the comparison group, the adjusted ORs for Blacks, Hispanics, Asians, and Other were respectively 1.20 (95% CI: 1.07-1.35), 1.52 (95% CI: 1.29-1.80), 1.74 (95% CI: 1.45-2.09), and 1.45 (95% CI: 1.14-1.84).

Table 2. Racial/ethnic disparity patterns in nonadherence to antidementia medications by intervention status

	Intervention Group			Intervention Group		Comparison Group		Comparison Group
Characteristics	Year 2016			Year 2016 & 2017		Year 2016		Year 2016 & 2017
	OR	95% CI	OR		95% CI	OR	95% CI	OR	95% CI
Predisposing Factors
Race/Ethnicity
Blacks	1.64	1.47 - 1.82	1.64		1.48 - 1.81	1.2	1.07 - 1.35	1.21	1.08 - 1.35
Hispanics	1.53	1.35 - 1.73	1.55		1.38 - 1.74	1.52	1.29 - 1.80	1.58	1.36 - 1.83
Asians/Pacific Islanders	1.36	1.16 - 1.59	1.34		1.16 - 1.55	1.74	1.45 - 2.09	1.77	1.49 - 2.11
Other	1.27	0.97 - 1.65	1.27		0.98 - 1.65	1.45	1.14 - 1.84	1.48	1.17 - 1.87
Year 2017			1.052		0.995 - 1.111			1.43	1.35 - 1.53
Year 2017 × Race/Ethnicity
Blacks			0.73		0.65 - 0.82			1.1	0.97 - 1.24
Hispanics			0.74		0.63 - 0.87			0.91	0.80 - 1.03
Asians/Pacific Islanders			0.88		0.74 - 1.05			0.85	0.68 - 1.06
Other			1.12		0.77 - 1.62			0.79	0.57 - 1.09
Age	0.986	0.981 - 0.990	0.986		0.983 - 0.990	0.981	0.976 - 0.987	0.986	0.982 - 0.990
Male	0.97	0.90 - 1.04	0.945		0.895 - 0.998	0.95	0.88 - 1.04	0.9	0.85 - 0.96
Proportion Married-couple Families^a	0.47	0.27 - 0.82	0.44		0.28 - 0.70	0.39	0.17 - 0.87	0.3	0.16 - 0.56
Proportion Education >= High School^a	1.14	0.49 - 2.68	1.14		0.59 - 2.23	0.81	0.28 - 2.37	1.1	0.48 - 2.53
Per Capita Income (in $1,000)^a	1.002	1.001 - 1.004	1.002		1.001 - 1.003	1.0029	0.9998 - 1.0060	1.0025	1.0004 - 1.0045
Proportion No Insurance^a	5.72	1.94 - 16.88	3.76		1.70 - 8.30	2.9	0.74 -11.38	2.85	1.06 - 7.65
Enabling Factors
Metropolitan Statistical Area^a	0.916	0.838 - 1.003	0.96		0.90 - 1.02	1.02	0.91 - 1.14	0.99	0.91 - 1.07
Health Professional Shortage Area^a	0.97	0.86 - 1.08	0.91		0.84 - 0.99	0.98	0.86 - 1.12	0.91	0.83 - 1.01
Census Regions^a
Midwest	0.997	0.891 - 1.116	0.96		0.88 - 1.05	0.93	0.78 - 1.11	0.99	0.87 - 1.12
South	1.06	0.93 - 1.20	1.04		0.95 - 1.15	1.14	0.95 - 1.36	1.132	0.999 - 1.283
West	0.98	0.87 - 1.10	0.97		0.88 - 1.06	1.08	0.92 - 1.27	1.09	0.97 - 1.22
Need Factor
Risk Adjustment Summary Score	1.21	1.19 - 1.23	1.2		1.19 - 1.22	1.25	1.22 - 1.29	1.25	1.22 - 1.27

Reference groups: non-Hispanic Whites, year 2016, female, non-metropolitan statistical area, non-health professional shortage area, and Northeast region.

Across the two study years, disparities were only reduced for Blacks and Hispanics in the intervention group (Table 2): the ORs of the interaction terms show that differences in the odds of medication nonadherence were significantly lowered by 27% (OR: 0.73, 95% CI: 0.65-0.82) between Blacks and Whites and by 26% (OR: 0.74, 95% CI: 0.63-0.87) between Hispanics-Whites. By comparison, no significant reduction in the differences for Asians-Whites and Other-Whites was observed in the intervention group. Similarly, the disparities between Whites and racial/ethnic minorities in the comparison group were not significantly different across time.

Table 3 exhibits the regression results of the unadjusted and multivariate DID logistic models. The results indicate that MTM enrollment was associated with significantly lower racial/ethnic disparities in the likelihood of Nonadherence for Blacks and Hispanics. In the unadjusted analysis, ORs of the three-way interaction terms of MTM enrollment, the year 2017, and racial/ethnic minority dummies show that differences in the odds of medication nonadherence for Blacks-Whites and for Hispanics-Whites in the intervention group were respectively reduced over time by 35% (OR: 0.65, 95% CI: 0.56-0.76) and 19% (OR: 0.81, 95% CI: 0.67-0.97) more than the comparison group. The reduction in racial/ethnic disparities remained robust after adjusting for beneficiary characteristics: disparities among intervention group were reduced by 33% (OR: 0.67, 95% CI: 0.57-0.78) between Blacks-Whites and by 19% (OR: 0.81, 95% CI: 0.67-0.99) between Hispanics-Whites.

Table 3. The effects of MTM on racial/ethnic disparities in nonadherence to antidementia medications

	Unadjusted		Adjusted
Characteristics	OR	95% CI		OR 95% CI
Predisposing Factors
Race/Ethnicity
Blacks	1.49	1.34 - 1.66	1.24	1.11 - 1.38
Hispanics	1.85	1.54 - 2.22	1.6	1.38 - 1.86
Asians/Pacific Islanders	1.79	1.51 - 2.13	1.82	1.54 - 2.15
Other	1.61	1.27 - 2.05	1.5	1.18 - 1.90
MTM	1.49	1.40 - 1.60	1.48	1.38 - 1.58
Year 2017	1.47	1.38 - 1.56	1.44	1.35 - 1.53
MTM × Year 2017	0.74	0.68 - 0.81	0.73	0.67 - 0.79
MTM × Race/Ethnicity
Blacks	1.21	1.05 - 1.38	1.3	1.13 - 1.49
Hispanics	0.91	0.72 - 1.14	0.96	0.78 - 1.17
Asians/Pacific Islanders	0.71	0.56 - 0.91	0.72	0.57 - 0.91
Other	0.85	0.62 - 1.18	0.84	0.60 - 1.17
Year 2017 × Race/Ethnicity
Blacks	1.11	0.98 - 1.25	1.1	0.97 - 1.24
Hispanics	0.91	0.81 - 1.04	0.91	0.80 - 1.03
Asians/Pacific Islanders	0.85	0.68 - 1.06	0.85	0.68 - 1.06
Other	0.81	0.59 - 1.11	0.79	0.57 - 1.09
MTM × Year 2017 × Race/Ethnicity
Blacks	0.65	0.56 - 0.76	0.67	0.57 - 0.78
Hispanics	0.81	0.67 - 0.97	0.81	0.67 - 0.99
Asians/Pacific Islanders	1.03	0.74 - 1.44	1.04	0.75 - 1.44
Other	1.38	0.89 - 2.13	1.41	0.90 - 2.20
Age			0.986	0.983 - 0.989
Male			0.93	0.89 - 0.96
Proportion Married-Couple Families^a			0.37	0.24 - 0.57
Proportion Education >= High School^a			1.13	0.63 - 2.04
Per Capita Income (in $1,000)^a			1.002	1.001 - 1.004
Proportion No Insurance^a			3.32	1.67 - 6.60
Enabling Factors
Metropolitan Statistical Area^a			0.97	0.92 - 1.03
Health Professional Shortage Area^a			0.91	0.85 - 0.98
Census Regions^a
Midwest			0.97	0.89 - 1.06
South			1.08	0.99 - 1.19
West			1.02	0.94 - 1.11
Need Factor
Risk Adjustment Summary Score			1.22	1.21 - 1.24

Reference groups: non-Hispanic Whites, MTM non-enrollees, the year 2016, female, non-metropolitan statistical area, non-health professional shortage area, and Northeast region.

CI, confidence interval; OR, odds ratio; MTM, medication therapy management.

^a Indicates a county-level characteristic.

Several beneficiary and community characteristics exhibited significant relationships with medication nonadherence (Table 3). Age (OR: 0.986, 95% CI: 0.983-0.989), being male (OR: 0.93, 95% CI: 0.89-0.96), the proportion of married-couple families in the community (OR: 0.37, 95% CI: 0.24-0.57), and living in an HPSA (OR: 0.91, 95% CI: 0.85-0.98) were associated with lower likelihood of medication nonadherence. However, per capita income in the community (OR: 1.002, 95% CI: 1.001-1.004), the proportion of people in the community without health insurance (OR: 3.32, 95% CI: 1.67-6.60), and risk adjustment summary score (OR: 1.22, 95% CI: 1.21-1.24) were associated with a higher likelihood of medication nonadherence.

Our adjusted comparisons in the baseline year indicated that Blacks, Hispanics, and Asians were more likely to be nonadherent than Whites. The findings were consistent with results from the broader ADRD literature that used less restrictive outcome measures to examine racial/ethnic disparities in antidementia medication utilization [6, 19–21]. For example, a recent study analyzed a 20% random sample of Medicare claims for outcomes of different adherence thresholds that ranged up to 270 possession days and two claims in a year. The results showed that Hispanics were less likely to use antidementia medications at higher thresholds while Blacks and Asians were less likely to use the medications at all thresholds compared to Whites [6].

Across the two study years, our adjusted comparisons indicated that the Blacks-Whites and Hispanics-Whites differences were significantly reduced in the second year for the intervention group. In contrast, the gaps between Asians and Whites in the intervention group as well as between all minorities and Whites in the comparison group did not significantly change over time. The DID analysis further showed that the Blacks-Whites and Hispanics-Whites disparities in the intervention group decreased more over time than the comparison group. The findings, therefore, provided evidence that the Medicare MTM program was associated with reduced disparities in antidementia medication nonadherence for older Black and Hispanic patients with ADRD.

Our multivariate analyses revealed that several patient and community characteristics deserve more attention in ADRD-specific interventions. For instance, nonadherence was positively associated with living in a community with a higher uninsured rate and having a higher risk adjustment summary score. The community-level uninsured rate may be a proxy for the affluence of the area and the level of health care access. As noted earlier, the risk adjustment summary score is a proxy for health status. Health care access and disease burden are both well-documented determinants of adherence [35, 36]. Our results, therefore, may highlight the essential role of health care access in reducing medication nonadherence and the importance of timely screening and interventions.

Several other covariates that were also significantly associated with nonadherence included age, male, living in an HPSA, and the per capita income level in a community. Being older and being male both had a negative association with nonadherence, which was consistent with the findings of previous studies. Researchers noted that older age was associated with a reduced risk of nonadherence and that female patients were more likely to be nonadherent than male patients [35, 37, 38]. The gender difference in nonadherence was speculated to stem from women’s caring nature that placed others’ health care needs ahead of the women’s own [37, 38]. Nonadherence was negatively associated with living in an HPSA, likely because patients in these areas again had issues with healthcare access. The counterintuitive positive relationship between per capita income in the community and nonadherence warrants further investigator in the future.

Another community characteristic associated with lower nonadherence is living in a community with a higher proportion of married-couple families, which may be a proxy of a higher level of social support [39]. For ADRD patients, the main sources of social support may come from both caregivers and health care providers. Caregivers play a pivotal role in antidementia medication adherence due to the caregivers’ high level of involvement in ADRD patients’ medication management. A recent Lancet report stressed the need to incorporate family caregivers into ADRD interventions [40]. Taking care of an ADRD patient is stressful due to the patient’s psychiatric and behavioral symptoms [41]. For family caregivers, the stress is compounded by the fact that they not only lack the necessary skills to deal with the patients’ psychiatric and behavioral problems, but also have to juggle with their own work and family responsibilities outside of the caring schedule [42]. Caregiver education focusing on communication, behavioral and stress management skills is therefore essential in ADRD medication management [42–46]. In addition, existing literature suggested that racial/ethnic disparities in antidementia medication use might be due to cultural beliefs of dementia being fateful as a result of God’s will among racial/ethnic minorities [21, 22]. Thus, the cultural competency of frontline MTM service providers, such as pharmacists, is front and center in improving medication adherence across different racial/ethnic populations with ADRD.

Currently, there are not adequate standard procedures for the Medicare MTM services except that Part D plans are required to use a CMS-developed standardized form in comprehensive medication reviews [29]. It is largely at the plans’ discretion how the services are provided. CMS may consider incorporating caregiver and provider education in the MTM program requirements, particularly for challenging diseases such as ADRD. Given the disease’s long-term nature, equipping caregivers and frontline providers with the necessary knowledge and skills not only improves the quality of MTM services but also is a more sustainable and cost-effective approach for ADRD care.

To our best knowledge, this study is the first that examined the effects of the Medicare MTM program on racial/ethnic disparities in adherence to antidementia medications. It bridged the gap in the existing MTM and ADRD literature by providing initial evidence of the Medicare MTM effects on reducing racial/ethnic disparities in adherence for older Black and Hispanic ADRD patients relative to their White counterparts. Methodologically, the study applied rigorous methods by studying 100% Medicare population and using a DID design that incorporates propensity score-matched groups.

The nature of the data and study design warrants a discussion of limitations. First, county-level proxies were used for patient characteristics such as income and education because such variables are not available in Medicare claims and Master Beneficiary Summary Files. While this is a widely-used strategy, due to their aggregated nature, the proxies only approximated the individual characteristics. Second, the widely-accepted PDC measure computed the proportion of days covered by medication fills, not whether a patient actually took the medications in prescribed doses. In essence, PDC was only a proxy for adherence. Third, the comparison group was identified by a patient’s predicted probability of MTM enrollment. The prediction model did not differentiate what conditions and medications a patient had, which probably vary in the actual MTM eligibility criteria of different plans. Therefore, the predicted probability may not reflect the actual likelihood of MTM enrollment. However, because information on individual plans’ eligibility criteria was not available, using the threshold values most frequently used (i.e., three conditions and eight covered drugs) was the best possible option in the propensity score matching.

The Medicare MTM program was associated with reduced racial/ethnic disparities in adherence to antidementia medications between older Blacks and Hispanics and Whites. No effect of the program was detected on the disparities between Asians and Whites. Our results suggested that a lower level of health care access and heavier disease burden may lead to higher nonadherence whereas a higher level of social support may reduce nonadherence. Incorporating caregiver and provider education in the MTM program may facilitate more effective and sustainable service delivery. Future research is warranted to investigate the specific areas of training related to Medicare MTM services caregivers and frontline providers may need so that the MTM program’s benefits may be optimized.

Alzheimer’s disease

ADRD

Alzheimer’s disease and related dementias

AHRF

Area Health Resources Files

Asians

Asians/Pacific Islanders

confidence interval

CMS

Centers for Medicare and Medicaid Services

DID

Difference-in-Differences

FDA

Food and Drug Administration

HSPA

health professional shortage area

ICD-9-CM

International Classification of Disease, Ninth Revision, Clinical Modification

ICD-10-CM

International Classification of Disease, Tenth Revision, Clinical Modification MSA = metropolitan statistical area

MTM

medication therapy management

odds ratio

PDC

proportion of days covered

PQA

Pharmacy Quality Alliance

standard deviation

Whites

non-Hispanic Whites

Ethics approval and consent to participate

The Institutional Review Board at the University of Tennessee Health Science Center US approved the study (approval number: #20-07197-XM). All methods were carried out in accordance with relevant guidelines and regulations. The need for informed consent was waived by the Institutional Review Board of the University of Tennessee Health Science Center US, because of the retrospective nature of the study.

Consent for publication

Not applicable.

Availability of data and materials

Medicare databases are United States federal databases sponsored by the Centers for Medicare & Medicaid Services (CMS). These data are available to researchers through the Research Data Assistance Center (ResDAC) at the University of Minnesota, according to a strict protocol for data requests. Users of Medicare databases cannot disclose to, nor share the data with, individuals not listed in the Data Use Agreement. ResDAC can be reached via email at [email protected], or by phone at 888-973-7322.

Competing interests

XD, CCST, JYW, CKF, JWT, JAB, JG, RZ: None. MAC: Received funding from Carlos and Marguerite Mason Trust. JW: Received funding from AbbVie, Curo, Bristol Myers Squibb, Pfizer, and Pharmaceutical Research and Manufacturers of America (PhRMA), and serves on Health Outcomes Research Advisor Committee of the PhRMA Foundation.

Funding

Research reported in this publication was supported by the National Institute on Aging of the National Institutes of Health under Awards Number R01AG040146. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Authors’ contributions

XD, CCST, JAB, JG, and RZ: concept and design, analysis and interpretation of data, and writing (original draft and editing). JYW, MAC, CKF, and JWT: funding acquisition, concept and design, analysis and interpretation of data, and writing (original draft and editing). JW: funding acquisition, concept and design, analysis and interpretation of data, writing (original draft and editing), and project administration. All authors read and approved the final manuscript.

Acknowledgements

The authors would like to acknowledge the following individuals for assistance with the formatting of the final manuscript: Lorraine Todor, Oluwasefunmi Agbanigo, Hannah Foster, all PharmD students, and Yongbo Sim, a postdoctoral scholar, at the University of Tennessee Health Science Center College of Pharmacy.

National Institute of Neurological Disorders and Stroke. Focus on Alzheimer’s disease and related dementias. Bethesda, MD: National Institute of Neurological Disorders and Stroke. 2021. https://www.ninds.nih.gov/Current-Research/Focus-Disorders/Alzheimers-Related-Dementias. Accessed 12 September 2022.
Alzheimer's disease facts and figures. Alzheimers dement. 2021;17(3):327-406.
Food and Drug Administration. Food and Drug Administration FDA grants accelerated approval for Alzheimer’s drug. Silver Spring, MD: Food and Drug Administration. 2021. https://www.fda.gov/news-events/press-announcements/fda-grants-accelerated-approval-alzheimers-drug. Accessed 12 September 2022.
Fowler NR, Chen YF, Thurton CA, Men A, Rodriguez EG, Donohue JM: The impact of Medicare prescription drug coverage on the use of antidementia drugs. BMC Geriatr. 2013;13:37.
Koller D, Hua T, Bynum JP. Treatment patterns with antidementia drugs in the United States: Medicare cohort study. J Am Geriatr Soc. 2016;64(8):1540-1548.
Barthold D, Joyce G, Ferido P, Drabo EF, Marcum ZA, Gray SL, et al. Pharmaceutical treatment for Alzheimer's disease and related dementias: Utilization and Disparities. J Alzheimers Dis. 2020;76(2):579-589.
Smith D, Lovell J, Weller C, Kennedy B, Winbolt M, Young C, et al. A systematic review of medication non-adherence in persons with dementia or cognitive impairment. PLoS One. 2017;12(2):e0170651.
Schwalbe O, Scheerans C, Freiberg I, Schmidt-Pokrzywniak A, Stang A, Kloft C. Compliance assessment of ambulatory Alzheimer patients to aid therapeutic decisions by healthcare professionals. BMC Health Serv Res. 2010;10:232.
Zhang P, Chen L, Gu W, Xu Z, Gao Y, Li Y. In vitro and in vivo evaluation of donepezil-sustained release microparticles for the treatment of Alzheimer's disease. Biomaterials. 2007;28(10):1882-1888.
Borah B, Sacco P, Zarotsky V. Predictors of adherence among Alzheimer's disease patients receiving oral therapy. Curr Med Res Opin. 2010;26(8):1957-1965.
Medicare program; Contract Year 2015 Policy and technical changes to the Medicare Advantage and the Medicare Prescription Drug Benefit programs. Final rule. Fed Regist. 2014;79(100):29843-29968.
Centers for Medicare & Medicaid Services. Medication therapy management. Baltimore, MD: Centers for Medicare & Medicaid Services. 2021. https://www.cms.gov/Medicare/Prescription-Drug-Coverage/PrescriptionDrugCovContra/MTM. Accessed 12 September 2022.
Perlroth D, Marrufo G, Montesinos A, Olinger L, Fair S, Mercincavage L, et al. Medication therapy management in chronically ill populations: Final report. Baltimore Maryland: Acumen LLC; 2013. 242 p.
Peasah SK, Hammonds T, Liu Y, Campbell V, Manolis C, Good CB. Economic assessment of changes to an existing medication therapy management program of a large regional health plan. J Manag Care Spec Pharm. 2021;27(2):147-156.
Dong X, Tsang CCS, Zhao S, Browning JA, Wan JY, Chisholm-Burns MA, et al. Effects of the Medicare Part D comprehensive medication review on medication adherence among patients with Alzheimer's disease. Curr Med Res Opin. 2021;37(9):1581-1588.
Ai AL, Carretta H, Beitsch LM, Watson L, Munn J, Mehriary S. Medication therapy management programs: promises and pitfalls. J Manag Care Spec Pharm. 2014;20(12):1162-1182.
Viswanathan M, Kahwati LC, Golin CE, Blalock SJ, Coker-Schwimmer E, Posey R, et al. Medication therapy management interventions in outpatient settings: a systematic review and meta-analysis. JAMA Intern Med. 2015;175(1):76-87.
Pestka DL, Zillich AJ, Coe AB, Farris KB, Adeoye OA, Snyder ME, et al. Nationwide estimates of medication therapy management delivery under the Medicare prescription drug benefit. J Am Pharm Assoc (2003). 2020;60(3):456-461.
Gruber-Baldini AL, Stuart B, Zuckerman IH, Simoni-Wastila L, Miller R. Treatment of dementia in community-dwelling and institutionalized Medicare beneficiaries. J Am Geriatr Soc. 2007;55(10):1508-1516.
Mehta KM, Yin M, Resendez C, Yaffe K. Ethnic differences in acetylcholinesterase inhibitor use for Alzheimer disease. Neurology. 2005;65(1):159-162.
Zuckerman IH, Ryder PT, Simoni-Wastila L, Shaffer T, Sato M, Zhao L, et al. Racial and ethnic disparities in the treatment of dementia among Medicare beneficiaries. J Gerontol B Psychol Sci Soc Sci. 2008;63(5):S328-333.
Thorpe CT, Fowler NR, Harrigan K, Zhao X, Kang Y, Hanlon JT, et al. Racial and ethnic differences in initiation and discontinuation of antidementia drugs by Medicare beneficiaries. J Am Geriatr Soc. 2016;64(9):1806-1814.
Research Data Assistance Center. Data file search. Minneapolis, MN: Research Data Assistance Center. 2021. https://www.resdac.org/cms-data?tid_1%5B%5D=1. Accessed 12 September, 2022.
Health Resources & Services Administration. Area Health Resources Files. Rockville, MD: Health Resources & Services Administration. 2021. https://data.hrsa.gov/topics/health-workforce/ahrf. Accessed 12 September 2022.
Pharmacy Quality Alliance. Technical specifications for PQA-endorsed measures: 2018. Alexandria, VA: Pharmacy Quality Alliance; 2018.
Chronic Conditions Data Warehouse. Condition categories. Baltimore MD: Centers for Medicare and Medicaid Services. 2021. https://www2.ccwdata.org/web/guest/condition-categories. Accessed 12 September 2022.
Pharmacy Quality Alliance. About PQA. Alexandria, VA: Pharmacy Quality Alliance. 2021. https://www.pqaalliance.org/our-story. Accessed 12 September 2022.
Centers for Medicare & Medicaid Services. CY 2016 medication therapy management program guidance and submission instructions. Baltimore MD: Centers for Medicare and Medicaid Services. 2015. https://www.cms.gov/Medicare/Prescription-Drug-Coverage/PrescriptionDrugCovContra/Downloads/Memo-Contract-Year-2016-Medication-Therapy-Management-MTM-Program-Submission-v-040715.pdf. Accessed 12 September 2022.
Centers for Medicare & Medicaid Services. CY 2017 medication therapy management program guidance and submission instructions. Baltimore MD: Centers for Medicare and Medicaid Services. 2016. https://www.cms.gov/Medicare/Prescription-Drug-Coverage/PrescriptionDrugCovContra/Downloads/Memo-Contract-Year-2017-Medication-Therapy-Management-MTM-Program-Submission-v-040816.pdf. Accessed 12 September 2022.
Centers for Medicare & Medicaid Services. 2016 Medicare Part D medication therapy management (MTM) programs: Fact sheet. Baltimore MD: Centers for Medicare and Medicaid Services. 2016. https://www.cms.gov/Medicare/Prescription-Drug-Coverage/PrescriptionDrugCovContra/Downloads/CY2016-MTM-Fact-Sheet.pdf. Accessed 12 September 2022.
Centers for Medicare & Medicaid Services. 2017 Medicare Part D medication therapy management (MTM) programs: Fact sheet. Baltimore MD: Centers for Medicare and Medicaid Services. 2017. https://www.cms.gov/Medicare/Prescription-Drug-Coverage/PrescriptionDrugCovContra/Downloads/CY2017-MTM-Fact-sheet.pdf. Accessed 12 September 2022.
Daniel GW, Malone DC. Characteristics of older adults who meet the annual prescription drug expenditure threshold for Medicare medication therapy management programs. J Manag Care Pharm. 2007;13(2):142-154.
Austin PC. A comparison of 12 algorithms for matching on the propensity score. Stat Med. 2014;33(6):1057-1069.
Gelberg L, Andersen RM, Leake BD. The behavioral model for vulnerable populations: application to medical care use and outcomes for homeless people. Health Serv Res. 2000;34(6):1273-1302.
Briesacher BA, Gurwitz JH, Soumerai SB. Patients at-risk for cost-related medication nonadherence: a review of the literature. J Gen Intern Med. 2007;22(6):864-871.
Hughes CM. Medication non-adherence in the elderly: how big is the problem. Drugs Aging. 2004;21(12):793-811.
Puskas CM, Forrest JI, Parashar S, Salters KA, Cescon AM, Kaida A, et al. Women and vulnerability to HAART non-adherence: a literature review of treatment adherence by gender from 2000 to 2011. Curr HIV/AIDS Rep. 2011;8(4):277-287.
Vimalananda VG, Miller DR, Hofer TP, Holleman RG, Klamerus ML, Kerr EA. Accounting for clinical action reduces estimates of gender disparities in lipid management for diabetic veterans. J Gen Intern Med. 2013;28 Suppl 2:S529-535.
Oxman TE, Berkman LF, Kasl S, Freeman DH, Barrett J. Social support and depressive symptoms in the elderly. Am J Epidemiol. 1992;135(4):356-368.
Livingston G, Huntley J, Sommerlad A, Ames D, Ballard C, Banerjee S, et al. Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. Lancet. 2020;396(10248):413-446.
Pasqualetti G, Tognini S, Calsolaro V, Polini A, Monzani F. Potential drug-drug interactions in Alzheimer patients with behavioral symptoms. Clin Interv Aging. 2015;10:1457-1466.
Boada M, Arranz FJ. Transdermal is better than oral: observational research of the satisfaction of caregivers of patients with Alzheimer's disease treated with rivastigmine. Dement Geriatr Cogn Disord. 2013;35(1-2):23-33.
Brady R, Weinman J. Adherence to cholinesterase inhibitors in Alzheimer's disease: a review. Dement Geriatr Cogn Disord. 2013;35(5-6):351-363.
Callahan CM, Boustani MA, Unverzagt FW, Austrom MG, Damush TM, Perkins AJ, et al. Effectiveness of collaborative care for older adults with Alzheimer disease in primary care: a randomized controlled trial. JAMA. 2006;295(18):2148-2157.
Deudon A, Maubourguet N, Gervais X, Leone E, Brocker P, Carcaillon L, et al. Non-pharmacological management of behavioural symptoms in nursing homes. Int J Geriatr Psychiatry. 2009;24(12):1386-1395.
Nguyen K, Hoffman H, Chakkamparambil B, Grossberg GT. Evaluation of rivastigmine in Alzheimer's disease. Neurodegener Dis Manag. 2021;11(1):35-48.

Competing interest reported. XD, CCST, JYW, CKF, JWT, JAB, JG, RZ: None. MAC: Received funding from Carlos and Marguerite Mason Trust. JW: Received funding from AbbVie, Curo, Bristol Myers Squibb, Pfizer, and Pharmaceutical Research and Manufacturers of America (PhRMA), and serves on Health Outcomes Research Advisor Committee of the PhRMA Foundation.

Download PDF

Journal Publication

published 01 Mar, 2024

Read the published version in Exploratory Research in Clinical and Social Pharmacy →

Version 1

posted

You are reading this latest preprint version

Effects of Medicare Part D medication therapy management on racial/ethnic disparities in adherence to antidementia medications among patients with Alzheimer’s disease and related dementias: An observational study

Status:

Journal Publication

Version 1

Abstract

Background

Methods

Results

Conclusions

Figures

Background

Methods

Data Source

Study sample

Study groups construction

Outcome measure

Covariates

Empirical approach and statistical analysis

Results

Discussion

Conclusions

Abbreviations

Declarations

References

Additional Declarations

Status:

Journal Publication

Version 1