Abstract
Many Medicaid beneficiaries aged 22 to 64 with serious mental illness may be admitted to nursing facilities rather than psychiatric facilities as a result of Medicaid policies prohibiting coverage of inpatient psychiatric care in institutions of mental disease while requiring states to cover nursing facility care. Using nationwide Medicaid Analytic Extract claims from 2002, we found that nearly 16% of nursing home residents aged 22 to 64 had a diagnosed mental disorder, while 45.5% received antipsychotic medication, but these rates varied widely across states. Further research is necessary to determine whether, among the nation's youngest nursing home residents, care in nursing homes is potentially substituting for care in institutions for mental disease or community-based settings.
INTRODUCTION
Several federal laws have been enacted in the past 2 decades to ensure that people with serious mental illness are not institutionalized and can live in the community with appropriate treatment. The Americans with Disabilities Act and the Supreme Court ruling in Olmstead v. L.C. (1999) require states to offer services in the “most integrated setting” to individuals with disabilities, including those with serious mental illness. The Pre-Admission Screening and Resident Review (PASRR) program, mandated by the federal Omnibus Budget Reconciliation Act (OBRA) of 1987, requires state Medicaid agencies to screen individuals admitted to nursing homes for serious mental illness or mental retardation and determine whether a nursing home is the most appropriate placement for them and, if it is, their need for specialized mental health treatment.
Prior studies have generally found that state implementation of PASSR programs varies widely, leading to many nursing home residents with serious mental illness not receiving appropriate mental health services (CitationLinkins, Lucca, Housman, & Smith, 2006; CitationO'Connor, Little, & McManus, 2009; CitationOffice of Inspector General [OIG], 2001a). Nonetheless, adults between the ages of 22 and 64 who have serious mental illness and are Medicaid beneficiaries may be admitted to nursing homes rather than psychiatric facilities. That is because federal Medicaid policy prohibits coverage of inpatient psychiatric care in institutions for mental disease (IMDs)Footnote 1 for those in this age group, while it requires all states to cover nursing facility care (CitationOIG, 2001b; CitationSwan, 1987).
Documenting the number of nursing home residents aged 22 to 64 with severe mental illness is vital to monitor the appropriateness, access, and quality of care for this population, as well as the effectiveness of federal nursing home policy. Overall, about 13% of Medicaid-funded nursing home residents are between the ages of 21 and 64, with the majority of these residents (over three-quarters) between the ages of 45 and 64 (CitationBagchi, Verdier, & Simon, 2009). In the absence of appropriate community-based services and funding for those services, the predominantly near-elderly residents described in this paper will soon become elderly nursing home residents.
Previous studies have attempted to estimate the number of Medicaid beneficiaries younger than 65 with serious mental illness placed in nursing homes; however, efforts to do so were hampered by incomplete and inconsistent Medicaid Statistical Information System (MSIS) data across states (CitationOIG, 2001b). Recently, research-quality data have become available that provide more reliable and comparable state information on Medicaid beneficiary service use. This study uses Medicaid Analytic Extract (MAX) data from 2002, maintained by the Centers for Medicare and Medicaid Services (CMS), to describe the percentage of nursing home residents younger than 65 with documented mental disorders and the percentage of younger residents receiving antipsychotic medications. Use of data from 2002 permits assessment of states' early, post-Olmstead experiences and provides a baseline measurement to assess the impact of subsequent changes in the availability of community-based services for Medicaid beneficiaries younger than 65 with serious mental illnesses. We examine use of antipsychotic medication to identify the presence of serious mental health symptoms, recognizing that diagnosis data might be incomplete or inaccurate. The data contain some indications of state variation in potentially undocumented mental health needs among non-elderly Medicaid beneficiaries in nursing homes and clues about the degree to which nursing home care might be substituting for care in IMDs or home- and community-based settings.
METHODS
Participants
Data for this analysis come from 2002 Medicaid service claims in MAX files, which contain person-level data on long-term care, inpatient care, and other Medicaid-paid services. The MAX long-term care file contains claims for all nursing facility residents whose care was paid by Medicaid, representing about 6 of every 10 nursing facility residents (CitationJones, Dwyer Bercovitz, & Strahan, 2009). This analysis was restricted to Medicaid nursing home claims for beneficiaries younger than 65 with 6 or more months of continuous residence to focus on the population with long-term admissions.
We excluded data from three states with incomplete data or other limitations. Data from Arizona were not included because all long-term care Medicaid beneficiaries are enrolled in managed care plans, which do not submit claims data. Data from North Carolina indicated an implausibly high number of nursing facilities and so were excluded. Nevada data did not have any secondary diagnosis information for mental disorders and were excluded from the analysis.
Measures
Each claim in the MAX files may include up to five diagnoses (one primary and four secondary diagnoses). In order to improve the identification of relevant diagnoses, our analysis included diagnoses of primary and secondary mental disorders from the long-term care (LTC), inpatient (IP), and other (OT) Medicaid claims in the MAX files. Diagnoses from the IP and OT claims were included if they occurred in the same month as the LTC claim. Race and ethnicity data in the 2002 MAX data were recorded as categorical variables.
Because the statutory definition of mental illness for PASRR specifically excludes persons with dementia and related conditions (if their primary diagnosis is not mental illness), we excluded these diagnoses. We also excluded substance abuse disorders because our intention was specifically to focus on the population with mental disorders. To maintain consistency with prior studies of mental disorders using MAX data, we also excluded diagnoses of mental retardation (CitationBagchi et al., 2009). We included all other mental disorders corresponding to International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes ranging from 295 to 302 and from 306 to 314 drawn from the diagnoses fields of the MAX LTC, OT, and IP files.Footnote 2 Studies of Medicaid claims data for use in research on psychiatric disorders have found the primary and secondary diagnosis variables to be valid and reliable (CitationKnapp et al., 2006; CitationWalkup, Boyer, & Kellerman, 2000).
Antipsychotic medication was identified using the MAX prescription drug claims. The MAX prescription drug claims were supplemented with the First Data Bank's Therapeutic Classification System data to identify typical and atypical antipsychotic medications. We employed antipsychotic use as a proxy for moderate to severe behavioral symptoms due to psychosis.
Statistical Analysis
SAS version 9.1.3 (SAS Institute, Cary, NC) was used for all data analysis. Chi-squared tests were used to identify differences in the presence of mental disorders and the use of antipsychotic medication among Medicaid residents aged 22 to 64 across demographic categories of age, sex, and race.
RESULTS
In 2002, 134,420 Medicaid beneficiaries aged 22 to 64 resided in nursing homes for at least 6 months; about half of the sample was between 55 and 64 years of age (). Nearly 16% of residents aged 22 to 64 had a primary or secondary mental disorder diagnosis. A substantially larger percentage (45.5%) of all residents aged 22 to 64 received an antipsychotic medication during the year. Antipsychotic use was more common among older residents; 36.4% of residents aged 22 to 34 had antipsychotic drug claims, compared with 45% of those aged 35 and older (p < .001). Residents of Asian/Hawaiian or Pacific Islander ancestry had the lowest rate of diagnosed mental disorders (9.9%) and antipsychotic use (37%) compared with residents of other races (p < .001).
TABLE 1 Demographic Characteristics of Nursing Home Residents Aged 22–64
The size of the population of residents aged 22 to 64 as well as the number and percentage of residents with a diagnosis of a mental disorder and receiving an antipsychotic varied considerably by state (). The proportion of nursing facility residents in this age group with a diagnosed mental disorder ranged from 5.4% (Rhode Island) to 29.3% (Oklahoma). The percentage of residents receiving antipsychotic medication also varied substantially by state, from a low of 12.7% in the District of Columbia to a high of 70.1% in Illinois.
TABLE 2 Nursing Home Residents Aged 22–64 with Mental Disorders and Receipt of Antipsychotics
There was also wide state variation in the ratio of the number of residents receiving antipsychotic medication to the number of residents with a diagnosed mental disorder (). The average ratio across states was about three residents receiving antipsychotic medication for every resident with a diagnosed mental disorder. In five states, however, this ratio was six or more: California, Louisiana, Iowa, Rhode Island, and South Dakota. Although a broad array of ICD-9-CM codes was used to identify the presence of a mental disorder, the vast majority of residents with a mental disorder had depression (92.8%, ICD-9-CM code = 311). Only 5.4% of residents with a mental disorder had a diagnosis of schizophrenia (ICD-9-CM code = 295) (not shown).
States varied considerably in the number and percentage of facilities with at least one resident aged 22 to 64 with a diagnosed mental disorder or receiving an antipsychotic (). To distinguish between facilities with large or small percentages of younger residents with either a mental disorder or receiving antipsychotic medication, shows the percentage of facilities in the highest decile (nationally) of residents with a mental disorder or receiving an antipsychotic. The top decile of the percentage of younger residents with a mental disorder was represented by facilities with at least 45.5% or more of their younger-than-65 population with a mental disorder. Likewise, the top decile for antipsychotic receipt was represented by facilities with at least 85.3% of residents receiving an antipsychotic medication. Nationally, half of all nursing facilities had any residents aged 22 to 64 with a diagnosed mental disorder, and 77% of facilities had any residents aged 22 to 64 receiving antipsychotic medications. Oklahoma, Texas, and Minnesota, respectively, had the highest proportion of facilities representing the top decile of younger residents with a mental disorder, while Kansas, Illinois, and Connecticut, respectively, had the highest proportion of facilities representing highest antipsychotic use among younger residents.
TABLE 3 Facility Distribution: Facilities with Any Residents with Mental Disorders and Residents Receiving Antipsychotic Medication
The national distribution of facilities with varying percentages of residents with either a mental disorder or receiving an antipsychotic is displayed in . In approximately 1 in 10 facilities, at least half of the younger residents had a diagnosed mental disorder. In nearly 4 out of 10 facilities (38%), at least half of all residents aged 22 to 64 received an antipsychotic medication in the past year.
FIGURE 1 Facility distribution: Percentage of residents aged 22–64 with a mental disorder or receiving an antipsychotic (N = 13,849).
DISCUSSION
Nationally, just more than half of all facilities have a Medicaid resident younger than 65 with a diagnosed mental disorder; more than three-quarters of all facilities have at least one resident younger than 65 who received an antipsychotic. In the context of the PASRR program, the Medicaid IMD exclusion, and the 1999 Olmstead decision, all of which are intended to prevent institutionalization of people with mental illness who could be cared for in community settings, these findings suggest that some of these individuals may not be receiving appropriate treatment in the least restrictive setting. That is, barring significant medical or functional requirements, care of these non-elderly residents in nursing homes is potentially in conflict with at least one of these directives. Further research is necessary to determine whether, among the nation's youngest nursing home residents, care in nursing homes is potentially substituting for care in IMDs or community-based settings or whether antipsychotic medications are potentially being inappropriately prescribed. For example, we found that nationally at least 10% of facilities meet one of the criteria to be considered an IMD: more than half of all residents have a mental disorder diagnosis (CitationCMS, 2010).
In many states, the federal Medicaid IMD exclusion, in combination with Medicaid policy that guarantees coverage for nursing home care, provides an incentive for states to use nursing homes to care for beneficiaries younger than 65 with diagnosed mental disorders and/or severe behavioral symptoms. However, state policies can moderate this incentive. States with fewer younger residents with mental illness may have greater availability of community-based mental health services and more generous Medicaid coverage of community mental health benefits (CitationNational Health Policy Forum, 2008). States with greater numbers of younger people with mental illness in nursing homes may not have enough psychiatric inpatient facilities, or they may not be equipped to care for patients with long-term care needs.
In such circumstances, however, nursing homes might be inappropriate settings for adults with mental disorders. Most nursing facilities lack trained mental health professionals to meet the disparate care needs of elderly and non-elderly residents with mental disorders (CitationOIG, 2001a). A shortages of qualified professionals is one reason those with mental disorders may not be receiving adequate mental health treatment in nursing homes, even in nursing homes that can offer some of the specific services identified by PASSR screening (CitationShea, Russo, & Smyer, 2000). When mental health treatment is made available, it is often limited to medication therapy, which is easier to manage by staff not trained in mental health (CitationLinkins et al., 2001; CitationSnowden, Piacitelli & Koepsell, 1998).
For every Medicaid-funded resident between the ages of 22 and 64 with a diagnosed mental disorder, approximately three residents received antipsychotic medication. The high prevalence of antipsychotic use relative to diagnosed mental disorders might indicate that nursing home residents' mental disorders are underdiagnosed or misdiagnosed in order to avoid being excluded as an IMD. It could also indicate that many people who need specialized mental health treatment are inappropriately placed in nursing homes. Alternatively, antipsychotic medication might be overprescribed to make it easier for nursing home staff to manage the care of individuals who would otherwise demonstrate disruptive behaviors. The finding that most residents with mental disorders had depression makes the high rates of antipsychotic particularly troubling and raises the possibility that potentially inappropriate antipsychotic use, although typically evaluated for older nursing home residents, may also be an issue for younger nursing home residents as well.
It is encouraging, however, to find an apparent reduction in the proportion of nursing home residents between the ages of 22 and 64 with a mental disorder following the passage of OBRA. Overall, 16% of the Medicaid-covered nursing residents between ages 22 and 64 had a mental disorder in 2002, a reduction from the 23% to 24% of Medicaid residents younger than 65 with mental disorders in 1985, prior to the passage of OBRA (CitationCongressional Research Service, 1993). This finding is consistent with that of CitationMechanic and McAlpine (2000), who found a two-thirds reduction in the number of all nursing home residents younger than 65 with mental disorders between 1985 and 1995. Our estimate of residents younger than 65 with mental disorders is considerably lower than a previously reported estimate of 34.2% using 1999 MAX data and the same list of diagnoses (CitationBagchi et al., 2009). We attribute this difference to this study's emphasis on long-stay residents (6 months or longer), whereas Bagchi et al. used diagnoses from all long-term care claims in the last half of 1999, irrespective of length of stay. CitationGrabowski, Aschbrenner, Feng, and Mor (2009) found that among newly admitted nursing residents with either schizophrenia or bipolar disorder, 54% were between the ages of 18 and 64 (CitationGrabowski et al., 2009). Because the study by Grabowski et al. used a different administrative data set from 2005 (the nursing home Minimum Data Set [MDS]) and focused on new admissions, that study cannot be directly compared with our study; however, they also report variation across states in the rates of nursing home residents with mental illness.
One important limitation to this study was the lack of functional impairment information, which inhibited our ability to understand the extent of care needs beyond psychiatric conditions. Some residents might have, for example, extensive impairment in activities of daily living in addition to psychiatric diagnoses, making nursing home placement appropriate. The extent to which nursing home care is substituting for more appropriate placement in an IMD or in the community cannot be determined from MAX data because they do not contain information on physical or functional limitations in performing daily activities. However, MAX data might provide a starting point to investigate these issues in greater depth, through the analysis of resident-level MDS data to identify functional impairment in this population among facilities in the 90th to 100th percentiles of mental disorders or antipsychotic use.
We note the proportion of total Medicaid long-term care spending used for home- and community-based services increased from 15% in 1992 to 30% in 2002 (CitationReester, Missmar, & Tumlinson, 2004). Analyses of more recent MAX data would help to assess states' longer-term response to the Olmstead decision in the years subsequent to this study. Other useful avenues for research could explore reasons for state variation in documentation of mental disorders and antipsychotic use. For example, a study of comorbidities and functional status of Medicaid beneficiaries in nursing homes with mental health diagnoses, or facilities indicating a rate of antipsychotic prescriptions in the 90th percentile, would help determine which other physical health conditions make a nursing facility the most appropriate placement. Also, primary data could be collected to determine the extent to which mental illness is underdiagnosed or misdiagnosed in order to avoid IMD exclusion. Finally, qualitative studies of nursing facilities that have concurrent high rates of residents with mental disorders or behavioral symptoms and low rates of antipsychotic use could provide insight regarding best practices to care for residents with behavioral problems.
In addition to pressure from the Olmstead decision to provide community-based alternatives to institutional care, we note that institutionalized residents desire these options as well. In a sample of nursing home residents with mental disorders, more than 40% of residents and clinicians view community-based care as appropriate (CitationBartels, Miles, Dums, & Levine, 2003). Since the predominantly “young-old” nursing home residents in this study will be the elderly nursing home residents in the coming decades, they will increase the demand for community-based alternatives.
This article was prepared for the Substance Abuse and Mental Health Services Administration of the U.S. Department of Health and Human Services under contract no 280-03-1501. The authors thank Miki Satake, MA, and Deo Bencio, BS, who provided programming support for data analyses. Judith L. Teich, MSW, served as the government project officer and Jeffrey A. Buck, PhD, served as advisor.
Notes
1. IMDs are defined as any institution (hospital, nursing facility, or any other facility larger than 16 beds) that is “primarily engaged in providing diagnosis, treatment, or care of persons with mental diseases, including medical attention, nursing care, and related services. Whether an institution is an institution for mental diseases is determined by its overall character as that of a facility established and maintained primarily for the care and treatment of individuals with mental diseases, whether or not it is licensed as such.” (42 CFR SS 435.1009)
2. These diagnostic codes include schizophrenia, affective disorders, paranoid states, other nonorganic psychoses, neurotic disorders, personality disorders, sexual disorders, acute reaction to stress, adjustment disorders, depressive disorders, and conduct disorders.
REFERENCES
- Bagchi , A. , Verdier , J. and Simon , S. 2009 . How many nursing home residents live with a mental illness? . Psychiatric Services , 60 : 958 – 964 .
- Bartels , S. J. , Miles , K. M. , Dums , A. and Levine , K. 2003 . Are nursing homes appropriate for older adults with severe mental illness? Conflicting consumer and clinician views and implications for the Olmstead decision . The Journal of the American Geriatrics Society , 51 : 1571 – 1579 .
- Centers for Medicare and Medicaid Services. (2010). The state Medicaid manual. Part 4, Section 4390 http://www.cms.hhs.gov/Manuals/PBM/list.asp (http://www.cms.hhs.gov/Manuals/PBM/list.asp) (Accessed: 3 February 2010 ).
- Congressional Research Service . 1993 . Medicaid source book: Background data and analysis. 1993 update, Appendix E–Medicaid services for the mentally ill , Edited by: Klebe , Edward . Washington, DC : U.S. House of Representatives, Committee on Energy and Commerce, Subcommittee on Health and the Environment, Committee Print 103-A .
- Grabowski , D. C. , Aschbrenner , K. A. , Feng , Z. and Mor , V. 2009 . Mental illness: Variations across states . Health Affairs , 28 : 689 – 699 .
- Jones , A. L. , Dwyer , L. , Bercovitz , A. R. and Strahan , G. W. 2009 . The National Nursing Home Survey: 2004 overview. National Center for Health Statistics . Vital and Health Statistics , 13
- Knapp , P. K. , Hurlburt , M. S. , Kostello , E. C. , Ladd , H. , Tang , L. and Zima , B. T. 2006 . Child sociodemographic characteristics and common psychiatric diagnoses in Medicaid encounter data: Are they valid? . The Journal of Behavioral Health Services and Research , 33 : 444 – 452 .
- Linkins , K. , Lucca , A. , Housman , M. and Smith , S. 2006 . PASSR screening for mental illness in nursing facility applicants and residents , DHHS Pub. No. (SMA) 05-4039 Rockville, MD : Center for Mental Health Services, Substance Abuse and Mental Health Services Administration .
- Linkins , K. , Robinson , G. , Karp , J. , Cooper , S. , Liu , J. and Bush , S. 2001 . Screening for mental illness in nursing facility applicants: Understanding federal requirements , SAMHSA Pub. No. (SMA) 01-3543 Rockville, MD : Center for Mental Health Services, Substance Abuse and Mental Health Services Administration .
- Mechanic , D. and McAlpine , D. D. 2000 . Use of nursing homes in the care of persons with severe mental illness: 1985 to 1995 . Psychiatric Services , 51 : 354 – 358 .
- National Health Policy Forum . 2008 . Medicaid and mental health , Background paper No. 66 Washington, DC : National Health Policy Forum .
- O'Connor , D. , Little , F. and McManus , R. 2009 . Elders with serious mental illness: Lost opportunities and new policy options . Journal of Aging and Social Policy , 21 : 144 – 158 .
- Office of Inspector General . Younger nursing facility residents with mental illness: An unidentified population . Report OEI-05-99-00701 . 2001a . Washington, DC: U.S. Department of Health and Human Services
- Office of Inspector General . Younger nursing facility residents with mental illness: Preadmission Screening and Resident Review (PASSR) implementation and oversight . Report OEI-05-99-00700 . 2001b . Washington, DC: U.S. Department of Health and Human Services
- Reester , H. , Missmar , R. and Tumlinson , A. 2004 . Recent growth in Medicaid home and community based waivers , Washington, DC : Kaiser Commission on Medicaid and the Uninsured .
- Shea , D. G. , Russo , P. A. and Smyer , M. A. 2000 . Use of mental health services by persons with a mental illness in nursing facilities: Initial impacts of OBRA 87 . Journal of Aging and Health , 12 : 560 – 578 .
- Snowden , M. , Piacitelli , J. and Koepsell , T. 1998 . Compliance with PASARR recommendations for Medicaid recipients in nursing homes . Journal of the American Geriatrics Society , 46 : 1132 – 1136 .
- Swan , J. H. 1987 . The substitution of nursing home for inpatient psychiatric care . Community Mental Health Journal , 23 : 3 – 18 .
- Walkup , J. T. , Boyer , C. A. and Kellermann , S. L. 2000 . Reliability of Medicaid claims files for use in psychiatric diagnoses and service delivery . Administration and Policy in Mental Health and Mental Health Services Research , 27 : 129 – 139 .