Long-term services and supports assessments do not adequately address functional limitations associated with serious mental illness

Abstract

Background

Medicaid funded long-term services and supports (LTSS) can provide crucial wraparound supports to adults with serious mental illness (SMI) that have been shown to improve overall health, enhance quality of life, and promote longer tenure in home-and-community settings. In the United States, access to LTSS is determined by level-of-care (LOC) assessments, which are typically homegrown instruments that vary by state. Currently, it is unclear whether LOC assessments address functional limitations commonly experienced by adults with SMI.

Method

To examine this, we first identified publicly available LOC assessments nationwide with clearly defined scoring strategies. Second, we utilized the Health of the Nation Outcomes Scale (HoNOS), an evidence-based, functional assessment for adults with SMI, to comparatively analyze the extent to which state LOC assessments addressed functional limitations associated with SMI.

Results

Assessments from five states met our criteria: Hawai’i, Arizona, Mississippi, Missouri and Tennessee. Problems with activities of daily living (ADLs) were the primary driver of LTSS LOC scores and were substantially overrepresented compared to the HoNOS. Furthermore, eight of the 12 HoNOS functional limitation domains were either drastically underrepresented or not represented in any state assessment.

Conclusion

This analysis suggests that LTSS LOC assessments do not adequately account for the full range of functional limitations commonly associated with SMI and may be systematically excluding high-need Medicaid recipients with SMI who do not have ADL challenges. To ensure parity for Medicaid recipients with SMI, we recommend state Medicaid programs consider implementing a uniform, evidence-based LOC assessment to determine LTSS eligibility for people with a primarily diagnosis of mental illness.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12913-026-14112-1.

Introduction

Medicaid funded long-term services and supports (LTSS) in the United States have provided a crucial safety net for low-income individuals with disabilities for over 85 years. In 2020, LTSS provided support for nearly 5.6 million Medicaid members. LTSS benefits vary by state but often include caregiving support, access to medical equipment, nutritional assistance, medical transportation, and housing in long-term care facilities, group homes, or community care foster homes [1, 2]. LTSS is not episodic care for medical conditions and is not a federally mandated Medicaid service, such as physician services or inpatient hospital care. Rather, LTSS uses Medicaid waiver funds to support the long-term, daily needs of qualifying members experiencing psychiatric, intellectual, or physical disabilities.

Qualifying LTSS members can use LTSS funding to cover the cost of meals, in-home caregiving (assistance with grocery shopping, cleaning, personal care or medical care), in-home nursing, and some medical equipment. These items typically are not covered through Medicaid (without LTSS approval), and also are not typically covered by private health insurance or Medicare [3]. Thus, LTSS provides Medicaid funding and care coordination to support some of the most basic and crucial daily needs of Medicaid members with physical, intellectual, or psychiatric disabilities. The purpose of this study is to determine whether LTSS qualification processes address common functional limitations experienced by adults with serious mental illness (SMI). We do this by analyzing the content of publicly available LTSS eligibility assessments, which are typically called LTSS level-of-care (LOC) assessments. We are concerned with whether LTSS LOC assessments address functional limitations commonly associated with SMI among adults. To operationalize these functional limitations, we use the Health of the Nations Outcomes Scale (HoNOS) as a comparison tool because it was developed specifically to assess functional limitations associated with serious mental illness. HoNOS has multiple versions, and the version for older adults was utilized to encompass functional limitations experienced by both adults and older adults, as it is slightly more comprehensive. Below, we provide greater detail on LTSS, the eligibility process, and the importance of this study for supporting adults with psychiatric disabilities.

Background on LTSS and SMI

While LTSS is offered in some form in every state, in 2023 47 states and Washington DC provided the option of home and community-based services (HCBS) through LTSS, which allow Medicaid members to receive nursing, caregiving, and other LTSS benefits within private residences or community settings. In 2020, approximately 4.2 million Medicaid enrollees utilized HCBS provided through Medicaid state plans and waivers [4]. HCBS supports community integration and maintenance of social connections while reducing costs [5–9] and thus is generally preferential to LTSS offered in institutional settings such as long-term nursing homes [10]. However, LTSS services in institutional and home and community-based settings are both crucial to ensure the needs of adults with SMI are met, and LTSS are unique in the US health care system, as they focus on funding long-term residential supports, caregiving, meals, and home medical equipment.

Thus, LTSS can provide vital support for Medicaid recipients with SMI who face elevated risk of poor health, comorbidity with other chronic illnesses, housing insecurity, and premature mortality [11, 12]. People with SMI are more likely than the general population to receive Medicaid services [13–15], twice as likely as individuals without SMI to be diagnosed with cardiovascular disease, and 75% of people with SMI have at least one non-psychiatric chronic illness [16–18]. Up to 50% of unhoused individuals have SMI [19, 20], and older adults with SMI are a high-risk, high-cost population with one of the fastest growing rates of homelessness [21–23]. Additionally, people with SMI are less likely than adults without SMI to have intact family supports and social networks that can assist with health management and provide in-home supports [24]. These risk factors result in significant premature mortality among people with SMI, who die on average 13 to 30 years sooner than the general population [25].

The presence of intersecting health and social vulnerabilities among adults with SMI highlights the importance of examining structural or systemic factors that may result in reduced access to LTSS services in this population. One potential barrier to LTSS access among adults with SMI is that LTSS eligibility assessments do not fully address limitations commonly associated with SMI. LTSS functional assessments, typically called level-of-care (LOC) assessments, determine eligibility for LTSS services [26]. Typically, after a Medicaid member has been identified as potentially in need of LTSS, a qualified health professional conducts an LOC assessment and determines whether that Medicaid member requires LTSS to support health and safety. LOC scores may also guide payment authorization for the type and frequency of services that are made available to LTSS recipients [2].

Despite their importance in determining access to care, LOC assessments are surprisingly under researched. The Medicaid and CHIP Payment and Access Commission (2016) conducted a descriptive analysis of LTSS functional assessments nationwide and found that a majority of states had developed their own homegrown tools, which had substantial variation, and were modified iteratively over several years without clear rhyme or reason. Another study conducted an in-depth analysis of LOC assessments in six states and found they excluded functional deficits that are characteristic of dementia, such as impaired judgment, fluctuations in decision-making capacity, and impulsive, inappropriate, or disruptive behaviors [27]. The authors determined this omission in assessment criteria may reduce access to care for adults with dementia, a highly vulnerable population that often requires a high level of care [27–29].

The present study

To date, no research has examined the extent to which LOC assessments capture functional impairments that are common among adults with SMI. Functional limitations related to SMI may present differently from those that stem from cardiovascular, respiratory, osteopathic, or neurological illnesses, and yet, they can be equally grave, jeopardizing both safety and health. For example, impaired decision-making related to delusions could result in behavior that puts an adult with SMI at risk of self-harm or poor health, while their balance, strength, and mobility remain entirely intact [30]. Ensuring that LOC assessments adequately address the limitations of adults with SMI is not only ethically important, it is legally mandated by the Olmstead Act and Mental Health Parity and Addiction Act, which require that people with SMI have equal access to effective treatment for their health conditions compared to individuals with physical health conditions [31–34]. Thus, to fill this important research gap, we examined the extent to which statewide LTSS LOC assessments adequately capture functional limitations commonly associated with SMI.

Methods

Overview

We first identified LOC assessments used statewide for Medicaid funded LTSS programs that were (a) publicly available, (b) had a well-defined scoring strategy, and (c) were not specific to a sub-population other than people with SMI. This resulted in assessments from five states: Hawai’i, Arizona, Mississippi, Missouri and Tennessee. Next, we utilized domains from an evidence-based, mental health outcome measure called the Health of the Nation Outcomes Scale (HoNOS) to analyze how the five state LOC assessments addressed functional limitations associated with SMI. We cross-referenced items from each of the LOC assessments with the HoNOS domains using content analysis and then evaluated the extent to which each domain was over, under, or not at all represented in each LTSS LOC assessment compared to the HoNOS.

Background on the health of the nations outcome scale (HoNOS)

The HoNOS is an ideal comparison tool because it was specifically designed to assess functional limitations associated with SMI, and thus provides a comprehensive yet concise set of functional limitation domains associated with SMI such as problems with maintaining social relationships, housing, and living conditions, along with benchmarks for each [35]. The HoNOS was developed by the United Kingdom’s (UK) Royal College of Psychiatrists over several years of analysis and testing [35]. It has strong psychometric properties and is the most widely used routine service eligibility and outcome measure for mental health services in the UK [36]. It is also used regularly for admission, review, and discharge in inpatient and outpatient public-sector mental health services in Australia [37], and is used globally by mental health professionals to assess functional limitations common among people with SMI [36, 38, 39]. The HoNOS contains 12 domains, presented below in Table 1. Each domain is counted as a single item and rated on a scale from zero (no problem) to four (severe to very severe problem). Thus, each domain is equally weighted, and accounts for 8% of an individual’s overall score.

Table 1.

Domains from the health of the Nation outcomes scale (HoNOS65+)

Domain	Example
1 Overactive, aggressive, disruptive, or agitated behavior	Irritability, threats or verbal aggression, restlessness or wandering, physically aggressive to others or animals.
2 Non-accidental self-injury	Frequent thoughts or talking about self-harm or suicide.
3 Problem drinking or drug-taking	Psychological and behavioral impacts of drug (illicit and/or prescription) and alcohol use.
4 Cognitive problems	Problems of orientation, memory, language, thought disorder, and problem solving associated with any disorder: dementia, learning disability, schizophrenia, etc.
5 Physical illness or disability problems	Illness or disability from any cause that limits or prevents movement, impairs sight or hearing, or otherwise interferes with personal functioning.
6 Problems associated with hallucinations and/or delusions	Unusual and bizarre behavior associated with hallucinations or delusions.
7 Problems with depressed mood	Loss of interest or pleasure, lack of energy, loss of self-esteem, feelings of guilt.
8 Other mental and behavioral problems	Problems related to anxiety, trauma, eating, sleep, etc.
9 Problems with relationships	Problems in making or sustaining supportive relationships. Active or passive withdrawal from social relationships.
10 Problems with activities of daily living	Problems with basic activities or self-care such as eating, washing, dressing, toileting; complex skills such as budgeting, organizing where to live, occupation and recreation, mobility and use of transport.
11 Problems with housing and living conditions	Lacking basic necessities such as adequate heat, light, sanitation, cooking facilities. Problems with one or more aspects of housing, living conditions, or routine.
12 Problems with occupations and activities	Overall level of problems with the quality of meaningful occupations and activities.

The HoNOS has undergone mild modifications and has been adapted into six versions for unique sub-populations with SMI based on age, setting, or dual-diagnosis [40]. Although our focus is on all adults with SMI, we used the HoNOS 65+ (Older Adults) for two reasons: (1) it includes all the same domains as the original assessment, but it is slightly more comprehensive, and (2) more than half of Medicaid users covered by LTSS are 65 + and thus we wanted to be inclusive of this population in our analysis [41]. The HoNOS 65 + features the same 12 domains as the general population scale but includes additional qualifying criteria for physical and cognitive challenges. These additions include delusions associated with dementia under the cognitive problems domain, incontinence under the ADLs domain, and restlessness and wandering under the overactive, aggressive, disruptive, or agitated behavior domain [42]. All of the criteria and scoring strategy for each domain can be seen in Additional File 1: Complex Scoring Systems, and the HoNOS is freely available online.

Identification and screening of LOC assessments

Figure 1 outlines steps and outcomes of our identification and screening strategy of state LOC assessments following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [43] with details on each step included below. While this is not a systematic literature review, the PRISMA guidelines provide a neat outline of the LOC assessment identification and screening process. All 51 U.S states were included in the first step of the identification and screening process. Forty LOC assessments were identified using iterations of the following search terms: “LTSS, LOC assessment + LOC form + long-term care assessment + eligibility assessment + functional assessment + [name of state]” on search engines Google and Google Scholar. Two members of the research team performed online searches for each state. Ten additional assessments were identified by cross-referencing our results with the “State Functional Assessment Tools for Long-Term Services and Supports Eligibility Determination and Care Planning,” which is an online database of state functional assessment tools for LTSS [2, 44].

Fig. 1.

PRISMA flow diagram of search strategy for state LOC assessments

After identifying LOC assessments (n = 50), we screened and assessed their eligibility for our analysis based on two inclusion criteria: (a) the full assessment was publicly accessible including all items, and (b) the assessment was used to determine LTSS eligibility for any Medicaid beneficiary (rather than addressing a program or sub-population such as individuals with intellectual disabilities). We excluded 19 LOC assessments based on these criteria. Notably we did not find any LTSS assessments specific to SMI population in this search. Next, we only included assessments with a publicly accessible scoring strategy, which allows us to quantitatively evaluate the extent to which the assessment tool adequately addresses functional limitations associated with SMI. We excluded 27 assessments in this step, including those that were not evaluated based on an explicit scoring system (e.g., relying solely on clinician judgment) (n = 17), required a provider login to access the scoring system (n = 2), or used a computer-based scoring algorithm with rules not publicly available (n = 5). This resulted in five state LOC assessments in the analysis, including Arizona, Hawai’i, Mississippi, Missouri, and Tennessee. A list of these five state assessments with links to their LOC assessments and eligibility scoring strategy instructions can be found in Additional File 2: LOC Assessment Links.

Analysis

Next, we performed a content analysis of the LOC assessments for each of the five states, to determine the extent to which each assessment captured the 12 domains of functional limitations associated with SMI based on the HoNOS. We referenced the HoNOS 65 + glossary (available online from the official Royal College of Psychiatrists website) to determine how each item in the state assessment aligned with each domain in the HoNOS. We then allocated points from relevant items to each domain based on the state scoring system. For example, if a state assessment included 10 items (with one point for each item) and five out of those 10 items addressed ADL, we would assign 50% (5/10) to Domain 10 (which addresses ADL), for that state. Each item and its total possible points per domain were analyzed by at least two researchers. Discrepancies were resolved through discussion with input from the lead author, until consensus was reached. Strategies to address complex scoring systems, such as weighted items and lumped items (those that include multiple domains in a single item) are included in Additional File 1: Complex Scoring Systems. After mapping items and their total possible points to each domain, we calculated the percentages of points allocated to each domain for each state. Finally, we calculated average percentages across the five states for each domain. We determined “over” and “under” representation of each domain by comparing the average percentage of points allocated to each domain in the LOC assessments to the HoNOS benchmark of 8% for all domains. Domains with an average percentage of 0% were classified as “not represented,” domains between 1% and 7% were “under-represented” as they fell below the HoNOS benchmark of 8%. Domains between 9% and 49% were “over-represented” and domains over 50% were determined to be “highly over-represented” as they accounted for more than half of the total points.

Findings

Overview

Trends in over and underrepresentation by domain were largely consistent across the five states with substantial overrepresentation of problems with ADL across all states (see Table 2; Fig. 2 for a visual summary of results). Problems with ADL were the primary driver of LTSS eligibility, ranging from 47 to 60% of total possible points across states, compared to 8% on the HoNOS. Furthermore, eight of the 12 functional limitation domains associated with SMI were either drastically underrepresented or not represented in any state assessment. The four domains that were not represented in any assessment included problems with drinking or drug-taking (D3), problems with relationships (D9), problems with housing and living conditions (D11), and problems with occupation and activities (D12). Only one state, Missouri, included items for problems associated with hallucinations and delusions (D6) and problems with depressed mood (D7), and the proportion of the overall score associated with these items was vanishingly small, accounting for 1% of the total possible score for each domain. Two states, Missouri and Tennessee, included items related to “other mental and behavioral problems” (D8) which accounted for 2% and 3% of their total score respectively, and primarily addressed issues associated with paranoia, abnormal thought process, and inappropriate or unsafe behavior. Non-accidental self-injury was present in all five state LOC assessments but accounted for a very small proportion of the overall score across states (0.39%-5%). Three domains were slightly overrepresented compared to the HoNOS, and these had the greatest variation across states. Cognitive problems accounted for 12–18% of the total scores across states, with Hawaii having the largest proportion of its score associated with this domain. Problems related to physical illness or disability ranged from 16% of the total score in Hawaii to 6% of the total score in Arizona. Finally, problems with overactive, aggressive, disruptive, or agitated behavior was underrepresented in two states, Missouri and Tennessee, and accounted for 3% of the total score in both, while it was highest in Arizona, accounting for 18% of the total score.

Table 2.

Five state LOC assessments: total possible points and percentage allocation by HoNOS domain

Fig. 2.

Average percent per HoNOS domain across five state LOC assessments

Additional findings related to scoring strategy that May impact care

Three out of five state LOC assessments combined multiple domains of behavioral health issues into a single item, which reduced these items’ weight and impact on the total score. For example, Hawaii’s LOC assessment combines questions for self-injurious, cognitive, and aggressive behaviors into one item. Thus, if an individual has functional limitations or risk factors associated with each of these domains, it has relatively low impact on their overall score, accounting for only 13% of the total. Missouri’s assessment combines homicidal ideation (D1) with suicidal ideation (D2), and thus if a Medicaid member has both homicidal and suicidal ideation, these substantial risk factors account for only 0.84% of their overall score. Tennessee’s LOC assessment lumps aggressive behavior towards others (D1), self-injurious behavior (D2), and inappropriate or unsafe behavior (D8). When various behavioral health issues are combined into one item, this can limit their impact on overall score and thus can reduce the likelihood of achieving eligibility for LTSS, and for accessing a level of care that is consistent with their actual need.

In addition to combining items representing discrete functional domains, and excluding other domains entirely from the LOC assessments, we found that one out of five states had questions associated with SMI that were not scored, and therefore not explicitly determined in their level of care. For instance, Mississippi’s LOC assessment includes questions regarding “evidence of neglect by self or others”, “unsafe situations”, and “depression” yet these were not included in the LOC scoring or eligibility determination.

Finally, Missouri had seven items and Mississippi had three items, accounting for 7.8% and 7.2% of their total points respectively, that did not match any of the HoNOS domains. This was because they did not assess an individuals’ functional limitations but rather captured the need for specific interventions. For example, Missouri included “Restraints” as a scored item in their assessment, with responses ranging from none to max. Mississippi included “Turning and Positioning” and both states included the need for rehabilitative services, like occupational, physical, speech therapy, in their level of care scores. Inclusion of specific health services in the overall score was unique from other functional assessments, which determined LOC scores based on functional status, and then used the scores to determine the need for services.

Discussion

Results from this analysis suggest LTSS LOC assessments from the included states do not fully address functional limitations commonly associated with SMI and tend to over-emphasize ADL limitations. ADL limitations are an important health determinant and have been associated with increased all-cause mortality among people with SMI [45]. However, many people with SMI, including those who do not demonstrate problems with ADL, may have substantial health and social challenges that would benefit from LTSS. For example, a study by [46] on ADL performance in 558 individuals with SMI found low rates of ADL impairment across diagnoses of major depressive disorder, bipolar disorder, and schizophrenia.

It was particularly concerning that several functional domains commonly associated with SMI were missing entirely from LOC assessments, such as problems with substance use, relationships, housing and living conditions, and participation in occupations and activities. [47] estimate that more than half of people with SMI are affected by substance use disorders, which are an important determinant of one’s overall health, ability to participate in health services, and ability to live safely in home and community settings. Drake and colleagues (2020) also found, over a 16-year longitudinal study, that people with SMI and co-occurring substance use achieved improvement on a variety of recovery outcomes when wraparound supports were provided. This suggests that programs like LTSS can support health and well-being among people with SMI and co-occurring substance use.

Along with substance use, housing and living conditions were not included in LOC assessments and are increasingly viewed as a social driver of health that disproportionately affect people with SMI [48]. While several state Medicaid agencies have attempted implementation of housing and tenancy supports as part of a new emphasis on social determinants of health [49], the states included in this study do not include issues with housing as part of their determination of overall need for LTSS. This is concerning, given that deficits associated with housing and participation in occupations and activities have been associated with higher all-cause mortality among people with SMI [45].

In addition to housing challenges, past research has shown the adults with SMI often struggle with participation in occupations and activities and have low community integration in comparison to the general population [50, 51] found that adults with SMI spend 75% to 80% of their day at home. Disproportionate employment rates also show that 60% of people with bipolar disorder and 80% of people with schizophrenia are unemployed compared to only 6% of the general population [52, 53]. Similarly, adults with SMI are more than twice as likely to be lonely and have social networks that are 67% smaller [54].

Non-accidental self-injury, hallucinations and delusions, and depressed mood are also other important domains that were under-assessed. These areas were only addressed in two out of five states, accounting for a small percentage of the overall score (3% or less) despite high potential to significantly affect overall health. Research indicates a strong association between self-injury (or self-harm) and mental health conditions such as mood disorders and schizophrenia [55]. Furthermore, hallucinations, delusions, and depressed mood are highly characteristic of schizophrenia, which can result in considerable stress, deteriorating health, and decreased quality of life [56, 57].

Implications

Failure to assess these important functional domains can result in decreased access to essential community-based LTSS benefits. Several scholars have linked the absence of community-based supports with higher representation of adults with SMI in the criminal justice system and institutional settings [58–60]. In the U.S., people with mental illness are 10 times more likely to be incarcerated than to be hospitalized [61]. The reported that one third of all inpatient hospitalizations and one fifth of all emergency department visits were patients with mental illness, although adults with SMI represent a relatively small proportion of the population [62]. These studies reflect an overall overdependence on carceral and institutional care for people with SMI, and underscore the need for coordinated, community-based supports [63]. While LTSS would not ameliorate these systemic injustices overall, greater access to LTSS may enhance needed community-based supports for this population and thereby reduce reliance on institutional and carceral settings to provide care for adults with SMI with high needs. Although we did not find a single state that captured the needs of people with SMI in their functional assessment, there is evidence of increasing awareness of the integrated medical needs of adults with SMI. In 2024, the Centers for Medicaid and Medicare Services launched the Innovation in Behavioral Health (IBH) Model pilot that is focused on enhancing integrated care and support for physical, psychological, and social needs for people with SMI. Four states: Michigan, New York, North Carolina, and Oklahoma began Jan 1, 2025 [64]. Ideally, such an system-level intervention would include a coordinated, evidence-based assessment of functional challenges commonly associated with SMI, rather than a diversity of home-grown measures that do not fully capture the needs of individuals with SMI.

Limitations

This work it is not without limitations. First, our analysis was limited to five state LOC assessments; therefore, this is not a comprehensive picture of the national landscape. However, we found these five LOC assessments had relative uniformity in the functional domains they did and did not address. Additionally, these LOC assessments determine LTSS eligibility for over 5.6 million Medicaid recipients collectively, approximately 29% of whom have psychiatric diagnosis [14, 65, 66]. Thus, albeit limited in scope, this study has significant implications for people with SMI who may benefit from LTSS as a community-based support. This research may also prompt future examination of LOC assessments, ideally in collaboration with state Medicaid agencies, who have the capability to revise them and utilize evidence-based measures.

Second, it is possible that the LOC assessments were outdated. We minimized this limitation by checking the publication date of each assessment and utilizing the most recent version available. We did not find substantial differences in assessments year to year, and, as mentioned, there was substantial overlap and consistency across states. This suggests multiple states follow a similar assessment paradigm, and this study highlights consistent patterns in the exclusion of functional challenges commonly associated with SMI and overreliance on the ability to carry out activities of daily living to qualify for LTSS benefits.

Third, our decision to exclude states without explicit scoring systems reduced the number of states included in the analysis, which could have provided useful content that could have potentially been analyzed qualitatively. The reason we decided to focus exclusively on assessments with scoring systems is twofold. First, comparing the scoring system of an evidence-based assessment to the scoring systems of home-grown LTSS assessments allowed us to make determinations of over and under-representation of items that were grounded in past research. Second, state assessments typically had many pages of items (non-scored assessment links are available to view in Additional File 2) and without scoring systems it was impossible to determine which of these items were included in LTSS eligibility. For example, the Alaska assessment was 32 pages; however, without a scoring system the author team had no way of determining which of the items in these 32 pages were considered for LTSS eligibility. If we had performed a content analysis of all the items included in this assessment, even without the scoring comparison to the HoNOS, we may have invertedly included items that were not actually considered for LTSS eligibility and instead collected for clinical care or eligibility for other programs. Thus, analysis of additional (non-scored) items may have caused us to stray from our primary focus on LTSS eligibility and potentially misrepresent which items were included in decision-making. While focusing on the scored assessments reduced the number of states included, it added clarity and rigor to our analysis and ability to answer our primary question. Despite these limitations, this research contributes significantly to the literature, particularly given the paucity of examination of US Medicaid LOC assessments, which are crucial gatekeepers to access to care.

Recommendations

Overall, these findings point to a critical need for LOC assessments to address the functional limitations associated with SMI more comprehensively. To ensure Medicaid beneficiaries with SMI can receive necessary and appropriate support for community participation and overall health, we recommend utilizing a uniform, evidence-based assessment specifically designed to capture functional needs of adults with SMI. A more comprehensive and multifaceted LTSS assessments would ensure that recipients are being assessed appropriately so they can receive support that is tailored to their diverse needs. This is not without precedent. In our review, we found that states sometimes utilize unique assessments for populations with specific primary diagnoses, such as intellectual or developmental disabilities. Studies also suggest that performance-based measures, such as the UCSD Performance-Based Skills Assessment (UPSA) which requires the individual being assessed to demonstrate planning, finance, communication, mobility, and house management skills by completing relevant tasks, in addition to clinician judgment, may be useful for assessing functional needs in people with SMI [67–69].

While this study is based in the United States, it is not without global implications. Other countries examining strategies to improve long-term supports for adults with SMI may undergo a similar analysis of their own assessment tools for social and clinical supports, examining the extent to which they address the unique needs of adults with SMI. Implementing these recommendations may lead to an increased adults with SMI receiving care that reduces exacerbation of symptoms and disability while supporting overall quality of life. Necessarily, achieving these aims requires following evidence-based assessment with appropriate services, training, and community-based supports to meet the complex needs of adults with SMI [70].

Conclusion

Current LTSS assessments do not fully account for functional limitations commonly associated with SMI and over-emphasize ADL limitations. Inadequately assessing important functional domains can result in overlooking deficits that commonly affect adults with SMI and potentially result in decreased access to LTSS benefits. Further, we found that home-grown LTSS assessments often have convoluted scoring strategies that can complicate service provision and obscure the causes of disparities in access to care. Overall, state Medicaid agencies would benefit from utilizing an evidence-based, uniform assessment that includes functional limitations commonly associated with people who have a primary SMI diagnosis. This may allow for more equitable access to LTSS, greater parity for people with mental and physical health conditions, better ability to research LTSS utilization, and ultimately improve quality of care for people with SMI who are at high risk of poor health and early mortality.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

J.A conceptualized the paper, contributed to the data analysis, drafted, reviewed, and edited the manuscript. J.S contributed to the data analysis, prepared Tables 1 and 2; Figs. 1 and 2, and drafted, reviewed, and edited the manuscript. H.Z contributed to the data analysis, prepared figures, and drafted and reviewed the manuscript. H.P contributed to the data analysis. All authors read and approved the final manuscript.

Funding

Dr. Joy Agner’s contributions to this work were supported in part through funding from the National Institute of Mental Health (NIMH 1K01MH136343).

Data availability

This research utilized publicly available Medicaid Level of Care assessments which can be accessed through an online search. We have included a supplementary table (see Additional File 2: LOC Assessment Links) with links to the assessments along with details about the form version (if applicable) and date they were last updated and accessed.

Declarations

Ethics approval and consent to participate

This study reviewed and analyzed publicly available Level-of-Care assessments, was not human subjects research, and therefore did not require ethics approval from an institutional review board.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.