Abstract
Background
Assisted-living (AL) settings are an important residential care option for old and disabled Americans, but there are no national data characterizing medication use in AL.
Objective
To investigate medication costs and use of older adults living in the AL settings compared to those in the community, independent living, and nursing home settings.
Design
2015 National Health and Aging Trends Study; nationally representative cross-sectional study.
Paticipants
Respondents ≥ 65 years with Medicare Part D prescription drug coverage (n = 5980, representing 32.34 million older adults).
Measures
Total Part D medication costs; number of 30-day prescription fills; binary indicators for overall polypharmacy (≥ 5 and ≥ 10 concurrent medications), prescription fills of opioid and psychotropic medications including antipsychotics, benzodiazepines, gabapentinoids, antidepressants, and central nervous system-active (CNS-active) polypharmacy.
Results
Adjusting for demographics, the annual medication costs among AL residents, at $3890, were twice as high as those of their community-dwelling counterparts ($1932; p < .01). All medication outcomes except opioids were higher for older adults in AL compared to community settings. While the adjusted number of 30-day prescription fills among AL residents was slightly lower than that of nursing home residents (89.5 vs. 106.2; p < .05), AL residents experienced equivalent rates of overall polypharmacy ≥ 10 medications (30.2% vs. 23.5%), antipsychotics (30.8% vs. 27.8%), benzodiazepines (30.7% vs. 32.6%), gabapentinoids (21.2% vs. 16.1%), and CNS-active polypharmacy (26.0% vs. 36.9%; p > .05 for all). Patterns of use across settings were consistent when limited to older adults with dementia.
Conclusions
Older Americans in AL experience a prescription medication burden similar to those in nursing homes. AL settings have an important opportunity to ensure their medication-related clinical services and supports match the needs of their residents.
Supplementary Information
The online version contains supplementary material available at 10.1007/s11606-022-07434-3.
KEY WORDS: assisted living, medication use, antipsychotics, polypharmacy, dementia
INTRODUCTION
Nearly one million older adults in the USA live in assisted living (AL) settings,1,2 an important residential care option for those who might potentially benefit from assistance with functional activities but do not require 24-h skilled nursing services. The services offered and populations served vary substantially across states and product markets, but AL settings generally offer some combination of: housing services, transportation, assistance with instrumental activities of daily living (e.g., laundry, meal preparation), personal care (e.g., dressing, bathing, toileting), health-related services (e.g., medication management and administration, access to health and medical services), and 24-h supervision.3
In 2018, 61% of AL residents in the USA needed assistance with three or more activities of daily living, while two-thirds had two or more chronic conditions.1 AL residents resemble those in nursing homes in terms of acuity and services needed.4–6 However, unlike nursing homes, where the federal government—through the Centers for Medicare & Medicaid Services—plays a central role as both payor and regulator,7 AL communities operate in a patchwork state-level regulatory environment and are primarily financed by personal resources rather than government funded health plans.2
Medication use and administration in AL is a useful paradigm through which to consider quality and associated policy in these settings. In a report by the US General Accounting Office that considered AL settings in four states, medication-related issues were among the most frequently cited problems.8 While the Centers for Medicare & Medicaid Services has significant policy leverage to address prescribing among nursing home residents, such as through its 5-star rating system or the National Partnership to Improve Dementia Care in Nursing Homes,9 the fragmented AL policy landscape has few such levers and few states have a system to help consumers and regulators compare AL settings on quality indicators.
While medication use in AL is a high-priority quality-of-care issue, existing studies are limited in generalizability and generally represent single-site, program, or state evaluations and, more rarely, multistate samples.10–18 We know of only one study using a nationally representative sample to examine medication use in AL and nursing home settings, and this dates back to 1998,19 before the implementation of Medicare Part D prescription drug benefit in 2006. In this study, we provide the first nationally representative analysis of Medicare Part D medication costs and use among older adults in AL compared with other residential settings by combining the National Health and Aging Trends Study, a nationally representative study of age-eligible Medicare beneficiaries in the USA,20 with linked Medicare claims data. We characterized the overall extent of use, including measures of overall polypharmacy (≥ 5 and ≥ 10 medications), as well as potentially high-risk opioid and psychotropic prescribing. We chose select medication classes monitored in nursing facilities by the federally mandated Minimum Data Set 3.0 or that feature in the American Geriatrics Society Beers Criteria for potentially inappropriate prescribing to older adults.21 Finally, since 42% of AL residents have moderate or severe cognitive impairment,22 we examined prescription outcomes by residential settings among respondents with dementia. We hypothesized that older adults in AL settings had medication costs and use higher than those in community or independent living settings but lower than those in nursing homes.
METHODS
Data and Study Population
The National Health and Aging Trends Study (NHATS) is an ongoing longitudinal, nationally representative survey of Medicare beneficiaries ≥ 65 years that includes detailed information regarding residence and housing type, socioeconomic status, physical and cognitive capacity, and health.20 We focus on the 2015 survey, with an overall response rate of 73.2%. Proxy respondents (mostly family members) were interviewed when the sample person could not respond, and accounted for 4.4% of all respondents, and 3.4%, 5.2%, 32.3%, and 47.6% for those living in the community, independent living, AL, and nursing home settings, respectively. Respondents, as part of agreeing to participate in NHATS, consent to linkage of their survey data with Medicare claims data. Of 7859 living NHATS respondents with positive analytic weights, this analysis included those with any Medicare Part D prescription drug coverage in 2015, through either a standalone prescription drug plan or Medicare Advantage (n = 5980 [74.0% of all respondents]).
Measures
We followed Freedman and Spillman to distinguish four residential settings: community (combining traditional community housing and retirement or senior housing, which Freedman and Spillman considered separately), independent living, assisted living, and nursing home.23
As the first step in identifying these settings, we distinguished nursing homes from non-nursing home residential care settings, the latter of which are defined in NHATS as (1) a multi-unit building or retirement community with areas where the respondent either currently receives or could move to receive additional care when needed (e.g., help with medications or self-care, providing meals) or (2) a freestanding or attached home, which the respondent reports as a “group home or board and care home” or “assisted living facility or continuing care retirement community.”20
We then divided these non-nursing home residential care settings into AL or independent living based on type of facility and levels of care. AL comprised settings the respondent explicitly called “assisted living,” including the assisted living section of a multi-level facility (e.g., continuing care retirement community), as well as group home, board and care home, or adult family care home.23 In our independent-living group, we included non-nursing home residential care settings that were not identified as AL, as described above, which were primarily independent living areas of multi-level facilities such as retirement or senior housing and continuing care retirement community.
Remaining residential settings not captured by the three groups outlined above (nursing home, AL, or independent living) were considered community, including traditional community housing and retirement or senior housing communities where supportive services (e.g., help with medications or self-care, providing meals) were not available.
Prescription fills in 2015 were obtained from Medicare Part D claims data based on the generic name, fill date, and days’ supply. Total Part D medication costs and number of standard 30-day prescription fills were obtained from the Medicare Master Beneficiary Summary File Cost and Utilization Segment, and included prescription drugs as well as over-the-counter medications and supplements covered by Part D. The 30-day prescription fill data standardizes prescription fills, regardless of the number of claims (i.e., a single 90-day prescription fill and three 30-day prescription fills would be represented as the same amount of medication). In addition to total number of fills and costs, we derived measures of overall polypharmacy with ≥ 5 and ≥ 10 concurrent medications, requiring > 30 days of consecutive overlap.
After examining use overall, we created binary indicators for prescription fills of opioid or psychotropic medications, including antipsychotics, benzodiazepines (including non-benzodiazepine benzodiazepine receptor agonists [eszopiclone, zaleplon, and zolpidem]), gabapentinoids, and antidepressants (Appendix Table 1). Finally, we identified respondents that met the Beers Criteria definition of central nervous system–active (CNS–active) polypharmacy by receiving overlapping prescription fills for ≥ 3 opioid or psychotropic medications for > 30 consecutive days.21,24
Respondent characteristics including demographics (sex, age, race/ethnicity), socioeconomic status (marital status, education, family income, Medicaid enrollment), and health (cognitive, functional, and sensory impairment, clinical comorbidities) were obtained from the NHATS survey. We measured functional impairment based on reported difficulty with or receiving help in (1) activities of daily living (ADL), including getting out of bed, getting around one’s home or building, bathing, dressing, eating, and toileting, and (2) instrumental activities of daily living (IADL), including doing laundry, going shopping, preparing meals, handling banking, and managing medications. Auditory and vision impairment was present if respondents reported being unable to “hear well enough to carry on a conversation in a room with a radio or television playing” and “see well enough to read newspaper print,” respectively.
Probable dementia (hereafter, simply “dementia”) was determined based on a doctor’s diagnosis of dementia or Alzheimer’s disease; a score indicating likely dementia on a screening instrument administered to proxy respondents; or performance on measures of memory, orientation, and executive function.25 Respondents were asked whether they had been diagnosed with heart disease, diabetes, lung disease, stroke, cancer, arthritis, and hypertension; we created binary indicators for each. Depression or anxiety were assessed using the 4-item Patient Health Questionnaire for Depression and Anxiety. Insomnia was considered present for respondents who, for “most nights” or “always,” reported requiring more than 30 min to fall asleep or having trouble falling back to sleep. Chronic pain was present for respondents who reported being bothered by pain. For overall comorbidity, we counted the number of the above conditions present. Finally, we included whether respondents reported a hip fracture since age 50 or fall in the last year.
Statistical Analysis
We first compared characteristics for older adults across community, independent living, AL, and nursing home settings using adjusted Wald and Rao-Scott chi-square tests for continuous and categorical characteristics, respectively.
Next, we examined prescription medication fills by setting (reference: AL) using linear regression models for total medication costs and number of medication prescription fills and logistic regression models for each medication class and polypharmacy measure (i.e., ≥ 5 overall, ≥ 10 overall, and ≥ 3 CNS-active), adjusting for sex, age, and race/ethnicity. We calculated the mean adjusted outcomes in each setting while holding all other characteristics constant and then determined the marginal effect of residential setting on the outcomes of interest. We repeated the analysis limited to older adults with dementia. In this primarily descriptive analysis, we limited adjustment to demographic characteristics, which are associated with medication use (e.g., females have higher rates of psychotropic use26 but further adjusted for socioeconomic characteristics and clinical comorbidities in sensitivity analyses. We performed additional sensitivity analyses excluding respondents who died or were in hospice during 2015 because medication use might be different at end of life.
For all analyses, we used NHATS survey analytic weights to generate nationally representative estimates for older adults, accounting for differential probabilities of selection and nonresponse. Standard errors were adjusted to account for the complex survey design. To derive population estimates (in millions), we further adjusted the weighted totals to match, by age group, the Medicare population ≥ 65 years at the time the sample was drawn.27 Statistical significance was set at two-tailed P < .05. Analyses were performed using STATA, version 17.0.
RESULTS
Among the 5980 respondents aged 65 years and older also enrolled in Medicare Part D (representing 32.34 million older adults), 5321 (92.4%) resided in the community, 167 (2.3%) in independent living, 164 (2.3%) in AL, and 328 (3.1%) in nursing homes (Table 1). AL residents had the oldest average age of all settings, at 84.3 years (74.7 [community] vs. 81.8 [independent living] vs. 82.9 [nursing home]; p < .01). On measures of functional and sensory impairment, those in AL settings had a greater degree of impairment than those in the community or independent living but less than in nursing homes (e.g., ADL impairments: 0.8 [community] vs. 1.2 [independent living] vs. 3.0 [AL] vs. 5.1 [nursing home]; p < .01). The pattern was similar for health conditions. The largest absolute difference between AL residents and their community-dwelling counterparts was for dementia, the prevalence of which was 46 percentage points higher in AL than among community-dwelling older adults (7.5% vs. 18.4% vs. 53.4% vs. 72.2%; p < .01). AL and nursing home residents had similarly high prevalence of history of hip fracture (22.0% and 18.2%) or past-year fall (50.4% and 49.0%, respectively).
Table 1.
Characteristics of Adults Aged 65 Years and Older with Medicare Prescription Drug Coverage, by Residential Setting, 2015
| Community | Independent living | Assisted living | Nursing home‡ | pvalue | |
|---|---|---|---|---|---|
| Sample population | 5,321 | 167 | 164 | 328 | — |
| Weighted n, millions | 29.87 | 0.74 | 0.74 | 0.99 | — |
| Weighted % | 92.4 | 2.3 | 2.3 | 3.1 | — |
| Demographics | |||||
| Female, % | 58.4 | 71.9 | 65.5 | 75.9 | * |
| Age, years (SD) | 74.7 (6.9) | 81.8 (8.4) | 84.3 (8.8) | 82.9 (12.0) | * |
| 65–74, % | 56.7 | 19.9 | 18.8 | 22.4 | * |
| 75–84, % | 32.1 | 41.0 | 25.8 | 31.7 | |
| 85+, % | 11.2 | 39.2 | 55.5 | 45.9 | |
| Race/ethnicity | |||||
| Non-Hispanic white, % | 75.9 | 81.4 | 79.1 | 77.5 | † |
| Non-Hispanic black, % | 8.7 | 8.3 | 4.2 | 12.8 | |
| Other, % | 15.4 | 10.3 | 16.7 | 9.7 | |
| Socioeconomic status | |||||
| Married, % | 55.9 | 25.5 | 14.0 | –§ | * |
| Some college or above, % | 52.4 | 58.7 | 35.1 | 35.7 | * |
| Family income, $, mean (SD) | 56,939 (102,831) | 58,859 (176,037) | 28,053 (24,873) | 24,381 (25,979) | * |
| Medicaid enrollment, % | 14.1 | 20.6 | 29.6 | 56.9 | * |
| Functional and sensory impairment | |||||
| Activities of daily living, n (SD)‖ | 0.8 (1.5) | 1.2 (1.8) | 3.0 (2.4) | 5.1 (1.7) | * |
| Instrumental activities of daily living, n (SD)‖ | 0.9 (1.4) | 1.8 (1.7) | 4.1 (1.3) | 4.7 (0.6) | * |
| Auditory impairment, % | 11.7 | 14.0 | 22.5 | 28.0 | * |
| Vision impairment, % | 4.9 | 6.8 | 15.3 | 21.8 | * |
| Health conditions | |||||
| Number of conditions, n (SD)‖ | 3.4 (2.0) | 3.7 (2.1) | 4.4 (2.1) | 5.7 (2.4) | * |
| Dementia, % | 7.5 | 18.4 | 53.4 | 72.2 | * |
| Heart disease, % | 25.1 | 33.5 | 30.2 | 46.9 | * |
| Diabetes, % | 26.9 | 28.8 | 30.4 | 39.9 | |
| Lung disease, % | 18.0 | 15.5 | 19.6 | 28.5 | |
| Stroke, % | 9.3 | 17.3 | 30.5 | 38.1 | * |
| Cancer, % | 25.8 | 30.9 | 25.4 | 38.9 | |
| Arthritis, % | 57.9 | 63.5 | 59.1 | 84.3 | * |
| Hypertension, % | 66.5 | 67.5 | 60.8 | 72.0 | |
| Depression, % | 12.4 | 8.5 | 30.0 | 35.2 | * |
| Anxiety, % | 11.1 | 12.9 | 26.3 | 31.9 | * |
| Insomnia, % | 27.1 | 26.1 | 28.5 | 25.8 | |
| Pain, % | 55.4 | 49.8 | 44.4 | 54.9 | † |
| Hip fracture, % | 3.9 | 9.5 | 22.0 | 18.2 | * |
| Fall, % | 30.5 | 38.4 | 50.4 | 49.0 | * |
Authors’ analysis of data from the 2015 National Health and Aging Trends Study (NHATS) and Linked Medicare enrollment data. Adjusted Wald tests were performed to compare continuous characteristics and Rao-Scott chi-square tests were performed to compare categorical characteristics among older adults with Medicare Part D prescription drug coverage across residential settings
*p < 0.01
†p < 0.05
‡For respondents that were already nursing home residents at the time of the first NHATS wave in 2011 (262 out of 328), information about socioeconomic status, functional and sensory impairment, and health conditions was not available (i.e., n = 328 for demographic characteristics and n = 66 for other characteristics among nursing home residents)
§Estimates based on too few cases (<11) may not be reported, per NHATS
‖Limitations in activities of daily living range from 0 to 6, limitations in instrumental activities of daily living range from 0 to 5, and number of health conditions range from 0 to 12
Adjusting for demographics (sex, age, and race/ethnicity), the mean total medication costs for AL residents in 2015 were $3890 (95% confidence interval [CI]: $2945 to $4834), higher than for older adults in the community ($1932 [CI: $1531 to $2332]; p < .01) but lower than for those in nursing homes ($5217 [CI: $4327 to $6107]; p < .05; Table 2). The adjusted number of 30-day medication prescription fills for AL residents in 2015 was 89.5 (CI: 80.3 to 98.8), higher than for those in the community (51.3 [CI: 49.8 to 52.9]) or independent living (62.4 [CI: 53.1 to 71.8]) but lower than for those in nursing homes (106.2 [CI: 95.1 to 117.4]; p < .05 for all). The burden of overall polypharmacy among older adults in AL was as high as among those in nursing homes and significantly higher than among those in the community: 72.2% ( CI: 64.4 to 80.0%) of AL residents experienced polypharmacy of ≥5 concurrent medications, compared to 71.3% (CI: 64.1 to 78.4%; p > .05) in nursing homes and 44.1% (CI: 42.5 to 45.8%; p < .01) in community; while 30.2% (CI: 21.7 to 38.8%; AL) of AL residents experienced polypharmacy of ≥10 concurrent medications, compared to 23.5% (CI: 15.3 to 31.8%; p > .05) in nursing homes and 6.1% (CI: 5.3 to 7.0%; p < .01) in the community.
Table 2.
Medication Costs and Use of Adults Aged Sixty-Five Years and Older With Medicare Prescription Drug Coverage, by Residential Setting, 2015
| Community | Independent living |
Assisted living (reference group) |
Nursing home | |
|---|---|---|---|---|
| Sample population | 5,321 | 167 | 164 | 328 |
| Weighted n, millions | 29.87 | 0.74 | 0.74 | 0.99 |
| Weighted % | 92.4 | 2.3 | 2.3 | 3.1 |
| Overall medication use | ||||
| Total medication costs, $ | 1,932* (1,531–2,332) | 3,443 (2,223–4,663) | 3,890 (2,945–4,834) | 5,217† (4,327–6,107) |
| Number of standard 30-day prescription fills, n | 51.3* (49.8–52.9) | 62.4* (53.1–71.8) | 89.5 (80.3–98.8) | 106.2† (95.1–117.4) |
| ≥ 5 concurrent medications for > 30 days consecutively, % | 44.1* (42.5–45.8) | 50.5* (41.8–59.2) | 72.2 (64.4–80.0) | 71.3 (64.1–78.4) |
| ≥ 10 concurrent medications for > 30 days consecutively, % | 6.1* (5.3–7.0) | 10.4* (5.2–15.6) | 30.2 (21.7–38.8) | 23.5 (15.3–31.8) |
| Opioid and psychotropic medications, % | ||||
| Opioid | 31.6 (29.9–33.3) | 30.6 (23.0–38.3) | 32.8 (25.7–40.0) | 50.3* (44.4–56.2) |
| Antipsychotic | 2.5* (2.0–3.0) | –‡ | 30.8 (21.5–40.0) | 27.8 (16.6–39.1) |
| Benzodiazepine | 17.8† (16.2–19.4) | 15.4* (8.8–22.0) | 30.7 (20.8–40.7) | 32.6 (26.4–38.8) |
| Gabapentinoid | 12.2† (11.1–13.3) | 12.3 (6.1–18.4) | 21.2 (12.6–29.9) | 16.1 (11.2–21.1) |
| Antidepressant | 24.9* (23.3–26.4) | 35.1† (24.7–45.4) | 52.3 (42.9–61.7) | 67.2* (62.2–72.2) |
| Central nervous system-active polypharmacy§ | 4.1* (3.3–4.9) | –‡ | 26.0 (14.3–37.6) | 36.9 (27.9–45.9) |
Authors’ analysis of data from the 2015 National Health and Aging Trends Study (NHATS) and linked Medicare enrollment and Part D prescription drug claims data. We examined medication costs and use across residential settings (reference: assisted living) using linear regression models for continuous outcomes and logistic regression models for binary outcomes, adjusting for demographics (sex, age, race/ethnicity). We calculated the mean adjusted outcomes for older adults in each residential setting. Ninety-five percent confidence intervals were reported in brackets. Differences in adjusted medication costs and use between older adults living in the assisted living settings and other residential settings were tested by the marginal effects
*p < 0.01
†p < 0.05
‡Estimates based on too few cases (< 11) may not be reported, per NHATS
§Central nervous system-active polypharmacy required > 30 consecutive days of concurrent exposure to ≥ 3 medications from the following classes: opioids, antipsychotics, benzodiazepines (including non-benzodiazepine benzodiazepine receptor agonists), antidepressants (selective serotonin reuptake inhibitors and serotonin norepinephrine reuptake inhibitors), and antiepileptics21,24
AL residents had a lower prevalence of opioid and antidepressant fills than nursing home counterparts but comparable amounts of antipsychotic, benzodiazepine, gabapentinoid, and CNS-active polypharmacy. Thirty-one percent (CI: 21.5 to 40.0%) of AL residents had an antipsychotic fill compared to 27.8% (CI: 16.6 to 39.1%; p > .05) of nursing home residents and 2.5% (CI: 2.0 to 3.0%; p < .01) in the community. While opioid fills among community-dwelling respondents were similar to those in AL, all psychotropic classes and CNS-active polypharmacy were less likely. Sensitivity analyses showed consistent results with the main findings when additionally adjusting for socioeconomic status and clinical comorbidities (Appendix Table 2), and when excluding respondents who died or were in hospice (Appendix Table 3).
Among respondents with dementia, the mean total adjusted medication costs for AL residents were $3232 (CI: $2364 to $4101), not significantly higher than for those in the community ($2598 [CI: $2136 to $3061]; p > .05) but lower than those in nursing homes ($5042 [CI: $4083 to $6001]; p < .01; Table 3). As with respondents overall, the 85.6 (CI: 73.2 to 97.9) 30-day medication prescription fills for AL residents with dementia was higher than among those living in the community (64.7 [CI: 60.9 to 68.6]) but lower than those in nursing homes (101.6 [CI: 92.3 to 110.9]; p < .05 for both). The burden of overall polypharmacy with ≥ 5 concurrent medications among AL residents with dementia (75.6% [CI: 67.1 to 84.0%]) was as high as among those in nursing homes (73.1% [CI: 65.2 to 81.0%]; p > .05) and significantly higher than among those in the community (59.5% [CI: 54.5 to 64.6%]; p < .01).
Table 3.
Medication Costs and Use of Adults Aged 65 Years and Older with Dementia with Medicare Prescription Drug Coverage, by Residential Setting, 2015
| Community | Independent living | Assisted living(reference group) | Nursing home‡ | |
|---|---|---|---|---|
| Sample population | 618 | 27 | 89 | 220 |
| Weighted n, millions | 2.18 | 0.13 | 0.40 | 0.62 |
| Weighted % | 65.5 | 4.0 | 11.9 | 18.6 |
| Overall medication use | ||||
| Total medication costs, $ | 2598 (2136–3061) | 1966 (817–3114) | 3232 (2,364–4101) | 5042* (4083–6001) |
| Number of standard 30-day prescription fills, n | 64.7* (60.9–68.6) | 61.9 (39.9–83.8) | 85.6 (73.2–97.9) | 101.6† (92.3–110.9) |
| ≥ 5 concurrent medications for > 30 days consecutively, % | 59.5* (54.5–64.6) | 52.5 (28.0–77.0) | 75.6 (67.1–84.0) | 73.1 (65.2–81.0) |
| ≥10 concurrent medications for >30 days consecutively, % | 12.9 (9.5–16.2) | –§ | 21.7 (11.5–32.0) | 19.3 (11.3–27.2) |
| Opioids and psychotropic medications, % | ||||
| Opioid | 36.3 (32.1–40.6) | –§ | 31.3 (20.3–42.3) | 45.2 (36.2–54.3) |
| Antipsychotic | 12.0† (7.9–16.1) | –§ | 27.2 (17.0–37.5) | 28.0 (18.6–37.5) |
| Benzodiazepine | 19.9 (15.6–24.2) | –§ | 28.1 (17.1–39.1) | 32.6 (26.3–38.9) |
| Gabapentinoid | 11.8 (8.7–14.9) | –§ | 11.3 (3.8–18.9) | 13.8 (8.4–19.1) |
| Antidepressant | 34.3 (29.0–39.7) | 45.9 (30.0–61.8) | 46.8 (34.3–59.3) | 64.5† (57.1–71.8) |
| Central nervous system-active polypharmacy‖ | 7.1† (4.3-9.9) | –§ | 18.7 (7.8-29.5) | 30.3 (21.2-39.5) |
Authors’ analysis of data from the 2015 National Health and Aging Trends Study (NHATS) and linked Medicare enrollment, Part D prescription drug claims, and long-term care Minimum Data Set assessment data. We examined medication costs and use across residential care settings (reference: assisted living) using linear regression models for continuous outcomes and logistic regression models for binary outcomes, adjusting for demographics (sex, age, race/ethnicity). We calculated the mean adjusted outcomes for older adults in each residential setting. Ninety-five percent confidence intervals were reported in brackets. Differences in adjusted medication costs and use between older adults living in the assisted living settings and other residential settings were tested by the marginal effects
*p < 0.01
†p < 0.05
‡For respondents that were already nursing home residents at the time of the first NHATS wave in 2011, the data elements required for dementia classification (e.g., self-reported diagnosis, cognitive tests, proxy items) were not available. For these respondents, we determined the presence of dementia using the indicators for active Alzheimer’s disease and other dementias from the Minimum Data Set, which included assessment data for beneficiaries in both traditional and managed Medicare
§Estimates based on too few cases (< 11) may not be reported, per NHATS
‖Central nervous system-active polypharmacy required > 30 consecutive days of concurrent exposure to ≥ 3 medications from the following classes: opioids, antipsychotics, benzodiazepines (including non-benzodiazepine benzodiazepine receptor agonists), antidepressants (selective serotonin reuptake inhibitors and serotonin norepinephrine reuptake inhibitors), and antiepileptics.21,24
Among those with dementia, older adults in AL and nursing homes had similar prevalence of opioid and psychotropic use, including CNS-active polypharmacy, with the exception of antidepressant fills, which were lower in AL. Antipsychotic prescribing was similar in AL and nursing homes, at 27.2% (CI: 17.0 to 37.5%) and 28.0% (CI: 18.6 to 37.5%; p > .05), respectively, and significantly higher than in the community, at 12.0% (CI: 7.9 to 16.1%; p < .05). CNS-active polypharmacy was also lower in the community; otherwise, opioid and psychotropic prescribing to those with dementia did not vary markedly across settings.
DISCUSSION
This analysis, to the best of our knowledge, is the first to characterize Medicare Part D prescription medication costs and use in a nationally representative sample of older adults across four different residential settings including assisted living, including beneficiaries in both traditional and managed Medicare. We found that in all measures of costs and use, older adults in AL settings obtain more prescription fills than their community-dwelling counterparts, with the exception of opioids. While the volume of 30-day prescription fills is slightly lower than that of nursing home residents, AL residents experienced equivalent rates of overall polypharmacy with ≥ 5 and ≥ 10 medications. Furthermore, the exposure of AL residents to antipsychotics, benzodiazepines, gabapentinoids, and CNS-active polypharmacy was as high as that among nursing home residents. These overall patterns of use were consistent when limited to older adults with dementia.
According to the 2010 National Survey of Residential Care Facilities, which would include those settings considered as AL in our analysis, virtually all (99.7%) offered some form of medication management to residents.28 Because AL settings and residents have been difficult to reliably identify in administrative data, however, far less is known about the quality of medication use in AL, unlike the closely regulated and monitored nursing home environment. We demonstrated that older adults in AL have a volume of prescription medication fills closer to counterparts in nursing homes than to those in community or independent living. In addition, the burden of overall polypharmacy of both ≥ 5 and ≥ 10 medications—consistently associated with bad outcomes for older adults29—in AL was equivalent to that in nursing homes.
We also examined psychotropic prescribing, as this has been an important quality-of-care indicator and the focus of significant regulatory and research attention in nursing homes since the Omnibus Budget Reconciliation Act of 1987, which included provisions prohibiting use of psychotropic medications as chemical restraints for resident behavior.30 The most recent concerted federal attention focused on antipsychotic prescribing is the National Partnership to Improve Dementia Care, which began in 2012 and is ongoing.9 We found that the rates of antipsychotic use among AL and nursing home settings were equivalent (30.8% and 27.8%, respectively) and far higher than the 2.5% in the community. While the gap was smaller when limited to respondents with dementia because of higher community use, the prevalence of AL was still significantly higher than that of the community. Overall, use of most classes examined in AL settings were equivalent to those in nursing homes, with the exception of less opioid and antidepressant use. Our findings are consistent with a prior multistate analysis comparing antipsychotic and antianxiety prescribing that found equivalent rates of use in AL and nursing homes14 and further emphasize the need to examine antipsychotic use in AL settings.31
The high rates of prescription medication fills in AL are particularly concerning because these residents have burdens of cognitive, functional, sensory impairment, and health conditions approaching those of nursing home populations, yet AL staffing does not provide a comparable degree of clinical support. In 2016, 99.1% and 98.3% of nursing homes employed full-time registered and licensed practical nurses, respectively, compared to just 39.4% and 35.7% of AL settings; nursing homes provided an average of 1.39 nursing hours per resident per day in contrast to just 0.37 h in AL settings.6 Furthermore, states vary widely in rules and requirements governing how AL staff may support medication management.32 For example, in 2014, only 26 out of 37 states require general medication-related training for unlicensed assistive personnel (e.g., medication technician/aide, certified nursing assistant, or personal care aide) who may administer medications.
States and state-based AL associations have recognized the need for attention to prescribing quality. The states of Ohio, Minnesota, and Oregon require AL quality reporting that includes medication-related measures.33–35 In other states, this impetus has come from the AL community itself: for example, facility associations in New Jersey and Wisconsin have launched voluntary quality assurance programs that include standards related to medication administration.36,37 In addition to these state-based initiatives, national AL accreditation and quality programs are now also addressing medication prescribing, including the National Center for Assisted Living’s National Quality Award Program and The Joint Commission’s new Assisted Living Community accreditation program, which was just launched in 2021.38,39 Both of these national initiatives include medication measures, such as off-label antipsychotic drug use. AL facilities may perceive medication quality among residents—and the staffing to support this quality—as an important feature to attract and retain residents, thereby potentially obviating the need for regulatory attention.
This study has several limitations. First, we measured prescription fills, which may overestimate use (if respondents use less than prescribed or use on an as-needed other than regular basis18), underestimate use (if respondents pay out of pocket or the prescription is covered by another payor such as Veterans Health Administration), or both. Second, we cannot account for prescribing indication, which limits the ability to comment on appropriateness. Third, while NHATS is nationally representative, respondents from the 2015 survey residing in AL lived in 32 rather than all 50 states.
While the population of older adults in the USA has grown, the number residing in assisted living settings has as well. We demonstrate that AL-residing older adults experience a prescription medication burden more like nursing home residents than their community-dwelling counterparts. The growing state and national industry-led emphasis on facility quality should consider the high medication-related need demonstrated by this national analysis. Providing services and supports related to medication use among AL residents may be an important opportunity for facilities to support quality care while also differentiating themselves in the residential care market.
Supplementary Information
(DOCX 42.7 kb)
Acknowledgments
The National Health and Aging Trends Study is sponsored by the National Institute on Aging Grant NIA U01AG032947 through a cooperative agreement with the Johns Hopkins Bloomberg School of Public Health. Funding for this work was provided by R01 AG056407 (Lei and Maust).
Declarations
Compliance with Ethical Standards
This study was approved by the Institutional Review Board of the University of Michigan Medical School.
Conflict of Interest
The authors declare that they do not have a conflict of interest.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Caffrey C, Sengupta M, Melekin A. Residential Care Community Resident Characteristics: United States, 2018. Hyattsville (MD): National Center for Health Statistics. NCHS Data Brief, no. 404, 2021 Sep. Available from: https://www.cdc.gov/nchs/products/databriefs/db404.htm. Accessed 2021 Sep 21. [PubMed]
- 2.National Center for Assisted Living. Facts & Figures [cited 2021 June 10]. Available from: https://www.ahcancal.org/Assisted-Living/Facts-and-Figures/Pages/default.aspx.
- 3.Hawes C, Phillips CD, Rose M, Holan S, Sherman M. A national survey of assisted living facilities. Gerontologist. 2003;43(6):875-82. [PMID: 14704387]. 10.1093/geront/43.6.875. [DOI] [PubMed]
- 4.Harris-Kojetin L, Sengupta M, Park-Lee E, Valverde R. Long-term care providers and services users in the United States, 2013 Overview. Hyattsville (MD): National Center for Health Statistics. Vital and Health Stat 3(37), 2013 Dec. Available from: https://www.cdc.gov/nchs/data/series/sr_03/sr03_037.pdf. Accessed 2021 Dec 5.
- 5.Harris-Kojetin L, Sengupta M, Park-Lee E, et al. Long-term care providers and services users in the United States: Data from the National Study of Long-Term Care Providers, 2013-2014. Hyattsville (MD): National Center for Health Statistics. Vital and Health Stat 3(38), 2016 Feb. Available from: https://www.cdc.gov/nchs/data/series/sr_03/sr03_038.pdf. Accessed 2021 Dec 5. [PubMed]
- 6.Harris-Kojetin LD, Sengupta M, Lendon JP, Rome V, Valverde R, Caffrey C. Long-term care providers and services users in the United States, 2015-2016. Hyattsville (MD): National Center for Health Statistics. Vital and Health Stat 3(43), 2019 Feb. Available from: https://www.cdc.gov/nchs/data/series/sr_03/sr03_43-508.pdf. Accessed 2021 Sep 21. [PubMed]
- 7.Centers for Medicare & Medicaid Services. CMS Survey and Certification Group 2016/2017 Nursing Home Action Plan: Action Plan for Further Improvement of Nursing Home Quality. 2019 Available from: https://www.cms.gov/Medicare/Provider-Enrollment-and-Certification/CertificationandComplianc/Downloads/2016-2017-Nursing-Home-Action-Plan.pdf. Accessed May 4, 2021.
- 8.US General Accounting Office. Assisted living: Quality-of-care and consumer protection issues in four states. Washington, DC: US Government Printing Office, 1999 April. Available from: https://www.gao.gov/assets/hehs-99-27.pdf. Accessed 2021 Sep 21.
- 9.Centers for Medicare & Medicaid Services. CMS announces partnership to improve dementia care in nursing homes 2012 May [cited 2021 April 27]. Available from: https://www.cms.gov/newsroom/press-releases/cms-announces-partnership-improve-dementia-care-nursing-homes.
- 10.Sloane PD, Zimmerman S, Brown LC, Ives TJ, Walsh JF. Inappropriate medication prescribing in residential care/assisted living facilities. J Am Geriatr Soc. 2002;50(6):1001-11. [PMID: 12110058]. 10.1046/j.1532-5415.2002.50253.x. [DOI] [PubMed]
- 11.Sloane PD, Gruber-Baldini AL, Zimmerman S, et al. Medication undertreatment in assisted living settings. Arch Intern Med. 2004;164(18):2031-7. [PMID: 15477439]. https://10.1001/archinte.164.18.2031. [DOI] [PubMed]
- 12.Kronhaus A, Fuller S, Zimmerman S, Reed D. Prevalence and medication management of dementia by a medical practice providing onsite care in assisted living. J Am Med Dir Assoc. 2016;17(7):673.e9-.e15. [PMID: 27346654]. 10.1016/j.jamda.2016.04.028. [DOI] [PubMed]
- 13.Sloane PD, Zimmerman S, Reed D, et al. Antibiotic prescribing in 4 assisted-living communities: incidence and potential for improvement. Infect Control Hosp Epidemiol. 2014;35(S3):S62-S8. [PMID: 25222900]. 10.1086/677821. [DOI] [PubMed]
- 14.Thomas K, Wretman CJ, Sloane PD, et al. To What Extent Do Local Nursing Home Prescribing Patterns Relate to Psychotropic Prescribing in Assisted Living? J Am Med Dir Assoc. 2020. [PMID: 33382990]. 10.1016/j.jamda.2020.11.037. [DOI] [PMC free article] [PubMed]
- 15.Beeber AS, Zimmerman S, Wretman CJ, Palmertree S, Patel K, Sloane PD. Potential Side Effects and Adverse Events of Antipsychotic Use for Residents With Dementia in Assisted Living: Implications for Prescribers, Staff, and Families. J Appl Gerontol. 2021. [PMID: 34160299]. 10.1177/07334648211023678. [DOI] [PMC free article] [PubMed]
- 16.Dys S, Tunalilar O, Carder P. Cognition-Enhancing, Antipsychotic, and Opioid Medication Use Among Assisted Living and Residential Care Residents in Oregon. J Am Med Dir Assoc. 2021. [PMID: 33516674]. 10.1016/j.jamda.2020.12.030. [DOI] [PubMed]
- 17.McNabney MK, Samus QM, Lyketsos CG, et al. The spectrum of medical illness and medication use among residents of assisted living facilities in central Maryland. J Am Med Dir Assoc. 2008;9(8):558-64. [PMID: 19083289]. https://10.1016/j.jamda.2008.03.003. [DOI] [PMC free article] [PubMed]
- 18.Carder P, Zimmerman S, Wretman CJ, Preisser JS, Dys S, Sloane PD. Pro Re Nata Prescribing and Administration of Psychotropic Medications in Assisted Living: A 7-State Study. J Am Med Dir Assoc. 2021. [PMID: 34919835]. https://10.1016/j.jamda.2021.11.009. [DOI] [PMC free article] [PubMed]
- 19.Briesacher B, Stuart B, Doshi J. Medication use by Medicare beneficiaries living in nursing homes and assisted living facilities. Prepared for Office of Disability, Aging and Long-Term Care Policy, US Department of Health and Human Services, 2002 June. Available from: https://aspe.hhs.gov/sites/default/files/migrated_legacy_files//40606/meduse.pdf. Accessed 2021 Sep 21.
- 20.Kasper JD, Freedman VA. National Health and Aging Trends Study User Guide: Rounds 1-10 Final Release. Baltimore: Johns Hopkins University School of Public Health, 2021 Available from: https://nhats.org/sites/default/files/2021-07/NHATS_User_Guide_R10_Final_Release.pdf. Accessed September 21, 2021.
- 21.2019 American Geriatrics Society Beers Criteria® Update Expert Panel. American Geriatrics Society 2019 updated AGS Beers Criteria® for potentially inappropriate medication use in older adults. J Am Geriatr Soc. 2019;67(4):674-94. [PMID: 30693946]. 10.1111/jgs.15767. [DOI] [PubMed]
- 22.Zimmerman S, Sloane PD, Reed D. Dementia prevalence and care in assisted living. Health Aff. 2014;33(4):658-66. [PMID: 24711328]. 10.1377/hlthaff.2013.1255. [DOI] [PubMed]
- 23.Freedman VA, Spillman BC. The residential continuum from home to nursing home: size, characteristics and unmet needs of older adults. J Gerontol B Psychol Sci Soc Sci. 2014;69(Suppl_1):S42-S50. [PMID: 25342822]. 10.1093/geronb/gbu120. [DOI] [PMC free article] [PubMed]
- 24.Maust DT, Strominger J, Kim HM, et al. Prevalence of central nervous system–active polypharmacy among older adults with dementia in the US. JAMA. 2021;325(10):952-61. [PMID: 33687462]. 10.1001/jama.2021.1195. [DOI] [PMC free article] [PubMed]
- 25.Kasper JD, Freedman VA, Spillman BC. Classification of persons by dementia status in the National Health and Aging Trends Study. Baltimore (MD): Johns Hopkins University School of Public Health, 2013 July. Available from: https://nhatspubdemo.westat.com/scripts/documents/NHATS_Dementia_Technical_Paper_5_Jul2013.pdf. Accessed 2021 Sep 21.
- 26.Maust DT, Gerlach LB, Gibson A, Kales HC, Blow FC, Olfson M. Trends in central nervous system–active polypharmacy among older adults seen in outpatient care in the United States. JAMA Intern Med. 2017;177(4):583-5. [PMID: 28192559]. 10.1001/jamainternmed.2016.9225. [DOI] [PMC free article] [PubMed]
- 27.Freedman V, Spillman B. Making National Estimates with the National Health and Aging Trends Study. Baltimore (MD): Johns Hopkins University School of Public Health, 2018 Jan. Available from: https://nhats.org/sites/default/files/2021-01/Making_National_Estimates_in_NHATS_Technical_Paper_2018_0.pdf. Accessed 2021 Sep 21.
- 28.Dwyer LL, Carder PC, Harris-Kojetin LD. Medication management services offered in US Residential Care Communities. Seniors Housing & Care Journal. 2014.
- 29.Maher RL, Hanlon J, Hajjar ER. Clinical consequences of polypharmacy in elderly. Expert Opin Drug Saf. 2014;13(1):57-65. [PMID: 24073682]. 10.1517/14740338.2013.827660. [DOI] [PMC free article] [PubMed]
- 30.Elon R, Pawlson LG. The impact of OBRA on medical practice within nursing facilities. J Am Geriatr Soc. 1992;40(9):958-63. [PMID: 1512394]. 10.1111/j.1532-5415.1992.tb01996.x. [DOI] [PubMed]
- 31.Zimmerman S, Scales K, Wiggins B, Cohen LW, Sloane PD. Addressing antipsychotic use in assisted living residents with dementia. J Am Geriatr Soc. 2015;63(9):1970-1. [PMID: 26390002]. 10.1111/jgs.13625. [DOI] [PubMed]
- 32.Carder PC, O'Keeffe J. State regulation of medication administration by unlicensed assistive personnel in residential care and adult day services settings. Res Gerontol Nurs. 2016;9(5):209-22. [PMID: 27054368]. 10.3928/19404921-20160404-03. [DOI] [PubMed]
- 33.Ohio Department of Aging Office of the State Long-Term Care. Assisted Living (Residential Care) Facilities [cited 2021 Sep 21]. Available from: https://ltc.ohio.gov/wps/portal/gov/ltc/care-information/care-options/assisted-living.
- 34.Shippee T, Akosionu O, Brasure M, Beebe T. Literature review and environmental scan: Identifying quality measures in assisted living. A report to the Minnesota Department of Human Services. University of Minnesota School of Public Health, 2019 June. Available from: https://mn.gov/dhs/assets/UMN-assisted-living-quality-report_tcm1053-393870.pdf. .
- 35.Oregon Department of Human Services. Quality Measurement Program Report for Oregon Community-Based Care Facilities: 2020 Annual Report. 2021 Available from: https://www.oregon.gov/dhs/PROVIDERS-PARTNERS/LICENSING/CBC/QMCReports/Quality-Measurement-Report-2020.pdf. .
- 36.Health Care Association of New Jersey Foundation. Advanced Standing Program 2012 [cited 2021 September 15]. Available from: https://www.hcanj.org/files/2013/09/advanced-standing-010213.pdf.
- 37.Wisconsin Coalition for Collaborative Excellence in Assisted Living [updated 2021 Sep; cited 2021 September 15]. Available from: https://wcceal.chsra.wisc.edu/.
- 38.National Center for Assisted Living. National Quality Award Program [cited 2021 September 15]. Available from: https://www.ahcancal.org/Quality/National-Quality-Award-Program/Pages/default.aspx.
- 39.The Joint Commission. The Joint Commission launches new Assisted Living Community accreditation program 2021 May [cited 2021 September 15]. Available from: https://www.jointcommission.org/resources/news-and-multimedia/news/2021/05/the-joint-commission-launches-new-assisted-living-community-accreditation-program/.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
(DOCX 42.7 kb)