 |
This Web site is a component of the SAMHSA Health Information Network. |
|
Screening for Mental Illness in Nursing Facility Applicants:
Understanding Federal Requirements
VIII. State Variations in PASRR Implementation
Medicaid is a State/Federal program that provides each State a great deal of flexibility in the design and administration of its program within broad Federal guidelines. Neither OBRA 1987 nor the resulting 1990 and 1992 regulations mandated a process for States to implement preadmission screening or stipulated the use of specific screening tools. The 1990 and 1992 regulations granted States great flexibility to implement even the most basic operational aspects of PASRR, such as the PASRR screening criteria and the definition of mental illness. Therefore, a great deal of variation exists in how PASRR is implemented across the States. The following section describes the various ways in which States have interpreted and implemented the program, including screening, specialized services, and alternative placements.
Screening Process
Level I Screens
The Bazelon Center (1996) found that approximately 70 percent (22 of 31) of the States responding to questions on Level II screening rely on Federal criteria specified under PASRR regulations to determine whether an individual is suspected of having a mental illness and qualifies for a Level II screen. States that did not use Federal criteria (9 of 31) reported they applied broader definitions to determine the presence of mental illness (see Table 2 in Appendix).
States also differ in determining who is responsible for conducting the Level I screen and who is qualified to administer it (Table 3 in Appendix). According to the Bazelon Center study (1996), only 2 of the 31 reporting States retained responsibility for the reviews. The majority (45 percent) use a mix of State agencies and private providers. The other States use an array of contracting arrangements, including a mix of individual private health care providers (16 percent), private agencies (13 percent), treating physicians (6 percent), nursing facilities (6 percent), and a combination of nursing facilities and hospitals (6 percent).
With regard to personnel qualified to administer the Level I screen, there is variation in terms of professional background (e.g., registered nurse, social worker, doctor, mental health professional, or discharge planner) and organizational affiliation (e.g., nursing facility; State Agency on Aging; or a State department, hospital, or independent entity contracted to conduct PASRR screens). For example, in Washington State staff at the admitting NF complete Level I screens, while in Maine screens may be completed by the hospital discharge planner, social worker, registered nurse, psychologist, psychiatrist, doctor, or nursing facility staff (Borson, Loebel, Kitchell, Domoto, & Hyde, 1997; Maine Department of Human Services, 1997). According to the Society for Social Work Leadership in Health Care (SSWLHC) report (1995), the amount of flexibility in who is authorized to complete a Level I screen is correlated with the complexity and bureaucracy involved in a State's PASRR process.
States also vary considerably in the type and complexity of screening tools used for Level I screens. For example, Washington State relies on three separate sources of data in Level I: admission and medical records, staff interviews with an aide or nurse (depending on whether the individual is being discharged from a hospital or currently resides in the NF), and a complete resident examination (including a psychiatric diagnostic interview, the Mini-Mental State Examination, the Hamilton Depression Rating Scale, and the Brief Psychiatric Rating Scale) (Borson et al., 1997). In Maine, Level I screens are much less complex; screeners are required to complete only one instrument, the Med '96 Module V (Maine Department of Human Services, 1997). Some States, such as Oklahoma, include the Minimum Data Set (MDS) as an instrument for the Level I screen since the MDS already is part of the assessment process for individuals applying for NF placement (Oklahoma Department of Mental Health and Substance Abuse Services [DMHSAS], 1997).
Level II Screens
As with the Level I screen, States vary in their implementation of the Level II screen. Definitions, responsibilities, and personnel qualifications for conducting the screen vary, as do the screening instruments used.
Not all individuals identified as meeting a State's criteria for mental illness are referred for Level II screens. Hospital discharges are exempt from PASRR prior to an NF admission when they meet certain criteria and their attending physician certifies (prior to admission) that fewer than 30 days of convalescent care is needed for the condition for which they were hospitalized. SMHAs can make determinations in advance for certain categories that have been approved by CMS and included in the State plan. The approved categories are based upon the fact that certain diagnoses, severity of illness, or need for a particular service indicate that NF admission is appropriate. For MI, a category can indicate that specialized services are not needed only for provisional emergency admissions (limited to 7 days) and short-term respite stays. All other categorical determinations that a person needs NF services must still have an individualized Level II evaluation for specialized services. There cannot be an advance determination that specialized services are needed. Other examples of categorical determinations are persons with a terminal illness and severe medical conditions prohibiting mental health treatment, including coma. The number of advance determinations made by category (aka categorical determinations) established by States varies, ranging from none to eight (Snowden, Piacitelli, & Koepsell, 1998; SSWLHC, 1995). States vary in the criteria used to make categorical determinations as well.
With regard to the determination of need for NF-level care, CMS permits each State to develop its own medical necessity criteria for nursing facility admissions and Federal guidelines require the SMHA to use the State criteria when making its PASRR determinations. The Bazelon Center (1996) found that 91 percent (30 of 33 responses) of the States defined specialized services as 24-hour inpatient psychiatric care (Table 6 in Appendix). The other three States defined specialized services as a broad spectrum of rehabilitative services designed to develop skills necessary for living independently in the community.
States vary regarding who is responsible for conducting Level II screens and who is qualified to administer the screen (see Table 7 in Appendix). Most State Medicaid agencies authorize private entities to complete Level II screens. The Bazelon Center (1996) found that 52 percent of States (26 States) contract with private entities, 24 percent use community mental health agencies, and 15 percent delegate screening to an independent State agency. In most cases, the referring individual contacts the designated entity only if a Level II screen is required (SSWLHC, 1995). For example, in Oklahoma, the Level II screens can only be completed by a CMHC that is not owned by the State. Individuals identified as requiring a Level II screen are referred to the Oklahoma Health Care Authority PASRR Unit, which in turn refers them to a State psychiatric consultant. The consultant refers individuals to an authorized CMHC, which completes the screen and sends it to the Oklahoma Department of Mental Health and Substance Abuse Services for review (Oklahoma DMHSAS, 1997). Often, the Level II screen involves input from interdisciplinary teams, including physicians, social workers, and licensed professional counselors or family therapists (Oklahoma DMHSAS, 1997). In Washington State, the Mental Health Division contracts with certified mental health professionals to conduct Level II PASRR screens. Licensed psychiatrists are required to review and sign each screen (Snowden et al., 1998). In Maine, Level I screens that meet the criteria are submitted to the Mental Health Authority, which is the final authority on whether individuals should be referred for a Level II screen. The Mental Health Authority then refers the individual to an independent assessor from a local CMHC to conduct the Level II screen (Maine Department of Human Services, 1997).
The way in which screens are conducted also differs among the States. For example, a few States, such as Hawaii, require discharging hospitals to gather and send all required information for individuals requiring Level II screens to the State Mental Health Agency or contracted agency for review. Other States, such as Kentucky, require a face-to-face psychiatric assessment conducted by the agency contracted to perform Level II screens (SSWLHC, 1995).
Screening tools also vary across States. Some States, such as Nebraska, require as many as nine different forms to be completed; other States only require one or two. Although some States do not require the use of standard forms for the Level II screen, most States do collect common data elements (SSWLHC, 1995) as required in Federal regulations. For example, Maine and Oklahoma have similar requirements for Level II screens, including the mandatory information on medical history, neurological systems, comprehensive drug history, psychosocial evaluation, psychiatric history, and an evaluation of the need for specialized services (Maine Department of Human Services, 1997; Oklahoma DMHSAS, 1997).
Specialized Services
Another area of State variation concerns the type and availability of alternative placement options, as well as the mental health services available in nursing facilities. While PASRR regulations mandate that States arrange for the provision of specialized mental health services to individuals requiring clinical intervention for acute manifestations of mental illness, CMS allows each State to create its own requirements for the kinds of services that would qualify as "specialized." For example, Maine and Oklahoma specify a range of specialized services, such as special staffing; diagnostic assessment by an interdisciplinary team that includes a psychiatrist; emergency detention (provided only for patients in danger of harming themselves or others); intensive one-to-one supervision; seclusion or physical restraint (if providing less intensive treatment fails); psychotropic medication; group therapy; individual therapy; psychiatric testing; recreation therapy; and neurological exams (Maine Department of Human Services, 1997; Oklahoma DMHSAS, 1997). Some States contract with entities or mental health professionals for a variety of reasons. For example, several services, such as emergency detention, are difficult for nursing facilities to provide and still comply with the nursing home residents' rights regulations. Because many facilities choose not to make the necessary modifications, the State is unable to contract with the facility to provide certain specialized services. Another reason States might choose not to provide specialized services in nursing facilities is because of the risk that a facility might unintentionally become an IMD and lose its Medicaid payment for persons age 22-64. A third reason is that nursing facilities do not have mental health professionals on staff and the general staff does not have the skills required to provide adequate care for persons with intensive mental health needs. Also, many States believe that the many younger people who require specialized services may pose a risk to their elderly and frail residents. Some States mandate that these services be provided at inpatient psychiatric facilities; other States have authorized community mental health centers or other outpatient facilities to provide them in the form of day treatment. These individuals residing in nursing facilities at the time of treatment receive outpatient specialized mental health services provided by an agency contracted through the State.
Community-Based Alternatives
Studies of PASRR implementation suggest that increasingly States rely on community-based alternatives for providing specialized mental health services. The Bazelon Center study (1996) found that 25 percent of individuals receiving specialized treatment in 1993 were referred to psychiatric hospitals, compared with 51 percent in 1991. The number of individuals receiving services from nonpsychiatric inpatient placements and nursing facilities increased during the same period. Alternative placements in community-based programs increased from 19 percent in 1991 to 22 percent in 1993, while the percentage of individuals remaining in nursing facilities and receiving specialized services from contracted agencies increased from 30 percent to 53 percent, respectively (Bazelon Center, 1996). There is one caveat to interpreting these data, however. Texas accounted for 76 percent of individuals receiving treatment from nursing facilities in 1993. Hence, the increase in referrals to community-based alternatives may be less substantial than the percentages suggest. However, Snowden and colleagues (1998) also found fewer referrals to inpatient psychiatric units in a study of PASRR implementation in Washington State, reporting that only 0.8 percent of Medicaid recipients were referred to inpatient psychiatric care in 1992 and 1993, while 74 percent of individuals were referred to non-nursing long-term care facilities during this same period.
According to the Bazelon Center study (1996), 36 percent of States (18) believe that PASRR had resulted in increased funding for community-based care; 39 percent (19) indicated no increased funding. Hence, reliance on community-based care varies across States and may relate to the State's own policies for increasing reliance on less restrictive settings for treatment. For example, although some States--such as Texas--are increasing reliance on alternative community-based options for specialized services, some State Mental Health Authorities still prefer referrals to more traditional inpatient facilities. For example, Maine's description of its OBRA program specifically designates psychiatric units of community hospitals as the best location for residents discharged from nursing facilities as a result of PASRR (Maine Department of Human Services, 1997). Oklahoma advises referring entities to place individuals in psychiatric units for specialized services since treatment is often short-term (Oklahoma DMHSAS, 1997).
Outcomes of PASRR Screening Process
An estimated 5.4 percent of all U.S. adults are considered to have a serious mental illness; further, it is estimated that a total of 9 percent of adults have mood disorders or SMI and experience functional impairment (U.S. Department of Health & Human Services, 1999).
No detailed studies address whether the PASRR program is achieving appropriate treatment for individuals with mental illness. However, a number of studies address the extent to which the PASRR screening process identifies people with serious mental illness.
Level I Screen
A recent Office of the Inspector General report (PASRR Implementation and Oversight, 2001) found States conducted Level I PASRR screens for only 47 percent of sampled residents with a serious mental illness. Of the screens completed, 16 percent were dated more than 2 months after admission to the facility. Compliance varied widely by State; the proportion of residents receiving Level I screens ranged from 11 to 100 percent.
In their review of PASRR, Borson and colleagues (1997) report that around 7 percent of individuals who received Level I screens ultimately were referred for Level II screens. This figure is consistent with other studies. For example, in the Bazelon Center study (1996), the number of reviews conducted ranged from 14,314 per State in 1991 to 19,775 per State in 1993. Of the States reporting both the number of reviews conducted and the number referred for Level II screens, average referral rates were 6.2 percent in 1991 and 5.8 percent in 1993. Referral rates varied greatly among the States, with Washington State reporting the highest referral rate in both 1992 and 1993--63.4 percent and 92.1 percent, respectively (Bazelon Center, 1996).
The purpose of the Level I screen is to identify NF applicants and residents who are suspected of having SMI. The Level I often is administered by hospital discharge personnel and NF staff who are not mental health professionals, and thus, Federal regulations require the diagnosis of SMI not to be included in the Level I, but to be part of the actual PASRR evaluation, which is the Level II. By regulations, the Level I cannot require a person to meet the definition of mental illness in order to be triggered to have a Level II PASRR evaluation. The Office of the Inspector General report (2001), however, recounted finding States that did not comply with this requirement and only referred for a Level II persons who met all 3 parts of the PASRR definition of mental illness.
Snowden and colleagues (1998) found that of individuals identified by Level I screens, the majority (59.7 percent) were found to have schizophrenia or schizoaffective disorder; 20.1 percent of individuals exhibited major depression. Borson, Loebel, Kitchell, Domoto, & Hyde (1997) examined diagnosis by age and found that 26 percent of individuals receiving Level I screens were younger than 65. Of these individuals, 57 percent were diagnosed with psychosis and 23 percent with dementia or mental retardation with predominantly psychiatric or behavioral presentation. The major diagnoses were more evenly distributed for individuals over 65 years old: 21 percent were found to have psychosis and 26 percent to have dementia or mental retardation.
Level II Screen
The intent of PASRR screening is to ensure that NFs only admit individuals with serious mental illness who actually need NF services and to assure that States provide needed specialized mental health services to residents who need them. When the PASRR program was enacted, by April 1, 1990, States were required to conduct a Level II PASRR evaluation on each nursing facility resident who had mental illness or mental retardation and determine whether they required nursing facility services and specialized services. Except for certain long-term residents, any resident with SMI or MR who did not need NF services was required to be discharged. Forty-six States found significant numbers of residents with serious mental illness and mental retardation who were determined not to need NF services and who submitted requests to CMS (previously HCFA) for additional time to develop alternative programs in which to place them.
The Bazelon Center (1996) found that the average number of Level II screens conducted per State in 1991 and 1993 was 1,009 and 923, respectively. Most States reported identifying very few NF applicants with serious mental illness who were not appropriate for NF care. The percentage of individuals found to be inappropriate for NF care averaged 12 percent in 1991 and 9 percent in 1993. Interestingly, Illinois found 80 percent of applicants inappropriate for NF care in 1991 and 31 percent in 1993 (Bazelon Center, 1996), which may reflect the fact that Illinois initially had a high proportion of persons with SMI residing in nursing homes who subsequently transitioned from these facilities or were diverted out through the PASRR program. Other studies have confirmed Bazelon's findings of an average diversion rate (those found to be inappropriate for NF care) of less than 10 percent. For example, the SSWLHC study (1995) found an average diversion rate of 6 percent for the 20 States reporting data between 1992 and 1994. An even earlier survey of State Mental Health Authorities by the National Association of State Mental Health Program Directors reported a diversion rate of 9 percent of 32,171 screens in 1989, 5 percent of 76,471 screens in 1990, and 10 percent of 43,853 screens in 1991 (SSWLHC, 1995).
Eight percent of nursing facility applicants who received a Level II preadmission screen in 1991 were found to need specialized services, 8 percent in 1992, and 7 percent in 1993. Of the nursing facility residents who received an annual resident review, 5 percent were found to need specialized services in 1991, 4 percent in 1992, and 7 percent in 1993 (Bazelon Center, 1996). No information was available to compare the percentage of individuals identified as needing both specialized services and nursing facility care with the percentage of individuals referred to alternative placements for specialty services. Such information would be useful, as it would provide a perspective on how often nursing facilities are responsible for arranging mental health services for individuals with serious mental illness.
Treatment Issues
Several studies have documented the inadequacy of mental health treatment for nursing home residents with serious mental illness prior to enactment of PASRR (Borson et al., 1997; Burns et al., 1993; Emerson Lombardo, 1994), but no studies have been published concerning the adequacy of mental health services for this population since PASRR implementation. Using type of treatment recommended and nursing facility compliance with recommendations as proxies for adequacy of services, Snowden et al. (1998) found that among the Medicaid NF residents receiving PASRR screens between 1992 and 1993 in Washington State, mental health services other than medication therapy were limited. The majority (87 percent) of residents screened in the study were already receiving medication therapy. Of these, 59 percent received recommendations for new treatment. Therapy was the most common new treatment recommended (39.8 percent), although it was the least common treatment already being provided. This finding is confirmed by other studies that report widespread use of psychotropic medications in nursing homes (Abrams et al., 1992; Beers et al., 1988). Snowden et al. (1998) identified that consultation was another highly recommended new treatment (27 percent of individuals received recommendations for this service).
Information on whether nursing facilities are complying with Level II screening recommendations also indicates that PASRR has had only limited impact on the appropriateness of mental health services for individuals with SMI in nursing facilities. One of PASRR's intents was to ensure that such individuals received treatment other than medication for their illness (U.S. Congress, 1987). Phillips, Hawes, Morris, Mor, and Fries (1994) compared pre- and post-OBRA 1987 data from more than 250 nursing facilities in 10 States. Their results suggest that OBRA 1987 has positively affected treatment alternatives in NFs, as more nursing facilities were providing psychological therapy and behavior management programs in 1993 than before OBRA 1987. Although this study was unable to separate out the effect of PASRR from other components of OBRA 1987, such as the Resident Assessment Instrument (RAI) and OBRA 1990 that added a requirement for NFs to provide or arrange for mental health services, the findings indicate that nursing facilities are beginning to provide more appropriate mental health services for their residents. Such improvements appear to be limited, however, as Snowden et al. (1998) found that individuals were much less likely to comply with nonpharmaceutical services. Indeed, while 94 percent of residents complied with recommendations for medication therapy, only 30.5 percent followed recommendations for new nonpharmaceutical treatments. Only 52 percent of residents recommended for therapy and 7 percent recommended for consultation complied with orders for these two most commonly recommended nonpharmacological treatments. Overall, individuals were much less likely to comply with a recommendation for a new treatment (34.9 percent) than to continue an existing treatment (91.4 percent), and the majority of existing treatment was pharmaceutical. Snowden and colleagues hypothesized that it might be easier for nursing facilities to provide medication treatment since medication management does not require NF staff to have a particular knowledge of mental health issues. Hence, while the PASRR program may help identify individuals needing mental health services, the program appears to have limited impact on the nature of treatment received.
Table of Contents | Previous | Next
|
|
