Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Mar 1.
Published in final edited form as: J Am Pharm Assoc (2003). 2024 Jan 3;64(2):463–470.e2. doi: 10.1016/j.japh.2023.12.023

The Effects of Medication Therapy Management on Medication Utilization among Medicare Beneficiaries with HIV/AIDS

Xiangjun Zhang 1, Junling Wang 2
PMCID: PMC10942721  NIHMSID: NIHMS1964831  PMID: 38182001

Abstract

Background

Medicare Part D plans are required to provide Medication therapy management (MTM) programs to eligible beneficiaries to optimize medication utilization. MTM programs’ effects on medication utilization among older persons living with HIV/AIDS (PLWHs) remain unclear.

Objective

This study examined the effects of the Medicare MTM programs on medication utilization among PLWHs.

Methods

This study analyzed 2017 Medicare databases linked to the Area Health Resources Files. Recipients and non-recipients of the MTM services were compared on their medication utilization: adherence to antiretroviral medications, drug-drug interactions, and concurrent use of opioids and benzodiazepines. The intervention group comprised recipients of the medication therapy management services, and the control group was non-recipients meeting the eligibility criteria. A propensity score with a ratio of 1:2 between the intervention and control groups was used to identify study groups with balanced characteristics. A logistic regression was used to control for patient/community characteristics.

Results

After matching, the intervention and comparison groups comprised 3,298 and 6,596 beneficiaries for the antiretroviral adherence measure, 809 and 1,618 for the drug-drug interactions measure, and 691 and 1,382 for the concurrent use of opioids and benzodiazepines measure. The intervention was associated with higher odds of adherence to antiretroviral medications (adjusted odds ratio = 1.15, 95% Confidence Interval = 1.04 – 1.26), and no concurrent use of opioids and benzodiazepines (adjusted odds ratio = 1.255, 95% Confidence Interval = 1.005 – 1.568). The study groups did not differ on no drug-drug interactions (adjusted odds ratio = 0.95, 95% Confidence Interval = 0.74 – 1.20).

Conclusions

Medicare MTM programs effectively improved medication utilization among people living with HIV/AIDS. Future studies should explore the long-term effects of the program.

Keywords: HIV, Medication Therapy Management, Highly Active Antiretroviral Therapy, Drug Interactions, Opioids, Benzodiazepines

Introduction

HIV/AIDS exerts a tremendous disease burden in the United States in general and for Medicare in particular. Approximately 1.2 million people were living with HIV/AIDS in the U.S., with 36,136 new infections in 2021.1 By 2020, the number of people living with HIV/AIDS (PLWHs) covered by traditional Medicare plans has more than doubled since 1997.2 As a result, Medicare covered approximately 28% of PLWHs in 2020.2 As of 2020, about 77% Medicare beneficiaries with HIV/AIDS initially qualified for Medicare through the disability pathway, among whom, some qualified due to end-stage renal disease and others amyotrophic lateral sclerosis.2 The proportion of PLWHs who qualified for Medicare based on age increased from 14% in 2015 to 23% in 2020 due to their increased lifespan thanks to effective HIV treatment.2 As the second largest government payer after Medicaid, Medicare pays a hefty price tag for HIV: Medicare spent $11.3 billion for HIV treatment and care in the fiscal year 2022, accounting for 39% of the total federal HIV care expenditure.2 Covering essential services, including preventive services, prescription medications, physician services, and inpatient and outpatient care, Medicare plays a significant role in HIV prevention and care.2

Advances in antiretroviral therapy have contributed to lower mortality and an extended lifespan; however, some chronic conditions are now more prevalent among older PLWHs.3,4 For example, according to a previous study that used a Medicare sample of 1996–2015, hypertension prevalence in older men and women with HIV was 72.0% and 83.1%, respectively, much higher than 61.0% and 71.6% among their HIV-negative counterparts.4 Further, older PLWHs had 12.7 and 1.3 times the odds of having chronic hepatitis and hypertension than their HIV-negative counterparts, according to a previous study that investigated a 2016 Medicare sample.5 Adherence to antiretroviral medications is necessary to achieve and maintain viral suppression. Thus, antiretroviral medication adherence is the key to the success of HIV treatment.

Although generally more adherent to antiretroviral medications than their younger peers, older PLWHs face many challenges in their daily medication management, such as cognitive impairment, mental health issues, and lack of social support.6 Further, many older PLWHs experience medication issues, including polypharmacy, drug-drug interactions (DDI), and adverse drug events due to the high burden of non-HIV chronic conditions.6 For instance, HIV-induced chronic inflammation and peripheral neuropathy can cause chronic pain, so chronic pain is prevalent among PLWHs: previous studies estimated this prevalence at 25% – 83%.7,8 To treat chronic pain, long-term opioids are more frequently prescribed to PLWHs than to people without HIV.9 Since the long-term use of benzodiazepine is ubiquitous among older PLWHs, concurrent use of opioids and benzodiazepines (COB) is also a legitimate concern among this population.10 This is also concerning because people who receive opioids and benzodiazepines concurrently show an increased risk of an emergency room visit or hospital admission for opioid overdose than those prescribed opioids but not benzodiazepines.11 Overdose death rates among individuals prescribed opioids and benzodiazepine concurrently were reported as ten times higher than the ones who received opioid prescriptions alone.12

To optimize medication therapeutic outcomes and reduce the risk of adverse drug events, the Centers for Medicare and Medicaid Services (CMS) initiated the medication therapy management (MTM) program in 2006 and has required all Part D plans to implement the program.13 Based on CMS guidelines, plans set their eligibility thresholds for MTM programs.14 For example, the majority of plans applied the MTM’s general eligibility criteria to enroll beneficiaries who had at least 3 chronic conditions, took 8 or more Part D-covered medications, and incurred a minimum annual medication expense of $3,919 in 2017.14,15 With the core MTM service, comprehensive medication review, medication consultations were provided to patients on a yearly basis. During these interactive consultations with patients, pharmacists or other qualified providers systematically assess medication therapies and identify medication-related issues.14 A written summary in a standardized format with a recommended medication action plan is provided to patients and prescribers after the consultations.14

Previous research revealed that MTM receipt was associated with improved medication adherence, lower healthcare expenses, and a racial/ethnic disparity reduction in nonadherence to medications.16–18 However, MTM effects on medication utilization among older PLWHs remain unclear. This study examined whether receiving MTM services was associated with improved medication utilization among older PLWHs in Medicare. This study’s medication utilization measures included adherence to antiretroviral medications, drug-drug interactions, and concurrent use of opioids and benzodiazepines. By providing empirical evidence of MTM’s effectiveness on medication utilization, this study may guide policy changes regarding the expansion of MTM among older PLWHs in Medicare.

Methods

Data sources and sample selection

This cross-sectional study analyzed 2017 Medicare data linked to the Area Health Resources Files.19,20 The Medicare databases used were the Master Beneficiary Summary File, Part D MTM Data File, Part D Event File, and Parts A and B claims.19 Beneficiaries’ information was derived from the Medicare databases, including demographic characteristics, chronic disease diagnosis, prescription utilization information, and MTM enrollment status. Further, community-level characteristics of interest were obtained from the Area Health Resources Files.20

Inclusion criteria were: (1) aged 65 years and above in 2017; (2) were alive in 2017; (3) had continuous coverage of Medicare Parts A, B, and D in 2017; (4) used antiretroviral medications in 2017.

Conceptual framework

Andersen’s Behavioral Model of Health Services Use was applied as a conceptual model in selecting factors associated with the study outcome, which were all measures of medication utilization.21 This model categorized factors determining individuals’ health service utilization into predisposing, enabling, and need factors.21 This study included individual-level and community-level sociodemographic characteristics as predisposing factors. The predisposing factors at the individual level included age, gender, and race/ethnicity. The community-level factors, including the percentage of married-couple families, the percentage of people with high school and above degrees, the per capita income, and the percentage of people without health insurance in a community, were also considered as predisposing factors. The enabling factors referred to metropolitan statistical area, Health Professional Shortage Area, and census regions in this study. These factors reflect resources that are necessary for individuals to access health care. The study analysis included a risk adjustment summary score as a need factor accounting for patients’ health status. CMS originally developed the risk adjustment summary score to predict beneficiaries’ healthcare expenses using their demographic and diagnostic information.22 Since the risk adjustment summary score calculation covered patients’ major medical conditions, additional diagnoses were not included as need factors.23 An individual with a higher score was projected to incur higher healthcare costs due to worse health status.

Intervention and comparison groups

MTM recipients in 2017 consisted of the intervention group of the study. The comparison group comprised beneficiaries eligible for MTM but were not enrolled in the program in 2017. According to CMS guidelines, the MTM eligibility thresholds for 2017 were 3 chronic diseases, a minimum of 8 Part D-covered medications, and incurring at least $3,919 for Part D-covered medications in 2017.14 Although these were established as minimum thresholds by CMS, most Part D plans adopted these thresholds.15 For example, over 84% of Part D plans adopted 3 chronic diseases, and 66% of programs set 8 Part D-covered medications as their thresholds to enroll beneficiaries in the MTM program in 2017.15 Therefore, beneficiaries who exceeded these thresholds, including 3 chronic diseases, 8 Part D-covered medications, and an annual cost of $3,919 for medications, but were not enrolled in the MTM program, were included in the comparison group.

Regarding data sources for MTM eligibility determination, prescription records and annual medication costs of eligible beneficiaries were obtained from the 2017 Part D Event File.24 Patients’ medical conditions were determined based on beneficiaries’ Parts A and B claims data using the ICD-10 diagnosis codes of a list of 25 chronic diseases for Medicare beneficiaries in a previous study.25 A comparable intervention and a comparison group were identified using a propensity score matching strategy to ensure balanced characteristics between the two groups. Beneficiaries were matched on their predicted likelihood of receiving MTM in a logistic regression, including the variables suggested by the conceptual framework. The nearest-neighbor matching mechanism without replacement was applied to obtain the most balanced groups.26 Three propensity score matchings were performed, one for each measure. Based on the sample size differences between the intervention group and the comparison group before matching, a ratio of 1:2 between the intervention group and the comparison group was applied to all 3 measures to achieve the best matching.

Study outcomes

The study outcome was medication utilization, including adherence to antiretroviral medications, no DDI, and no COB. The 3 measures were developed by the Pharmacy Quality Alliance, a national quality organization dedicated to advancing medication utilization by developing performance measures for Medicare Part D.27,28 Adherence to antiretroviral medications was defined as greater than or equal to 90% of the proportion of days covered by the prescriptions of interest.27,28 The proportion of days covered was determined by the number of days covered by one or more prescriptions divided by the total number of days starting from the first prescription date to the end of the year.27,28 The medication safety measure DDI was applied in detecting beneficiaries’ concurrent use of two medications not recommended by pharmacology or practice due to adverse effects or unfavorable interactions. Further, COB was defined as 30 or more cumulative days of overlapping supply days for an opioid and a benzodiazepine during the measurement year.28

Statistical analysis

The Medicare databases were accessed through the CMS Virtual Research Data Center desktop, and its associated SAS software (9.4, Cary, NC) was used to perform all the data analyses. The Institutional Review Board at the corresponding author’s institution granted the exempt status of the study on February 20, 2020 (approval number 20–07197-XM).

Beneficiaries’ characteristics were compared between the intervention and comparison groups before and after propensity score matching for all 3 medication utilization measures. Chi-square tests were performed to compare all categorical variables, and t-tests were performed for continuous variables. Further, Chi-square tests were conducted to compare the percentages of beneficiaries who appropriately used medications regarding the 3 measures between the intervention and comparison groups. A 2-tailed alpha value of 0.05 or below determined the statistical significance.

MTM effects on medication utilization were assessed in the adjusted multivariate analysis incorporating all factors defined by the conceptual model. The estimated coefficient of MTM receipt captures the MTM effects on a beneficiary’s medication utilization. An odds ratio over one would indicate that MTM receipt was associated with improved adherence to antiretroviral medications and safety. The observations from the same county were clustered to account for their potential correlations.

Results

For the antiretroviral adherence measure, 3,659 and 13,105 beneficiaries were identified for the intervention and comparison groups before the matching. After matching with a ratio of 1:2, 3,298 beneficiaries remained in the intervention group and 6,596 in the comparison group, with a total sample size of 9,894. Table 1 presents beneficiaries’ characteristics by study group before and after the matching for the antiretroviral adherence measure. Before the matching, the intervention group exhibited a lower proportion of male beneficiaries, a lower proportion of the uninsured at the county level, and a higher mean risk adjustment summary score than the comparison group (p < .01). The geographic distributions also were different between the two groups (p < .01). After the matching, the intervention and comparison groups did not differ significantly for any characteristics (all p > .05).

Table 1.

Baseline Characteristics for Intervention and Comparison Groups Before and After Propensity Score Matching among Medicare Beneficiaries Who Used Antiretroviral Medications in 2017

Characteristics Before Matching After Matching
Intervention Comparison Intervention Comparison
(n=3,659) (n=13,105) (n=3,298, 33.3%) (n=6,596, 66.7%)
Number % Number % Number % Number %
Predisposing Factors
 Age, mean (SD) 70.18 (4.56) 70.28 (4.69) 70.25 (4.62) 70.14 (4.64)
 Maleb 2,659 72.67 9,880 75.39 2,450 74.29 4,900 74.29
 Race/Ethnicity
  Non-Hispanic Whites 1,698 46.41 6,246 47.66 1,569 47.57 3,138 47.57
  Blacks (African Americans) 1,322 36.13 4,740 36.17 1,242 37.66 2,484 37.66
  Hispanics 471 12.87 1,538 11.74 372 11.28 744 11.28
  Asian/Pacific Islanders 62 1.69 175 1.34 37 1.12 74 1.12
  Other 106 2.90 406 3.10 78 2.37 156 2.37
 Proportion of Married-Couple Families, mean (SD)a 0.68 (0.09) 0.68 (0.09) 0.68 (0.09) 0.68 (0.09)
 Proportion of High School Education and Above, mean (SD)a 0.86 (0.05) 0.86 (0.05) 0.86 (0.05) 0.86 (0.05)
 Income Per Capita (in $1,000), mean (SD)a 59.77 (29.53) 59.84 (29.90) 60.32 (30.08) 59.65 (29.64)
 Proportion of the Uninsured, mean (SD)ab 0.097 (0.046) 0.104 (0.048) 0.098 (0.046) 0.097 (0.044)
Enabling Factors
 Metropolitan Statistical Areaa 3,402 92.98 12,155 92.75 3,104 94.12 6,208 94.12
 Health Professional Shortage Areaa 3,531 96.50 12,589 96.06 3,226 97.82 6,452 97.82
 Census Regionsab
  Northeast 1,192 32.58 3,421 26.10 1,023 31.02 2,046 31.02
  Midwest 395 10.80 1,378 10.52 347 10.52 694 10.52
  South 1,305 35.67 5,478 41.80 1,204 36.51 2,408 36.51
  West 767 20.96 2,828 21.58 724 21.95 1,448 21.95
Need Factor
 Risk Adjustment Summary Score, mean (SD)b 2.79 (1.87) 2.23 (1.56) 2.54 (1.55) 2.49 (1.60)
Open in a new tab

SD = standard deviation.

a

County-level characteristics.

b

Characteristic significantly differed between the intervention and comparison groups before matching (P < .01).

All characteristics were similar between the intervention and comparison groups after matching (P > .05).

Before the matching, the intervention and comparison groups comprised 908 and 2,670 beneficiaries who met the DDI measure criteria among beneficiaries prescribed antiretroviral medications in 2017 (Appendix Table 1). After matching with a 1:2 ratio, the intervention group comprised 809 beneficiaries, and the comparison group included 1,618 beneficiaries (Appendix Table 1). The final study sample for this measure was 2,427 beneficiaries. Three characteristics exhibited significant differences between the intervention group and the comparison group before the matching: proportion of the uninsured at the county level, census regions, and risk adjustment summary score (p < .05). After the matching, the two groups did not show significant differences among all characteristics (p > .05).

A total of 802 beneficiaries who met the COB measure criteria among those who used antiretroviral medications were identified for the intervention group and 2,471 for the comparison group before the matching (Appendix Table 2). After matching with a ratio of 1:2, the final sample comprised 2,073 beneficiaries; the intervention group included 691 beneficiaries, and the comparison group consisted of 1,382 beneficiaries (Appendix Table 2). The intervention group and the comparison group differed in risk adjustment summary score and census regions before the matching (p < .05). However, after the matching, the intervention and comparison groups did not show significant differences for any characteristics (p > .05).

Figure 1 illustrates the proportions of beneficiaries with appropriate use of medications based on the unadjusted analysis. The proportion of beneficiaries who adhered to antiretroviral medications in the intervention group was 75.44%, significantly higher than the comparison group (73.09%, p = .01). The proportions of beneficiaries who did not have DDI issues were similar between the two groups (85.78% for intervention and 86.59% for comparison, p > .05). Further, the intervention group exhibited a significantly higher rate than the comparison group regarding the proportion of beneficiaries who did not use opioids and benzodiazepines concurrently (78.87% for intervention and 74.38% for comparison, p = .02).

Figure 1.

Figure 1

Open in a new tab

Proportions of Beneficiaries Who Safely Used Medications in Intervention and Comparison Groups among People Living with HIV/AIDS in Medicare

The adjusted analysis demonstrated that MTM receipt was associated with higher odds of adherence to antiretroviral medications and safe medication utilization regarding COB (Table 2). In contrast, MTM recipients and non-recipients did not differ in their odds regarding DDI. The adjusted odds ratio for MTM receipt in adherence to antiretroviral medications was 1.15 (95% Confidence Interval or CI = 1.04 – 1.26). This result indicated that the odds of adherence to antiretroviral medications among MTM recipients were 15% higher than among MTM non-recipients. Regarding the DDI measure, MTM recipients and non-recipients did not exhibit significant differences in their medication utilization with an adjusted odds ratio of 0.95 (95% CI = 0.74 – 1.20). For the COB measure, the adjusted odds ratio of 1.255 indicated that the odds of safe medication utilization for MTM recipients were 25.5% higher than for MTM non-recipients (95% CI = 1.005 – 1.568).

Table 2.

MTM Effects on the Medication Utilization among Medicare Beneficiaries Who Used Antiretroviral Medications in 2017

Characteristics Adherence to Antiretroviral Medications Safe Use Regarding Drug-Drug Interactions Safe Use Regarding Concurrent Use of Opioids and Benzodiazepines
(n = 9,894) (n = 2,427) (n = 2,073)
Odds 95% Odds 95% Odds 95%
Ratio Confidence Ratio Confidence Ratio Confidence
MTM Receipt 1.15 1.04 – 1.26 0.95 0.74 – 1.20 1.255 1.005 – 1.568
Predisposing Factors
 Age 0.998 0.988 – 1.008 1.030 1.004 – 1.057 1.04 1.01 – 1.06
 Male 1.22 1.08 – 1.37 1.22 0.94 – 1.58 1.30 1.06 – 1.60
 Race/Ethnicity
  Blacks (African Americans) 0.61 0.54 – 0.69 1.45 1.11 – 1.89 2.26 1.79 – 2.84
  Hispanics 0.77 0.66 – 0.89 1.35 0.84 – 2.17 1.31 0.92 – 1.86
  Asian/Pacific Islanders 0.70 0.47 – 1.06 0.88 0.28 – 2.77 1.29 0.48 – 3.44
  Other 0.88 0.63 – 1.21 1.39 0.66 – 2.96 1.53 0.61 – 3.84
 Proportion of Married-Couple Familiesa 1.44 0.62 – 3.35 0.53 0.08 – 3.38 0.23 0.03 – 1.72
 Proportion of High School Education and Abovea 0.36 0.06 – 2.26 6.23 0.17 – 224.39 2.18 0.08 – 56.69
 Income Per Capita (in $1,000)a 1.0001 0.9979 – 1.0022 0.997 0.993 – 1.001 0.999 0.994 – 1.005
 Proportion of the Uninsureda 1.30 0.20 – 8.46 0.50 0.01 – 22.23 0.31 0.01 – 16.56
Enabling Factors
 Metropolitan Statistical Areaa 1.17 0.94 – 1.45 1.07 0.65 – 1.76 0.72 0.46 – 1.12
 Health Professional Shortage Areaa 1.31 0.99 – 1.72 0.60 0.27 – 1.32 0.96 0.51 – 1.82
 Census Regionsa
  Midwest 1.11 0.92 – 1.33 1.38 0.88 – 2.16 0.77 0.51 – 1.15
  South 1.07 0.90 – 1.27 1.26 0.84 – 1.90 1.06 0.73 – 1.54
  West 1.12 0.95 – 1.31 1.11 0.77 – 1.60 0.92 0.64 – 1.34
Need Factor
 Risk Adjustment Summary Score 0.86 0.84 – 0.89 0.82 0.77 – 0.89 0.936 0.879 – 0.998
Open in a new tab

MTM = Medication Therapy Management.

a

County-level characteristics.

Reference groups: comparison group (MTM non-recipients), female, non-Hispanic White, non-Metropolitan Statistical Area, non-Health Professional Shortage Area, and Northeast region.

A few other factors in the adjusted model also were significantly associated with the study outcomes. The odds of adherence to antiretroviral medications in males were 22% higher than in females (OR = 1.22, 95% CI = 1.08 – 1.37). Black race, Hispanic ethnicity, and risk adjustment summary score were negatively associated with adherence to antiretroviral medications. For instance, Black and Hispanic patients had 39% and 23% lower odds of medication adherence than Non-Hispanic Whites (Whites), respectively (for Black: OR = 0.61, 95% CI = 0.54 – 0.69; for Hispanic: OR = 0.77, 95% CI = 0.66 – 0.89). In terms of the DDI measure, older age was associated with higher odds of appropriate medication utilization (OR = 1.030, 95% CI = 1.004 – 1.057), and Black beneficiaries had 45% higher odds of proper medication utilization compared to White beneficiaries (OR = 1.45, 95% CI = 1.11 – 1.89). An elevated risk adjustment summary score was associated with lower odds of safe medication utilization regarding DDI (OR = 0.82, 95% CI = 0.77 – 0.89). For the COB measure, older age, male gender, and Black race exhibited higher odds of safe medication utilization. For example, Black beneficiaries had 126% higher odds of safe medication utilization regarding COB compared to White beneficiaries (OR = 2.26, 95% CI = 1.79 – 2.84). A higher risk adjustment summary score was related to lower odds of safe medication utilization regarding COB (OR = 0.936, 95% CI = 0.879 – 0.998).

Discussion

This study analyzed 2017 Medicare databases to assess the effects of MTM among PLWHs in Medicare. MTM recipients exhibited a higher likelihood of adherence to antiretroviral medications and a lower chance of COB among older beneficiaries with HIV/AIDS in this study. In this study, MTM recipients and non-recipients of older PLWHs did not exhibit significant differences in DDI.

Previous studies found that MTM participants reported more significant improvements in medication adherence or higher reduction in medication nonadherence to diabetes, hypertension, and hyperlipidemia medications than the comparison groups.16,29 The study results are consistent with previous literature regarding the effects of MTM in improving medication compliance.

The Pharmacy Quality Alliance DDI measure was developed as a population-level performance indicator rather than one that guides clinical assessment for a particular situation.30 This measure is not an all-inclusive list specified by the Pharmacy Quality Alliance and only includes a small portion of antiretroviral medications.30 For example, only one antiretroviral medication was included as a target medication, representing one class of seven available antiretroviral medication classes.28,31 Therefore, the results relative to DDI in this study were not comparable with previous literature, which mainly investigated DDI of antiretroviral medications.32,33 MTM effects relative to COB remained unknown in the existing literature.

The prevalence of antiretroviral medication adherence in this study was high compared to multiple previous studies among older PLWHs, which reported a range of 32% to 53%.34,35 It could be explained by the differences in populations studied, information sources, and measure identification methods. The prevalence of DDI in this study falls in the range of 7% to 70% reported in some previous studies for PLWHs.33,36 For COB, the prevalence findings in this study could not be compared with previous literature due to the lack of comparable results. Given the medication utilization challenges older PLWHs face, effective MTM services may be particularly beneficial in addressing medication-related issues and improving medication utilization for this population.

Although this study and previous literature attested to MTM’s effectiveness on medication utilization, the generally low MTM enrollment rate prevented MTM from achieving its full potential. For example, the average MTM enrollment rate was 10% among Medicare Advantage drug plans and stand-alone prescription drug plans in 2012.37 The generally low MTM enrollment rate may be due to plans’ restrictive eligibility criteria. Moreover, health plans could decide whether they target beneficiaries with any chronic conditions or specific conditions.15 The majority of Part D plans targeted beneficiaries with specific chronic conditions for MTM intervention in 2019.38 For example, only 11.9% of MTM programs included HIV/AIDS as one of their targeted diseases in 2019 because HIV/AIDS was not on the list of nine conditions that CMS recommended MTM programs to target.38 In addition, MTM service costs were included in plans’ annual bids for a contract as part of the administrative costs, so Part D plans lack incentives to expand MTM services beyond meeting CMS’ basic requirements.39 Reforms are needed to further develop MTM services among older PLWHs, such as adopting less restrictive eligibility criteria, including HIV/AIDS as a targeted disease for MTM intervention, and reimbursing plans for their expanded MTM services.

A few individual-level characteristics also exhibited significant associations with the study outcomes. Older age was associated with fewer DDI and COB issues in this study. It may be because prescribers are less likely to prescribe opioids and benzodiazepines to patients of very old age for medication safety concerns.40,41 This study exhibited that males generally were more likely to adhere to antiretroviral medications and be free from COB issues than females. Previous studies also reported that women were generally less adherent to their antiretroviral medications and were more likely to report lifetime prescription opioid use than men.42,43 Blacks and Hispanics exhibited a lower likelihood of adherence to antiretroviral medications than White beneficiaries. The results were consistent with previous research that reported minorities were generally less likely to achieve optimal adherence to antiretroviral medications than Whites among PLWHs.44 Conversely, Black beneficiaries exhibited a higher likelihood of being free from DDI issues than White beneficiaries. Lower adherence among Black patients than White patients may mean fewer medications being taken among Black patients, rendering a lower risk of DDI among this population.44,45 This study found that White beneficiaries were more likely to report COB issues than Black beneficiaries. This result was consistent with a previous study that reported that the White group was among the highest-represented groups prescribed both benzodiazepines and opioids during their first visit to primary care doctors.46 Further, an individual with a higher risk adjustment summary score was less likely to adhere to their antiretroviral medications and use medications safely. A higher risk adjustment summary score signified more chronic conditions and complex regimens, which were barriers to achieving optimal medication management.6

This study has several limitations. First, the findings generated from a Medicare sample of older PLWHs may not be generalized to other populations, such as younger PLWHs, PLWHs covered by other insurance programs, and the uninsured. Nevertheless, Medicare plays an increasing role in HIV care, and the Medicare databases allowed for a reasonably large sample size for the analyses. Second, a few county-level variables were included in the analyses since the individual-level characteristics were unavailable in the databases. These county-level variables may not accurately represent individuals’ characteristics. Third, beneficiaries’ prescription records were used to measure medication utilization. Although patients’ prescriptions may not equal their actual medication utilization, prescription records are valid proxies and have been adopted by CMS and Pharmacy Quality Alliance as core measures.27,47 Fourth, the general eligibility criteria for MTM enrollment were applied in the study to identify a comparison group. The criteria may not accurately represent beneficiaries’ eligibility to MTM programs for individual plans because the criteria vary across plans. However, the eligibility thresholds of 3 or more chronic conditions and 8 or more Part D medications were adopted by most MTM programs in 2017.15 Finally, the study outcomes of medication utilization were not equivalent to patients’ health outcomes. Future studies should investigate the long-term effects of MTM services on patients’ health outcomes, such as mortality among PLWHs in Medicare.

Conclusions

This research assessed the effectiveness of MTM services in adherence to antiretroviral medications and appropriate use of medications regarding DDI and COB among older PLWHs. This study provided critical evidence that supports the expansion of the MTM program in PLWHs who face many challenges in their daily medication management. Expanding MTM programs could benefit older PLWHs by optimizing antiretroviral medication adherence and lowering the COB risk. Future studies should investigate the long-term effects of MTM services on patients’ health service utilization, healthcare expenses, and health outcomes in PLWHs in Medicare.

Key Points.

What was already known

  • Advances in antiretroviral therapy have contributed to lower mortality and an extended lifespan, but some chronic conditions are now more prevalent among older people living with HIV/AIDS.

  • Centers for Medicare and Medicaid Services initiated the medication therapy management (MTM) program in 2006 to optimize medication therapeutic outcomes and reduce the risk of adverse drug events.

  • Previous research revealed that MTM receipt was associated with improved medication adherence, lower healthcare expenses, and a racial/ethnic disparity reduction in nonadherence to medications.

What this study adds

  • This study provided critical evidence that supports the expansion of the MTM program in PLWHs who face many challenges in their daily medication management.

  • This study found that the odds of adherence to antiretroviral medications among MTM recipients were 15% higher than among MTM non-recipients.

  • The odds of safe medication utilization regarding concurrent use of opioids and benzodiazepines for MTM recipients were 25.5% higher than for MTM non-recipients. Regarding drug-drug interactions, MTM recipients and non-recipients did not exhibit significant differences in their medication utilization.

Acknowledgments

The authors would like to acknowledge the clerical support from Chi Chun Steve Tsang, PhD, a Postdoctoral Scholar at the Department of Clinical Pharmacy & Translational Science, University of Tennessee College of Pharmacy. The authors also would like to acknowledge two Doctor of Pharmacy Students at the University of Tennessee Health Science Center College of Pharmacy, Clayton Hausberger and Prakash Kafle, for providing valuable insights for the introduction and discussions of the manuscript.

Declaration of funding

Research reported in this publication was supported by the National Institute on Aging of the National Institutes of Health under Award 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.

Disclosures of conflicts of interest:

Xiangjun Zhang has no financial disclosures to report. Junling Wang received funding from AbbVie, Curo, Bristol Myers Squibb, Pfizer, and Pharmaceutical Research and Manufacturers of America (PhRMA) and serves as the Co-Chair of the Value Assessment-Health Outcomes Research Advisory Committee of the PhRMA Foundation.

Acronyms and abbreviations

PLWHs

Persons/people living with HIV/AIDS

DDI

Drug-drug interactions

COB

Concurrent use of opioids and benzodiazepines

CMS

Centers for Medicare & Medicaid Services

MTM

Medication Therapy Management

OR

Odds ratio

CI

Confidence interval

Whites

Non-Hispanic Whites

Appendix Table 1. Baseline Characteristics for Intervention and Comparison Groups Before and After Propensity Score Matching among Medicare Beneficiaries Who Met Drug-Drug Interactions Measure Criteria and Used Antiretroviral Medications in 2017

Characteristics Before Matching After Matching
Intervention Comparison Intervention Comparison
(n=908) (n=2,670) (n=809, 33.3%) (n=1,618, 66.7%)
Number % Number % Number % Number %
Predisposing Factors
 Age, mean (SD) 70.44 (4.70) 70.51 (4.90) 70.46 (4.72) 70.49 (4.97)
 Male 672 74.01 2,011 75.32 619 76.51 1,238 76.51
 Race/Ethnicity
  Non-Hispanic Whites 437 48.13 1,307 48.95 406 50.19 810 50.06
  Blacks (African Americans) 322 35.46 986 36.93 287 35.48 552 34.12
  Hispanics 108 11.89 267 10.00 80 9.89 182 11.25
  Asian/Pacific Islanders and Other 41 4.51 110 4.12 36 4.45 74 4.57
 Proportion of Married-couple Families, mean (SD)* 0.68 (0.10) 0.69 (0.09) 0.69 (0.09) 0.68 (0.09)
 Proportion of High School Education and Above, mean (SD)* 0.86 (0.06) 0.86 (0.05) 0.86 (0.05) 0.86 (0.05)
 Income Per Capita (in $1,000), mean (SD)* 57.53 (26.91) 58.05 (27.45) 58.37 (27.96) 58.10 (26.96)
 Proportion of the Uninsured, mean (SD)*# 0.097 (0.046) 0.105 (0.048) 0.099 (0.048) 0.098 (0.046)
Enabling Factors
 Metropolitan Statistical Area* 841 92.62 2,455 91.95 747 92.34 1,498 92.58
 Health Professional Shortage Area* 874 96.26 2,558 95.81 779 96.29 1,561 96.48
 Census Regions*#
  Northeast 282 31.06 682 25.54 235 29.05 470 29.05
  Midwest 108 11.89 323 12.10 103 12.73 206 12.73
  South 324 35.68 1,102 41.27 289 35.72 578 35.72
  West 194 21.37 563 21.09 182 22.50 364 22.50
Need Factor
 Risk Adjustment Summary Score, mean (SD)# 3.04 (2.04) 2.49 (1.76) 2.70 (1.66) 2.62 (1.68)
Open in a new tab

SD = standard deviation.

*

County-level characteristics.

#

These characteristics significantly differed between the intervention and comparison groups before matching (P < .05).

All characteristics were similar between the intervention and comparison groups after matching (P > .05).

Appendix Table 2. Baseline Characteristics for Intervention and Comparison Groups Before and After Propensity Score Matching among Medicare Beneficiaries Who Met Concurrent Use of Opioids and Benzodiazepines Measure Criteria and Used Antiretroviral Medications in 2017

Characteristics Before Matching After Matching
Intervention Comparison Intervention Comparison
(n=802) (n=2,471) (n=691, 33.3%) (n=1,382, 66.7%)
Number % Number % Number % Number %
Predisposing Factors
 Age, mean (SD) 69.39 (4.20) 69.34 (4.13) 69.39 (4.20) 69.24 (4.06)
 Male 506 63.09 1,640 66.37 453 65.56 906 65.56
 Race/Ethnicity
  Non-Hispanic Whites 318 39.65 1,018 41.20 289 41.82 569 41.17
  Blacks (African Americans) 345 43.02 1,078 43.63 300 43.42 580 41.97
  Hispanics 118 14.71 295 11.94 87 12.59 203 14.69
  Asian/Pacific Islanders and Other 21 2.62 80 3.24 15 2.17 30 2.17
 Proportion of Married-couple Families, mean (SD)* 0.68 (0.09) 0.68 (0.09) 0.68 (0.09) 0.68 (0.09)
 Proportion of High School Education and Above, mean (SD)* 0.86 (0.06) 0.86 (0.05) 0.86 (0.06) 0.86 (0.05)
 Income Per Capita (in $1,000), mean (SD)* 56.85 (24.76) 56.78 (25.96) 57.12 (24.97) 56.49 (25.84)
 Proportion of the Uninsured, mean (SD)* 0.10 (0.05) 0.10 (0.05) 0.10 (0.05) 0.10 (0.05)
Enabling Factors
 Metropolitan Statistical Area* 729 90.90 2,250 91.06 634 91.75 1,268 91.75
 Health Professional Shortage Area* 774 96.51 2,380 96.32 667 96.53 1,330 96.24
 Census Regions*#
  Northeast 219 27.31 550 22.26 176 25.47 352 25.47
  Midwest 103 12.84 300 12.14 85 12.30 170 12.30
  South 300 37.41 1,071 43.34 277 40.09 554 40.09
  West 180 22.44 550 22.26 153 22.14 306 22.14
Need Factor
 Risk Adjustment Summary Score, mean (SD)# 3.10 (1.90) 2.57 (1.70) 2.79 (1.55) 2.71 (1.54)
Open in a new tab

SD = standard deviation.

*

County-level characteristics.

#

These characteristics significantly differed between the intervention and comparison groups before matching (P < .05).

All characteristics were similar between the intervention and comparison groups after matching (P > .05).

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Contributor Information

Xiangjun Zhang, Department of Clinical Pharmacy & Translational Science, University of Tennessee Health Science Center College of Pharmacy.

Junling Wang, Department of Clinical Pharmacy & Translational Science, University of Tennessee Health Science Center College of Pharmacy.

Data availability statement

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 resdac@umn.edu or by phone at 888–973-7322.

References

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

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 resdac@umn.edu or by phone at 888–973-7322.

RESOURCES