State Variation in Antipsychotic Use among Assisted Living Residents with Dementia

Abstract

Objectives:

Over two-thirds of assisted living (AL) residents have dementia or cognitive impairment and antipsychotics are commonly prescribed for behavioral disturbances. As AL communities are regulated by state-level policies, which vary significantly regarding the care for people with dementia, we examined how antipsychotic prescribing varied across states among AL residents with dementia.

Design:

This was an observational study using 20% sample of national Medicare data in 2017.

Setting and Participants:

The study cohort included Medicare beneficiaries with dementia aged 65 years or older who resided in larger (25+ bed) ALs in 2017.

Methods:

The study outcome was the percentage of eligible AL person-months in which antipsychotics were prescribed for each state. We used a random intercept linear regression model to shrink estimates toward the overall mean use of antipsychotics addressing unstable estimates due to small sample sizes in some states.

Results:

A total of 20,867 AL residents with dementia were included in the analysis, contributing to 194,718 person-months of observation. On average, AL residents with dementia were prescribed antipsychotics during 12.6% of their person-months. This rate varied significantly by state, with a low of 7.8% (95%CI 5.9%–10.3%) for Hawaii to a high of 20.5% (95%CI 16.4%–25.3%) for Wyoming.

Conclusions and implications:

We observed significant state variation in the prescribing of antipsychotics among AL residents with dementia using national data. These variations may reflect differences in state regulations regarding the care for AL residents with dementia and suggest the need for further investigation to ensure high quality of care.

Brief summary:

Among assisted living residents with dementia in the US, state-level prevalence of antipsychotic use varied widely from 7.8% to 20.5% in 2017. These variations may be attributable to different state regulations pertaining to antipsychotic use in assisted living settings.

INTRODUCTION

Antipsychotic medications are associated with serious adverse effects, including fall-related injuries, cardiovascular events, and mortality, among persons with dementia.¹ Despite the U.S. Food & Drug Administration Black Box Warnings about these harms,^2,3 antipsychotics are still prescribed to many individuals living with dementia for the treatment of behavioral disturbances.^4,5

Assisted living (AL) provides long-term care to older and disabled adults in the US, the majority of whom are female (67%), non-Hispanic white (89%), age 85 years and older (55%), and have dementia or cognitive impairment (71%).⁶ Although antipsychotic prescribing in nursing homes across the United States is subject to federal oversight, by contrast, individual states regulate AL communities. State-level policies vary considerably with respect to the care of AL residents with dementia and the management of antipsychotic prescribing.^7–9 For instance, in 2015, Kaskie, Nattinger & Potter reported that several states strictly prohibited the use of chemical restraints, including psychotropic medications. Other states articulated it was an AL resident’s right to be free from chemical restraint, and nearly half the states permitted their use under the direction of a physician or during an emergency.¹⁰ Thirteen states had no direct policy regulating psychotropic medication use in AL settings. Despite these documented variations in states’ AL regulations, there is no evidence on between-state variations in antipsychotic prescribing in AL settings using national data. To address this gap in knowledge, we conducted this study to examine how antipsychotic prescribing among AL residents with dementia varied across states.

METHODS

Study Design and Data Sources

This was a one-year observational study based on a 20% representative sample of Medicare beneficiaries aged 65 years or older in 2017. We used the Medicare Master Beneficiary Summary File (MBSF; demographics and enrollment) and the Enrollment Data Base (EDB; residential ZIP codes) linked to Medicare Part D event files (prescription drug claims), Medicare Provider Analysis and Review (MedPAR; inpatient, skilled nursing facility stays), outpatient, carrier (visits and diagnoses), hospice and home health claims. The Brown University institutional review board approved the study.

Study Population

The study population included Medicare-enrolled AL residents with dementia. We identified person-months in which beneficiaries were enrolled in Medicare fee-for-service Parts A, B and D and resided in an AL setting as the highest level of care for at least one day during that month. We determined AL residency based on ZIP+4 codes recorded in the EDB that corresponded to licensed ALs with 25+ beds in 2017 (see Table S1 for licensure terms used to identify AL).¹¹ Person-months in which a beneficiary made the transition between AL and other settings (e.g., nursing homes) were excluded.

A beneficiary was determined to have dementia based on diagnosis codes in Medicare claims for inpatient, outpatient, home health, hospice and skilled nursing facility services during 2016 and 2017 (Table S2).¹² We excluded individuals younger than 65 years, those without a dementia diagnosis, and those with Huntington’s disease, schizophrenia or Tourette syndrome (Table S2), consistent with federal nursing home antipsychotic quality measures.¹³ We further aggregated individual level records to state-level data to examine state-level antipsychotic prescribing among AL residents with dementia in 2017.

Outcome

The study outcome was the percentage of eligible AL person-months in which antipsychotics were prescribed in each state during 2017. Antipsychotics were identified from Medicare Part D prescription drug claims (Table S3) and were examined as a single therapeutic class.

Statistical Analysis

Considering the multilevel nature of our data - residents were nested within ALs and ALs were further nested within states, we initially considered a 3-level model to examine antipsychotic use across states. However, due to small sample sizes in some ALs in some states that may lead to unstable estimates, we employed a two-level random hierarchical intercept linear regression model with individuals (level 1) clustered within states (level 2), to shrink estimates toward the overall mean use of antipsychotics.

All analyses were conducted using SAS version 9.4 (SAS Institute Inc) and R version 4.0.4 (February 2021).

RESULTS

There were 20,867 AL residents with dementia who met our inclusion criteria: they had a mean age of 85.4 (standard deviation [SD]: 7.4 years), 71.2% were female, and 93.9% were White. These individuals contributed 194,718 months of person-time.

AL residents with dementia had a prescription filled for an antipsychotic during 12.6% of their person-months. This rate varied significantly from a low of 7.8% (95%CI 5.9%–10.3%) for Hawaii to a high of 20.5% (95%CI 16.4%–25.3%) for Wyoming (Figure 1). The majority of states (two-thirds) had an average exposure rate between 10% and 15%.

Figure 1.

Percentages of person-months with antipsychotic claims among assisted living residents with dementia aged 65 years or older in 2017 (n=194,718).

* Findings were based on a random intercept linear regression model that was applied to shrink estimates towards the overall mean of antipsychotic use, thereby accounting for instability in the estimates of states with smaller sample sizes.

DISCUSSION

Despite the Black Box Warning, antipsychotics were prescribed among people with dementia who reside in AL settings; state level rates are generally consistent with prior estimates from seven states (New Jersey ~7%, Louisiana ~8%, Oklahoma ~11%, Arkansas, New York, Pennsylvania and Texas ~12%).¹⁴ Future research is needed to explain the observed state-level variations in antipsychotic use among AL residents.

The variation in antipsychotic exposure may be a function of variability in state regulations that impact antipsychotic use in AL (e.g., chemical restraints, staff training requirements for dementia care, level of care permitted, monitoring programs).^9,15 It could also be that the variability we observed is a function of local practice patterns; however, prior work has not found a relationship between antipsychotic use in local ALs and NHs, nor do these rates in AL correspond to state rankings in antipsychotic use among NH residents during the same time period.^4,14 Future work is needed to understand the drivers behind the variability we observed across states and if it is a function of multiple factors, such as inconsistent state regulations, local antipsychotic prescribing patterns, and availability of providers with experience in the care of older adults with dementia.

There are important limitations of this study to note. First, as we relied on ZIP codes to identify AL residents, we only included AL facilities with 25+ beds as smaller ALs are likely to share ZIP codes with non-AL addresses. However, according to 2016 data, while our approach may only identify ~39% of all AL facilities, facilities with 25+ beds comprise ~84% of all licensed beds nationally.¹⁶ Second, our study sample may have included beneficiaries residing in independent living on the same campus that shared a ZIP code with an AL.¹¹ Third, our approach was based on fee-for-service Medicare beneficiaries and did not capture AL residents who were not enrolled in the Medicare program (e.g., younger than age 65) or who were enrolled in Medicare Advantage - the managed care program within Medicare which accounted for 40% of Medicare beneficiaries residing in ALs in 2017. Fourth, we do not have access to clinical information, e.g., dementia severity, in our administrative claims data. However, we have excluded residents with Huntington’s disease, Tourette’s syndrome, and schizophrenia, which are indications for antipsychotic treatment. Additionally, no medications have been approved by the U.S. FDA for the treatment of dementia-associated neuropsychiatric symptoms, including agitation, and the Black Box Warning for antipsychotics in dementia is silent with respect to dementia severity. Future work would have to determine whether dementia severity, and more specifically, behavioral symptoms vary across states. Fifth, our study was based on 2017 data; findings may not be generalizable to recent years when state-level antipsychotic related policies were implemented in AL,⁹ or during or after COVID-19 pandemic. Further studies are warranted to examine how antipsychotic use has changed in response to new state-level antipsychotic-related policies and the pandemic. Finally, our analysis did not take into account potential within-state factors that may contribute to the observed variations in antipsychotic use, such as different licensure requirements within a state for the care of residents within AL communities,^7,17 and varying local practice patterns related to medication prescribing.^9,14,18

CONCLUSIONS AND IMPLICATIONS

Our study provides new insights into state variation in the prescribing of antipsychotics among AL residents with dementia using national data. While the Black Box Warning suggests that antipsychotics should not be prescribed for older adults with dementia, there may be circumstances during which their prescribing is clinically appropriate. Further research is needed to determine how variation in the use of antipsychotics may relate to health outcomes for AL residents with dementia. Further exploration of the relationship between state laws governing dementia care and appropriate medication prescribing for the population living with dementia in AL is also warranted to ensure resident safety.