State Medicaid Financing and Access to Large Assisted Living Settings for Medicare–Medicaid Dual-Eligibles

Abstract

Assisted living has become more widely used by dual-eligible Medicare beneficiaries as states try to rebalance their long-term services and supports away from institutional (nursing home) care. In an analysis of 2014 Medicare data for 506,193 adults who live in large (25+ beds) assisted living communities, we found wide variability among states in the share of assisted living residents who were dual-eligible, ranging from 6% in New Hampshire to over 40% in New York. This variation is strongly correlated with the Medicaid support for assisted living care: In states with a Medicaid state plan option covering services in assisted living or both a state plan and waiver, the percent of assisted living residents with dual-eligibility was more than 10 percentage points higher than in states with neither a state plan nor waiver. Findings provide a basis for understanding the role of Medicaid financing in access to assisted living for duals.

Introduction

More than 800,000 Americans reside in assisted living settings, which are residential settings that provide or coordinate 24-hour supervision and assistance, activities, and health-related personal care services (Harris-Kojetin et al., 2019; Hawes et al., 2003; Mollica, 2009). Assisted living settings vary in size, although large settings (25+ beds) made up 40% percent of all assisted living settings and comprised 85% of all licensed beds in 2016, nationally (Harris-Kojetin et al., 2019). Most residents finance their care with personal resources (Genworth, 2019; Mollica, 2009; Stevenson & Grabowski, 2010). However, as states rebalance their long-term care services away from institutional care toward home and community-based services (HCBS), there is an increasing number of residents relying on Medicaid to finance personal care and other supportive services in assisted living. In 2014, over 330,000 Medicare and Medicaid beneficiaries (“duals”) received services in assisted living in 48 states, resulting in more than $10 billion in spending (U. S. Government Accountability Office, 2018). Information about the demographic and health characteristics of duals in assisted living, as well as how states’ Medicaid programs enable their access to care, is needed to inform policy and the delivery of care to this growing and increasingly vulnerable population.

Access to assisted living for duals varies from that of other Medicare beneficiaries for several reasons. First, a private bed in assisted living costs $48,612 per year (Genworth, 2019), on average, making it cost prohibitive for many duals. Second, while 48% of assisted living settings were Medicaid-certified (National Center for Assisted Living, 2019), settings are less often available in counties with lower educational attainment and median household incomes (Stevenson & Grabowski, 2010; Fabius & Thomas, 2019; Cornell et al., 2020). As a result, duals living in low-income counties may have limited access to assisted living due to affordability and availability. Third, payment rates from Medicaid are often lower than that of private pay residents. Therefore, assisted living providers might limit the number of dual-eligible residents they serve (or not serving this population at all).

While Medicare beneficiaries in assisted living may temporarily receive Medicare-financed services such as home health (Nazareno et al., 2019), Medicare does not pay for HCBS provided in assisted living. To enable duals to receive services in assisted living to assist with health and functioning needs, states can offer services under (1) a Medicaid state plan, which allows for personal care and other services to be covered for all Medicaid beneficiaries; (2) a Medicaid HCBS waiver (e.g., 1915(c) that may include specialized assisted living waivers; 1,115 waivers), which requires individuals to meet certain state criteria based on geography or level of need and can cap enrollment; or (3) both a state plan and waiver. As such, states differ in their Medicaid financing of services in assisted living (Carder et al., 2015). In 2014, four states had only a Medicaid state plan available to cover services in assisted living, 32 states and the District of Columbia offered Medicaid waiver options only, and 10 states offered both waivers and state plans. Five states did not use Medicaid programs to pay for services in assisted living settings (Carder et al., 2015).

Federal regulation does not allow for Medicaid financing of room and board—these payments must come from residents’ incomes or other state funds (Lepore et al., 2017; U. S. Government Accountability Office, 2018). To assist with costs, states may set a cap on room and board charges for Medicaid beneficiaries that is equal to the amount a beneficiary receives as Supplemental Security Income (Carder et al., 2015). States can also offer an optional state supplement (OSS) to individuals who receive full Medicaid benefits and reside in a licensed assisted living setting. The availability of Medicaid financing and room and board assistance in a state are likely essential factors that impact whether duals can access assisted living (Carder et al., 2015; U. S. Government Accountability Office, 2018).

New Contribution

In the present study, we leverage a new methodology (Thomas et al., 2016) to identify Medicare beneficiaries residing in large assisted living settings with two objectives: (1) compare the characteristics of duals to nonduals among assisted living residents and (2) examine the relationship between Medicaid financing for services in assisted living and states’ proportions of residents in large settings who are duals. We hypothesize that states with more generous Medicaid financing (i.e., waiver and/or state plan), and room and board policies (i.e., OSS payment and/or room and board caps) will have higher proportions of residents in large assisted living settings who are duals compared to states without these supports. Our study provides important information to policymakers and providers on the frontlines of long-term services and support (LTSS) rebalancing about the role of Medicaid financing in access to large assisted living settings for dual-eligibles, as well as the characteristics of this growing population.

Conceptual Framework

To better understand the characteristics of duals, as well as how the proportion of dual-eligible assisted living residents may vary by state and the role of Medicaid financing and assisted living room and board policies, we propose a conceptual model based on the socioecological model (Bronfenbrenner, 1977; Greenfield, 2012) to describe factors influencing a person’s entry to assisted living. The model emphasizes cross-level influences in which community and policy levels can shape individual behavior (i.e., access to assisted living). We propose that assisted living entry depends on factors nested within the three levels. First, an individual’s level of need (e.g., chronic conditions) and the availability of informal support as well as finances likely influence whether a person decides to transition to assisted living from a traditional community-based living arrangement or skilled nursing facility. Second, community-level factors include the availability of assisted living as well as assisted living providers’ characteristics, such as their willingness to accept Medicaid, and may enable or prevent older adults from accessing assisted living. Third, policy-level factors, like federal and state regulations (e.g., Medicaid waiver and state plan availability, assisted living room and board policies) have a direct impact on other levels, because Medicaid financing options may or may not be available for assisted living.

Method

This study uses a retrospective cohort study design to compare the characteristics of Medicare beneficiaries residing in large (25+-bed) assisted living settings who are duals to nonduals and to examine the variability of the prevalence of duals in large assisted living settings across states. A national list of licensed or certified assisted living providers was collected from state agencies. We used beneficiaries’ nine-digit ZIP codes reported in the Medicare Master Beneficiary Summary File and the Residential History File to identify cohorts of Medicare beneficiaries residing in large assisted living settings on December 31, 2014 (n = 506,193). Residents in Alaska and Hawaii were excluded due to differences in Medicare practices in assisted living. Additionally, too few assisted living settings were in Washington DC to include in the study. Dual eligibility was assigned to individuals if they were Medicare and Medicaid enrolled for the month of December. State policies (i.e., Medicaid coverage of services, room and board caps and subsidies) were compiled from the Compendium of Residential Care and Assisted Living Regulations and Policy (Carder et al., 2015).

Statistical Analyses

We first compare demographic, entitlement, and health characteristics of assisted living residents in large settings, by dual-eligibility status. Chronic health conditions are reported for Medicare beneficiaries enrolled in Traditional, Fee-for-Service (FFS) Medicare for the entire year (n = 362,461). Bivariate comparisons and tests for statistical significance using Pearson’s chi-square are reported. Next, we report the percent of assisted living residents who are dually eligible by state and describe the interstate variability of this statistic by comparing it to the percent of community-dwelling Medicare beneficiaries who are duals in a state. Last, we perform ordinary least squares regression to estimate the association between state policies and states’ adjusted percentage of duals in assisted living. We also calculated an adjusted percentage of duals in assisted living, as follows: [state observed assisted living dual percent] / [state percent dual in the community] * [national percent dual], where the “state percent dual” and “national percent dual” are the state and national percentages of Medicare beneficiaries not presently residing in an institutional setting (i.e., nursing home or assisted living) who are dual-eligible. The rationale for using community-dwelling dual eligibles as a benchmark is to adjust for state levels of poverty in the community. Nursing home residence was determined through the presence of a skilled nursing facility Medicare claim or Minimum Data Set assessment. To examine other aspects of state support for assisted living, we include variables indicating the state’s maximum OSS room-and-board supplement, in dollars, as well as an indicator for whether the state has a cap on room and board charges for Medicaid residents of assisted living. Statistical analyses were completed using Statistical Analysis Software.

Results

Table 1 presents descriptive characteristics of duals and nonduals residing in large (25+-bed) assisted living settings in 2014. Of the 506,193 Medicare beneficiaries, we were able to identify in large assisted living settings in December 2014, 18% were duals. Black, Hispanic, and other races and ethnicities (Asian American and Native American) were more likely to be duals than Whites. Duals were more likely to qualify for Medicare due to disability rather than age. The average length of dual eligibility was 11.4 months among duals. Among Medicare FFS beneficiaries, compared with nonduals, duals had higher rates of all chronic conditions, except for arthritis.

Table 1.

Characteristics of a Cohort of Residents in Large Assisted Living Settings, by Dual-Eligibility Status^a (N = 506,193).

	Dual		Nondual
	n	%	n	%
Characteristic	89,636	18%	416,557	82%
Age group (years)
<65	26,220	29.3	10,118	2.4
65–74	23,076	25.7	64,716	15.5
75–84	20,413	22.8	117,680	28.3
85+	19,927	22.2	224,043	53.8
Sex
Male	33,346	37.2	136,469	32.8
Female	56,290	62.8	280,088	67.2
Race
White	68,178	76.1	394,903	94.8
Black	10,741	12.0	7,919	1.9
Hispanic	6,631	7.4	5,727	1.4
Other	4,086	4.6	8,008	1.9
Original reason for Medicare entitlement
Age	47,158	52.6	387,925	93.1
Disability	42,011	46.9	28,212	6.8
Months of dual eligibilityb, M (SD)	11.4		0.05
Medicare advantagec	26,382	29.4	117,350	28.2
Health characteristics (Medicare Fee-For-Service beneficiariesd
At least three chronic conditionse	53,917	60.2	239,110	57.4
At least six chronic conditionse	30,109	33.6	119,183	28.6
Alzheimer’s disease or related dementia	23,667	26.4	83,760	20.1
Arthritis	27,768	31	134,623	32.3
Chronic obstructive pulmonary disorder	16,829	18.8	39,522	9.5
Congestive heart failure	18,272	20.4	74,041	17.8
Depression	28,386	31.7	70,506	16.9
Diabetes	29,312	32.7	74,977	18.0
Hypertension	48,259	53.8	214,465	51.5
Nursing home stay	2,894	3.2	2,020	0.5

Note. Pearson chi-square tests conducted for all comparisons. All comparisons significant at p < .001 level. Source: Authors’ analysis of data from the 2014 Master Beneficiary Summary File and Residential History File.

^aDual-eligibility: at least 1 month of Medicare-Medicaid dual eligibility in the year.

^bMedicaid eligibility standard deviation: 2.0 for duals and 0.6 for nonduals.

^cMedicare Advantage: at least one month of Medicare Advantage coverage in the year.

^dChronic conditions reported for n = 362,461 Medicare Fee-ForService beneficiaries only.

^eNumber of conditions calculated based on a list of 22 conditions reported in the Chronic Condition Warehouse.

We present the share of assisted living residents who are dual versus the share of community-dwelling individuals who are duals, by state, in Figure 1. The unadjusted percent of assisted living residents who are duals range from 6% in New Hampshire to 44% in New York. States above and to the left of the plotted line have a higher share of their assisted living residents who are duals than their community residents. For example, in Utah, 12% of assisted living residents were dual-eligible, versus 8% of community residents. In other states, differences were more pronounced. In Maine, 42% of assisted living residents were duals, versus 28% of community residents. In New York, 44% of assisted living residents were duals, versus 23% of community residents (see the Supplemental Appendix Tables 1 and 2, available online, for state percentages of duals in assisted living and community-dwelling residents).

Figure 1.

Percent of dual-eligible community-dwelling and assisted living residents, by state.

In the adjusted regression (Table 2; Model 3), states with a Medicaid state plan and waiver have a 12.9 percentage point difference in the share of duals among their residents in large assisted living settings than states with no Medicaid assisted living support (p = .003). States with only a state plan have roughly a 14.0 percentage point difference in the share of residents in large assisted living settings who are duals than states that offer no support (p = .009). Relative to states with no support, states with waivers had 7.6 percentage points higher prevalence of duals (p = .044). Although not statistically significant, we found a positive association between the state’s OSS maximum benefit and the percentage of duals in assisted living, and a negative association with room and board cap. These results were robust to using a more restrictive criteria that excluded beneficiaries whose nine-digit ZIP codes did not exactly match that of the assisted living community’s mailing address, as well as criteria that exclude partial dual-eligibles who often have less access to Medicaid HCBS services in assisted living (see Supplemental Appendix Tables 3-6).

Table 2.

Association of Medicaid Financing Options With Percentage of Duals in Assisted Living.

	n (%) or M (SD)	Regression of percent dual
		(1) Percent dual	(2) Adjusted percent duala	(3) Adjusted percent duala
State Medicaid financing options
No Medicaid AL support	5 (10%)	Om.	Om.	Om.
State plan and waiver	10 (21%)	13.17*** (4.376)	12.65*** (4.138)	12.89*** (3.501)
State plan only	4 (8%)	10.04* (5.360)	14.25*** (5.068)	13.97*** (4.294)
Waiver only	29 (60%)	2.88 (3.87)	7.11* (3.61)	7.63** (3.095)
State room and board option
Optional state supplement maximumb	$289 (320)			0.00299 (0.00365)
State has room and board cap for Medicaid AL residentsc	15 (31%)			−1.32 (2.41)
Constant		11.23*** (3.308)	9.740*** (2.812)	10.14*** (3.70)
Observations		48	48	48
R 2		0.266	0.221	0.243

^*p < .1.

^**p < .05.

^***p < .01.

Note. 506,193 Medicare beneficiaries residing in large AL in the United States (excluding AK, HI, DC) on December 31, 2014. The dependent variable in regression models is the state’s percent of Medicare beneficiaries in large AL who are dual eligible. AL = assisted living; Om. = omitted category; SP = state plan option. Source: Authors’ analysis of data from the 2014 Master Beneficiary Summary File, Residential History File, and Carder et al. (2015).

^aAdjusted for percent total state Medicare population that is dual eligible and not living in an institutional setting.

^bA total of 28 states have no optional state supplement.

^c22 States have no room and board cap for Medicaid AL residents.

Discussion

We describe the characteristics of a cohort of dual-eligible Medicare beneficiaries residing in large (25+ beds) assisted living settings and examine the role of state Medicaid financing and room and board regulations in a state’s share of assisted living residents who are duals. Our hypotheses are partially supported—states with Medicaid financing for assisted living services have a significantly higher share of residents in large settings who are dual-eligible relative to states that offer no Medicaid financing for services in assisted living. There was no significant association between the proportion of assisted living residents who were duals and state support for room and board. We also find substantial differences in sociodemographic and health characteristics by dual-eligibility status: Duals represent 18% of assisted living residents in our cohort and are more often younger, minorities, and disabled, with higher rates of chronic conditions and cognitive impairment.

The associations observed between Medicaid financing options and the presence of duals in large assisted living settings is consistent with the design of these benefits. For example, state plans are entitlements that must provide services to all beneficiaries who meet eligibility criteria (Meucci et al., 2018). Beneficiaries do not have to meet nursing home level of care to qualify for services, so duals with fewer functional or cognitive limitations likely require fewer services in assisted living—providers may be incentivized to accept these individuals. Conversely, the coverage restrictions allowable by waivers, including the ability for states to restrict enrollment for HCBS to a certain number of beneficiaries (Meucci et al., 2018), may result in waiver waitlists that delay or prevent entry to assisted living. In 2016, 75% of states reported 1915(c) or 1,115 waiver waitlists, totaling 656,195 individuals, including those waiting for services in assisted living (Watts & Musumeci, 2018). For duals, waiting for services may result in an increased level of care that exceeds what can be provided in assisted living. Additionally, residents who become dual eligible while in assisted living may face nursing home entry if waiver services cannot be approved in a reasonable amount of time. Moving forward, states might consider implementing approaches to prioritizing waitlists to address the needs of individuals who are at greater risk for nursing home placement or losing their housing, which may directly affect assisted living residents.

We observed a positive (though not statistically significant at the .05 level) association between the state’s maximum allowed OSS benefit and the share of assisted living residents who were dual in the state. A positive relationship would be expected, as the supplement could enable duals to afford room and board. The effect of caps on room and board charges also did not reach a level of statistical significance but had a negative association with the percent of assisted living residents who were duals in the state. Perhaps it is the case that these caps, while intended to make assisted living more affordable, decreases the willingness of providers to serve duals. One room and board option we did not examine is family supplementation. There are varying ideas about what family members should be responsible for, since family supplementation is counted as unearned income and can lead to a reduced SSI (Supplemental Security Income) payment or the loss of SSI and Medicaid, and the implementation of this option varies by state (Carder, et al., 2015; Smith et al., 2010). If the share of duals in assisted living continues to grow, states will likely experience an increasing demand for support for room and board. This additional need may require states to reevaluate the availability of support options as well as the current allocation of resources to room and board assistance, both of which will likely affect the ability of duals to access assisted living.

Relative to nonduals, duals in large assisted living settings have higher rates of disability and chronic conditions and are more likely to have received care in a nursing home. Given that these characteristics make duals in assisted living a particularly vulnerable population, consideration around how best to oversee the care provided in assisted living settings is needed. A 2018 U.S. Government Accountability Office report found that states varied in methods used to monitor the health of Medicaid beneficiaries in assisted living and recommended improving the way in which they oversee critical incidents. If assisted living settings are to continue to serve as an option for duals, policy makers, states, and providers will need to work collaboratively to develop ways to systematically collect and report the health and welfare of dual- eligible residents (U. S. Government Accountability Office, 2018). Given that additional oversight may put burden on assisted living providers who care for duals (Stevenson & Grabowski, 2010; U. S. Government Accountability Office, 2018), special consideration is needed around if and how this may differentially affect large and small assisted living settings with different levels of resources to meet regulatory demands. It is possible that regulatory burdens associated with caring for duals may dissuade providers who can afford the lost revenue (potentially higher quality settings) from accepting duals (Stone & Reinhard, 2007).

Over the past decade, states increasingly used their waivers or state plans to finance HCBS using capitated managed Medicaid LTSS (MLTSS). As of January 2019, 24 states were operating regional or statewide MLTSS programs (Snyder et al. 2019). In this arrangement, Medicaid programs contract with managed care organizations to administer LTSS in an effort to ensure quality and increase efficiency (Medicaid. gov, 2019). We did not examine whether waivers or state plans that are for MLTSS differentially impact the share of assisted living residents who are duals across states relative to other waivers. Therefore, it is unclear whether MLTSS programs increase access for duals as managed care companies aggressively work to provide care in lower cost, less restrictive settings, or if they decrease access because payment rates may not be directly set by state Medicaid agencies, and can be negotiated between managed care organizations and assisted living settings (Stone & Reinhard, 2007). Moving forward, states must consider the access implications of MLTSS for duals who seek care in assisted living.

Findings should also be interpreted in consideration of other factors, such as the HCBS Settings Rule (Centers for Medicare and Medicaid Services, 2014; Coulter & Hummel, 2017). Medicaid eligibility pathways, assisted living licensing, and reimbursement policies. First, CMS published the final HCBS Settings rule in 2014 in efforts to promote community integration, individual privacy and choice, and ensure that HCBS were being delivered in “home-like” environments. As a result, some assisted living settings may have become ineligible to receive Medicaid reimbursement for services (Centers for Medicare and Medicaid Services, 2014), including some of the settings we identified in 2014. Those at risk of becoming ineligible include settings with “institution-like qualities,” such as secured dementia/memory care units, separate or converted sections of skilled nursing facilities, and settings that are in rural and isolated areas (Coulter & Hummel, 2017). Settings opting not to make recommended changes, or lacking the resources to do so, would no longer be able to serve duals. It is unclear whether these changes would result in more duals remaining in the community in other settings, or transitioning into institutions, given the acuity and need we observed in this study. This is further complicated by the COVID-19 pandemic, which has disproportionately impacted older adults, racial and ethnic minorities, and residents in long-term care settings (Laurencin & McClinton, 2020; Ouslander & Grabowski, 2020; Temkin-Greener et al., 2020). Assisted living settings may have difficulty maintaining home-like environments as they implement social distancing efforts (American Geriatrics Society, 2020; Zimmerman et al., 2020). Additionally, assisted living settings that serve more dual-eligibles may be particularly affected by COVID-19, because they have higher rates of racially/ethnically diverse residents, have fewer resources, and are located in areas of greater socioeconomic disadvantage relative to settings that receive the majority of their revenue from private-pay residents (Cornell et al., 2020).

Second, 33 states expand Medicaid coverage to individuals with high medical expenses (“medically needy”) who would otherwise be ineligible for Medicaid because their incomes exceed eligibility limits. Relative to nursing home residents requiring more assistance with daily tasks related to health and functioning, assisted living residents with fewer functional limitations may have difficulty incurring enough medical expenses to meet spend-down requirements (Carder et al., 2014; Smith et al., 2010). States may also implement the special income rule, which allows people who require an institutional level of care to qualify for Medicaid with incomes up to 300% of the SSI level ($2,199 per month for individuals in 2015; Watts et al., 2016). In these states, individuals may qualify for Medicaid by having their income administered through a qualified income trust (Woodcock et al., 2011). These eligibility paths may influence whether Medicaid beneficiaries are able to enter assisted living, or whether they are able to qualify for Medicaid and waiver or state plan financed services once residing in assisted living.

Third, because assisted living settings are licensed at the state level (Carder et al., 2015), there are likely state-by-state differences in whether a setting is either eligible to accept duals or willing to comply with Medicaid’s requirements. Medicaid may set stringent standards, such as only contracting with settings that offer single-occupancy units (Carder et al., 2015). States may also require settings to reserve a number of units for Medicaid-eligible residents. For instance, assisted living settings licensed after 2001 in New Jersey must designate at least 10% of their units to serve Medicaid-eligible residents (Carder et al., 2015). State Medicaid programs looking to increase access to assisted living for duals must consider the regulations that govern assisted living.

Fourth, Medicaid reimbursement policies likely influence assisted living providers’ decisions to serve duals. Providers may be less willing to provide services during a spend-down period to those eligible for Medicaid through the “medically needy” category, as payment cannot be guaranteed and collection may be difficult (Carder et al., 2014; Smith et al., 2010). In HCBS waiver programs, payment methodologies vary and may relate to providers’ willingness to accept duals. For example, flat rates reimburse providers the same daily amount, motivating them to admit individuals with fewer care needs and discharge duals once their needs exceed the allotted daily rate (Mollica, 2009). Tiered rates reimburse providers based on the needs of individuals by constructing payment levels based on the functional, cognitive, or behavioral impairments of residents, incentivizing providers to serve residents with high levels of need (Mollica, 2009). Finally, FFS rates are determined by the amount of services identified in a care plan or assessment (Mollica, 2009). States’ understandings of the implications of these different payment methodologies are important as they may influence decisions to admit and retain duals requiring long-term care.

Our study has several limitations. First, we are only able to identify a cohort of Medicare beneficiaries residing in large (25+ beds) assisted living settings; therefore, it is unclear whether findings hold true in smaller settings where it is reported that a higher proportion of residents are duals (Han et al., 2017; Park-Lee et al., 2011). However, the greatest number of duals reside in large settings (Park-Lee et al., 2011). Additional work is needed to understand the characteristics of duals in smaller settings and how Medicaid policy impacts assisted living access. Second, we are unable to determine the timing of Medicaid eligibility in our sample. Our criterion for dual eligibility may include people who qualify through a state’s medically-needy category of eligibility. While the average length of time of Medicaid eligibility in our sample is 11.4 months, and we restricted the sample to individuals who were not in a nursing home on December 31, 2014, future research should focus on the needs of individuals on spenddown and whether they vary from that of traditional Medicare beneficiaries or duals who have been Medicare–Medicaid eligible for a longer amount of time. Third, we use Medicare administrative data, including ZIP codes where beneficiaries receive their mail, to identify our cohort of assisted living residents. Therefore, our sample is limited to Medicare beneficiaries and will not capture beneficiaries who have not updated their addresses to reflect their place of residence. Our reliance on ZIP codes to identify our sample may also mean that we inadvertently include beneficiaries who reside in other types of settings, such as less restrictive independent living settings with fewer services, that are colocated with a licensed assisted living setting and, therefore, share the same nine-digit ZIP code. Still, there is no national publicly available individual-level data to analyze for assisted living settings, and our methodology yields the best results to date. Fourth, our data and findings reflect the characteristics of duals in large assisted living settings in 2014. As assisted living markets and Medicaid policies are constantly changing across states, future work should conduct analyses that reflect more recently available data. Fifth, our comparison of duals in assisted living to community-dwelling duals limits our ability to account for a small number of beneficiaries who only would have enrolled in Medicaid if generous assisted living benefits were available. Eligible beneficiaries may not enroll in Medicaid until there is a benefit available of great value—such as HCBS coverage in assisted living—that would allow for avoiding nursing home placement (Pezzin & Kasper, 2002) Finally, due to the cross-sectional nature of our analyses, our results cannot be interpreted as causal. Despite these limitations, the present study yields important findings for understanding the association between Medicaid financing and duals residing in large assisted living settings.

Conclusion

Our study suggests that state Medicaid financing for services in assisted living is associated with greater access to assisted living for duals. Findings lay the groundwork for new and impactful research to understand the mechanisms that enable improved access to assisted living for duals. As there is tremendous variation in state policies concerning assisted living, and the process by which Medicare beneficiaries enter assisted living is multifaceted, results of the present study warrant further investigation into policies that enable duals to enter assisted living and the impact of these policies, and providers’ responses to these policies, on duals’ outcomes. As states work to rebalance Medicaid LTSS spending away from nursing homes toward HCBS, assisted living is becoming an option for duals who need assistance with daily activities but desire to remain in a community setting. More research is needed to understand the experiences of duals residing in assisted living, including small assisted living settings that were not included in our study. In addition, efforts should be made to further examine the implications of these findings, as well as how policies and regulations contribute to health outcomes and quality of life for duals in assisted living.