Clinical Services Research

VOL. 18, NO. 4, 1992

Schizophrenia Bulletin

CHAIRPERSON Clifford Attkisson MEMBERS Judith Cook Marvin Kamo Anthony Lehman Thomas H. McGlashan Herbert Y. Meltzer Michael O'Connor Donald Richardson Abram Rosenblatt Kenneth Wells Janet Williams PRINCIPAL NIMH STAFF LIAISON Ann A. Hohmann

1' I

iinicaI Services Research .-

clinical services research focuses on the effectiveness of clinical interventions in Community settings. clinical services research begins where clinical research itself leaves off. It is concerned with the application of clinical knowledge, gained in a controlled research environment, to the larger, relatively uncontrolled environment in which the mentally ill actually function. Its goal is toso improve the quality of care that everyday clinical practices in the provision of mental health services consistently meet existing te-of-the-art criteria. INationaJ titute of Mental Health 1991, p.9]

a

This report recommends priorities for research on clinical services for individuals who suffer from severe and persistent mental disorders. These research priorities emerged from a review of the literature and an analysis of the implications of current knowledge for future research directions. The following questions served as the basis for the Panel members' deliberation: • How accurate are clinical diagnoses in practice, and how can accuracy be improved? • How can a treatment plan best be designed to match an individual's needs for service? • How can treatment and rehabilitation services be coordinated? • How effective are services provided to individuals with severe mental illness? After considering these and other questions about the care of those with severe mental illness, the Clinical Services Research Panel developed priorities for research. The priorities are organized within the major substantive areas identified by the Panel: Characteristics of the Target Population, Assessment Research, Treatment Services Research, -

561

Rehabilitation Research, and Outcome Research. These areas mirror the intervention process of case identification, assessment of need, provision of treatment and rehabilitative services, and documentation of service effectiveness. The first section, "Characteristics of the Target Population," includes a discussion of the symptomatology, diagnostic variability, cultural factors, and demographics of severe mental illness. The second section, "Assessment Research," focuses on assessment of severe mental illness in terms of diagnosis, psychological and physical characteristics, and service needs. The next two sections, "Treatment Services Research" and "Rehabilitation Research," deal with existing service approaches and promising interventions. Treatment and rehabilitation services are given equal weight, a structure that reflects the emergent import of rehabilitation as an integral part of services provision. The final section, "Outcome Research," deals with how to document the effectiveness of services to persons with severe mental illness.

Characteristics of the Target. Population Definitions and Diagnoses. Who are the people with "severe mental illness"? They are people who suffer from long-term, persistent, and disabling mental disorders that have a profound impact on their functioning and the lives of those around them. This definition can be further explained in at least three ways. The first relies on a conventional approach to classification and diagnosis, despite controversies about the validity of the categories. The other two methods rely on assessments of severity of pathology and degree and duration of disability or functional impairment.

SCHIZOPHRENIA BULLETIN

562

As defined in the current standard

nosology (DSM-III-R; American Psychiatric Association 1987), schizophrenia and most of the disorders included in its differential diagnosis are among the mental disorders that plague severely mentally ill people. Schizophrenia, the prototypic severe mental illness, often affects thought processes, perception, sense of self, and volition. The resulting disturbances usually disrupt routine daily functioning in all areas, including work, social relationships, and selfcare. The illness may also increase the individual's vulnerability to other mental and physical illness. Many of these characteristics are also present in people with schizophrenia-related disorders such as persistent psychotic affective disorders or schizoaffective disorder. Although the course of illness may vary somewhat with special diagnoses, all of these illnesses are persistent and disabling. Psychoactive substance use disorders (abuse and dependence) are reported to coexist with the severe mental disorders with increasing frequency and severity and can significantly affect life functioning and clinical care. These diagnoses must be given special attention as comorbid conditions in any research plan to improve the care of people with severe mental illness. The multiaxial system of DSM-III-

R, which includes information about related physical disorders, psychosocial stressors, and level of adaptive functioning, ensures that certain factors beyond just the diagnosis are considered in the evaluation of an individual with severe mental illness. A recent report on physical disorders that may produce or exacerbate coexisting mental disorders revealed that they are of surprising frequency and often go undetected in many mental health facilities and settings (Koran et al. 1989).

A second method of diagnosing people with severe mental illness is by the severity of their psychopathology; a number of instruments are available for this purpose (e.g., the Brief Psychiatric Rating Scale IBPRS; Overall and Gorham 19621). Finally, level of disability or functional status may be used to define severe mental illness. l)isability or functional status can be judged by a trained professional or by the client's perception of his or her own functioning. Demographics and Epidemiology. For purposes of this report, the mental disorders that are considered severe include those defined by DSM-III-R

as schizophrenia and schizophreniform disorder, delusional disorder, psychotic disorders not elsewhere classified (including schizoaffective disorder), manic disorder, and major depressive disorder. The best available U.S. estimates of the community prevalence of these disorders derive from the National Institute of Mental Health (NIMH)sponsored Epidemiologic Catchment Area (ECA) Study (Regier et al. 1984; Robins et al. 1984). This household and institution survey was conducted in five communities (New Haven, Connecticut; Baltimore, Maryland; St. Louis, Missouri; the Piedmont area of North Carolina; and Los Angeles, California), involved over 18,500 household face-to-face interviews conducted by lay interviewers, and used a computer-based diagnostic algorithm to derive DSM-iI (American Psychiatric Association 1980) diagnoses. The following lifetime prevalence figures were reported: schizophrenia/schizophreniform disorders, 1.5 percent; bipolar disorder, 0.8 percent; and major depression, 5.8 percent. These rates are standardized to the age, sex,

and ethnic distribution of the 1980 adult household U.S. population. These rates reflect some overlap of diagnoses in that a small number of individuals may have two or all three of these disorders. Most important, these figures, which have been projected to a potential total of 15 million or more Americans, refer only to individuals who at one time or another in their lives have experienced a mental disorder. The definition of severe mental disorder in this report includes two additional factors: substantial disability induced by the disorder and an extended duration of such disability. Prior estimates of the number of severely mentally disordered people have been based on length, frequency, and duration of hospitalizations; have included the senile and primarily organic mental disorders of the elderly; and have ranged up to 2.4 million. The best estimate, based on criteria similar to those employed in this re-

-

-

port, is that people with severe mental disorders number about 1.7 million, of whom 900,000 are in institutions and 800,000 reside in community settings or are homeless (Goldman and Manderscheid 1987). These estimates, based on 1977 data, are therefore very conservative estimates for the 1990s. Delusional disorder (DD) and other psychotic disorders were not included in the ECA study. The best epidemiologic estimates for the presence of DD, summarized by Manschreck (1988), are from two Scandinavian studies that found rates of 0.024 and 0.03 per 100 persons. The community prevalence rates of such other psychotic disorders as schizoaffective disorder and brief reactive psychoses (including postpartum psychoses) are unknown but are generally thought to be less than that for schizophrenia. Reliable community epidemiologic data are needed on the prevalence of

'VOL. 18, NO. 4, 1992

563

DD and "other" psychotic disorders. Obsessive-compulsive disorder was thought to be rare (estimated at 0.05 percent) in the general population before the ECA study, which found a community lifetime rate 50 times greater, namely, 2.5 percent (Karno et al. 1988). The possibility that the population prevalences for other psychotic disorders may be considerably greater than suspected is supported by the fact that 2 to 6 percent of respondents at the Los Angeles ECA site reported various individual psychotic symptoms. Risk factors for these poorly understood disorders should be evaluated in future epidemiologic studies. Much attention is being given within health services research to th e aging of the general population. For example, it is projected that nearly 20 percent of the population will be over 65 by 2020. However, the consequences of this aging for public mental health service delivery are poorly usderstood. For example, the increased frequency of physical illness with aging may further limit functioning, and longstanding social supports that have maintained community functioning may no longer be available. Thus, the future of mental health services research will undoubtedly require greater attention and adaptation to the needs of aging and elderly severely mentally ill people. This issue should be addressed by collaborative work between mental health and geriatric research. Risk factors. The prevalence of schizophrenia has been found to be equal in men and women. Rates for schizophrenia do not differ by ethnicity but vary inversely and strongly by socioeconomic status (SES). ECA respondents in the lowest SES quartile had an estimated relative risk for schizophrenia almost eight times greater than those in the highest -

-

SES quartile (Holzer et al. 1986). It is not dear, however, whether the low SES associated with schizophrenia is the cause or effect of the disease. Fuhire research on clinical services to people with schizophrenia must take into account this disproportionate prevalence of the disease among the poorer members of society and address the potential etiologic significance of the life conditions of poverty. Although past debates focusing on social causation versus social drift as explanations for the association between low SES and schizophrenia have faded with advances in understanding the biologic basis of this disorder, the relevance of this association to clinical services research and planning should not be lost. For example, low SES may represent significant biologic risks for developing psychotic disorders that could be addressed by services. The role of better maternal and infant care programs for the poor in reducing the long-term prevalence of psychotic disorders among low-SES groups is unknown. Other biological factors that could explain the link between low SES and psychosis, such as head trauma associated with child abuse and exposure to environmental toxins, need to be examined. Opportunities to examine such issues may exist among cohorts of low-SES children involved in social experiments during the past two decades. Bipolar disorder is equally prevalent among men and women, whereas major depression is more than twice as common in women than in men. Bipolar disorder does not vary by ethnicity, but major depression is less common in blacks than Hispanics and non-Hispanic whites. The young as well as women and nonblacks are at increased risk for major depression (Weissman et al. 1991). People with disrupted marital status, as well as the unemployed, are at higher risk for

major depression. SES has little or no relation to rates of bipolar or major depressive illness. People living alone, however, are twice as likely to suffer from major depression as those living with others (Weissman et al. 1991). The above-cited demographic risk factors for schizophrenia, bipolar disorder, and major depressive disorder call for strong sociocultural awareness and planning on the part of investigators studying treatment for these disorders. Ethnic minority populations. The ECA data cited above indicate relatively few and slight differences in rates of mental disorders among blacks compared with non-Hispanic whites or those of Hispanic origin. However, the geographically and socioeconomically limited scope of the study's black sample has been cited as not adequately representative of the large and diverse U.S. black population (Williams 1986). Indeed, despite the large size of the ECA household sample, it is not representative of the United States and only provides a best-yet, but still crude, estimate of the prevalence of selected mental dis- • orders among the U.S. population. The U.S. Hispanic population, rapidly growing and now exceeding 20 million, has been the focus of major epidemiologic investigation of mental disorder only in the Los Angeles ECA study. That study was essentially confined to the Mexican-American population in two catchment areas. The National Center for Health Statistics' Hispanic Health and Nutrition Examination Survey (1-IHANES; National Center for Health Statistics 1985), which induded persons of Mexican, Puerto Rican, and Cuban descent, did indude a depression section that used the Diagnostic Interview Schedule (DIS; Robins et al. 1981) and the Center for Epidemiologic Studies Depression Scale (CES-D; Radloff 1977). -

564

SCHIZOPHRENIA BULLETIN

Results have recently been published by MOICICk1 et al. (1989) and in journals such as the American Journal of Public Health (Trevino 1990). Unpublished analyses report rates of depression generally comparable to those found among Mexican-Americans in Los Angeles. Despite these data, the prevalence of nonmood severe mental disorders among most of the U.S. Hispanic population is unknown. Those of Central American origin, numbering in the hundreds of thousands, are mainly recent refugees from military and political conflict, and profound posttraumatic stress may, as in the case of Southeast Asian refugees, compound symptomatic expressions of severe mental disorder (Cervantes et al. 1989).

The Alaska Native and American Indian populations of approximately 1.5 million people are dispersed across the North American Continent. About half live on Federal Indian reservations in 33 States; the remainder live in urban areas and in small offreservation communities. Many of the 500 Native American tribes have maintained much of their traditional culture. However, the high reported rates of alcohol use, other substance abuse, and suicide (particularly among young men) suggest a high prevalence of severe mental illness, particularly mood disorders. No reliable epidemiologic data are yet available, although culturally and epidemiologically sophisticated efforts at assessing mental disorders among this population have been initiated (Manson et al. 1987; Yates 1987), and NIMH issued a program announcement in March 1990 supporting psychometric and epidemiologic work in this area. Reliable epidemiologic data on severe mental disorder are also lacking f or Americans of Asian origin. This is

-

a nationally, culturally, and linguistically diverse and growing population, heavily concentrated on the west coast. In addition to many hundreds of thousands of persons of Chinese, Japanese, Korean, and Filipino origin, almost a million Southeast Asian refugees have migrated to the United States in the past 15 years (Flaskerud and Anh 1988). Again, although reliable prevalence rates of severe mental disorder among the Asian population have not been determined, there are clinical reports that at least suggest high rates of severe mood disorders (Kinzie et al. 1984; Beiser 1988; Flaskerud and Anh 1988; Westermeyer 1988), with a major complicating factor of posttraumatic stress disorder. Homelessness and severe mental disorders. The first report from a probability sample of inner-city homeless adults on the prevalence of DSM-III-defined severe mental disorders revealed a strikingly higher prevalence of such disorders among homeless compared with domiciled Los Angeles residents (Koegel et al. 1988). The lifetime risks for schizophrenia, manic episode, and major depressive episode were 26, 18, and 3 times greater, respectively, for these disorders among the homeless. Schizophrenia is dramatically more prevalent among the long-term compared with the newly homeless, whereas major depression is more common among the latter group. Other studies (see Tessler 1990 and Institute of Medicine 1988 for reviews) estimate the prevalence of mental illness among homeless people to be between 28 and 37 percent. Some studies have shown that between 25 and 40 percent of the homeless have been psychiatric inpatients and that half of these persons are currently in need of treatment by the mental health system. Compared with the homeless who are not mentally ill, those with

-

severe mental disorders are more likely to have been homeless for a long time, to have limited contact with family and friends, to encounter barriers to employment, to be in poor physical health for which no treatment has been sought, and to have considerable contact with the legal system. Studies of service use show that although roughly one-third of the homeless acknowledge the need for mental health services, only about one-half are willing to receive services; and of those who need mental health services, only one-quarter will actually receive any. Clearly, there is a strong need to determine what types of mental health services are acceptable and under what conditions to homeless people with severe mental disorders. Institutionalized populations. The ECA study assessed mental disorders among approximately 2,000 residents of institutions in addition to the 18,571 evaluated in their homes. These institutions included prisons, mental hospitals, and chronic care settings. Eighty percent of the institutionalized population interviewed in the ECA study (60 percent of prisoners and 70 percent of nursing home residents) had histories of mental disorder, representing double or more the rates found in the household population. An important concern here is that people with histories of schizophrenia and bipolar disorder were found in substantial numbers in all three types of institutions surveyed (Robins et al. 1991). Again, the implication is that treatment resources and programs that target persons with severe mental disorders who are in institutions other than mental hospitals should also be the focus of research concerning appropriateness, effectiveness, and continuity of care. -

-

VOL 18, NO.4, 1992

Case Mix. In health policy research and evaluations, it is frequently important to control for clinical and demographic characteristics of patients or health care insurance beneficiaries when comparisons are made across groups. For example, in comparing hospital mortality rates, it is important to control for severity of illness of patients treated in different hospitals. Otherwise, differences in mortality rates may be incorrectly attributed to hospital characteristics (e.g., quality of clinicians) when these differences are in fact due to differences in characteristics of patients (e.g., older or sicker patients). In health policy research, such control is often referred to as adjustment for case mix or complexity (Pettengill and Vertrees 1982). In clinical research, such adjustment is referred to as control for initial clinical status or severity. Components of case mix may include aspects of health status, sociodemographic factors such as education (see Burkhauser et al. 1991, unpublished manuscript), prior experiences with health care (Lewis et al. 1989), and other individual and family characteristics. What is known. Initial severity of psychopathology and specific psychiatric symptoms are strong predictors of clinical outcomes for severely mentally ill patients and for mental health patients in general. Indicators of initial severity of illness and degree of comorbidity have important effects on rates of recidivism, length of inpatient stays, and probability and level of use of outpatient mental health services by general and patient populations (Essock-Vitale 1987; Vertrees et al. 1989). Global ratings of illness severity have been developed for public mental health settings (e.g., the Global Assessment Scale [GAS; Endicott

565

et al. 19761), and some preliminary measures of illness severity have been developed on the basis of medical records (Mezzich and Sharfstein 1985; Ashcraft et al. 1989; Horn et al. 1989). The multiaxial system of DSM-III and DSM-III-R incorporates important predictive factors into our standard diagnostic system. A DSM-III-R multiaxial evaluation involves recording mental disorders on Axes I and U, physical disorders and conditions on Axis ifi, severity of psychosocial stressors on Axis IV, and a global assessment of functioning rating on Axis V that reflects social, occupational, and psychological functioning. This Axis V scale is derived directly from the GAS. What needs to be known. Comprehensive models of severity of illness, or case mix, need to be developed for specific severe mental disorders; these models should be evaluated against a variety of outcomes, including clinical course of disorders, functional status, and use of services. For people with severe mental illness, it is especially important to include multiple domains of functional status and to allow comprehensive assessment of disabilities and strengths. Components of case mix that affect important policy outcomes, such as death and recidivism, should be identified; the extent to which they predict outcomes for different clinical populations should also be determined. Models of clinical case mix should be tested across health care settings that deliver care to severely mentally ill people and should be evaluated in other populations and settings (e.g., the homeless) that include large nuznbers of people with severe mental illnesses. In these settings, components of clinical case mix should be used to predict clinical outcomes and use of services.

Specific measures of case mix adjustment need to be developed. These measures should initially be developed from a comprehensive perspective (e.g., psychiatric diagnosis, severity of illness, extent of comorbidities; level of functioning). Summary measuies of case mix that can be derived from administrative and clinical records should be identified and tested against the more developed, more clinically oriented models. Where necessary, recommendations should be made to alter reporting requirements to obtain more relevant data on patients or beneficiaries on the important components of clinical case mix. Studies should compare the data on clinical status that are obtained through different informants (e.g., patients, families, and providers) to access the predictive validity of alternative sources of information. General Health Issues. People with psychiatric disorders in general and severe mental disorders in particular—especially those who also engage in substance abuse—have a high prevalence of comorbidities, including serious chronic diseases such as hypertension, cardiovascular disease, arthritis, and chronic lung disea.es (Koran et al. 1989; Wells et al. 1989b). These comorbid conditions have particular importance because they further limit the level of functioning of severely mentally ill people. These conditions may also complicate the clinician's evaluation or treatment of severe mental illness by masking or enhancing symptoms of psychopathology. One particular public health issue of importance to severely mentally ill people is risk of acquiring the human immunodeficiency virus (HIV), which itself leads to severe disability and can cause major psychiatric illness. Research is needed on whether people

-

SCHIZOPHRENIA BULLETIN

566

with severe mental illness are at increased risk for HIV infection, through either group living patterns or difficulty in understanding or practicing safe sexual practices. People with severe mental illnesses have the same health prevention needs as others. Yet research has not focused on the general well-being of severely mentally ill people, that is, primary prevention practices to remain healthy (e.g., avoiding smoking, using seat belts), secondary prevention to detect disease early, or tertiary prevention to minimize functional limitations due to physical disease. Future research should describe the extent of major medical comorbidities and the degree of involvement in preventive behaviors of persons with severe mental illness, as well as the service implications of these general health problems. Predicting the Course of Severe Mental Illness. The natural history or course of severe mental illness has been studied most abundantly for schizophrenia and schizoaffective disorders and for the mood disorders, especially depression. Data from longterm followup studies clearly document that a substantial proportion of patients from each diagnostic class are intermittently or continuously disabled for a lifetime. Within the past 25 years, 10 outcome studies, each including patients with a minimum average followup of 10 years and involving patients with schizophrenia from various North American sites, have found that roughly 40 percent of the collective sample died by suicide, or were chronically hospitalized or living in structured situations (McGlashan 1988). Their outcomes were described variously as poor, chronic, unimproved, marginal, or continuously incapacitated. The samples included inpatients and clinic outpatients from

rural to urban settings, from low to high SES, and from first to multiple hospital admissions, and patients who were both responsive and nonresponsive to modern short-term treatment interventions. No data exist on the natural history of schizophrenia identified in epidemiologic samples. For mood disorders, recent reports from Europe and North America have challenged the original Kraepelinian (1904) hypothesis of uniformly good outcome. Followups of 11 years or longer found that between 10 percent and 33 percent of patients with major mood disorders were chronically psychotic or functionally incapacitated or both (Kiloh et al. 1988; Lee and Murray 1988; Lehman et al. 1988; Opjordsmoen 1989). Finally, a review of 12 followup studies of patients with schizoaffective disorders found rather uniformly that outcomes ranged between schizophrenia and mood disorders (Samson et al. 1988). Clearly, all such disorders can cause significant disability for a lifetime. The vicissitudes of these processes among individuals are not yet clear. The schizophrenic process, while disabling and chronic, does not appear to progress toward dementia over the long term (except perhaps in rare cases). Deterioration of functioning does characterize the disorder in its earlier stages, but this process appears to level off approximately 5 to 10 years after onset (McGlashan 1988). Bipolar and unipolar depressive episodes, on the other hand, may become more frequent and more severe with advancing age (Angst et al. 1979). Patterns of the illness process have yet to be described for schizoaffective disorders. The long-term trajectories of these disorders across individuals are heterogeneous, of course, and identifying patients who will become chronically incapacitated is a crucial task for pub-

lic health. In this endeavor, good data exist only for schizophrenia, and even here the picture is clouded. A great deal of outcome variance in schizophrenia can be linked to nondiagnostic criteria. Current disability (i.e., established chronicity and poor response to available treatments) is the most powerful predictor of future disability. Extensive comorbid psychopathology, especially substance abuse, characterizes many of the disenfranchised severely mentally ill people (Koegel et al. 1988). Chronic We stress in the form of poor economic conditions or crowded urban living often overwhelms whatever limited coping resources these patients possess. Certain characteristics have emerged repeatedly as predictors of outcome in followup studies. Poorer outcomes, for example, are found in patients with schizophrenia who have a high genetic loading for schizophrenia, a low genetic loading for mood disorders, a premorbid history of social isolation or poor instrumental functioning, an illness that emerges insidiously without precipitating factors, and psychopathology that includes thought disorders, emotional blunting, and the absence of depression, confusion, anxiety, and subjective distress (McGlashan 1988). Some of the progression of the disorders can be predicted by adequate attention to the above parameters. Even for chronic schizophrenia, for example, one study found that predictors accounted for approximately onethird of the outcome variance across multiple outcome dimensions (McGlashan 1986a). However, prediction remains far from perfect. With rare exceptions, it is still impossible to say at onset which patients will benefit from treatment and rehabilitative efforts and which will require indefinite assistance. Accurate prediction based on early cross-sectional assessment -

-

-

VOL. 18, NO. 4, 1992

567

should be a goal of future natural history studies. Overall, more needs to be known about the natural history of all the psychotic disorders, not just schizophrenia. Determination of the course and outcome of epidemiologically identified cases is especially needed, for all the existing data come from clinical populations. Better prediction of the ultimate fate of people affected with these illnesses remains the most important goal. It is central to evaluating effective treatment planning, because treatment effect must be measured against no treatment. Types of Impairments Among Persons With Severe Mental illness. Vocational impairments: What is known. Few studies address this question; most research focuses on clients or particular types of service programs (e.g., community support programs, psychos'cial rehabilitation, inpatients about to be discharged), and none has surveyed the population at large (Anthony et al. 1988). Data on employment rates and vocational needs of service recipients are therefore sketchy. Anthony and coworkers (1972) reviewed the employment rates in eight studies for persons discharged from inpatient treatment and found that between 20 percent and 30 percent of ex-patients were able to work competitively at any one time. Reviews of studies conducted more recently suggest a lower rate of 10 to 15 percent (Anthony et al. 1978; Anthony and Jansen 1984). A nationwide survey of family members of young adults with severe mental illness (Unger and Anthony 1983) found a full- and parttime employment rate of 16 percent for an 18- to 35-year-old population. Results from a survey (Mulkern and Manderscheid 1989) of 1,053 clients enrolled in federally sponsored com-

-

munity support programs (CSPs) nationwide (weighted for individual site caseload size) in 1984 indicated that only 25 percent were employed on a part-time basis. This rate was similar to the 26-percent employment rate found in the survey of 1980 CSP clients (n 1,471) (Goldstrom and Manderscheid 1982). Similar results have been reported by Minkoff (1978). The rate of employment varied noticeably when data were disaggregated by community type: 10 percent were employed in rural areas, 26 percent in urban areas, and 26 percent in suburban areas. In urban and suburban areas, approximately 75 percent of clients with jobs were employed in unskilled occupations, while 55 percent of clients in rural areas worked in unskilled occupations. The average period of joblessness of the 1984 clients was 54.6 months, or slightly more than 4.5 years, and relatively few clients (6 percent across locations) were seeking employment at the time of the survey. Finally, Dion and Anthony's review of 20 experimental and quasi-experimental psychiatric rehabilitation studies (1987) found employment rates ranging from 11 to 60 percent for severely mentally ill people receiving some type of vocational rehabilitation. A meta-analysis of nine controlled experimental studies evaluating vocational training for severely mentally ill people (Bond 1986) found that the mean employment rate (including all forms of paid employment) was 51 percent for clients receiving the experimental treatment and 29 percent for those in control group conditions. Vocational impairments: What needs to be known. The research in this area is characterized by varying definitions of employment (full- vs. part-time, sheltered vs. competitive, supported vs. independent), so that comparing studies is often difficult. -

Also, the point at which ex-patients' employment status is assessed (i.e., 6 months after discharge vs. 1 year vs. 5 years) varies widely from study to study. In addition, there has been little attempt to characterize the vocational deficits of this population beyond their employment status. It is not known, for example, what percentages of severely mentally ill peopie need work skills training, jobfinding assistance, and different types of on-the-job support. Thus, while it is possible to describe the employment status of different populations, far less is known about their vocational needs. It is important to understand which severely mentally ill people are being served (and not served) by State departnients of vocational rehabilitation. As the largest provider of vocational services for the disabled, States need to know how many people with severe and persistent mental illness have ever received State vocational rehabilitation services and how many of those individuals are currently receiving such services. Federally supported employment initiatives to serve the most severely disabled have opened up these services to this group in unprecedented, but as yet undetermined, numbers. Understanding a concept like vocational status among the mentally ill now requires State vocational information. Residential needs: What Is known. Assuming that residential needs can be inferred from current living situations, the Panel examined studies that estimate the current living situation of populations of psychiatrically disabled people. Data from the 1984 CSP cohort (Mulkern and Manderscheid 1989) indicated that the majority, 57 percent, lived in private homes or apartments, while 26 percent lived in supported community living situations and 9 percent lived in a boarding -

-

568

house or hotel. Only 1 percent had no regular residence, 2 percent lived in a psychiatric hospital, 3 percent lived in a sldiled nursing care facility, 0.3 percent lived in jail, and another 1 percent lived in some other arrangement. These data show significant changes since 1980; a larger proportion of clients now lived in private homes and apartments and a significantly smaller proportion in structured community settings. Also, 78 percent of clients lived near their families, although this proportion was much larger in urban (85%) than suburban (65%) or rural (69%) areas. Similarly, estimates of the postdischarge living status of people with severe mental illness suggest that around two-thirds of deinstitutionalized patients return to live with their families (Lamb and Oliphant 1978; Goldman et al. 1981; Goldman 1982). Torrey and Wolfe (1986) used the findings of prior prevalence studies of particular residential arrangements for a projected 2 million people with severe and persistent mental illness. These researchers estimated that 40 pezut live with families, 15 percent live with foster families or in group homes, 15 percent live in nursing homes, 10 percent live alone, 10 percent live in hospitals, 8 percent live in public shelters or the streets, and 2 percent live in jail. Hatfield (1989) surveyed 228 families who were members of the Alliance for the Mentally ill of Maryland and found that 34 percent of mentally ill relatives lived with families, 26 percent lived on their own (although 13 percent received family assistance), 19 percent lived in a supervised community residence, 19 percent lived in a hospital, and 2 percent had unknown living arrangements. These figures contrasted significantly with the same family members' preferences for the living arrangements of their mentally ill relatives: 51 percent

SCHIZOPHRENIA BULLETIN

of the mentally ill were felt to need supervised community residences; 27 percent, out-of-home arrangement with personal care; 8 percent, an outof-home environment with no faintly care; 9 percent, living arrangements with family members; 3 percent, hospital care; and 1 percent were not characterized in this fashion. Finally, estimates of the number of homeless mentally ill individuals vary widely. Data from several cities have been used to derive estimates of the total population of homeless mentally ill as somewhere between 500,000 and 1,000,000 (Drake et al. 1989). Studies have used diverse measures and definitions of mental illness (Santiago et al. 1988), as well as samples from highly diverse populations (compare, e.g., Fischer et al. 1984; Lipton et al. 1986; and Vergare and Acre 1986). As samples have improved, there has been a reduction in the common esti-. mated prevalence of mental illness among the homeless (Katz 1986). However, if data cited earlier from Koegel et al. (1988) are extrapolated to the total homeless population, about 30 percent have experienced severe mental illness and 70 percent have experienced a substance use disorder at some time. Residential needs: What needs to be known. The effect of the client's living situation on community functioning is often confounded by the degree of support from family members. Because so many people with severe mental illness live with relatives, this problem is understandable but has not helped researchers to determine the relationship between certain living arrangements and clients' actual levels of residential independence. There is little available information on what types of independent living skills training are needed by what groups of people with mental illness, and whether those in more

independent residential situations have a higher level of independent living skills. It would be useful to know the nature of mentally ill populations found in public housing arrangements throughout the country. People with severe mental illness who are trying to live in the community, especially urban communities, are logical users of such systems, yet procedural and bureaucratic impediments often prevent them from taking advantage of these alternatives. A public housing census is needed, although the methodological problems here are staggering. Nevertheless, progress has been made in counting the homeless, and research designs and extrapolation formulas could be borrowed and adapted from this area. Social skills defldts: What Is known. Actual numbers or percentages of people with mental illness who have different types of social skills deficits (i.e., problems with eye contact, gestures, voice volume) have not been identified in the literature. Several investigations base the need for social skills training on rates of rehospitalization, noting that readmission is significantly less likely among those involved in social and leisure activities (Franidin et al. 1975; Miller and Wilier 1976; Willer and Biggin 1976; Klein et al. 1978). More recent studies (Linn et al. 1980; Farkas et al. 1986) report that schizophrenic patients with strong social skills have significantly longer remissions than those with relatively weaker social skills. Regarding the prevalence of different types of social skills deficits, Liberman (1982) notes that the wide variety of cognitive and attentional deficits in schizophrenia means that a more personalized approach may be necessary for such a characterization. Two studies have found a wide gap between mental health workers' and

569

VOL. 18, NO.4, 1992

their clients' assessments of the need for social skills training. Wasylenki and coworkers (1981) reported that even though 43 percent of an inpatient population identified a need for social and recreational training, only 12 percent received referrals for such services. Similarly, Lynch and Kruzich (1986) surveyed 30 mental health agencies about 286 clients' needs for social rehabilitation and found that two-thirds to three-quarters of the clients were rated as not in need of social skills assistance, while many of the clients (no percentage given) reported that they needed programs to enhance their social skills. Social skills deficits: What needs to be learned. As noted by Avison and Speechely (1987), the need for social skills training among people with mental illness is difficult to assess, given researchers' difficulties in dearly articulating the meaning and measurement of these skills. Clearly, social skills may be defined differently by and among both mental health workers and clients; consequently, it is difficult to say exactly how many people with psychiatric disabilities need social skills training. There is also a need to understand the degree to which deficit characteristics occur in the populations with severe and persistent mental illness. For example, how.many individuals living in unstable residential situations are also in need of vocational rehabilitation and social skills training? The research evidence (discussed later in this report) strongly suggests that comprehensive, multidimensional service delivery programs work best; thus, there is a need to understand the rehabilitation-related characteristics of this population to better develop and study services to them. Family Resources. When families of mentally ill people start experiencing

the typical cycle of doctor's appointments, hospitalizations, stabilization, release-decompensation, rehospitalization, stabilization, release—the socalled revolving-door syndrome—the gravity of mental illness becomes apparent and the need for family resources becomes critical. Perhaps the most significant resource development for families was the formation of a national grassroots network in 1979 called the National Alliance for the Mentally 111 (NAMI). The purpose of NAMI is to advocate for research and better services for severely mentally ill people. The burden of families of mentally ill people has been insufficiently assessed, but the limited reports and studies conducted indicate it is considerable. Sixty-five percent of discharged psychiatric inpatients, approximately 1.5 million per year, return to live with family members (Lamb 1982). Because of cost-driven policies of short-term hospitalization, it is estimated that one-fourth to onehalf of those discharged to family care remain severely disabled. Major research questions remain to be asked concerning the characteristics and conditions of care-giving families, the measurement of family stress and burden, the assessment of naturalistic coping and adaptation patterns, the timing and extent of "caregiver burnout," and the effectiveness of various methods of respite care for family members. During the 10 years of NAMI affiliates' participation in national workshops, a consensus has developed among professionals and family members about the importance of an educational approach as a valuable resource. When families are surveyed as to their needs, their list of priorities usually reads as follows: (1) reduction of anxiety about the patient, (2) learning to motivate the patient to do -

more, (3) understanding appropriate expectations, (4) assistance during times of crises, and (5) learning about the nature of mental illness (Hatfield 1983).

Research concerning the efficacy and long-term outcome of family educational programs in reducing patient morbidity and disability and family burden is urgently needed and requested by families. Families in NAMI chapters have also urged research on methods to facilitate the community readjustment of patients and the transition to the least restrictive environment. Prevention of relapse through early intervention is another area of research urged by families. In the past few years, early symptoms of relapse have been identified. Families often know the symptoms well and can assist professionals in early recognition. It appears that if early intervention with crisis therapy and pharmacotherapy is instituted, many relapses and hospitalizations can be prevented. Specific model programs need to be tested and evaluated. Other areas of research recently urged by NAMI families include the varieties and effectiveness of supportive living arrangements, vocational rehabilitation programs (discussed later in this report), and interactive pharmacotherapeutic and sociotherapeutic programs in longer term care of severely mentally ill people. A valuable resource for families would be a guide that helps them receive the most services from their limited personal resources. Mental illness is expensive, and rewards are few. Families become frustrated and angry when they find savings being depleted with few or no resulting benefits. Family support groups often serve as a lay referral system in which families share information on which practitioners to engage and which to avoid.

-

SCHIZOPHRENIA BULLETIN

570

Important goals for these resources include more selective use of mental health services and ways to ask relevant questions about the length, nahire, and cost of treatment that will be answered in dear, jargon-free language. The effectiveness of such fainily guidance resources, however, requires evaluative research and should not be assumed, taken on faith, or judged by popularity. Emerging major family problems. Two major problems confronted for generations by families of the mentally ill are only recently becoming the focus of research. These are intrafamihal violence and substance abuse by severely mentally ill persons. Studies are needed to assess the benefits of educating family members to anticipate and prevent or minimize aggression and violence. Traditional research definitions of dangerousness or violence based on criminal arrest records skew results to emphasize only the most extreme behavior. Less serious but fear-inspiring behavior, which frequently results in painful injuries, seldom results in formal charges, because such behavior is often tolerated as part of the illness. Family members who live with the periodic bizarre actions, and threatening, assaultive, or aggressive behavior of ill relatives often deny the increased risk of violence by severely mentally ill people in an effort to reduce stigma. There is additional painful stigma suffered by family members whose mentally ill relatives are in or have been in jails or prisons as a consequence of such behavior. Research needs to focus on the realities of occasionally unpredictable and frightening behavior of people with severe mental illness. Although the special problems of diagnosing and treating the increasing numbers of dually diagnosed young adults with severe mental illness and -

substance abuse have received recent research priority, the impact of this problem on the family requires special attention. Families of mentally ill people who are also substance abusers are directly exposed to the frustrations of people who must be designated as mentally ill or substance abusers to receive specialized professional care. The all-toocommon consequence is that those with both problems receive no treatment. Studies of early intervention, education, and prevention within the family context during the 1990s must place a high priority on the dualdiagnosis patient and the dual-burden family. Indeed, the previously emphasized issue of intrafamilial violence appears to reflect increasing rates of substance abuse by people with severe mental illness. This dual crisis is an ideal arena in which to strengthen the vital partnership between professionals and families of severely mentally ill individuals. Research Priorities. • Epidemiologic studies are needed on the prevalence, risk factors, and consequences of delusional disorders, schizoaffective disorders, brief reactive psychoses, and atypical psychoses among representative treated and untreated populations. • Studies of the life course of those with major severe mental disorders are required to discover predictors of clinical course and potential response to early intervention and later treatment, and to define the specific benefits of treatment versus no treatment at different stages of disorders. • Epidemiologic studies are urgently needed on disability resulting primarily from mental disorders. Such studies of representative treated and untreated populations should assess the types, durations, and pattern of dis-

ability; its mutability over time; socioeconomic and cultural influences; and the relationship of disability to diagnosis. • Epidemiologic studies of severe mental disorders have yet to be carried out among treated and untreated samples that fully represent the socioeconomic, geographic, cultural, andin particular—ethnic minority diversity of the U.S. population. Such studies should include a careful evaluation of the potential etiologic significance of low SES levels and component factors. Such new epidemiologic studies should also assess the co-occurrence and interaction of substance abuse and severe mental disorders. • Comprehensive models of severity of illness or case mix should be developed for specific severe mental disorders and should be evaluated against a variety of outcomes, induding clinical course of disorders, functional status, and use of services. Such models should include components affecting important policy outcomes (e.g., recidivism, death) and should be tested across diverse health care settings. • Development and use of uniform terms and definitions regarding employment and vocational status, skills, and deficits among people with severe mental disorders are urgently needed. Data are lacking on what kinds of and how many vocational rehabilitation services are being provided to how many persons in State vocational rehabilitation programs. • There is a strong need for studies to determine what types of independent living skills training are needed by which people with severe mental disorders and to establish the relationship between actual living situations and independent living skills. • A public and low-income housing census is needed to determine the diversity and distribution of living

VOL. 18, NO. 4, 1992

arrangements used by people with severe mental disorders, mduding assessment of costs, impediments, discrimination, and adequacy of such housing. • Research is needed on the extent and nature of intrafamilial violence experienced by those caring for mentally ill relatives. The risk factors and early indicators for such acts of violence should be carefully assessed. • The burden on families caring for relatives with diagnoses of both substance abuse and severe mental disorders should be evaluated, with attention given to techniques of family adaptation reported to be effective.

Assessment Research Before clinical services research for people with severe mental illness can be conducted, the clinical status and needs for services of this population must be described. For example, for clinical efficacy studies the clinical status must be described at baseline and then reassessed during and after treatment. For program evaluation purposes, the needs for services of the target population must be assessed, as well as the response to the services delivered. Further, some clinical services research focuses exclusively on describing the clinical status or service needs of those with severe mental illness and is thus a research subject of some clinical and policy interest in its own right. The task of assessing the clinical status and service needs of people with severe mental illness is complicated by two cardinal features of such illnesses, as discussed earlier: the diagnostic heterogeneity and the lirnitations in functioning for this population. Thus, there are many aspects of clinical status and needs for services that potentially are important. The first section briefly describes some of

571

the relevant domains. Assessment of forensic issues is considered in detail in the Service Systems Research report (Steinwachs et al. 1992, this issue). The section on outcomes addresses the research needs related to assessment of changes in clinical status and service needs over time. Here the focus is on research needed to assess severely mentally ill people at a particular point in time, for example, the baseline assessment. Seven broad areas of clinical status and service needs are described that play a major role in services research for severely mentally ill people: diagnosis, psychosis and general psychopathology, general health status and functioning, quality of life, rehabilitation status (social skills, skills of independent living, and subsistence needs), major comorbidities, and family burden. The current state of research varies widely across these areas. Considerable work has been done in some areas, such as assessment of diagnosis and psychosis; yet even the most fully developed assessment techniques have not been tested across a wide range of populations of severely mentally ill people or across locations in which such people receive services. For some areas, such as general health status and functioning, considerable work has been done for general populations or patients who are not severely mentally ill, but that work has not been rigorously applied to severely mentally ill people. On the other hand, relatively little research has been conducted in some areas, such as major cornorbjd medical conditions. As a result of the uneven development of assessment tools across these domains, researchers are far from having either standardized operational definitions of these domains or standardized methods of assessing them for people with severe mental illness. Further, little is known about how assessment

techniques should be modified for particular subpopulations of severely mentally ill people such as the homeless or ethnic minority groups, or across types of treatment settings, such as rehabilitation and inpatient treatment programs. The Panel's recommendations for needed research are based on the underlying assumption that it is now time to work toward these objectives: operational definitions, standardized assessment techniques, and knowledge of appropriate modifications for special populations and specific types of treatment settings. In working toward these objectives, researchers are likely to rely heavily on the self-reports of severely mentally ill people, in part because patient-perceived outcomes are generally increasing in importance in health services and policy research. One important limitation of such measures is that the validity of self-reports in psychotic populations has seldom been addressed. Although the patient's perspective must be emphasized, careful attention must be paid to validation of the patient report to obtain research results that may influence clinical practice and health policy. Research Needs: What Is Known. Below is a summary of what is known about assessment in each of the seven content areas identified earlier. No attempt has been made to be comprehensive; rather the approach is to highlight the overall strengths and deficiencies in current methods of assessment, and the major research needs. Psychiatric diagnosis. Assessment of diagnosis is important both from clinical and policy perspectives. In many public mental health care settings, eligibility for services is determined in part by diagnosis. Considerable progress has occurred during the

-

SCHIZOPHRENIA BULLETIN

572

past two decades in the assessment of specific psychiatric diagnoses. In the United States, the line of development of specific diagnostic criteria is represented by the Feighner Criteria (Feighner et al. 1972), the Research and Diagnostic Criteria (RDC; Spitzer et al. 1978), and DSM-III and its revisions (American Psychiatric Association 1981). Structured diagnostic interviews have been developed to assess disorders according to these criteria. These include the DIS, the Schedule for Affective Disorders and Schizophrenia (SADS; Endicott and Spitzer 1978), and the Structured Clinical Interview for DSM-iI-R (SCID;

Spitzer et al. 1989). Other diagnostic interviews are currently av' 1 " 12 or under development (P .ie 1988). Increasingly, these interview schedules have been subjected to tests of reliability and validity. They cover a wide range of diagnostic categories, permitting the assessment of primary psychiatric diagnoses and comorbidities, and they offer the advantage of comparable assessment of diagnoses across studies. A common problem among these measures is reliance on patient self-report of the occurrence and timing of symptoms and episodes. This problem may be of greater importance for assessment of people with severe mental illness, especially for psychotic or acutely ill individuals. The accuracy of DIS-based diagnoses of schizophrenia, in particular, has been questioned (Anthony et al. 1985), but comparable validity and reliability studies for the SCID and other measures have not yet been published. Currently, no "gold standard" exists for evaluating such measures. The SCID has the advantage of a clinician's assessment, while the DIS has the advantage, especially for largescale research, of lower cost because of assessment by lay persons.

In clinical services research studies, it has not always been possible to ascertain diagnosis by a structured psychiatric interview. The reasons include the burden placed on the interviewee, which may be too great for severely mentally ill patients, and the unavailability of interviewees (for example, in retrospective studies of discharged patients). In such circumstances, studies have relied on diagnoses in medical records, reconstruction of diagnoses based on symptoms and history of illness in the medical record, or reports by the treating provider of the diagnoses or symptoms of the patient at various points in treatment (e.g., intake, discharge). To the extent that such techniques have been studied, they are thought to have lower reliability and validity than studies based on independent assessment with structured diagnostic interviews. There has been little research on the relationship between such measures and assessment with diagnostic interviews, and only limited identification of techniques to enhance the reliability and validity of information from the medical record. Although it is commonly accepted that most psychiatric disorders have features that remain the same across cultures (Draguns 1980; Escobar et al. 1986), a range of informed opinion among cross-cultural scholars of psychopathology varies with this majority view. Westermeyer (1985) argues that diagnosis across cultures specifically requires that pathologic delusions be distinguished from experiences that are part of normative cultural belief systems. Kinzie and Manson (1987) regard the subjective experience of affective symptoms to be particularly subject to cultural and subcultural influence, while Kleinman (1987, 1988) and Fabrega (1989) consider the diagnostic categories of Anglo-American psychiatry to be

-

strongly ethnocentric and laden with specific values that render them only variably and partially appropriate for use among ethnic minority and nonEuro-American populations. Even those who emphasize the universality of major psychopathology acknowledge the difficult methodologic problems in assessing psychiatric diagnoses among ethnic or cultural groups other than the dominant Anglo-American population (Beiser 1987). This problem can lead to cultural differences in the symptomatic expression of underlying psychopathology (Escobar et al. 1986; Lawson 1986), in the experience or interpretation (attribution) of symptoms (Angel and Thoits 1987; Jenkins 1988a, 1988b), or as a result of language barriers (Marcos et al. 1973; Jones and Gray 1986; Malgady et al. 1987). In addition, some psychiatric syndromes are culture bound or are defined only for a specific culture (Westermeyer 1985); there is debate about the extent to which such syndromes should be included in current nomenclature systems (Prince and Tcheng-Larouche 1987). Culturally influenced symptoms among minorities in the United States have been inadequately delineated (Jones and Gray 1986; Lopez and Hernandez 1987; Sue and Sue 1987). The DIS has been translated into Spanish and tested and applied in a MexicanAmerican community population (Burnam et al. 1983), but that translation required a number. of (mostly minor) modifications for use in a Puerto Rican community (Bravo et al. 1987), and the appropriateness of using such structured interviews across a variety of cultures has not yet been rigorously established by empirical investigation. Similar issues apply to virtually all aspects of clinical status and service needs reviewed in this section.

VOL. 18, NO. 4, 1992

Outstanding research issues related to methods of assessing diagnoses for severely mentally ill people prompt a number of questions. What are the reliability and validity of psychiatric diagnoses when assessed among vari ous populations of people with severe mental disorders? What sources of information (patient, family, friends, providers) are required to obtain reliable diagnoses in this population? How does reliance on patient selfreport affect the reliability and validity of diagnostic assessment? Can the reliability and validity of assessment techniques be improved for schizophrenia? In addition, there are questions related to the development of cost-efficient methods of screening large populations to identify cases of severe mental illness. Can brief screening instruments for severe mental disorders be developed? What are the reliability and validity of the medical record as a source of diagnostic inforination? How can the validity of such information be enhanced? Finally, there is a question related to the generalizability of diagnostic assessment techniques across major sociodemographic and cultural groups and to special populations: What is the degree of bias in assessing diagnoses across various groups, such as men versus women, underserved ethnic minorities versus majority nonHispanic white populations, and among those in chronic care institutions and among the homeless? Psychosis and general psychopathology. Similar issues pertain to the assessment of psychotic symptoms or other aspects of psychopathology. A limited number of measures of psychotic symptoms exist. Probably the most commonly used and wellstandardized self-report measures of psychotic symptoms are the Minnesota Multiphasic Personality Inventory (Hathaway and McKinley 1956)

573

and the Symptom Checldist-90 (Derogatis and Cleary 1977). Both assess active symptoms of paranoia and schizophrenia, and both require sustained concentration. In recent years, much attention has been given to negative symptoms of schizophrenia (Andreasen 1985; Bilder et al. 1985; Carpenter et al. 1985; Cornblatt et al. 1985; Crow 1985; Gibbons et al. 1985; Goldberg 1985; Lewine 1985; Mayer et al. 1985; Pogue-Geile and Harrow 1985; Sommers 1985; Strauss 1985; Zubin 1985). These symptoms appear to be highly relevant to clinical services research, for they may account for the "chronicity" and difficulty of treating severe mental illness. Further, they may be closely related to functional status and thus may serve as a focal point for the treatment concerns of both clinical and rehabilitation services. The most widely used measure is the Scale for Assessment of Negative Symptoms (SANS; Andreasen 1984a). In addition, measures of psychotic symptoms have been derived from clinical interview measures, including the Change Version of the SADS (SADS-C; Endicott and Spitzer 1978), the BPRS, and the Scale for Assessment of Positive Symptoms (SAPS; Andreasen 1984b). These measures have the advantage of additional face validity from the clinician interview, but they are more costly to administer. Other aspects of psychopathology that may be important to assess among severely mentally ill people include level of depression and anxiety, presence of other DSM-III-R Axis I disorders (discussed above), and presence of personality disorders (DSM-iI-R, Axis II). Levels of depression and anxiety are important for persons with both mood disorders and schizophrenia. These affects are associated with the clinical course of both types of disorders; for example, the

presence of an affective component is associated with a favorable course of schizophrenia. Depression is also important because, in addition to quality-of-life concerns, suicide is a • significant risk among severely mentally ill people. There are currently many self-report measures (e.g., Radloff 1977) and observer ratings (e.g., Hamilton 1967) of depressed and anxious affect. The reliability and validity of these measures, however, have not been rigorously studied in representative samples of people with schizophrenia or other nonaffective severe mental illnesses. Severe mental illness in the absence of significant psychiatric comorbidity, especially personality disorders, may be quite rare (McGlashan 1987; Fyer et al. 1988). In relation to assessment of Axis I disorders, however, assessment of personality disorders is in its infancy. The development of measures of personality disorders has been difficult in "normal" populations or among patients with less severe mental illnesses, and even more so among severely mentally ill people. A number of questions related to the assessment of psychosis and other dimensions of psychopathology can be raised here: What is the balance of negative and positive symptoms of schizophrenia that is optimal for purposes of clinical services research? What sources of information are needed to obtain reliable and valid measures of psychotic symptoms? What is the validity of information on psychosis collected in a diagnostically mixed severely mentally ill population? What is the validity of psychosis-related information obtained from the medical record? What does information on psychosis add beyond the information on diagnostic status? Can valid and reliable measures of depression, anxiety, and personality disorders be developed for

574

SCHIZOPHRENIA BULLETIN

people with schizophrenia or other severe mental illnesses? General health status and functioning. In the field of health status assessment, two general lines of instrument development can be seen: illnessspecific measures, such as those for diagnosis of psychopathology discussed above, and general health status and functioning measures. The latter permit comparisons of multiple domains of health and functioning across illnesses and provide an overview of comprehensive domains of daily functioning and health. In the field of general health services and policy research, there is now an emphasis on the use of standardized measures of health status, for which validity and reliability have been extensively assessed, as a means of evaluating the performance of health care delivery systems (e.g., a patient outcome methodology). Several alternative standardized assessment batteries have been developed (see reviews in Brook et al. 1979; McDowell and Newell 1987). All rely primarily on patient-reported information, which may be a limitation for those with severe mental illness. The measures that have been developed for general or medically ill populations have not been extensively tested for use with people who are severely mentally ill. Wells and coworkers (1989a) provide an example of such an application for depressed patients. Otherwise, studies of people with severe mental illness have focused on either one or two specific types of limitations in functioning or on the related concept of quality of life. The domains of health assessed vary across the major alternative sets of instruments but typically include general health perceptions; physical functioning, or the capacity to perform exercise and physical activities necessary for daily living; role func-

tioning, or the ability to fulfill major life tasks such as jobs, school, and housework; social functioning, or the ability to maintain and develop relationships with friends and family members; and a variety of other dimensions, such as extent of bodily pain, quality of sleep, and level of energy and fatigue. While theoretically these domains may all be viewed as relevant to assessing the health status of people with severe mental illnesses, concerns remain about the sensitivity of existing measures (developed for general or medically ill populations) to the lower ranges of health and functioning of severely mentally ill people. There is also particular concern about the omission of measures of instrumental functioning—that is, the ability to perform basic activities of daily living, such as cooking or reading a telephone book. Several questions remain about assessment of general health status and functioning of those with severe mental illnesses: What are the reliability and validity of such comprehensive assessment tools in this population? How well do they measure the lowest levels of functioning for this population? How might existing comprehensive batteries be made suitable for assessing people with severe mental illnesses? Are measures of general health status and functioning unique in relationship to measures of psychopathology and quality of life? What domains of general health status and psychopathology are most important to assess? How do general health status and functioning compare with the experience of persons with other chronic and disabling conditions? How useful are generic measures of health and functioning for assessing morbidity across treatment settings? How valid and reliable are these measures for assessing the homeless and

people from ethnic or minority backgrounds? Quality of life. The concept of qua!ity of life has become an increasingly important component of research on severe mental illness. An overview of the literature suggests that this concept consists of access to resources and opportunities, fulfillment of life's roles and level of functioning, and sense of well-being or life satisfaction. Within any of these dimensions, quality of life in several domains may be assessed (housing, finances, work, social relations, family relations, safety and legal issues, health, leisure, religion, etc.). Quality-of-life assessment has been recommended as a central focus of evaluation of services for severely mentally ill people (Schulberg and Bromet 1981) and for particular programs such as the NIMH CSP (Tessler and Goldman 1982). Alternative instruments are available for assessing these domains (see review in Lehman and Burns 1990). Major research issues related to assessment of quality of life pose the following questions: What is the relationship of quality of life to psychopathology and generic measures of health status and functioning? What dimensions of quality of life are most salient for people with severe mental illnesses? Which are most important to assess for which research purposes? How are the reliability and validity of these measures affected by social class, ethnic status, and housing status? How stable is quality of life over time, and how sensitive to change are measures of quality of life? Can brief screening versions of quality-of-life measures be developed? Rehabilitation status. Disability is a hallmark of severe mental illness. Several components of clinical status and needs for services discussed above include some aspect of disability. Moreover, severe limitations are the

-

VOL. 18, NO.4, 1992

575

primary focus of the clinical and research discipline of rehabilitation. Within this discipline, there is an established literature on the assessment of functioning and needs for services from a rehabilitation perspective. The specific components of health status addressed are social skills, vocational status, and residential placement and independent living skills. There is considerable diversity in the conceptualization and definition of social skills for people with severe mental illness. Liberman and coworkers (1986a) suggest that social skills vary along three dimensions: topographical, or specific verbal and nonverbal communication skills; functional skills, or the ability to elicit positive social interactions; and information processing, or the ability to receive social information and formulate a response. As noted by Conger and Conger (1986), there is no empirically established list of the social skills required for daily life. While various authors have recommended that studies assess motoric, physiologic, and cognitive aspects of social skills, such a comprehensive assessment has seldom been attained. Assessment techniques have included self-report questionnaires, structured interviews, naturalistic observations, and role-play situations. While each method has its advantages and disadvantages, observations of persons in natural environments are considered most accurate but, for ethical and economic reasons, are seldom used. Liberman and coworkers (1986a) noted that role playing is the main method of assessing skills in the context of social skills training programs. Anthony and Farkas (1982) reviewed 42 assessment instruments in the social skills training literature and identified several shortcomings in the methods of social skills assessment: a lack of specificity in the description of -

instruments used, the population for which they are appropriate, their validity, and the reference used to obtain the instrument; a lack of standardized measures for the chronic mental patient; a lack of instruments that directly assess client benefits (of social skills training) by using the dient as a data source; and the uncertain relationship of skills assessed in structured treatment settings to the natural environment. Other authors have suggested that research should focus more on describing the relationship of environinental reinforcers to actual skill level; some note that existing assessment techniques have only face validity and lack psychometric validation. Further, the specific assessment techniques used vary widely across settings. It is recommended that research on the assessment of social skills examine the relationships among components of social skills and move toward a more standardized operational definition of these skills. The psychometric properties of alternative social skills assessment techniques should be rigorously studied so that their relative advantages and disadvantages for particular research applications can be understood. Further, studies should examine the relationship between measures of social skills, in the rehabilitation sense, and measures of social functioning, in either the qualityof-life or the general health status sense. Studies of vocational status of people with severe mental illness have used a variety of definitions of employment and a variety of measures of vocational status (Anthony and Farkas 1982). Employment status has typically had parameters defined by temporal characteristics (recency, fulltime status), location of employment (community or workshop based), and method of payment (minimum wage

or piece rate). Some investigators have created gradations of employment levels by combining types of employment with rate of pay. Vocational status assessment techniques have included basic rates of employment applied at a programmatic level and individual assessments of employment status or of the attributes and skills known to be associated with subsequent employment status. There is little research on the specific vocational impairments of people with severe mental illness or on the need for specific types of vocational rehabilitation services. In general, relatively few of the measures that have been developed are in widespread use, and their reliability and validity among the severely mentally ill population have not been rigorously studied. Studies of attributes associated with likelihood of employment, or "employability," have relied on measures such as the Preliminary Diagnostic Questionnaire (Moriatry et al. 1988)

-

-

and Community-Oriented Program Evaluation Scale (COPES; Moos and

-

Otto 1972). (See the 1984 review by Anthony and Jansen.) Across studies, assessment based on functioning in worklike settings seems to be most predictive of later employment, emphasizing the importance of such measures. Vocational assessment of people with severe mental illness has to become as sophisticated as that used in labor force participation research generally. Research should work toward a standardized operational definition of vocational status and skills. The relationship of employment to labor market features, historical trends, and regional characteristics needs to be understood. Measures of vocational status need to be developed that take into account the level of economic depression of the local environment, especially for assessing severely men-

-

SCHIZOPHRENIA BULLETIN

576

tally ill individuals from ethnic minority backgrounds. Measures of employment should be expanded to include the various forms of labor in the secondary labor market, including informal payment arrangements, under the-table income, panhandling, illegal earnings, and other types of getting by financially. There is a need for multidimensional assessment of the vocational impairments of severely mentally ill people and for the development of measures of needs for vocational rehabilitation services. The rehabilitation literature focuses on three components of residential status and independent living skills: community tenure, or amount of time spent in the community versus institutions (Carling et al. 1988); residential status, or actual residential setting described by degree of independence or extent of financial assistance required (Dion and Anthony 1987); and independent living skills, or instrumental role performance, as well as degree of community participation. Instrumental role performance skills refer to the ability to perform basic activities of living such as cooking, cleaning, shopping, and budgeting (Cohen and Anthony 1984). Measures of community tenure are usually quite straightforward, assessing time spent in various types of locations. However, they often ignore qualitative features of the settings. Measures of residential status have been developed but not widely used across a variety of treatment settings or studies. Measures of independent living skills are diverse and have largely been developed for use at a particular treatment setting. One exception is the computerized Time-Sample Behavioral Cheddist under development by Paul (1987). Other measures of independent living skills assess community participation such as friendship and support networks and leisure activi-

ties. Because these are self-report measures, they are conceptually related to quality-of-life measures. There has been relatively little research on which aspects of residential status and independent living skills are most critical to assess and on the extent to which the status of people with severe mental illness is a result of the environment (including residential options) versus personal limitations and strengths. Future research should focus on establishing the reliability and validity of assessments of residential status and independent living skills, determining their predictive value, and developing standardized instruments that can be used across settings. Research should be conducted to define the acceptability or appropriateness of residential status for individuals at different levels of independent living skills and to establish the desirability of the residential placements to clients. The relationship among social skills, vocational status, residential status, and independent living skills should be determined, as well as the relationship of such measures to general health status and functioning, quality of life, and indicators of clinical status (diagnosis and severity of psychopathology, extent of comorbidity). Major medical comorbidities. People with severe mental disorders commonly have major chronic physical illnesses and vice versa. These conditions are important to assess because they can interfere with the effects of psychiatric treatments and rehabilitation, affect outcomes of care, and determine whether the individual decides to use services and which services are used. Yet studies of severely mentally ill people have rarely assessed the presence or absence of major comorbidities, much less assessed them with the rigor with which -

-

-

-

they would be evaluated in general health services research. Most studies rely on diagnoses in medical records, which are unreliable sources of information on medical conditions, selfreported information by individuals on the presence or absence of chronic medical conditions, and extensive medical histories or physical exarninations and laboratory tests. Most largescale surveys have relied on self-report measures of the simple presence or absence of conditions, because of the greater costs of interviews, examinations, and laboratory tests. An example of this approach is the work of Wells et al. (1988b) based on data from the NIMH ECA project. An example of a more intensive approach applied to patients with severe mental illness receiving care in the public sector was provided by Karno et al. (1988). Questions remain about the validity of data reported by people with severe mental illness regarding their physical symptoms and specific illnesses; this issue has not been systematically examined. Suggested directions for research include the following: testing the relationship between self-reported measures of chronic medical conditions and data from more intensive medical evaluations; developing structured interview and examination protocols, such as that developed for the RAND Medical Outcomes Study (Wells et al. 1989b), to assess medical status of psychiatric patients; determining which medical conditions are most important to assess for severely mentally ill people in a variety of settings; and examining the extent to which measures of general health status and functioning (especially physical functioning) capture health status variations due to chronic medical illnesses for those who are severely mentally ill. Family burden. People with severe mental illnesses can place a substantial

VOL. 18, NO. 4, 1992

burden on their family networks. Estimates of the proportion of those with severe mental illness living with families range from about 25 to 65 percent. The stress-induced family burden includes financial problems and psychological distress due to interacting with or caring for the person with a severe mental illness. Family burden is often divided into two dimensions: objective burden, or the economic and social impact of the severe mental illness on finances, household stress, and restriction on activities; and subjective burden, or the psychological distress (i.e., depression and anxiety) related to either coping with the behavioral patterns or symptoms of the family member with severe mental illness or caring for that individual. Descriptive studies have found a high level of both objective and subjective burden associated with caring for severely mentally ill people (Grad and Sainsbury 1968; Creer and Wing 1974; Hatfield 1978; Falloon et al. 1982; Cook 1988). Particular attention has been given to the level of burden experienced by parents of those with severe mental illness. Components of this burden that have been assessed include degree of disruption of future orientation (Falloon et al. 1982; Cook and Cohler 1986), conflicts related to interference in normal development of the child (Dincin et al. 1978; Hatfield 1978; Falloon et al. 1981), and conflicts in family relationships resulting from psychiatric disability. Components of subjective distress that have been studied include interference in social lives (Vaughn et al. 1984), physical fatigue, embarrassment (Creer and Wing 1975; Leff 1976), and an,dety and depression. The concept of family burden is becoming more carefully delineated, but this area of research is relatively new. Thus, there is a need for research on the reliability and validity of alter-

577

native measures of family burden and on identification of components of the family's experience that are most critical to measure, either in terms of societal costs or effects on the interaction between the course of the severe mental illness and the level of stress in the family environment. To date, there has been little research on positive aspects of the family experience for severely mentally ill people or on the contributions of families to the quality of life of severely mentally ill family members, that is, the role of the family as a source of support or care (Freedman and Moran 1984). Further, research has not systematically compared burdens experienced by families with severely mentally ill members with those experienced by families with other types of problems, such as substance abuse, severe physical illness, or mental retardation, or with those of families caring for members who are not seriously ill. Further, little research has been done on how family burden should be conceptualized or assessed for people with different diagnoses or different levels of psychopathology, or across age or developmental groups. For example, is the concept of family burden different for families with depressed members and families with schizophrenic members? How do the relevant aspects of family burden differ for young children with severe mental illness and older children with severe mental illness? Research is needed on how characteristics of the family, such as size and composition, affect both the assessment of family burden and the level of burden. For example, are certain types of burden especially likely for small families? Research is needed that will clarify how measures of family burden may be useful in the context of planning service provision. For example, how should information on family burden

be used in formulating a treatment plan? What services should be offered to family members who are not themselves severely mentally ill? This is the context in which measures of family burden become measures of clinical status and service needs. Overall Research Priorities. A number of research priorities emerge when the state of the art is examined for the separate components of clinical status and service needs reviewed above. • Future research on assessment of persons with severe mental illness should establish and empirically test operational definitions of the various domains of clinical status and service needs. • Future research should rigorously test the psychometric properties of assessment methods, using comparable standards for acceptable measures across domains. These studies should test the reliability and validity of measures for severely mentally ill people and their family members. • Research is needed to determine which sources of information are necessary to obtain reliable and valid data on domains of clinical status and service needs for people with severe mental illnesses. • The relationships among measures across domains should be examined and attempts made to eliminate redundancies. • It would be desirable to use some well-established measures across studies to allow comparisons of findings. This would facilitate the development of standardized, comprehensive measures of clinical status and service needs for people with severe mental illnesses. • Measures of comorbidity that are valid, reliable, and useful for largescale research on people with severe mental illness should be developed.

SCHIZOPHRENIA BULLETIN

578

• Rigorous assessment research is especially needed in the area of rehabilitation, which is identified here as a priority area for clinical services research on people with severe mental illnesses. • There is a need for assessment measures to describe client characteristics in treatment settings, especially the public mental health sector. These measures should be brief yet comprehensive and also have acceptable reliability and validity. • The generalizabiity of assessment methods to ethnic minority, poor, and homeless populations should be addressed. • The research on assessment should be based on representative populations of people with severe mental illness so that the generalizability of models and measures will be improved and use of measures across settings will be encouraged.

Treatment Services Research Statement of the Problem. This section deals with treatment services research, which differs from rehabilitation services research mainly in the relative emphasis on clinical outcomes (symptoms, relapse, psychological well-being, and problem behaviors such as suicide and substance abuse) versus rehabilitation outcomes (functional status, access to resources, and quality of life). These areas also differ in the focus on clinical providers and settings (physicians, nurses, social workers, psychologists, hospitals, and clinics) versus rehabilitation providers and settings (rehabilitation specialists, occupational therapists, rehabilitation centers, and vocational programs). Treatment services research differs from service systems research mainly in the level of services studied. Treatment research focuses on specific clinical modalities and settings, whereas

-

service systems research examines the interactions of services and the factors that influence these interactions. The first major problem is that little is known about which treatments work and which severely mentally ill patients benefit in various outcomes and circumstances. Even when the efficacy of individual clinical modalities is known, empirical information is usually lacking about how to combine or sequence them to achieve optimal outcomes. This picture is further complicated by the heterogeneity of the population, of potential providers and settings for treatment delivery, and of the relevant outcomes. These are questions of treatment effectiveness. The second major problem is how to ensure that clinicians apply proven treatments in clinical practice. Do providers use the most effective treatments? If so, for which patients, under which circumstances, and with what results? If not, why not? What can be done to speed the transfer of new treatment modalities and technologies into clinical practice? Can standards of treatment be established for these patients? These are questions of treatment delivery effectiveness and qualityof care. Some additional problems must be faced by mental health services research on severe mental illnesses. The recruitment and retention of qualified personnel to deliver treatment to persons with severe mental illnesses pose major impediments to effective treatment. Undergraduate and graduate training programs typically fail to train professionals to work with this patient population. Further, such work is associated too often with inferior professional status, and promising young clinicians are unlikely to choose it as their focus. Those who do choose to work with this patient population run a high risk of burnout -

from the difficulties inherent in the work, lack of information about what is new and most effective, and lack of systemic supports that encourage professional growth and self-esteem. Research on such issues would help to ensure the development of a better trained, motivated, and expanded work force to serve severely mentally ill people. The complexity of the treatment services required by these patients challenges the understanding of how to promote continuity and coordination of care. There is a growing but limited research base on how to achieve these goals under optimal circumstances. Far less is known about how to promote continuity and coordination of care in practice settings. In addition, little is known about the care of severely mentally ill people outside specialty mental health settings. Our understanding about the treatment of severely mentally ill people with concurrent conditions—especially substance abuse, serious physical illnesses, and mental retardationremains severely limited. Effective treatments for the additional problems created by such comorbidities must be found. Social circumstances also complicate the development and delivery of services to people with severe mental illness. Homelessness and cultural and ethnic differences influence the effectiveness of treatment for severe mental illnesses in ways that remain poorly understood. Finally, much of the research on people with severe mental illnesses has been in the context of program and treatment evaluation. This type of research may lack the level of generalizability needed for major policy decisions. Large-scale, multisite mental health services studies of people with

-

VOL 18, NO. 4, 1992

579

severe mental illnesses have not been done. Review of the State of Knowledge. This section examines the state of research in five categories of knowledge about treatments for people with severe mental illnesses: treatment settings, treatment modalities, continuity and integration of care, practice issues, and special populations. Treatment settings. Hospitals. A considerable body of research exists on the in-hospital treatment of people with severe mental illnesses. Before deinstitutionalization, clinicians tried to shape various social and environmental aspects of the inpatient ward to achieve maximum treatment effectiveness. Although the studies of milieu inpatient therapy for schizophrenic patients generally failed to show evidence of effectiveness (Van Putten and May 1976), these studies mainly involved chronically ill patients in nonintensive milieus (Margreaves and Shumway 1989). However, more intensive hospital milieus may benefit schizophrenic patients whose illness has not yet become chronic (Carpenter et al. 1977; Magaro et al. 1984; Hargreaves and Shumway 1989). Two studies of extended hospitalizations for severely mentally ill people deserve special mention. Paul and Lentz's classic study (1977) of social learning among chronic State hospital patients demonstrated remarkable therapeutic effects of behaviorally oriented milieus among patients previously regarded as suitable only for custodial care. Unfortunately, this study has never been replicated. Ellsworth and associates (1979) conducted an impressive study of the relationship between characteristics of Veterans Affairs hospital wards and patient outcomes. The more effective wards mixed acute-care and chronic-

-

care patients, fostered open interaction between patients and staff, used optimal medication practices, discouraged social isolation, and maintained stable staffing. Despite various methodologic limitations (Hargreaves and Shumway 1989), these two studies epitomized the era of research that sought to characterize the optimal milieu for extended hospitalizations for people with severe mental illnesses. The era of deinstitutionalizatjon brought about marked changes in the role of hospitalization in the treatment of those with severe mental illnesses and also a dramatic shift in the focus of hospital care research. It was no surprise that researchers turned from their concerns about hospital milieu to study the impact of length of stay on outcomes. Several well-controlled studies of length of hospital stay failed to find any advantages to longer hospitalizations (more than 30 days) versus shorter hospitalizations for patients with acute severe illnesses (Braun et al. 1981; Talbott and Click 1986; Hargreaves and Shumway 1989). Another series of studies examined the impact of length of stay on severely mentally ill patients (Test and Stein 1978; Braun et al. 1981) and found that early hospital discharge of long-stay patients is possible if adequate community services exist. Although symptom outcomes were no better with shorter than with continued hospitalization, some improvement in social adjustment occurred with shorter hospitalization if active community treatment was delivered (Hargreaves and Shumway 1989). Beyond these length-of-stay studies, researchers have not examined other aspects of the role of the hospital under currently prevailing conditions of short-term stays. What distinguishes effective from ineffective short hospitalizations? Does longer term hospital-

-

ization benefit some severely mentally ill people? If so, what distinguishes a successful long-term hospitalization? Finally, as noted by Magaro and colleagues (1984), there is still a need for hospital care, because better community facilities often do not exist and because patients with severe mental illness need so much at Once: psychiatric attention, medical care, socialnetworks, shelter, food, income maintenance, and rehabilitation. For some patients, such multiple needs can be addressed most effectively in a centralized facility. Furthermore, many severely mentally ill patients are nonresponsive and resistant to even the best of therapeutic efforts. Such patients nevertheless require care in long-term facilities where structure is provided that compensates for chronic, irreversible disabilities. Studies that determine which patients need total structure, not only for survival but also for a decent quality of life, are crucial. Day treatment and partial hospital-

Researchers have examined the effectiveness of day treatment and partial hospitalization programs both as alternatives to hospitalization and as posthospitalization treatment settings. Several studies have shown that these generally are effective alternatives to hospitalization, producing comparable symptomatic outcomes and at times superior social outcomes (Hargreaves and Shumway 1989). Research on aftercare alternatives has shown some advantages for day treatment programs over medication maintenance only, particularly for lower functioning patients (Margreaves and Shuniway 1989). Day treatment appears to increase patients' likelihood of continuing on medication and at times promotes better social outcomes. Community-based care. Perhaps one of the most active areas of mental ization.

SCHIZOPHRENIA BULLETIN

580

health services research in recent years has been that of community-based alternatives to hospital-based care, induding day treatment and partial hospitalization. Experimental homebased care has been found to be at least as effective as, and at times more effective than, hospitalization for patients who could be managed at home (Pasamanick et al. 1967; Rittenhouse 1970; Coates et al. 1976). The NIMH CSP initiatives during the past decade have provided considerable impetus to the development of communitybased services (Stroul 1984). Well-designed comparisons of Programs of Assertive Community Treatment (PACTs) have demonstrated that this type of proactive community-based treatment can yield better outcomes than can hospital-based treatment at approximately comparable costs (Stein and Test 1980; Test and Stein 1980; Weisbrod et al. 1980; Hoult et al. 1983). These programs typically include aggressive clinical and social case management, training in community living skills, and crisis intervention. Outcome gains appear to require indefinite continuation of the program (Stein and Test 1980). Treatment modalities. Somatic therapies. Pharmacotherapy has been the mainstay for the treatment of severely mentally ill adults since the 1950s. Dramatic progress has been achieved in the development and effective use of a wide range of drugs for schizophrenia, depression, and mania—the disorders that will be considered here. The neuroleptic drugs, such as chiorpromazine, haloperidol, and longacting forms of these classes of compounds, have proven to be effective in alleviating the symptoms of schizophrenia and acute mania and to play a role in the treatment of psychotic depression. Lithium carbonate has proven to be invaluable not only in the treatment of hypomania and acute

mania but, most important, in the prevention of recurrent bipolar and unipolar disorders. Tricydic antidepressants (TCAs), monoamine oxidase inhibitors (MAOIs), and the new generation of antidepressants, which includes trazodone, maprotiline, and fluoxetine, are the mainstays in the treatment of urnpolar depression and the depressed phase of bipolar disorders. Electroconvulsive therapy (ECT) also remains a most effective treatment when used for severe depression and very occasionally for acute manic states that do not respond rapidly enough to neuroleptic drugs. Despite its effectiveness in the treatment of severe depression and despite the explication of clear standards of care (American Psychiatric Association 1990), ECT remains controversial because of past abuses that have been vividly portrayed in the media. Research clarifying the effectiveness and risks of ECT, current treatment practices, and methods for educating consumers, families, and the public regarding its use would be quite helpful. Similarly, psychosurgical treatments of certain extremely disabling mental disorders may deserve focused and prudent attention in research. These treatments have permitted psychiatrists and other physicians who work with mentally, ill people to respond to the vast majority of acute psychiatric crises that develop in those with severe mental illness and to help prevent relapse. There are two broad areas of deficiencies with existing somatic treatments. First, approximately 20 to 30 percent of schizophrenic patients and perhaps 10 to 20 percent of manic patients respond poorly to agents of the types mentioned above. These patients remain disabled by unremitting positive psychotic symptoms or highly labile mood disorders. It is also -

well known that the neuroleptic drugs are less effective in reducing negative symptoms than positive symptoms. Thus, many schizophrenic patients have persistent defects in motivation, initiative, spontaneity (Henn and Nasrallah 1982), and the capacity to experience pleasure. Second, it is well known that neuroleptic drugs produce a wide range of extrapyramidal side effects, including acute dystoniás, akathisia, Parkinsornan rigidity, and, after many years, tardive dyskinesia. These side effects, plus the limited efficacy alluded to above, have greatly contributed to the problems in compliance that are characteristic of many people with severe mental illness and that inevitably contribute to an increase in the rate of relapse. Innovative dosing strategies that may reduce the risk of these side effects need further examination; these approaches include noncontinuous and low-dose strategies that have yielded positive preliminary results and are the focus of large-scale investigations (Carpenter 1986). Lithium carbonate produces fewer serious side effects than the neuroleptics, but tremor, thyroid dysfunction, and polyuria nevertheless limit the ability of some patients to remain on lithium for prophylactic purposes. TCAs have a number of side effects, including anticholinergic, cardiovascular, and gastrointestinal problems, that limit their use in the elderly. While they are effective in more than 90 percent of patients with major depression, they have a relatively slow onset that leaves patients suffering for an extended time and increases the risk of suicide between initiation of treatment and eventual response, which comes 3 to 6 weeks later. The rest of this section will review some of the novel developments in the treatment of schizophrenia and the major affective disorders.

VOL. 18, NO.4, 1992

For the first time since the introduction of chiorpromazine, a more effective antipsychotic drug has been developed for people with severe mental illness. This agent, dozapine, was discovered in the midsixties, but because of a serious side-effect profile, its introduction into the United States was delayed until pivotal research (Kane et al. 1988; Meltzer et al. 1989) demonstrated that this drug was effective in treating schizophrenic patients who were minimally responsive to all other available antipsychotic drugs. Clozapine is the prototype of atypical antipsychotic drugs and is the forerunner of the nct generation of antipsychotic drugs. The following dozapinelike drugs are being or will be tested in the United States: melperone, amperozide, and risperidone. There are other novel antipsychotic drugs from a number of major pharmaceutical cornparties. The aim of most of these compounds is to show advantages in extrapyrantidal side effects and negative symptoms. They must show effectiveness in neuroleptic responders before they can be tested for clozapinelike advantages in the treatment-resistant group. Mood disorders are among the most prevalent psychiatric disorders seen in the general population, and epidemiologic studies suggest they are increasing in frequency. These diiorders are accompanied by substantial morbidity and mortality, and yet this class of illness is not viewed as degenerative. Despite the well-documented efficacy of TCAs in the treatment of depression, controlled studies suggest nonresponse or partial response rates of 20 to 35 percent (Klein et al. 1980). Although lithium continues to be the treatment of choice for bipolar affective disorders, its average failure rate in preventive treatment studies is 33 percent (Prien and Gelengery 1989). Therefore, both kinds of mood dis-

581

turbance are examples of nondegenerative diseases accompanied by significant morbidity and mortality as well as by substantial rates of nonresponse to conventional pharmacotherapy. These nonresponse rates, combined with the epidemiologic data, suggest that a large percentage of the general population have refractory mood disturbance and that the development of viable treatment alternatives will decrease human suffering. Following the unsuccessful use of TCAs in the treatment of depression, MAOIs have come to be viewed by many as the second-line treatment. If patients fail to respond to monotherapy with these or other unimodal agents, augmentation (or partial response) is an alternative. Recent data suggest that patients with intractable depressions benefit from augmentation with either lithium or thyroid hormone. Others have suggested that stimulants may work in elderly, medically ill patient subgroups. The current controversy in the field is whether these regimens possess independent efficacy or just hasten the effect of the conventional treatment without improving overall response rate. A growing body of data suggests that rapid-cycling manic depression (RCMD) accounts for a significant proportion of the 20 to 40 percent of patients who fail to benefit from lithium therapy. The occurrence of RCMD among bipolar patients varies between 13 and 20 percent (Dunner and Fieve 1974). Because frequency of recurrence is thought to contribute substantially to morbidity, improving the clinical management of this refractory patient population could reduce the frequency of treatment resistance. Anticonvulsants such as carbamazepine are believed to have particular efficacy in the management of RCMD (Post et al. 1983). Although carbamazepine has shown marked acute anti-

manic properties, its acute antidepressant properties are less impressive, and its prophylactic efficacy has recently been questioned (Frankenburg et al. 1988; Prien and Gelengery 1989). Although several authors have suggested valproate as an alternative to carbamazepine in the management of the lithium-resistant rapid cycler, no trials have systematically studied large homogeneous samples of RCMD subjects in a controlled setting (Calabrese and Delucchi 1990). Generally, anticonvulsants as a new dass of psychotropics appear topossess, at best, potent antimanic and mixed-state properties in both acute and prophylactic settings, but minimal to moderate antidepressant properties. Psychotherapy. The group and individual psychotherapies represent the first and oldest forms of psychosocial treatment for people with severe mental illness. They are based on the psychoanalytically informed conflictdynamic model of psychopathology. The assumptions of this model are that conflict and defense patterns are learned in the context of development and interpersonal relations and that these patterns can be unlearned in the context of a corrective and insightfully interpretive therapeutic relationship. It has become increasingly apparent in recent decades, however, that the conflict-dynamic model falls far short of accounting for much of the psychopathologic phenornenology of severe mental illness. Certainly the psychosocial treatments constructed from this model (e.g., intensive individual psychotherapy and investigative group therapy) have generated disappointing results (McGlashan 1984, -

1986b).

In keeping with Zubin and Spring's (1977) vulnerability-stress hypothesis,

the newer psychosocial group and individual treatments of people with

SCHIZOPHRENIA BULLETIN

582

severe mental illness are based on a deficit-compensatory model of psychopathology. This model postulates the existence of specific deficits in anatomic brain structure and physiologic functioning that lead to predictable aberrations in various components of the individual's integrated psychological systems or with integration itself. Symptomatology comes from the deficits plus the individual's psychological attempts to adapt to and compensate for the deficits. Being structural, such deficits cannot be reversed, but compensatory psychological strategies can often make a difference. In the past two decades, much has been learned about modulating stress levels in milieus of care for people with severe mental illness, such as families and residential facilities. These ideas have found their way back into group and individual treatment in the form of strategies that explicitly incorporate soothing techniques or foster supportive social networks. These may be called the newer "supportive psychotherapies." However, almost no research hasexamined these newer supportive psychotherapeutic strategies. Stanton and coworkers (1984) investigated supportive versus exploratory psychotherapy with schizophrenic patients and found only slight trends for better outcomes for the supportive group in recidivism and role performance. Some studies of outpatient therapy have suggested that group approaches may be more effective than individual techniques with chronically ill patients (Hargreaves and Shumway 1989). However, Keith and Matthews (1984) have described the research results of group psychotherapy for these patients as disappointing. The therapeutic potential of peer support in the form of social networks has received some validation by research (Anthony and Blanch 1988).

Poor social support may be correlated with increased mortality and heightened psychological distress. Programs designed to replace or augment natural support networks can decrease recidivism, although it is difficult to separate the specific contributions of peer support in these multimodal treatments. Network therapy (family therapy plus active social network construction), as tested in retrospective control group designs, can decrease recidivism and the total number of days hospitalized. Again, however, the peer group contribution is unclear. Studies should aim to isolate network effects so they can be tested. Studies should also try to understand why self-help groups among people with severe mental illness are so rare compared with self-help groups among other disabled populations. More recently, researchers have begun to explore techniques to address the psychological and cognitive deficits inherent in severe mental illness. Spaulding and coworkers in the United States (1986) and Brenner and coworkers in Europe (1989) have begun work on innovative cognitive techniques. Brenner, for example, outlines dysfunctions that patients with schizophrenia demonstrate in various stages of information processing and has developed group techniques to help patients master the steps of information processing. Spaulding's strategies derive from the "hierarchy collapse" model of schizophrenia, which postulates that under stress, the vulnerable individual's hierarchy of appropriate responses collapses because of inherent neurologic deficits or reduced tolerance to heightened arousal. He has devised interventions to remediate these deficits. These strategies are quite new and have not yet been put to any rigorous empirical tests. Sociotherapies. Two types of sociotherapies have been applied in con-

-

trolled research to people with severe mental illnesses: social skills training and family interventions. The research on social skills training is covered in the "Rehabilitation Research" section. The bottom line of much of this research is that social skills training effectively increases specific skills in the treatment context but that significant problems exist in generalizing these skills to the patient's natural environment (Dion and Anthony 1987). The tremendous growth of support and advocacy groups for families of the mentally ill over the past several years, most notably NAMI, has shifted research on the role of families in the development and course of severe mental illnesses. There has been a move away from viewing families as pathogenic systems and toward viewing them as secondary victims of these illnesses and sources of major support and assistance to patients. The conditions and burdens facing the families of those with mental illness have been documented frequently in recent years (Grad and Sainsbury 1968; Creer and Wing 1975; Doll 1976; Kreisman et al. 1979; Williams et al. 1986; Hatfield 1987). Measures of family burden have been introduced to studies as one of the central outcome variables (Test and Stein 1980; Hoult et al. 1983). Another line of research has focused on"expressed emotion" in the family environment and its impact on illness relapse. The finding of an association between expressions of criticism and hostility by family members toward the patient and patient relapse has been replicated in several studies (Brown et al. 1958, 1972; Brown 1959; Goldstein and ROdniCk 1975; Vaughn and Leff 1976; Goldstein and Doane 1982; Vaughn et al. 1984). Currently, the interaction between emotional climate and the vicissitudes of psychopathology is perceived more broadly

VOL 18, NO.4, 1992

(Hatfield 1987), rather than as evidence that family dynamics cause psychopathology. With the stress-vulnerability concept of schizophrenia, it is not surprising that a stressed family environment may be related to illness symptoms. This phenomenon may generalize to other environments frequented by people with severe mental illnesses (i.e., hospital wards, residential care settings, neighborhoods, day treatment programs, and work settings), but whether it does indeed generalize has not yet been studied. Related is the question of the role that patients' symptoms and behavior may play in generating or maintaining a negative emotional environment. The interaction of patient and environmental factors in the course of severe mental illnesses needs further study. During the past decade, various similar family interventions have been described that provide support, information, practical advice, and training in how to cope with the challenges posed by severe mental illnesses. Leff and colleagues (Berkowitz et al. 1981; Leff et al. 1982) in Great Britain developed a program for families of persons with schizophrenia consisting of educational sessions, multifamily groups, and meetings with individual families and patients. A randomized evaluation of this approach versus routine clinical care found a clear reduction in relapse for the experimental group versus controls during the ensuing 9 months (9% vs. 50%). Goldstein and associates at the University of California at Los Angeles (Goldstein and Kopeikin 1981; Kopeikin et al. 1983) developed and evaluated a short-term, crisis-oriented family therapy program in the aftercare of acute schizophrenia. They found that patients receiving the family intervention plus a standard dose of fluphenazine decanoate had fewer relapses during the next 6 months than

583

-

did patients who did not receive the family intervention or who received a low-dose neuroleptic. Falloon and colleagues (1982) implemented a behavioral family program for outpatients with schizophrenia and their families. This homebased intervention also reduced relapse rates compared with those of a more standard treatment regimen. Anderson and coworkers at Pittsburgh have developed an extensive family intervention program consisting of family sessions without the patient, a survival skills workshop, family sessions with the patient, and a maintenance phase over an extended period (6 to 12 months; Anderson et al. 1986). Evaluations of these family interventions combined with antipsychotic medication have demonstrated a positive effect of the family program on reductions in symptomatic relapses. Combinations of medication, family program, and social skills training for the patient yielded the best result; medication plus either family intervention or patient skills training produced the next best result, and medication alone yielded the poorest result (Hogarty et al. 1986). Other approaches include multiple family groups (McFarlane 1983) and supportive family counseling (Bernhelm and Lehman 1985), to mention a few. These approaches vary in format, duration, and setting, and they have not been systematically compared. Continuity and Integration of care. Continuity of care. This concept is basic to most current thinking about services for severely mentally in people. As Bachrach (1981) notes, the concept has been in the lexicon of medical care for at least four decades and remains a concern in many medical fields dealing with persistent illnesses. The growth of interest in the concept parallels an increasing focus

-

on and optimism about the treatment of chronic medical conditions, as well as the increasing complexity of the relationship between patients and service providers. Definitions of the concept vary considerably (Rogers and Curtis 1980). Some emphasize the general feeling of enduring responsibility and caring toward the patient on the part of the provider as the essence of continuity of care (Gordis and Markowitz 1971; Marsh 1972). Others also stress the mutuality of continuity of care, namely, that both provider and patient have responsibilities to maintain the contract of care (McWhinney 1975). Rogers and Curtis (1980) state: 'The essence of continuity of medical care rests with the information concerning the coherent relationship between provider and consumer. This implies a knowledge of the sequence of a consumer's contacts with the medical care system" (p. 124). They also note that continuity of care has been used to refer both to the longitudinality of a service relationship, emphasizing the locus of responsibility for care over a long period of time, and to the coherence of care over a shorter episode of care. The literature also reflects the wide view that continuity of care is a multidimensional concept. Hennen (1975) identifies five major dimensions: chronological, geographic, interdisciplinary, interpersonal, and informational. Rogers and Curtis (1980) add to these the dimensions of accessibility and stability. In her review of the concept, Bachrach (1981) notes seven dimensions: temporal (longitudinal), individual (individually tailored care), cross-sectional (comprehensiveness of service plan), flexibility, relationship (primary provider-consumer relationship characterized by familiarity and doseness), accessibility, and communication (between consumer and pro•viders and among providers).

SCHIZOPHRENIA BULLETIN

Given the diversity and complexity of meanings of the concept, it is not surprising to find that widely accepted methods for measuring continuity of care are yet to be developed. Bass and Windle (1973) developed a service center questionnaire to measure continuity of care, defined according to two dimensions: lack of obstacles to client movement among an agency's services on the basis of therapeutic needs, and administrative mechanisms linking present and past care. More recently, Tessler and coworkers (1986) designed a provider questionnaire that assesses continuity of care along three dimensions for severely mentally ill patients being discharged from a hospital: discharge planning, successful and rapid transfer, and implementation of individualized service plans. In an application of this measure to assess the relationship between continuity of care and patient complaints about treatment and subsequent adjustment, Tessler (1987) found that the strongest correlate of these patient outcomes was an index of the total number of services recommended but not received. One must look to the nonxnental health literature to find other attempts to operationalize continuity of care, and here also the measures are quite limited. Percentage of office visits with the same personal physician has been used as an index of continuity of care (Breslau and Reeb 1975; Aylett 1976; Rogers and Curtis 1980). Return for followup care has also been used (Boyle and Rockhold 1979). More generally, Rogers and Curtis (1980) have proposed looking at five measurable components in a continuity-of-care model: provider characteristics, consumer characteristics, encounter types, knowledge base, and the "continuity environment," which isdefined in terms of the chronology of care, site of care, specialized

versus holistic care, relationships among key people, information system accessibility, and mobility versus stability. They state that these components can be measured with such quantifiable variables as number of encounters, number of providers involved in the care, number of sites of these encounters, and aspects of information recorded. In summary, the concept of continuity of care, though widely accepted as important to understanding the quality of services for severely mentally ill people, remains a complex and largely unmeasured variable. As Rogers and Curtis (1980) state, "It seems unlikely that continuity lof carel can be measured in a global sense, nor can all the dimensions ... be accurately defined. It is therefore important to select and agree upon specific areas of continuity of care which are easily measured, yet have significance when. related to outcomes studies" (p. 125). Case management. Case management, much like continuity of care, now has wide conceptual acceptance as a bulwark of effective services for persons with severe mental illnesses. However, the great diversity of definitions and the lack of widely accepted measures have impeded systematic research. An extensive theoretical and anecdotal literature exists and has been reviewed elsewhere (Harris and Bergman 1988). Conceptually there are two overarching issues that emerge in defining case management: First, what are its dimensions? Second, should case management be viewed as a service system activity or as the function of specific service providers, that is, case managers? A wide variety of potential activities have been subsumed under the term case management: outreach, client assessment, case planning, referral to service providers, advocacy for

the patient, direct casework, development of natural support systems, reassessment, advocacy for resource development, monitoring of quality, public education, and crisis intervention (Intagliata 1982). Various models of case management have been proposed: minimal, coordination, comprehensive (Intagliata 1982), clinical (Lamb 1980; Harris and Berman 1988; Kanter 1989), and rehabilitation-oriented (Goering et al. 1988). Thesemodelscanbeprovidedbyindividuals, teams, or systems (Intagliata 1982; Torrey 1986; Harris and Bergman 1988). The great variability in the definition of case management contributes to much of the confusion that exists in the field and poses a major challenge to the research community. The practical effects of this conceptual ambiguity are far-reaching. An ill-defined term can create a false sense of agreement among providers and between providers and patients, families, and regulatory bodies. This perceived agreement of common purpose and responsibility often evaporates when more specific expectations are expressed about who will do what; conflict is often exacerbated in the face of the high service needs of severely mentally ill people. Considerable disagreement exists as to which model of case management works and who should provide it (Lamb 1980; Intagliata 1982; Hargreaves et al. 1984). The multiple meanings also make it difficult to interpret the few existing empirical studies of case management, which do not always define the nature of the case management studied. Even when they do, there is a tendency to assume that the same phenomenon is under study when studies are being compared. Most studies of case management provide at best a brief description of the case management model employed without any objective measures of -

-

-

VOL. 18, NO. 4, 1992

what services actually were provided. Hargreaves and coworkers, in particular, have been working to develop measures of case management. In 1984, they reported on a record review technique to measure three dimensions of case management activity: structure (identified case manager, caseload); actions (evaluation, written plan, referrals, advocacy, monitoring, case conference, level of effort); and context (patient willingness, voluntary patient status, system problems, and catchment status). More recently, Hargreaves has been developing the Mental Health Work Environment Scale with the following dimensions: innovation orientation, involvement in job, clarity, cohesion of work group, supervisory support, outreach orientation, team functioning, results of team orientation, interest in severely mentally ill clients, tolerance for aberrant behavior, clinical and case management autonomy, case management importance, case management functions performed, referral advocacy, interagency orientation, client empowerment philosophy, model treatment philosophy, importance of psychotherapy, medication attitudes, and continuity of care (Hargreaves 1989). In summary, the development of measures of case management is in its infancy. This lack of measures makes it difficult to interpret the research. The importance of this concept in the minds of clinicians, agency directors, and mental health authorities, combined with the ambiguities in the definition of the concept, different expectations about its primary functions (improvement of quality of care, improvement of patient outcomes, cost containment), and the lack of high-quality research, forms the setting for much misinterpretation that can have farreaching policy consequences.

585

-

The empirical literature on the effectiveness of case management is both scant and mixed in its results. The best known experimental study is the evaluation of assertive community treatment (essentially intensive clinical case management) by Stein and Test (1980) in Madison, Wisconsin. In a 14-month randomized comparison of hospital treatment plus standard aftercare versus an innovative, assertive outreach and community-based treatment program, the patients in the experimental program had lower rates of hospitalization, higher levels of functioning, greater life satisfaction, and no differences in levels of family burden versus the control cases. Costs for the two programs were essentially equivalent (Weisbrod 1983). These results have been replicated in Australia (Hoult et al. 1983). In the Wisconsin study, the experimental case management occurred within the context of considerable local control of the mental health system by the investigators. There have been some other studies with promising results for case managelnent. Bond and coworkers (1988) compared assertive case management (again essentially an intensive clinical case management model) with standard aftercare for severely mentally ill patients at risk for hospitalization in a randomized experiment in three community mental health centers (CMHCs) in Indiana. In the 6-month foflowup period, the assertive case management patients had fewer hospital days and fewer rehospitalizations than controls. However, no group differences were found on quality of life, medication compliance, involvement in CMHC programs, or contacts with the legal system. Direct and indirect treatment costs for the case management group varied by a factor of 250 percent across the three CMHCs. In Canada, Goering and coworkers

-

(1988) compared a rehabilitationoriented case management program with a standard aftercare program following inpatient treatment. Case management patients were compared with matched controls, who received standard aftercare before initiation of the new case management program. At 2-year followup, patients in the case management group had better occupational functioning and were more likely to live in independent living settings and be less socially isolated. No differences were found in hospitalizations. Other studies of case management have yielded less favorable results. In a randomized comparison of case management (type not dearly specified) and standard community-based care without formal case management, Franklin and coworkers (1987) found that the case-managed patients at 12-month followup had received more services, their maintenance cost more, and they had more hospitalizations but had no better quality of life than the standard treatment group. Borland and coworkers (1989) followed a cohort of young severely mentally ill patients over a 7-year period in a time-series design with 2 years of preintensive case management and 5 years with case management. They found that with case management, patient hospital days were reduced by 75 percent, but the reduction in inpatient costs was offset by a 183-percent increase in structured residential care days. No significant changes in level of functioning or in total costs of services were noted. In summary, the research on case management is extremely limited, and studies vary in the types of case management examined, comparison group conditions (also not often well defined), and results. -

Linking clinical and rehabilitation services. Although the proper coordi-

SCHIZOPHRENIA BULLETIN

586

nation of clinical and rehabilitation services seems essential for people with persistent and disabling illnesses, such linkage remains problematic for many reasons. Clinical services refer to somatic and psychotherapeutic treatments f or mental disorders. Reha-

bilitation services are interventions that provide for a more supportive environment and for enhancement of functional skills. The goal of clinical services is to treat patients to alleviate the symptoms of their illnesses. The goal of rehabilitation is to help clients overcome the disabilities caused by their illnesses so as to maximize their capacity to fulfill life's roles. These two broad types of services arise from different traditions, use different terminologies, are delivered by different sets of service providers, and have typically been at odds. What they have in common is the people who need their services. Despite the historical differences between these two fields of services, recently there have been notable efforts to integrate clinical and rehabilitation concepts and approaches (Anthony and Liberman 1986; Myerson and Fine 1987; Liberman 1988). These moves toward more effective collaboration reflect the maturity of the two fields and the realization that clinical and rehabilitation services must be coordinated to provide the best service. There is very little empirical literature on linking dinical and rehabilitation services for severely mentally ill populations. One study exemplifies what might be accomplished with such linkage. Hogarty and coworkers (1986) randomized patients with schizophrenia among four treatment groups with various combinations of fluphenazine decanoate (FD), a clinical service, and two rehabilitation services: family psychoeducation (FT), an environmental support service, and social skills train-

ing (SST), a functional skills enhancement service. Each of the four groups received one of the following treatments: FD only, FD and F, FD and SST, or FD and FT and SST. During the first 12 months of followup, the relapse rates were 41, 19, 21, and 0 percent, respectively. While this study focused primarily on symptom outcomes, it indicates that combining clinical and rehabilitation services will improve outcomes.

Practice issues. Treatment in genMost severely mentally ill people have some contact eral medical settings.

with the general medical sector. For example, the California Homeless Study found that 60 percent of the homeless who had severe mental illnesses had contact with the general medical sector in 6 months (Vernez et al. 1988). Further, about one-half of psychiatric hospitalizations for the elderly or disabled (e.g., Medicare hospitalizations) occur in medical wards or acute-care general medical hospitals (Taube et al. 1985). In addition, a large proportion of the nursing home population has some form of severe mental illness (Goldman et al. 1981). Some of this contact with the general medical sector occurs within a primary care relationship. Regier and coworkers (1982, 1985) have defined a primaiy care provider as the first point of entry for patient contact and as the agent having responsibility for providing continuous and coordinated care. Given the difficulty that many people with severe mental illnesses have in establishing ongoing relationships, most encounters with the general medical sector may not, in fact, occur in the context of a primary care relationship as defined above. Nevertheless, this encounter may be the only point of contact with the health care delivery system. If so, it would

be extremely important to develop an awareness within the general medical sector of the important role primary care providers have in detecting and treating or providing referral for people with severe mental illness. The literature on patterns of effectiveness of treatment by general medical clinicians of persons with severe mental illnesses is very spotty (Hanldn and Oktay 1979; Wilkinson 1985). While there is now a substantial literahire on treatment of affective disorders or depressive symptoms (Regier et al. 1978; Orleans et al. 1985; Schurman et al. 1985; Wells 1985; Wells et al. 1986; Elkin et al. 1989) and on patterns of prescription of psychotropic drugs in general medical settings (Verbrugge 1984; Wells et al. 1988a), almost none of this literature focuses on patients with schizophrenia or schizophreniform disorders, or on use of neuroleptic agents. As a result, virtually no estimates are available on how the general medical sector provides care to patients who have severe mental illnesses (other than depressive disorders). Quality of care research. In considering the current state of quality-ofcare research for people with severe mental illness, two conceptual paradigms are useful. First, Donabedian (1980) has separated studies of quality of care into studies of the structure of care (e.g., provider systems, characteristics of the treating facility, and organization of services), process of care (specific clinical interventions), and the outcomes of care. Second, Berwick (1989) has identified four components of health services research that relate to the quality of care received by patients. The first is effectiveness of care, or "what works"; the second is the appropriateness of care, or "using what works"; the third is execution of care, or "doing well what works"; and the fourth is examination -

VOL 18, NO. 4, 1992

587

of the purpose of care, or "the values that underlie action." The first paradigm focuses on important components of the process of health care delivery; the second focuses essentially on different stages in the application of acceptable standards of clinical care to practice. Progress with both of these paradigms is needed to advance quality-ofcare research for the severely mentally ill population. Mental health services research on the quality of care for severely mentally ill people is inconsistent in quality and addresses only a few of the aspects of the two paradigms above. Wells and Brook (1989) provide an overview of the studies of the quality of mental health care in general. McGlynn and coworkers (1988) provide recommendations for quality-of-care research for severely mentally ill populations in the context of a mental health policy perspective. There are numerous studies of the issue of effectiveness for specific patient populations. Most studies of the quality of care (appropriateness, execution) received by severely mentally ill people derive from program evaluations of CMHCs or State hospital inpatient treatment; most of these studies were conducted before 1980. In addition, there are several studies of the quality of drug-prescribing practices in primary care clinics. There have been only a few attempts to develop standards of care for severely mentally ill people (Tischler and Reidel 1973; Pincus et al. 1983), and these attempts have generally not led to the development of new or rigorous research. Advancement in quality-of-care research requires careful synthesis of literature on the effectiveness of care, development of specific criteria for acceptable care through such techniques as decision analysis or consensus conferences (Pincus et al. 1983), -

-

testing of standards by examining process-outcome links through descriptive studies of quality of care, development of measures of process and outcomes based on these criteria, validation of measures, and then descriptions of quality of care in particular research applications (Wells and Brook 1988). Provider's evaluation. Once a person enters care, the provider's evaluation of the clinical status and needs for services (clinical, rehabilitative, and subsistence services) is the first step in instituting appropriate care. A misdiagnosis or incomplete diagnosis could lead to inappropriate treatment (e.g., antidepressant medication rather than antipsychotic medication). Ultimately, errors in evaluation may affect clinical outcomes—an important area for research. There are five components within the provider's evaluation of a person with severe mental illness: (1) the comprehensiveness of the evaluation, or the range of patient needs assessed; (2) the depth of evaluation of a particular need; (3) the reproducibility (reliability) of the evaluation across providers and assessment occasions; (4) the consistency of the evaluation over time, that is, whether the provider continues to reassess the patient; and (5) the validity of each component of the evaluation relative to independent measures of need, such as a research assessment of need (see the "Characteristics of the Target Population" section). The literature has focused on whether primary care providers detect serious emotional problems or psychiatric disorders in their patients (Goldberg et al. 1982; Kessler et al. 1985). These studies suggest that for twothirds to one-haif of patients with emotional problems treated in primary care settings, the treating provider is unaware of the problem. Von -

-

-

-

Korff et al. (1987) found a higher rate of detection of major depression, but in that study, providers were specifically aware that depression was being assessed. A few studies have focused on patient characteristics associated with level of awareness of emotional problems, and found gender, SES, and severity of psychopathology to be important factors (Schwab et al. 1967; Schottstaedt et al. 1971; Nielsen and Williams 1980; Seller et al. 1981; Cavanaugh 1983; Cleary 1989). In a study of English family practitioners, Shepard and coworkers (1966) found that patient turnover and physician attitudes about psychogenic factors explained 51 percent of the variation in level of detection of general psychological distress. Marks and coworkers (1979) found that the physician's level of concern and authoritarianism explained a high proportion of variation in accuracy of detection of psychological distress. There have been few studies of detection or assessment of emotional conditions or specific psychiatric disorders by mental health specialists. Schulberg and coworkers (1985) found that primary care clinicians underdiagnosed while providers In a CMHC overdiagnosed cases of DSM-III depressive disorders (independently assessed using the DIS). Wells and associates (1989b) found that patients with current major depression or dysthymia were much more likely to have their depression detected if they were seen by a mental health specialist rather than by a general medical clinician. These researchers also observed that among patients visiting general medical clinicians, detection of current depression was significantly greater when care was financed by fee-for-service arrangements than by prepayment (Wells et al. 1989a). -

-

-

-

-

588

A few studies and surveys have been conducted to determine the accuracy with which practicing clinicians use the diagnostic manual. On the basis of a survey conducted in 1980 and repeated in 1982, Lipkowitz and Idupuganti (1985) concluded that DSM-1II did not have a major effect on the diagnostic practices of psychiatrists. Although most of the 341 respondents were able to note at least one of the DSM-II1 criteria for schizophrenia, hardly anyone was able to list all of them. Upton and Simon (1985) conducted a study in 1982 to determine the effect on diagnostic practice of a statewide succession of DSM-III training programs in New York. In a series of 89 randomly selected charts with diagnoses of schizophrenia, expert reviewers were able to confirm only 16 of the diagnoses; the experts were much more likely than the treating physicians to diagnose affective disorders. Documentation of the DSM-ilI criteria was not present in 80 percent of the charts reviewed. The authors concluded that "Most often, once written, the diagnosis of schizophrenia became irrevocable and apparently was never reconsidered" (p. 371). Jampala and coworkers (1986, 1988) conducted a survey in 1983 and 1984 of 1,000 randomly selected psychiatrists and all graduating psychiatric residents in the United States to determine their views and use of DSM-III. The researchers found that by mid-1984, about 50 percent of the 557 responding psychiatrists and 75 percent of the 498 responding fourthyear residents had received some formal training in the use of DSM.-III, and that such formal training was associated with "voluntary enthusiastic use of DSM-III." However, 41 percent of psychiatrists in private practice responded that they would stop using DSM-III if it were not required. (It

SCHIZOPHRENIA BULLETIN

may be that fewer members of this group had received DSM-ffl training.) The researchers also found that the survey respondents could not accurately list the DSM-iI criteria for major depression, mania, or schizophrenia, leading the researchers to conclude that DSM-III is not being adequately used. A few studies have examined the quality of detection of physical illness in patients who have psychological problems and are treated in public mental health care settings or in primary care clinics (LeBruzza 1981; Hoffman and Koran 1984; Koran et al. 1989). These studies suggest that a high proportion of patients treated in the public mental health sector have undetected physical illnesses and that these illnesses may be causing or contributing to the patients' psychiatric conditions. For example, Katerndahl (1981) examined the assessment of contributing organic factors for depressed patients seen by family practitioners. While nearly 50 percent of the patients had a nonpsychiatric disorder that could cause depression, and more than 40 percent were on medications that could cause depression, only 7.6 percent of charts indicated that the physician was aware of these associated factors. These studies are quite limited for drawing inferences about the quality of assessment of people with severe mental illness. None of these studies examined the comprehensiveness of the initial evaluation. While some examined issues related to the depth and validity of the assessment, few did so rigorously or examined other issues, such as reliability of the provider's assessment. Rather, there is a tendency in the literature to give the provider full credit if any general physical or emotional problem is noted as assessed. Because most studies were lirnited to one site or based on nonrepre-

sentative sampling techniques, there are few estimates of how components of the provider's evaluation differ among treatment settings or types of provider. Another body of research has focused on efforts to improve the provider's assessment. Part of this research has been aimed at developing brief versions of assessment measures that can be used in office settings. For example, Greist and coworkers (1984) developed a computerized, selfadministered version of the DIS. Zimmerman and coworkers (1986) developed a short, self-administered questionnaire to assess current major depression. Burnam and coworkers (1988) developed a very brief screening instrument for current depressive disorders. Studies, however, have not addressed the extent to which such measures can be implemented in practice settings, their aceptability to patients and providers, or their validity when administered by office staff versus research teams. The literature is mixed regarding the value of providing feedback to providers on the basis of results of selfadministered screening instruments for psychological distress. Johnstone and Goldberg (1976) and Hoeper and cbworkers (1984) found that such feedback had very little effect in medical outpatient clinics. Linn and Yager (1980, 1982), however, found higher rates of diagnoses of depression and higher total medical costs after such feedback. In that study, however, feedback was provided along with specific suggestions for change in treatment. None of these studies addressed the value of providing feedback on the clinical status and service needs (outcomes) of people with severe mental illnesses. The existing literature, while representing a beginning for studies of the clinical assessment of people with Se-

-

-

-

589

VOL. 18, NO.4, 1992

vere mental illness, has either not f ocused on that population or has addressed only a very narrow range of research issues related to the broader concept of patient evaluation. Research is needed that specifies the components of a complete evaluation, including the way in which that evaluation should reflect different types of providers and settings. This research is similar to the development of standards for acceptable care, a component of quality-of-care research discussed in the next section. To study evaluation, however, the provider's activities need to be measured in terms of the provider's objectives inproviding care. Some sources of data on the provider's assessment, such as the medical record, are thought to be unreliable or of uncertain validity. As a result, measures of the provider's assessment must be developed that capture aspects of the evaluation according to the conceptual framework, training, and treatment orientation of that provider. Otherwise, the context and purposes for the evaluation may be unclear. To date, few studies have focused on the psychometric properties of the provider's assessment or of the appropriateness of the measurement to the particular treatment setting. The process of evaluation is affected by the patient's presentation and understanding of the illness and symptoms as well as the provider's style and practice constraints. Yet the role of these factors in clinical evaluation of people with severe mental illness has not been studied. For example, what factors related to seeking help affect the patient's awareness of symptoms and ultimately the provider's assessment? In addition, because evaluation is often the result of coinmunication among providers, patients, and family members, research may productively focus on the role of interper-

sonal communication in attribution of symptoms and ultimately the clinical evaluation. Evaluations cost money. Yet to date there has been little research on the cost-effectiveness of complete evaluations or of the implementation of programs to improve evaluation of people with severe mental illness. Presumably, better evaluations lead to better outcomes. Even for detection, the most studied aspect of evaluation, however, virtually nothing is known about the link between process of care and health outcomes. Research is needed to determine how components of the clinical evaluation affect outcomes and which ones are therefore most clinically important. Prior research indicates that studies of evaluation should be conducted to reveal deficiencies in care; however, little is known about how to improve deficiencies in care (i.e., the literature is mixed on the value of providing feedback). If evaluation is shown to be linked to outcomes, demonstration programs on alternative approaches to improving evaluation are needed. A number of research priorities for improving provider evaluation emerge from the foregoing considerations: development of standards for completeness, depth, reliability, and accuracy of clinical evaluations of people with severe mental illness, with special attention to variations across treatment settings and types of providers and patients; development of research that describes patterns of evaluation across settings and types of patients; development of cost-effectiveness research related to the clinical evaluation of patients; development of research on the links between initial evaluation, evaluation over time, and other important treatment decisions, such as the decision to hospitalize or to start or change a medication regimen; development of research that -

demonstrates the effectiveness of alternative programs to improve clinical evaluations; and development of research on the psychometric properties of measures of clinical evaluation, including their reliability and validity in relation to independent measures of the patient's clinical status. Special populations. As discussed in the "Assessment Research" section, a number of problems exist with cultural biases in many standard assessment procedures. Such biases in assessment and diagnosis have many implications for the quality of care received by persons from cultural minorities. It follows that inaccurate assessments and diagnoses will lead to inappropriate treatments for minority groups; however, no such quality-ofcare studies have been conducted for people with severe mental illness. Early studies of mental health care revealed dear differences in the type of treatment received by patients with the same diagnosis depending on their socioeconomic background (Hollingshead and Redlich 1958). Similarly marked discrepancies in the types of care received by severely mentally ill people from different socioeconomic and cultural backgrounds probably persist today, but again, little if any research exists on this point. Beyond the issues of cultural bias in assessment processes and in the piescription of treatments, cultural issues bear on a variety of other important mental health services treatment areas, including different cultural views of mental illness and treatment, use of culturally varied healing methods, variations in support networks, and larger cultural issues such as alienation. Research on treatment of mental disorders among members of ethnic minority groups has focused on several issues: bias and its effect on outcome in the psychotherapy of minor-

-

590

ity patients by nonminority clinicians (Sattler 1977; Abramowitz and Murray 1983; Sue 1988), culturally adapted treatment techniques for specific cultural or subcultural groups (Marcos et al. 1973; Comas-Diaz 1981; Munoz 1981; Szapocznik et al. 1981; Canino and Canino 1982; Acosta et al. 1983; Rogler et al. 1987; Sue 1988), different rates of utilization of mental health services by members of ethnic minorities (Keefe 1982; Vernon and Roberts 1982; Griffith 1985; Dworkin and Adams 1987), and different responses and dose requirements in psychopharmacologic treatment (Lin and Finder 1983). These studies have produced inconsistent and inconclusive findings, largely because of small, nonrandom samples and inconsistent methodologies, including diagnostic and symptom assessments and clinical course and outcome measures. Most of the studies have not used blind-outcome ratings, control groups, and longitudinal designs for followup, or they have had other methodological limitations. The ECA study cited earlier did indicate that Hispanic (Los Angeles Mexican-Americans) community residents with and without recent mental disorders used mental health services significantly less than did nonHispanic white residents. Black community residents were, on the other hand, similar to non-Hispanic whites in their rates of use of mental health services. The ECA study unfortunately did not determine the reasons for such mental health visits or the perceived outcomes or benefits received. Thus, even these data from large and (within the communities studied) randomly chosen samples are of limited value in understanding ethnic influences on mental health care utilization. Other subgroups of people with severe mental illnesses are prone to

SCHIZOPHRENIA BULLETIN

inaccuracies in diagnosis and treatment. It has become increasingly clear that comorbid problems such as substance abuse, mental retardation, and medical illnesses frequently go undiagnosed (Gift et al. 1978; Black et al. 1985; Lehman et al. 1989). There is thus reason to believe that these problems in assessment lead to inappropriate or incomplete treatment. Again, almost no health services research exists in this area. While there is a developing body of research on the homeless mentally ill, various pitfalls remain in this literature. There are problems with the definition of homelessness, sampling bias, lack of accurate prevalence data, difficulties of case identification, and limitations in projection of service needs from prevalence data (Breakey 1987). It is also apparent that many of the homeless mentally ill suffer from comorbid conditions, particularly substance use disorders. The literature still lacks any well-detailed and wellquantified documentation of service needs among the homeless mentally ill, and evaluation studies of services for this group are just beginning (Breakey 1987). Finally, a significant proportion of severely mentally ill people lives in various types of institutional and group settings. Patients living in such enclaves may have special treatment needs. For example, little is known about public health issues such as drug abuse and sexual behaviors that pose potentially serious risks to patients in these group settings. The needs arising from the social contexts of these new community-based institutional settings must be examined. Research Priorities. This section will be organized according to the five subsections in the preceding section on the current state of treatment knowledge.

Treatment settings. • In this age of brief hospitalizations, what is the most appropriate use of short-term hospitalization and what should occur during brief hospitalizations for people with severe mental illness? • What, if any, should be the role of longer term hospitalization? For whom? For what purpose? When should hospitals be used for long-term asylum? What should occur during long-term hospitalization? • Given a variety of treatment settings (day hospitals, partial hospitals, home-based care, clinics, residential programs, and mobile community facilities), how should alternatives to traditional hospital care be used most effectively and efficiently? What services should be delivered? In which settings? For which patients? At what cost? • What are the special treatment and public health needs of severely mentally ill people living in group enclaves in the community? Treatment modalities.

Somatic

therapies.

• Which people with which types of severe mental illness should receive various types of pharmacotherapy? • How can the newer types of pharmacotherapeutic agents be transferred most efficiently and safely into clinical practice? • What is the impact of various types of pharmacotherapy on patient outcome and on patients' needs for and ability to use psychosocial interventions? • What is the appropriate role for ECT in the treatment of severe mental disorders? • What is the relative effectiveness of standard, low, and noncontinuous dosages of conventional antipsychotic agents? How do these agents differen-

591

VOL. 18, NO.4, 1992

tially interact with psychosocial and rehabilitation interventions? Which outcomes are positively and negatively affected by these alternative strategies? • How are alternative somatic therapies (i.e., alternative neuroleptic dose regimens, ECT) used in various practice settings (private vs. public; inpatient vs. outpatient; specialty vs. general practice)? Psychotherapies.

• Are the newer forms of psychotherapy (i.e., techniques emphasizing supportive and cognitive-behavioral dimensions) effective for those with severe mental illnesses7 if so, what should be their role in treatment relative to pharmacotherapies, rehabilitation approaches, and so on? • How should the newer concepts about the psychology of schizophrenia be brought to bear on treatment (Strauss 1989), given the growing evidence of biological etiologies and emphasis on biological treatments? • What are the optimal characteristics of therapeutic milieus for severely mentally ill people, regardless of setting, and how can such milieus be promoted in different settings? • Is peer support treatment effective for people with severe mental illness? If so, what therapies are needed with psychological and social interventions? Combining treatment and rehabilitation modalities. • How do different medication regimens affect a person's ability to benefit from psychosocial treatment and rehabilitation? • How should psychosocial treatment and rehabilitation efforts be timed in the process of recovery from an episode of disabling mental illness?

• Should treatment and rehabilitation services be kept at separate sites to promote recovery? Does geographic separation or integration of these services make any difference? • How can psychosocial and rehabilitation services be linked most effectively to inpatient settings to promote successful transition to the community whenever possible, or to optimize functioning in the institution when necessary? • What service linkage models work best in promoting effective interaction of dinical and rehabilitation services? • What are the actual relationships between clinical outcomes (symptoms, psychological well-being) and rehabilitation outcomes (functional status, access to resources and opportunities)? Continuity and Integration of care. Much research is also needed to implement effective combinations of treatment services. Specifically, many issues and questions remain unanswered about continuity of care and case management. • Much work is needed in the measurement area. Operational measures of the various hypothesized dimensions of continuity of care should be developed. Measures that assess these dimensions from several perspectives (patient, provider, and family) may be important. • With this portfolio of measures, relationships between outcomes and different dimensions of continuity of care could be explored. Which dimensions of continuity affect which patient outcomes? How do these various dimensions relate to other types of outcomes, such as family burden and costs of care? Do the relationships between dimensions of continuity and outcomes vary according to patient characteristics?

• What types of services and service systems enhance which dimensions of continuity of care? • On the basis of the relationships noted in the two previous bulleted items, what alternative methods of delivering services to severely mentally ill people are most likely to optimize the critical dimensions of continuity of care and outcomes? For which patients? Hypothesized àlternalives should be compared experimentally to assess the impact on dimensions of continuity and, in turn, on cost-effectiveness. • What factors promote continuity of care at the provider level? At the patient and family level? At the system level? Are there negative effects at these different levels (e.g., staff burnout, patient perceptions of intrusiveness, or unforeseen system expenses)? • Methods are needed for measuring the types and quantities of case management provided and for describing comparison conditions. Even standard community aftercare can involve certain elements of case managelnent; these need to be well characterized. • Different types of case management need to be compared—for example, clinical case management versus the minimal and coordination models,. clinically based versus rehabilitationbased models, team versus individual case management models, and provider-based versus family-based models. Such studies will directly address the questions of the qualifications and tasks of case managers. • Which outcomes are affected by which models of case management? For which patients? • Are there appropriate phases of case management? That is, how do patients' needs for different types and Intensities of case management change over time? How should the responsibility of case management evolve? -

-

SCHIZOPHRENIA BULLETIN

592

• Are formal case managers necessary for good case management? Practice issues. The paucity of effectiveness studies on the treatment of severely mentally ill people has inhibited the development of studies on their care. A wide variety of practice issues needs to be addressed.

• Standards for quality of care for people with severe mental illnesses must be developed. These criteria should initially be based on findings from clinical efficacy and other studies, supplemented by consensus development, decision analysis, and related techniques. • Research should examine the values that underlie care for people with severe mental illnesses. Such studies may focus on the ethical implications of public fund allocations, of priorities set by treatment programs or individual providers, and of attitudes toward providing care for people with severe mental illness. Such research might focus especially on hard-to-treat or hard-to-reach populations, such as the homeless, violent offenders, or those afflicted with the neuropsychiatric complications of HIV infection. • What is the level of quality of care provided to people with severe mental illness in representative treatment settings, including general medical hospitals, State psychiatric hospitals, nursing homes, board-and-care homes, and public outpatient clinics? • What characteristics of patients, providers, and particular communities and health care delivery systems affect or are associated with variation in the quality of care for people with severe mental illnesses? In addition to these quality-of-care research issues, a variety of questions needs to be addressed about the care of severely mentally ill people in general medical settings.

• What types of health care encounters do individuals with severe mental illnesses have in the general medical sector? • What are the attitudes of general medical clinicians toward providing general health and mental health care to individuals with severe mental illnesses? What are the attitudes of these patients and their families toward receiving care in the general medical sector? • What are the treatment approaches used in various types of general medical settings in providing care to people with severe medical illnesses? • Do general medical providers assume a primary care role for people with severe mental illnesses? Should primary care be defined differently for this group? To what extent do those with severe mental illnesses receive coordinated and continuous care through general medical clinicians? • What is the role of the family's primary care provider in assisting relatives to engage a severely mentally ill family member in treatment? Can this role be enhanced? • How are services that are delivered by general medical clinicians integrated into services provided by the specialty mental health sector and integrated into social services? What are referral patterns? What is the level of access to mental health specialty care for severely mentally ill individuals who are treated in the general medical sector? • Which outcomes of care are achieved for severely mentally ill individuals who are treated in the general medical sector? • What are the characteristics of severely mentally ill persons who receive treatment in the various components of the general medical sector? How do the profiles of health needs -

differ among individuals in these different settings? • To what extent are new interventions implemented in general practice? • How can effective, new treatments be promoted in practice and monitored for adequacy of application? • How can standards of care for people with severe mental illnesses be established and monitored? Special populations. • How do the types and quality of treatment services provided to people with severe mental illnesses vary with patients' demographic, religious, ethnic, and cultural characteristics? • How do affected patients' and families' views of mental illness and treatment vary according to demographic, religious, ethnic, and cultural characteristics? • Which adaptations to standard treatment services are needed to improve care for severely mentally ill people from cultural and ethnic minorities? • Where do people with dual disorders (severe mental illnesses plus substance abuse, mental retardation, or other serious and persistent medical conditions) receive treatment and how adequate is this treatment? • Which treatments are most effective for people with dual disorders? In what settings should these treatments be provided? • What are the service needs of homeless mentally ill individuals? Which homeless mentally ill people need which services? • Which types of services are homeless mentally ill individuals receiving? Which homeless receive which services? • What are the most effective methods for delivering services to homeless mentally ill individuals, and how can these services be adapted to meet the -

-

VOL 18, NO. 4, 1992

needs of various subgroups in this population? Priority areas. Among the many research issues and questions identified above, the following deserve special priority: • Research on how to integrate and sequence clinical (somatic, psychological, and social) and rehabilitation modalities to optimize outcomes. • Development of measures to describe aspects of care for people with severe mental illnesses, especially measures of case management and continuity of care. • Research on the effectiveness of different models of case management. • Research on the proper role of hospitals in the care of severely mentally ill people in an age when community-based care is the valued norm. • Research on the relationship of dimensions of continuity of care to treatment interventions (e.g., case management) and outcomes. • Effectiveness research leading to the development of standards of care for severely mentally ill people and application of these standards to quality-of-care research. • Assessment of the impact of cultural and ethnic issues on the treatment of minority patients and development of ways to improve treatment for these groups. • Research on methods to speed the transfer of new treatment knowledge and treatment techniques for severely mentally ill people into general practice. • Research on methods to recruit and retain well-trained and motivated personnel to work with severely mentally ill individuals. • Research on the role of families and consumers in developing and delivering effective care. • Research on more effective methods for treating people with other

593

medical, psychiatric, and social conditions, in addition to a severe mental illness, especially substance abuse, mental retardation, homelessness, and serious physical illnesses, including HIV infection. RehabHltatlon Research The aim of rehabilitation is to assist the individual in learning or relearning physical, emotional, and intellectual skills necessary to function in his or her environment with the least amount of assistance from the helping professions (Dion and Anthony 1987). The foundations of psychiatric rehabilitation are skills training, drug therapy, and community support (Anthony and Parkas 1982). Skills training and community support, in conjunction with drug therapy when appropriate, form the basis of the definitions of rehabilitation used by the Panel in its consideration of rehabilitation services research. Skills training approaches are used to teach clients success in a particular environment. Drug therapy is used to reduce symptomatology and prevent relapse. Community support is provided to enable clients to function in specific community settings. Attempts to study these three dimensions are hindered because most rehabilitative treatments consist of integrated services aimed at remediating a series of deficits. The field of rehabilitation needs to develop assessment models that take account of the resulting varied client changes (Cook et al., in press). In addition, there is concern with understanding the effects of both outcome and process variables in psychiatric rehabilitation research. The process dimension focuses on measuring changes in factors such as skill acquisition, training setting functioning, and client understanding. Here,

the focus is on the specifics of how the intervention was carned out and whether new skills were acquired. Often, studies of rehabilitation focus solely on outcomes, neglecting this important dimension of research; however, outcome research is largely meaningless without an accurate and understandable process description. The outcome dimension centers on changes in the client's application of acquired skills beyond the training environment, as well as changes in benefits to the client. Both types of outcome involve the direct measurement of change in material circumstances such as employment, residentiai independence, income level, and perceived quality of life. Currently, a major impediment to the development of knowledge about rehabilitation services is the absence of outcome models that would allow the exploration of changes in clients that are both anticipated and unexpected and positive and negative. In particular, models are needed that can account for treatments that influence outcomes in different ways, such as when skills training enhances work quality while increasing anxiety about task performance. Using both process and outcome variables continues to be a major challenge to the field of rehabilitation research. This challenge is complicated by the use, in many studies, of idiosyncratic, untested instrumentation that precludes replication. Another difficulty centers on measuring goals that may be interrelated. It is difficult, for example, to sort out the interlocking effects of a psychosocial rehabilitation model (Dincin 1975) that provides vocational, educational, social, residential, and medication management services to clients living in the community. -

594

In addition to interactive effects of different rehabilitative interventions, there are the difficulties of studying models that involve unique supportive subcultures, such as the individual Fairweather Lodges (Fairweather et al. 1969). In these approaches clients live, work, and recreate together in the community, involving milieu factors that are highly complex and perhaps unique to each group. Studying these types of programs involves complex methodological and design issues. Social Skills Training. The effectiveness of social skills training during the past 15 years has been documented in more than 50 research studies (reviews indude Hersen and Bellack 1976; Wallace et al. 1980; Brady 1984; Morrison and Bellack 1984; Liberman et al. 1986a). These studies provide evidence that social skills training is a feasible treatment strategy, especially for concrete behavioral skills such as eye contact, voice volume, and appropriate gestures (Wallace et al. 1980), which are the focus of most training programs and research studies. There are four focal points in social skills training models (Uberman et al. 1986a; Morrison and Bellack 1987): (1) improving basic psychobiological and cognitive functions by, for example, training to lengthen attention span or to improve ability to engage in higher order abstractions; (2) training in the receiving, processing, and sending skills of communication, which involves coaching, reinforcement, modeling, and homework; (3) reprogramming the individual's natural environment so that acquired skills will meet with support and favorable feedback—induded are working with families and analyzing the client's unique social environmental needs; and (4) improving the understanding of social perceptions; targets are aspects such as understanding and ma-

SCHIZOPHRENIA BULLETIN

nipulating social norms, and negotiating personal and interpersonal judgments, a particular problem for people with schizophrenia (Platt and Spivack 1972, 1974; Platt et al. 1975; Platt and Siegel 1976). Several literature reviews of studies in these areas have reached condusions about what is known. First, the psychiatrically disabled can be trained in behaviors that will improve their social skills in specific situations (Liberman et al. 1986a). Second, persons so trained show moderate to substantial generalization of trained behaviors to novel situations, depending on the complexity of the behaviors (Goldsmith and McFaII 1975; Kelly et al. 1980; Liberman et al. 1984). So, for example, changes in voice volume are more likely to be maintained than behavior changes. Finally, comprehensive and intensive social skills training can reduce clinical symptoms and relapse rates (Liberman et al. 1984; Wallace and Liberman 1985). While most researchers agree that social skills training is effective in reducing targeted social deficits, some reviewers question whether the generalization of social skills to the natural environment has been established (Dion and Anthony 1987; Morrison and Bellack 1987). There is some evidence that generalization follows intensive, long-term training in which a number of partners are used to maximize generalization and assignments are completed outside the dassroom (Liberman et al. 1986b). In other cases, systematic homework assignments, backup reminders, and praise from staff for completion have been shown to promote generalization (Martinez-Diaz et al. 1983). Another important issue is durability of social skills training. There is a dearth of information about which services contribute to the maintenance

-

-

-

of acquired skills in the natural environment. This lack is probably due to the difficulty and expense of conducting this type of long-term followup. Wallace and coworkers (1980) report that some studies have assessed the durability of outcomes and found that treatment gains were maintained for up to 6 months (Foy et al. 1975; Hersen et al. 1975; Bellack et al. 1976; Hersen and Bellack 1976); however, one report showed that after only 12 weeks, appropriate behavior rapidly deteriorated to prebaseline levels (Matson and Stephens 1978). Some have suggested that the involvement of significant others in the training program might promote maintenance as well as increase generalization. There also is evidence that social skills training reduces symptoms and increases program participation. Bellack and coworkers (1984) report that patients who were in a day program that included social skills training had more durable symptom reductions in a 6-month period than patients who were in the day program but did not receive social skills training. It even has been shown that social skills training alleviates depression for unmedicated depressed outpatients (McKnight et al. 1984), has diniâal effects much like those of antidepressant medication, and is associated with a lower dropout rate from treatment in general (Bellack et al. 1983). What needs to be learned. Researchers should conduct more studies that determine which services are most likely to promote generalization of acquired social skills to novel situations. Studies of the durability of newly learned skills in the natural environment are also important. This research must examine very long-term maintenance issues, including the impact of refresher services. Researchers also should conduct assessments of any concurrent

VOL. 18, NO.4, 1992

changes in behavior, especially if the individual's rehabilitation services in this other area include social skills training. These studies of concurrent behavior changes should clarify whether other aspects of a person's functioning affect his or her ability to acquire and maintain social skills. The concurrent changes in work and ability to live independently should be examined when a person is receiving social skills training. One uninvestigated aspect of social skills training is the nature of relationships with trainers and its effects on durability and generalizability. Attention should also be given to the role of social skills training in teaching people with severe mental illnesses how to negotiate safe sex and use contraception to prevent pregnancy. Research is needed on the effectiveness of social skills training in promoting safer sexuality, generalization to novel situations, and maintenance of learned interpersonal skills. In general, research on effectiveness must become more sophisticated. Evaluation studies need to determine whether rigid or flexible structuring of a social skills training curriculum is better. Another issue is the extent to which staff adhere strictly to a training curriculum or engage in innovation. Finally, effectiveness studies need to address the possibility that different types of social skills training may be needed by individuals at various stages of their illness and rehabilitation. Vocational Rehabilitation. Various models have been used to secure and maintain paid employment for persons with mental illnesses. Controlled, randomized evaluation studies exist but are scattered and contain design problems inherent in vocational evaluation research. There is enough evidence to conclude that these models

595

work well for some, but not all, people. Models differ in their reliance on community job placements as a mechanisin of rehabilitation. Some, such as transitional and supported employment, stress arrangements with real employers and the presence of nonhandicapped coworkers (Roussel and Cook 1987). Others, such as Fairweather Lodges and Soteria houses, involve a group of mentally ill persons living and working together in a consumer-owned business (Fairweather et al. 1969). Still others, such as the Job Club methcid, concentrate on teaching vocational skills such as job finding or other specific competencies (Azrin and Philip 1979). As several reviewers have remarked, interest in hospital-based vocational programs has waned in the past two decades because inpatient admissions are shorter. A number of studies have concluded that there is little relationship between posthospital employment and successful participation in hospital-based vocational services, and that such programs may even increase institutional dependency (Bond and Boyer 1988). This may mean that the most successful vocational programs are those that include integration into the community (Douzinas and Carpenter 1981). The Job Club method, a structured approach to securing employment, was developed for low-income populations (Azrin and Besalel 1980) and later adapted for use with psychiatric populations (Azrin and Philip 1979; Kramer and Beidel 1982; Jacobs et a]. 1984). In this model, intensive job search and interview skills preparation is taught for half a day, 5 days a week. While initial research showed some promise for this method, subsequent studies indicated that patients with a psychiatric disability took longer to find jobs with the training

than without (Jacobs et al. 1984) and they were unlikely to retain the jobs they did find (Kramer and Beidel 1982). This finding may indicate the need for a method that offers more long-term supports to individuals once they become employed. Experimentally controlled studies of sheltered workshop performance of people with mental illnesses have had minimal success. Two controlled studies indicated no significant differences between treatment and control groups with this method (Weinberg and Lustig 1968; Griffiths 1974). A number of studies have indicated that people with mental illnesses fare less favorably in sheltered workshops than do those with other disabilities. Clients with mental illnesses earn lower wages, present more behavior problems, and show poorer workshop attendance than workers with other disabilities (Olshansky 1973; Olshansky and Beach 1974, 1975; Whitehead 1979; Ciardiello 1981). A small number of studies have suggested that work in a real employment setting is more effective in development of selfesteem, competence, and vocational skills (Rapp 1979; Conte 1983). A meta-analysis of experimentally controlled studies of five transitional employment programs located within. psychosocial rehabilitation programs found a significant 2-score, indicating the effectiveness of this treatment model in securing community employment (Bond 1986). Other research has indicated the effectiveness of this method in obtaining community employment and holding it for relatively brief periods (from 6 to 42 months) (Fountain House 1985; Rutman and Armstrong 1985; Cook et a]. 1989). The Federal initiative for supported employment has created possibilities for research on this method, which invo'ves finding clients nonsegregated, nonsheltered employment at

-

596

minimum wage or above and then providing services to help them maintain employment. Prior studies of time-limited transitional programs, which have had modest success in psychiatric populations, indicate that supported employment, or the provision of ongoing services, is needed because mentally ill clients require ongoing agency involvement to maintain community jobs (Cook and Razzano, in press). The Fairweather Lodge model promotes the philosophy that clients need living skills, a job, and support from family or friends to become contributing members of society (Stroul 1986). By addressing the areas of living, working, and socializing, this model allows clients to reside together in the community and start a consumerowned business for which they all work. This model has had very encouraging results in community tenure and employment. In comparison with other models, Lodge members have higher employment rates and spend more time in the community (Fairweather et al. 1969; Fairweather 1980). This model has not been replicated outside a Lodge setting, and it is not clear to what extent this model can be implemented in enough different settings (i.e., rural and urban) to make it Some researchers have focused on trends apparent in survey studies of ability to work and have attempted to draw conclusions from similar findings and conclusions (Anthony and Jansen 1984). Some features such as diagnosis, ability to function in other life areas, and independent living are consistently found to be unrelated to employment and need further investigation. Other researchers have focused on specific features of the employment process, such as appropriate vocational assessment techniques for

SCHIZOPHRENIA BULLETIN

persons with psychiatric disabilities (Bond and Friedmeyer 1987) and the reasons why mentally ill workers leave their jobs (Cook, in press). Attempts to understand employment as a process rather than simply an outcome are necessary to the development of effective vocational rehabilitation service delivery models. What needs to be learned. There is a need for more research into the relative effectiveness of each of the vocational models discussed here, particularly studies addressing which types of services are most beneficial for which types of clients. Given the likelihood of client self-selection for different models, randomized studies should be augmented by approaches using quasi-experimental designs. Vocational rehabilitation services that are embedded in models providing many other services should also be investigated. For example, it is important to know how the other goals of psychosocial rehabilitation programs and Fairweather Lodges affect delivery of services aimed at vocational outcomes. In general, evaluation studies indicate that multidimensional treatment models offer advantages over one- or two-service models. Next, it needs to be established which combinations of services delivered to particular clients are most effective; a focus on the case mix is necessary. The current emphasis on supported and transitional employment leads to questions about which types of clients benefit most from ongoing versus time-limited vocational services. Information about methods that encourage growth without unnecessary or harmful transitions focuses on aspects of employment that are therapeutic versus those that are not. The problem of providing ongoing support without encouraging unnecessary dependence is another topic for investigation. Yet another focus is on the

relative cost-effectiveness of the two vocational methods. Very little is known about the effects of race, sex, class, and age on employment among the mentally ill. Since employment of the handicapped also depends on employment and hiring trends and other labor market features, research needs to focus on the effects of these features on different types of vocational programs. Since State vocational rehabilitation systems have the responsibility and the funding to provide vocational services to people with psychiatric disabilities, there is a need for research addressing how well State systems serve this population. Studies are needed of who is referred and accepted for vocational services, as well as their outcomes after rehabilitation. Methods of educating State rehabilitation personnel to understand the unique needs of those with severe mental illnesses and the effects of these methods on client outcomes deserve greater attention. The Americans with Disabilities Act (ADA), passed in 1990, provides a context for research in several areas, including employment and schooling. Under this legislation, employers are mandated to provide "reasonableaccommodation" to an individual's disability in the form of more flexible. leave practices, reorganized job tasks, and adjusted work schedules. As of July 26, 1992, this legislation applies to all employers with 25 or more workers. One ambiguous provision in this legislation specifies that employers must make only those accommodations that do not cause undue harm to their organizations. The way in which this harm clause is interpreted by the courts as cases are filed and adjudicated will be the focus of studies of community integration. Other potential research topics concern public awareness of the new law -

-

597

VOL. 18, NO.4, 1992

and its relevance for people with psychiatric disabilities. As supported employrnent and natural supports are used increasingly in aiding efforts to work, the concept of reasonable accommodation may be used to convince employers to allow innovations (Cook et al., in press). Under this legislation, such accommodations indude use of coworker job coaches, peer coaching, and flextime. Employer and nonhandicapped coworker attitudes toward these accommodations should be investigated as part of this line of inquiry. Another potential issue for investigation is the nature and resolution of complaints filed with the agency charged with responsibility for enforcing the ADA—the Equal Employment Opportunity Commission (EEOC). The types of inequities faced in the labor force by those with mental illnesses can be studied, as well as the outcomes achieved. While a number of disabilities will be covered in addition to mental illness, the highly stigmatized nature of psychiatric disability may influence whether and how this new law affects access to employment and schooling (Mancuso 1991). Cross-disability comparative studies on EEOC enforcement and variable definitions of accommodation will shed light on the unique nature of stigma for persons with psychiatric disabilities. Residential and Independent Living Services. Despite catchrnent area studies and extrapolated estimates from nonrepresentative samples, there is little generalizable information about where people with psychiatric disabilities live. In the 1950s and early 1960s, deinstitutionalization resulted in a proliferation of community-based residential services. The trend was toward a transitional program structure organized developmentally or

functionally. The continuum of services includes quarterway houses, halfway houses, family foster care, crisis residences, and Fairweather Lodges (Carling et al. 1988). This transitional model has come under recent criticism because its overstructured and timelimited nature has encouraged the establishment of a diverse range of residential services regardless of need. Three main questions have been addressed in the evaluation of residential programs (Carling et al. 1988). The first and most frequently asked question is whether community-based programs are more successful than inpatient treatment in helping individuals to live independently. Although several hundred studies have been conducted on traditional residential programs (Gibson et al. 1972), community support systems (Weinman et al. 1970; Lamb and Goertzel 1972; Stein et al. 1975), crisis resolution programs (Pasamanick et al. 1967; Langsley et al. 1971; Rutman 1971; Brook 1973; Polak and Kirby 1976), and day treatment or outpatient services, only a handful of methodologically sound studies address this question (Carpenter 1978; Test and Stein 1978; Braun et al. 1981; Dellario and Anthony 1981; Kiesler 1982). The main conclusion of these studies is that community-based treatment is virtually always as effective as or more effective than hospital-based treatment in helping people with mental illness reenter the community. The second question addressed in the evaluation of residential services is how a particular type of residential service works in different rehabilitation programs. A review of research on halfway houses (Comets et al. 1979) concluded that evidence on the effectiveness of this type of residence in reducing hospitalizations, improving economic self-sufficiency, reduc-

ing symptoms, and promoting community integration was ambiguous at best. These findings lead to the conclusion that while community-based residential programs may be better than inpatient treatment, they do not necessarily lead to successful longterm community adjustment (Anthony and Blanch 1988; Carling et al. 1988). The third question commonly addressed in evaluation studies of residential programs is what characteristics of residential programs (e.g., facility size, presence of roommates, number of activities offered) are related to client outcomes. The main conclusion of this type of evaluation is that characteristics of the environment are more important than characteristics of the individual in predicting community integration (Segal and Avirarn 1978; Hull and Thompson 1981; Coumos 1987; Anthony and Blanch 1988). Environments in which the individual can participate in many different activities and have opportunities for social interaction are most likely to lead to positive outcomes, while highly structured or overstimulating environments can lead to relapse (Cournos 1987). Although it is estimated that more than 40 percent of people with severe mental illnesses live in private homes or apartments, the possibility of using normal housing situations rather than mental health programs has received relatively little attention (Tessler and Goldman 1982; Blanch et al. 1988), perhaps because mental health professionals and consumers have very different ideas about what types of living situations are possible and desirable. Several studies have indicated that professionals favor transitional, highly staffed residential programs, while clients prefer to live on their own with flexible supports (Daniels -

SCHIZOPHREN$A BULLETIN

598

and Carling 1986; Solomon et al. 1988b).

Recently, a new model of supportwe housing services (National Institute of Mental Health 1987a, 1987b; Carling et al. 1988) has been used in an approach that focuses on consumer goals and preferences, an individualized and flexible rehabilitation process, and normal housing. One impediment to the use of normal housing is the shortage of low-cost housing and its effects on those with severe mental illness (Anthony and Blanch 1988). People with severe mental illnesses are often unable to compete for housing or use rent subsidy or voucher programs (Boyer 1987). Frequent changes of residence and homelessness often result from this shortage of lowincome housing (Appleby and Desai 1987).

What needs to be learned. Research in this area should include large-scale surveys of current housing arrangements of people with severe and persistent mental illnesses. There is not yet an accurate understanding of how many people live with relatives, how many live in long-term custodial care (State hospitals, nursing homes, board-and-care homes), how many are incarcerated, how many are homeless, how many live in residential programs, and how many live independently. Such information would contribute substantially to policy formation and serve as a foundation for future studies. Once identified, the residential patterns of these samples need to be followed over time along with sample members' preferences about where they want to live as their needs and illness situations change. In addition, research needs to begin measuring the degree of consumer choice in housing transitions. Too often, studies of residential instability have viewed flux as if it were an outcome of client levels

of functioning without understanding how it often results from a philosophy of transitional housing services. Studies addressing how consumer preferences are matched to what is available will be instructive for public policy development. Also, it will be important to determine the reasons why consumers' preferences are not matched to their housing arrangements. Identifying systemic, economic, cultural, and gender-based factors in this gap between preferred and obtained housing will provide a starting point for understanding housing as a process that changes over time along with a person's illness, functioning level, economic situation, and local housing options. To aid in this research, valid and reliable instrumentation is required, especially in eliciting preferences from consumers and dassifying living situations accurately. Outcome measures that apply to a wide range of people, not just those in residential programs, need to be developed and used. Research designs that gather sample sizes large enough to permit multivariate statistical modeling also need to be developed. Samples should include seldom-studied groups such as those who receive no mental health services and those who are incarcerated. Extending the concept of natural supports to the residential area, there is a need for studies of informal support used by persons with psychiatric disabilities who live in the community. A related area, community resistance to integration, should be studied in acknowledgment of the stigma issues raised repeatedly in this report. The meaning of community integration to all parties involved—consumers, professionals, neighbors, community organizers, and local politicians—should be studied and compared with the degree of integration achieved. -

-

-

Evaluations of the supported housing model in areas with little lowincome housing are also needed to determine whether the model is replicable in large cities as well as areas with housing shortages. In addition, research detailing the effects of our current low-cost housing crisis on people with mental illnesses would allow better service development. An emphasis on the programmatic features that might affect outcome (e.g., house atmosphere, programmatic goals, staff-client relationships) is needed in evaluation studies. Studies exploring these features, their interrelationships, and their relationships to successful outcomes, defined by clients as well as professionals, should constitute the next wave of residential research. The importance of environmental features over client characteristics in residential research suggests that investigators should pay more attention to where and how participants in residential programs spend their time, rather than focusing exdusively on specific programmatic interventions. Such a focus on the multidimensional nature of residential programs requires separating the evaluation of housing and treatment goals. Focusing on measures of recovery or improvement assumes that residential services have (or should have) a therapeutic impact. This may or may not be the case, and thus far it remains an empirical question. If some people show better outcomes when living in more structured residential settings or in transitional housing programs, research should address the service needs and outcomes of those who are unable to benefit from a supported housing approach. Along these lines, research attention should be paid to the housing needs of people who are unable to live independently in the community

VOL 18, NO.4, 1992

or who require large amounts of support and structure to do so. Finally, we need better understanding of the residential service needs of people who are not receiving housing services, particularly those whose families serve as de facto providers of such services. The residential service needs and current living situations of groups of persons typically excluded from community-based housing programs, such as forensic and AIDS patients with severe mental illnesses, should be studied. For-profit housing models need to be studied and compared with their nonprofit counterparts, including board-and-care, halfway, single room occupancy (SRO), supported, and transitional models. Minority and Cross-Cultural Issues in Rehabilitation Literature. Few studies have compared minority and nonminority populations of psychiatrically disabled persons. There is, however, some research suggesting that African-American rehabilitation clients tend to be younger, are more likely to live with family, have less education, and are less likely to seek work or enter vocationai training (Atkins and Wright 1980; O'Sullivan et al. 1989). Some researchers have found that those African-American clients who are served receive less evaluation, less vocational training, less group therapy, and little residential support, the latter probably because of their tendency to live with family (Atkins and Wright 1980; Solomon 1988a). Some have claimed that these findings indicate underservice of black clients by State vocational rehabilitation agenaes. Other researchers point out, however, that these differences disappear when variables such as education, mental retardation, public aid status, gender, or weekly earnings are controlled (Bolton and Cooper 1980).

599

The debate over this issue is unresolved. Vocational services to Mexican-Americans are limited because of their lower levels of formal education, the unavailability of bilingual rehabilitation counseling, and the need to coordinate rehabilitation services with other medical and mental health services (Sparks 1991). Other researchers point out the unique strengths that minority persons bring to the rehabilitative relationship. African-American consumers follow up better with recommended treatment regimens in some studies of psychiatric services (Solomon 1988a) and hold favorable attitudes toward mental health centers. This point is especially true of clients who are more politically and racially conscious (Gary 1987). Recent research updating earlier studies of community mental health services to minorities has focused on ethnic-specific mental health centers. These centers target a particular ethnic or racial population, using staff from their respective communities (O'Sullivan et al. 1989). Research on these centers indicates the importance of several variables worth exploring: cultural perceptions of the psychotic disturbance, cultural styles of helpseeking behavior, cultural appropriateness of services, ethnicity of the mental health provider, and mental health specialty of the provider. Other studies confirm a pattern of use in which African-Americans and Native Americans are overrepresented compared with their numbers in the population (consistent with earlier studies), whereas Mexican-Americans and Asian-Americans underuse mental health services. Studies that have investigated variables that facilitate the cultural match of agency to consumers agree that this factor alone cannot solve rehabilitation difficulties (Danek and Lawrence 1985). Inservice

and preservice training to increase rehabilitation staff awareness of ethnic issues is most effective in dealing with this issue (Bolton and Cooper 1980).

Some issues have different ethnic perspectives but also some common elements: acculturation; individuation versus interdependence; beliefs about causation and definition of illness; expectations of consumers, their farmlies, and their communities; and use of bilingual and bicultural staff. Acculturation may be defined as a process of changes in language, attitude, beliefs, values, and customs that can be ongoing or confined to the past and that can occur abruptly or developmentally (Cuellar and Arnold 1988). External community acculturation, such as the proportion of Hispanics to Anglos in the community, may influence the use of mental health services (Wells et al. 1987). Appropriate measures of acculturation would aid in the interpretation of standard assessments in this research and improve understanding of the role of acculturation in rehabilitation. Some cross-cultural research has targeted individuation versus interdependence, a dichotomy that is thought to permeate the expectations of mental health and rehabilitation professionals as well as consumers (Lefley 1985). Cross-cultural studies indicate that consumers may view interdependence as beneficial and esteem enhancing (Lin and Kleinman 1988; Naranjo 1988). It also allows the family and the community to have an active role in the rehabilitation process, which is consistent with the role of families in many cultures worldwide. Large, supportive networks mitigate family burdens by providing resources for caregiving and economic sharing, and they buffer the effects of disruptive behavior (Fujiwara and Christian 1988). It is in this respect that Western

SCHIZOPHRENIA BULLETIN

600

values, ideals, and treatment may adversely affect client progress. What needs to be learned. Some research indicates that nonindustrialized countries have better rehabilitalion outcomes (Lefley 1987b). Further exploration is needed in this area, preferably longitudinal studies and a search for prognostic indicators. More research should also be conducted on ininority-nonminority comparisons and on State vocational services to these groups. Such studies will help identify any differences in rehabilitation outcomes, as well as population features beyond minority status that account for these differences. In addition, within-group comparisons of minority populations are needed to reveal population diversity related to other factors, such as education and family income. More research also is needed on ethnic- and cultural-specific mental health centers and their effects on service use. Other important topics are the questions of client-clinician ethnic and cultural matching and training for clinicians in multicultural competence and sensitivity. More studies are needed to identify which services respond most effectively to the outsider status of minorities with severe mental illnesses. One process studied in the race and ethnic relations field but seldom employed in mental health rehabilitation is that of acculturation. More studies are needed that connect the degree of an individual's acculturation across a variety of dimensions with his or her use of rehabilitation services. Similarly, interdependence (vs. individuation)—a process typically maligned in Western cultural values but important in minority culturesshould be explored. One example is the therapeutic use of interdependence in programs in which minority family members take an active part in a

client's rehabilitation and spread the burden over a larger network. Linking Clinical and Rehabilitation Services. The literature investigating combinations of rehabilitation and therapeutic services is vmy sparse but suggestive. In one study, drug treatment was randomly assigned along with structured occupational therapy and intensive group therapy (Linn et al. 1979) and in another with social learning intervention and milieu therapy (Paul and Lentz 1977). In matched control group studies, drug treatment was investigated in combination with a skills development intervention (Vitab 1979) and a community-based psychosocial program (Mosher and Menn 1978). Although this research showed definite advantages for clients who received both rehabilitation and drug treatment in social functioning, employment, and independent living, the combination resulted in little difference in symptom levels and hospital readmission rates. While drug treatment may reduce symptomatobogy, it does not appear to have a positive impact on a client's living skills unless it is combined with rehabilitation interventions (Engelhardt and Rosen 1976). Since drug treatment alone does not foster the growth of skills, it may be that drug treatment negatively affects the development of certain skills (Dion and Anthony 1987). Future studies should be directed at the possible negative, as well as positive, interaction of drug treatment and skills training. Rehabilitation and the Family. Studies of the needs of family members indicate their desire for crisis services and assistance in dealing with the long-term problems of severe mental illness in a family member (Cook and Pickett 1987). In addition, several investigators have reported that family

members would like to learn basic case management techniques in partnership with rehabilitation professionals (Spaniol et al. 1987). Yet families are often ignored or neglected by professionals. Several surveys of parents belonging to advocacy groups suggest that families feel hostility toward mental health professionals because families are excluded from the process of making decisions about the treatment of their relatives (Yess 1981; Hatfield et al. 1982). This hostility may be intensified when professionals deliver rehabilitation interventions, because counselors often suggest changes that are not always viewed by families as being in their best interests or the client's (Cook 1988). Suggestions that the client live apart from the family or risk losing benefits by seeking full-time employment are two examples of conflicts between families and rehabilitation professionals. As already discussed, cross-cultural issues may affect the family's involvement in the rehabilitation process (Lefley 1985). Studies of family satisfaction with the services they and their ill relatives receive often fail to differentiate between psychotherapeutic and rehabilitation interventions. In one study (Spaniol et al. 1987), however, the only client service with which families were moderately satisfied was drug treatment, while relatives rated practical advice, illness knowledge, and treatment coordination as the most important services. The exdusion of families from rehabilitation planning is unfortunate, given their long-term and intimate knowledge of the ill individual (Cohler et al. 1991). As some have noted, mental health professionals come and go in the lives of clients, but families remain the one consistent element. Families can sometimes identify areas of interest that predate the ill-

-

VOL. 18, NO.4, 1992

ness as well as provide talents and strengths that can be used in rehabilitation efforts. There also is a need for better usderstanding of how families can speed achievement of rehabilitation goals by providing social support and reinforcement of acquired skills. Given the importance of social environment in generalization and durability of learned skills (Liberman et al. 1986b), collaboration with the families allows rehabilitation professionals to manipulate the clients' environments to foster acquired skills. What needs to be learned. There is a need to study family expertise regarding thementally ill and to deter.. mine how this knowledge and its accompanying care and concern can be employed in rehabilitation interventions. Studies of the use of familyprofessional collaboration, in the sense of shared responsibility or partnership, also are needed to determine effective ways for the two groups to interrelate. Families display different degrees and kinds of involvement with their ill members; that involvement should become an important variable in rehabilitation effectiveness research. At a broader level, research attention should be directed to families' social change efforts and resulting modifications in the perception of mental illness, as well as in policy and some funding structures. NAM!, for example, has lobbied for services, induding rehabilitation, for family members (Cook and Cohler 1986). Families have tended to lobby more strongly for residential services than for other rehabilitation goals. Parents often are involved in providing residential services, as evidenced by the number of family-run transitionai and permanent residential programs. If attention turns to providing vocational and educational programs, It

601

remains to be seen whether families will be able to mold service delivery and outcomes in these areas as well. Integrating the Consumer Perspective and Self-Help Movement. Primary mental health consumers have a voice distinct from the needs of family and clinicians, which must be recognized in any efforts at rehabilitation (Chamberlin et al. 1989). The need for peer support has been identified by consumers and increasingly receives recognition as a necessary component of adjustment (Chamberlin 1979; Leete 1988). The need for such support is confirmed by research showing that many consumers who are deinstitutionalized spend most of their time alone (Solomon et al. 1980). In addition, it is suggested that relationships formed at the hospital serve as a cog in the social network of the mental health service consumer, and that rehospitalization often may be due to a need for companionship (Solomon et al., inHaths and Bergman 1985) Also, researchers have acknowledged that disconnection from the social support network plays a key role in becoming homeless and prolongs homelessness (Segal et al. 1977). Consumer service deliverers may reeducate other staff and ancillary service providers, destigmatizing people with mental illnesses (Solomon et al. 1988). Consumers may improve their own self-esteem and job satisfaction, while the self-worth of the clients they serve may also increase. Consumer staff may become more empowered to create more reciprocal social networks among their clients (Hammer 1981). This approach also creates an opportunity for consumer service deliverers to experiment with mentorship and role modeling with their clients. Consumers being served are likely to identify more closely with a crisis counselor who has been there

-

and to benefit from the aid of a person who has personally experienced the system, understands the subjective nature of psychiatric disabilities, and is empathic. It has been suggested that consumers can provide more meaningful and relevant expert consultationto community care providers than professional staff can (Levine and Kennedy 1985). Finally, unexpected and negative experiences can also occur in consumer service delivery programs, such as job-related stress and decompensation (Van Tosh 1989). Programming to meet peer support needs is emerging in many forms, from consumer drop-in centers to consumer-run businesses (Stroul 1988). Federally funded programs to develop consumer job coaching are now under way (Cook et al., in press). Although many consumer service delivery programs reduce the length of hospitalization, it is often difficult to isolate the effects of peer support on client outcome, and evaluations of peer support programs are scarce (Anthony and Blanch 1988). Previous evaluations of interventions using consumers as staff indude studies of consumer case management services. Most notably, the Colorado Consumer Case Management Project (Sherman and Porter 1991) trained and placed 17 consumers as part of a citywide case management program in Denver, with 60 percent of the employed consumers continuing as case managers for more than 12 months. Other case management models using consumer service providers include programs for the homeless in Cuyahoga County, Ohio (Solomon 1988b) and Philadelphia, Pennsylvania (Kravita 1987). Mobile outreach teams staffed by consumers serving the homeless have been deployed In New York City and, although anecdotal information is available (Putnam et al.

SCHIZOPHRENIA BULLETIN

602

1986), rigorous outcome data are lack-

rng for this and many other programs. What needs to be learned. More research is needed on the consumer perspective and its incorporation into all aspects of rehabilitation, from philosophy and structure to nature of services and service providers. Along these lines, more attention should be paid to the role of the client in setting rehabilitation goals. Well-designed empirical evaluations are needed that recognize the multiple perspectives on process and the multidimensional nature of rehabilitation outcomes among consumers. Creative service delivery programs must be matched to rigorous, sensitive research designs to gauge the effects of consumer-focused and consumerdelivered services. Attention should be paid to the possible differences between consumers' perspectives and the services provided to them in various rehabilitation areas. For example, consumer service delivery may vary by setting in vocational and residential rehabilitation, because dealing with mental health service providers as a consumer case manager may require different skills than dealing with community employers as a consumer job coach. Attention should be paid to the unique features, if any, of approaches integrating consumers into different goal areas. Place of Habilitatlon Services in Rehabilitation Research. Given the primary rehabilitation emphasis on acquired skills and ability to perform in the community, it is important to remember that some clients show only modest improvement (McGlashan 1984). In addition, some severely mentally ill people never learn certain skills before the onset of their illness; thus, their treatment becomes one of habilitation rather than rehabilitation.

The difference between habilitation and rehabilitation has not been fully explored in severe mental illnesses. It becomes relevant, however, when researchers consider interventions and the most effective services for the most severely and chronically impaired. We need better understanding of how to serve individuals with unremitting symptomatology or dual diagnoses that impair habilitative interventions. Attention should be paid to the possible effects of age and gender on the need for habilitation versus rehabilitation. For example, women may be learning some job skills for the first time rather than relearning previously acquired employment habits and values. Similarly, men may be learning homemaking and self-care living skills for the first time. In addition, differences in education may affect the habilitation process; at least one survey found that rehabilitation clientele read and do math far below their ageappropriate levels (Cook et al. 1987). Habilitation approaches recently adopted for psychiatric populations include programs for remedial education and postsecondary training (Unger et al. 1987; Cook et al., in press). Along with gender, age is a crucial feature in the habilitation-rehabilitation process that is poorly understood and underresearched, despite the aging of the first cohorts of deinstitutionalized populations (Lefley 1987a). As families (especially parents) age and die, increasingly larger cohorts of middle-aged and older people with severe mental illnesses are requiring rehabilitation services, including residential and social skills training (Colder et al. 1991). What needs to be learned. Research must address the question of habilitation for people with severe and persistent mental illnesses, including

further development of the concept of habilitation and its applicability for the psychiatrically disabled. Studies identifying the need for habilitation rather than rehabilitation approaches must address the roles of factors such as gender, education, and age in determining who needs which emphasis. Finally, more research is needed in the area of rehabilitation and habilitation for older people with mental illnesses, and a better understanding of the needs and roles of older family members must be gained. Knowledge is needed about how services can be tailored to meet the requirements of older people with severe mental illnesses to enable investigators, program planners, and policyinakers to help this growing population. Outcome Research: The Effects of Caring One of the most critical elements of clinical services research is the study of the impact of the services provided—specific treatments, broad social interventions, or both. Such studies are not easily accomplished. To begin with, severe mental illnesses are manifested in a variety of ways, not only in profound deteriorations in intellectual and emotional functioning, but also in marred social and vocational adaptation, the destruction of well-being and personal potential, and massive impediments to a normal existence in the community. It is difficult to assess such factors, both before and after the provision of service interventions. Moreover, even when useful measurement techniques are available, researchers must find study designs that permit definitive findings and conclusions regarding outcome. Despite the difficulties, efforts to evaluate the outcome of service interventions are critical. Seven major principles of outcome assessment must

VOL 18, NO.4, 1992

be considered for the advancement of research in this area. Prindple One: Multidimensionality of Outcome. To understand the full impact of a service intervention, multiple dimensions of outcome must be measured. While multiple measures of outcome are usually included in mental health services research projects, the conceptual focus of outcome assessment is typically quite narrow. One helpful framework for categorizing the different measurement areas that must be included is proposed by Hargreaves and Shumway (1989): For mental health target populations there are four relevant domains of service goals. They might be called clinical, rehabilitative, humanistic, and public safety goals. Clinical goals are to improve or cure an illness or disorders, reducing or eliminating its signs and symptoms. Rehabilitative goals are to restore or improve social and vocational functioning. Humanistic goals are to maximize patients' and family members' sense of well-being and personal fulfillment. Public safety goals are to prevent injury to others and to the patient, whether from assaulfive or self-destructive behaviors that arise out of the illness, or from destructive side effects of the services themselves. [p. 2551 These four domains—clinical, rehabilitative, humanitarian, and public safety—cover most, if not all, of the outcomes relevant to mental illness and health. Each domain contains several dimensions that need to bemeaswed. And for areas of specific importance to the research question, measures from multiple perspectives within a domain are important (Cook and Campbell 1979). Clinical domain. The clinical outcome domain indudes various aspects of syndromal psychopathology and the course of these phenomenologies over time. This frequently includes

603

signs and symptoms, their aggregation into diagnostic entities, the amount and severity of symptomatology, and the amount and intensity of interventions (e.g., hospitalization, medication, type and amount of aftercare) required to control and treat these manifestations of illness (Strauss and Carpenter 1972). The dimensions of clinical outcome would probably indude what Ware (1989) subsumes under the physical and mental "generic mental health concepts." His examples of the latter include constructs such as anxiety/depression, behavioral/ emotional control, and cognitive functioning. Clinical outcome dimensions have been the most studied among the four domains. They are both categorical (e.g., diagnosis) and dimensional (e.g., severity of thought disorders). Many reliable and useful assessment instruments measure this vast array, ranging from structured diagnostic interviews such as the DIS, SADS, and the SC!!), to symptom rating or enumeration scales, such as the Symptom Checklist-90 (Derogatis and Cleary 1977) and CES-D. In fact, it now appears that mental health dm1cal outcome assessment enjoys excessive attention and instrumentation and that initiatives for further development may best be targeted toward other outcome domains. However, it is important to note that the general health status of severely mentally ill people has been largely ignored in studies of treatment and rehabilitation outcomes (for an exception, see Stine et al. 1988). Traditional health status measures such as mortality, morbidity, life satisfaction, and lifespan have equal applicability to the measurement of mental health outcomes. Severe mental illness may indeed have an impact on length of life as well as on the physical well.. being of the individual in both direct

and indirect ways. This is a highly underresearched and underused form of outcome measurement in mental health that could have important policy implications. Rehabilitative domain. Rehabilitation focuses on adaptation and functional capacity, not illness and psychopathology. Rehabilitation strategies target a client's strengths, talents, and coping strategies. Two broad outcome realms are relevant: social and instrumental functioning. In measuring the outcomes in the rehabilitative domain, it is imperative to examine the interrelationships among outcomes; changes in one outcome can have effects on other outcomes. Social functioning outcome dirnensions are varied and include frequency of interpersonal contacts, closeness of these contacts, and degree of involvement in a meaningful social network—work, school, family, peer group, community, or all of these. Implicit in this dimension are constructs as basic as primitive social skills or as advanced as capacity for generativity. The client's marital stahis and parental status reveal a great deal about his or her capacity for intimacy and the complexities of which he or she is capable in interpersonal relationships. Clients with severe mental illness are compromised socially. Their relationships appear to be fleeting, superficial, and suffused with conflict and anxiety. Yet many clients with severe mental illnesses exist within and are sustained by a loose-knit social network of similarly disenfranchised peers. Studies of the nature, constitution, aeation, and evolution of these networks are necessary because, in addition to family and treatment networks, these networks are the client's community. Because these networks do exist, self-help organizations become more of a possibility.

-

604

Instrumental functioning outcome dimensions encompass the client's problem-solving capacities, ability to learn and to work, and the amount of initiative, drive, and frustration tolerance he or she brings to these pursuits. These variables are among the most straightforward and easy to assess. As outlined by Dion and Anthony (1987), for example, relevant measures indude level of education (grade level, courses completed, degrees received, grades achieved, achievement test scores, etc.), level of employment (sheltered to competitive, part-time to full-time), level of work complexity (often reflected in earnings), productivity, satisfaction, and perseverance. However, it is important when choosing these measures of instrumental functioning to take into account that many persons with severe mental disorders are in the lower socioeconomic strata. So, for example, measures of economic resources should include income from sources such as panhandling, and employment should include secondary labor force participation. Perhaps the most uniform consequence of severe mental illness is a vastly compromised capacity for productive activity. If there is a distinctive characteristic of schizophrenia, it is unemployment. Virtually all clients with severe mental illness who are in rehabilitation programs are unable to work at some time, and most clients with chronic psychotic disorders are unable to work all the time, at least in competitive employment. For these clients, vocational rehabilitation efforts are always welcome, but for most of them the outcome cannot be described in terms of the usual marketplace. For the person with severe mental illness, supported employment is required. Relevant measures of the amount of support necessary to maintain a client in a wage-earning position are needed. It is also important to

SCHIZOPHRENIA BULLETIN

measure types of supports and their cost-effectiveness. However, in assessing work history or current employment status and complexity, researchers should take into account the job market features of the area in which the individual lives. For example, the probability of a severely mentally ill person finding employment in a geographical area with a high rate of unemployment is lower than in an area where the job market is better. Local market conditions also need to be taken into account when housing outcomes are being examined. For both markets, it is important to account for race- and gender-specific conditions in the local area when assessing instrumental functioning outcomes. Humanitarian domain. Subjective well-being, consumer satisfaction, and quality of life constitute the ultimate humanitarian goals; a few instruments exist to assess these dimensions for severely mentally ill adults (Weissman and Bothwell 1976; Dicta and Dvoskin 1980; Baker and Intagliata 1982; Bigelow et al. 1982; Heinrichs et al. 1984; Diamond 1985; Burns and Lehman 1988; Lehman 1988; Zastowny and Lehman 1988; Greenfield and Attkisson 1989). However, most instruments need to be refined and elaborated to measure the depth and complexity of severe mental illness. They also need further psychometric testing. The development of instruments to measure the quality of life of families and family satisfaction with providers and services has just begun; considerably more work is needed (Struening, unpublished manuscript 1991; Tessler et al. 1991). Perhaps the reason why humanitarian outcome dimensions are the least developed is that humanitarian goals for people with severe mental illness are the most neglected in our society. The thousands of mentally ill home-

-

less individuals bear grim witness to our neglect of humanitarian goals. The causes of this neglect are many and have been well delineated by Torrey (1988). This neglect of humanitarian goals is also reflected in the selective attention policymakers give to the results of psychiatric research. For example, the clear therapeutic advance documented by clinical trials of chiorpromazine was hailed as a cure and initiated the deinstitutionalization of patients from State institutions. More recently, cutbacks in State funding of care for indigent individuals in the public mental health care system have been justified by policymakers who cite research showing that shorter and less elaborate treatment programs are as effective as their longer and more traditional counterparts in terms of clinical and rehabilitative outcomes. What policymakers seem to overlook in the psychiatric research is the power of mental illness to destroy lives in spite of a wide variety of psychiatric and psychopharmacologic treatments. In the future, the presentation of psychiatric research must address policymakers' and the public's misconceptions about the chronic and devastating nature of severe mental illnesses and the needs of those who suffer from them. It is important for the public to understand that there are very real limits to therapeutic strategies with severe mental illness and that humanitarian considerations demand we provide supportive asylum to people whose illnesses place them at the boundaries of these limits. Although the nature of this asylum is only beginning to be articulated, a community-based form of asylum is dearly a necessity. Some of these specifics are clear, at least in principle, and indude adequate food, clothing, and shelter. Other questions require -

VOL 18, NO. 4, 1992

careful study, such as how to maximize a patient's self-care capacities, how to recognize the limitations of these capacities in a patient, how then to tailor a living situation that matches these limitations, and how providing Such a match affects a patient's well-being and quality of life. Before realistic and relevant humanitarian outcome instruments can be developed, much work needs to be done to solve the measurement problems inherent in developing measures of this type. However, the primary task is to ensure that humanitarian goals are considered in research designs and that the people who suffer from devastating mental illnessesnot the illnesses or the institutions that provide care—be the focus of clinical services research. At the very least, humanitarian dimensions should be included in every study that also measures cost as an outcome variable. It is important to document the actual relationships between price of service and quality of life. Such research is vital to guide our society in forging newer, more enlightened communitybased asylums for our citizens with severe mental illnesses. Public safety domain. One of the founding principles of our Nation, individual liberty, is not always compatible with two others: life and the pursuit of happiness. In fact, personal happiness, which requires protection of life, can be at variance with absolute personal liberty. These are issues that achieve preeminent importance when the care of people with severe mental illnesses is being considered. The institutionalization of severely mentally ill people, particularly in hospital back wards, constituted a form of societal paternalism in which many persons suffered bleak, meaningless lives. With deinstitutionalization and the lack of a community support system, many former patients

605

and others with severe mental illnesses have been given nearly absolute individual liberty but at a very high price. Now that patients can be committed to treatment services only if they are extremely and imminently dangerous to themselves or society, our society allows individuals incapable of realistic planning to struggle through life and wander the streets. Like ships without rudders, homeless people with severe mental illness are free, but at significant risk to life and without much hope of happiness. One could argue that in the name of individual liberty we have created a system that is a much greater danger to life and the pursuit of happiness than the one it replaced. Furthermore, by moving from one extreme societal response to another in the care of severely mentally ill people, we have avoided the very difficult policy question regarding the appropriate balance between absolute liberty and paternalism. However, this should also be considered an empirical question: What are the societal, family, and individual consequences of different civil commitment statutes? Or to phrase the question in the terms used here, what is the appropriate balance between liberty and paternalism that will maximize individual and societal rights to physical safety and wellbeing? Mental health services researchers must address this question of balance by incorporating public safety and welfare outcome measures into studies of treatment and rehabilitation effectiveness. Although it is important to measure physical health, destructive behaviors, and self-destructive behaviors, it is equally important that measures reflecting individual, familial, and community perceptions of safety, stability, and well-being be developed. Most important is that the balance cannot be reached without

understanding that every severely mentally ill individual has unique treatment and rehabilitation needs. Principle Two: Multiple Perspectives. Important though it is to measure multiple dimensions of an individual's or a system's progress, it is also important to have many perspectives on that progress. Multiple perspectives must be considered in the formulation and assessment of mental health outcomes. Outcome assessment must involve observations of the patient, the family, the social network, and members of the patient's social sphere, in addition to the perspectives of the clinician, the human service worker, and the clinical scientist. Each views the clinical phenomena and the individual in a unique way and in a shared way. The comparison, blending, and reconciliation of these perspectives are the essential tasks in outcome assessment of the treatment of severe and persistent mental disorders. Principle Three: individual Utility Differences. Because perceptions of good and bad outcomes are individually defined and these perceptions or "utilities" can affect outcome measurement, they are a source of variability that need to be systematically measured and accounted for in outcome studies. Even within a diagnostic category such as schizophrenia, individuals vary considerably in their premorbid adjustment and the range and pattern of their level of adaptation and psychopathology ; these differences can affect their utilities for different outcomes. Other factors have strong effects on utilities. Culture, age, and gender can have an impact not only on the course of illness but also on what is defined by individuals, their families, and their community as desirable outcomes. For example, not all cultures

606

place equal emphasis on variables such as independence, social ties, and work skills in defining the functioning of their members. Different age groups and different genders may also have different definitions of desirable outcomes. It may be important to understand, for example, how the needs of an older black woman differ from the needs of a young white man. Certainly, few would argue with the view that reduction of pain and suffering is desirable; however, even concepts such as quality of life have various cultural and social meanings. It is critical that research include individual utilities, particularly in studies of cost, so that issues of quality of life can be addressed. Some research has been done on defining outcomes in terms of what patients or their families want to accomplish (e.g., goal attainment scaling). The decision analytic research literature in the genera] health services field has incorporated utilities in patient treatment models, particularly those induding cost-effectiveness analyses (Weinstein and Stason 1977; McNeil and Packer 1982; Berwick and Weinstein 1985). However, these models typically focus on acute illnesses or a single procedure. Models for chronic illnesses, such as manic depression or schizophrenia, are much more difficult to construct and require more sophisticated statistical techniques (e.g., Zarin and Pass 1987). Therefore, much more thought and work will be required to adequately incorporate utilities into outcome studies of mental health treatments and rehabilitation. Principle Four: Standardization of Measures and Design. Progress in the measurement of outcomes requires a standardization of measures so that comparisons can be made across sites, across studies, and across disorders.

SCHIZOPHRENIA BULLETIN

However, equally important is a sensitivity to the need for specificity of measures. For example, to fully understand the different dimensions of mental health service delivery outcomes among urban Puerto Ricans with severe mental illnesses, it may be important to include different types of symptoms, types of instrumental functioning, aspects of consumer satisfaction, and perceptions of safety than would be used in a study among severely mentally ill residents of rural Louisiana. Tension exists between standardization and specificity in measurement. A similar tension exists in research design. The ideal design for determining cause, the randomized clinical trial, is frequently not possible in mental health services research; the ideal design to understand individual perceptions and process—.ethnography—.is frequently too idiosyncratic to have external validity. However, for the field of mental health services to progress, there must be a standardization of protocols and a move to the highest level of research rigor possible. In areas where small-scale research has provided conflicting results—for example, research on case management—large, longitudinal, multisite studies are necessary. The balance of these tensions—between standardization and specificity of measurement and between causal and ethnographic designs—must be a priority in the design of studies of outcome. -

Principle Five: Longitudinal Designs. Outcomes exist in time. Crosssectional assessments of outcome describe the situation at a single point in time. They can be quite elaborate and informative, especially about the interrelationships among dimensions. But the target population in question suffers from disorders of chronicity,

making longitudinal assessment vitally important. Longitudinal studies can provide information on recidivism and degrees of improvement over time, as well as provide the power to determine, for example, the effectiveness of clinical treatment plus rehabilitation compared with clinical treatment alone. It is important to recognize that time itself (e.g., time since treatment, spacing of outcome measures) is an important variable that can affect whether interventions are considered successful. In fact, rehabilitation studies have found that outcomes can improve with time alone (Dion and Anthony 1987). The question of timespan between baseline, treatment, and followup measurements is critical for a valid understanding of outcome. Related to the need for longitudinal studies of outcome is the need for analytic techniques that can handle the types of problems inherent in longitudinal data, such as missing data, timevarying covariates, and correlated errors. It is also important to have analytic strategies that adequately represent the subtlety of the data and do not rely simply on change scores. Techniques must take into account varying levels of impairment at baseline that can resuit in different rates of change and ultimately potentially misleading outcomes. For example, those who improve most in a rehabilitation setting may still show poor outcomes and may never reach independent living status because they were more severely impaired at baseline. The development of techniques to handle these problems in the analysis of longitudinal psychiatric data has only begun (Laird and Ware 1982; Gibbons et a]. 1988; Waternaux et a]. 1989). -

-

-

Principle Six: Costs as an Outcome Measure. The costs of care are not an appropriate outcome measure un-

-

-

VOL 18, NO. 4, 1992

607

less combined with other measures of intervention effectiveness, as noted under principle three. This is an unportant issue because costs have received so much attention recently in services research, especially among program managers and policymakers. If cost of service is to be measured in future comparative treatment studies, the cost to patients, to families, and to society of the absence (or refusal) of such services must also be estimated and factored in. The final comment from a 5-year outcome study of continuous case management is illustrative (Borland et al. 1989). While the treatment approach reported here costs no less than standard care, neither does it cost more. The behavior of some patients suffering from chronic mental illness demands comprehensive treatment and support functions, regardless of whether the setting is the hospital or community. In choosing between the prevailing revolving door approach on the one hand, and persistent community care on the other, governmental policymakers and planners should not be guided purely by cost considerations. The alternatives are equally expensive. (p. 3761 Principle Seven: Relevance and Impact. Mental health services research needs to have an impact on workers at many levels, including researchers, practitioners, program managers, and policymakers who deal with the problems of severe mental illnesses. This means that in the design of the re. search, the choice of outcome measures, and the dissemination of the results, mental health services researchers should consider the information needs of three spheres of interest: science, clinical practice, and policy and legislation. The range of topics within mental health services research necessitates designs and measures that will provide -

information in these three spheres of interest. However, one sphere must be represented in all research: science. To advance the field and improve our knowledge about truly effective approaches to treatment and rehabilitation, future mental health services research must be based on the theoretical and empirical research that has preceded it. -

Recommendations. • Systematic reviews of available measures relevant to the four domains (clinical, rehabilitative, humanitarian, and public safety) are necessary. (See the following for past reviews: Comrey et al. 1973; Lyerly 1973; Waskow and Parloff 1975; Hargreaves et al. 1977; National Institute of Mental Health 1980, 1985.) Development of instruments in areas that are incompletely measured is critical. • The reliability and validity of instruments (current and future) in each of the four domains need to be determined for special and understudied populations on which they have not been tested (e.g., persons who are homeless, migrants, minorities, and rural residents). • Consensus on a basic set of measures to be used as test batteries in clinical and community settings is also needed. A balance must be sought between the cost of any battery proposed and the value of the information desired. • The validity and reliability of instruments used in general health care settings to measure health, mental health, and functioning, such as those used in the Medical Outcomes Study (Stewart et al. 1988) should be tested in mental health care settings and among clients in outpatient treatment and rehabilitation. • Testing the performance of i nstruments used in longitudinal studies,

which include repeated measures, is critical. Because many of the aspects of outcome that we wish to measure are not stable over time, we must seek to determine how reliable measures are, taking into account this inherent instability and at what point it is most appropriate to measure outcomes. Analytic techniques that can take into account the difficulties encountered in repeated measures are critical (e.g., missing data, irregularly spaced observations, correlated errors, timevarying covariates, and time-invariant covariates).

Summary of the Panel's Recommendations Characteristics of the Target Population. • Epidemiologic studies are needed to: determine the prevalence, risk factors, and consequences of severe mental disorders among representative treated and untreated populations; examine the prevalence and incidence of severe mental disorders within the full socioeconomic, geographic, cultural, and especially the ethnic minority diversity of the U.S. population; assess the co-occurrence and interaction of substance abuse and severe mental disorders; and assess the types, durations, and patterns of disability resulting from mental disorder. • Studies of the life course of people with severe mental disorders are needed to determine the predictors of clinical course, the potential response to early intervention and later treatment, and the benefits of treatment versus no treatment at different stages of disorder. Such studies are also needed to define comprehensive models of illness severity or case mix—models that can be evaluated against a variety of outcomes, including clinical course of

608

disorder, functional status, and use of services. • Studies of rehabilitative and habilitative needs of people with severe mental disorders are required to determine several factors: uniform terms and definitions regarding employment and vocational status, skills, and deficits; the types of independent living skills training needed by different subpopulations of people with severe mental disorders; and the diversity and distribution of living arrangements used by people with severe mental illness, including assessment of costs, impediments, discrimination, and adequacy. • Research is needed on the burden on families caring for relatives with diagnoses of both substance abuse and severe mental disorder. This includes research on the extent and nature of intrafamilial violence experienced by the caregivers as well as observations of family adaptations that have been reported effective. Assessment Research. • Research is needed to develop comprehensive measures of the dinical and rehabilitative status of people with severe mental illness. Psychometric properties of these measures should be rigorously tested across treated and untreated populations and in important subgroups, including ethnic minority cultures and the homeless. • Measures of clinical and rehabilitative status that meet rigorous psychometric criteria should be applied across studies to improve interstudy comparisons. • Measures of clinical and rehabilitative status should be developed for describing client status in service programs; such measures should be used to move toward uniformity in client data base systems across sites.

SCHIZOPHRENIA BULLETIN

Treatment Services Research. • Research is needed on integrating and sequencing clinical and rehabilitative modalities to optimize outcome. • Research and measures are needed to study the effectiveness of case management and continuity of care, different models of case management, the proper role of hospitals as a mode of intervention integrated with community-based care, standards of care and their application to quality of care, and methods for treating people with comorbid or co-occurring conditions. • The role of families and consurners in developing, delivering, and evaluating effective care should be studied. • Investigation of methods of improving the treatment and rehabilitation of cultural and ethnic minorities is needed, along with study of the treatment and rehabilitation of people. with severe mental disorders where violence or violence potential is a primary concern, the transfer of new treatment and rehabilitative knowledge into general practice, and the recruitment and retention of welltrained and motivated personnel to work with severely mentally ill people. Rehabilitation Research. • Methods of improving generalization, durability, and maintenance of social skills training are needed. Such studies would focus on interaction of other aspects of adaptive functioning with individual capacity to benefit from skills training; qualitative aspects of the relationship between trainers and recipients of training; effectiveness of safe sex training for reducing disease transmission and pregnancy rates; and identification of approaches to social skills training that are most effective with differing subpopulations.

• Vocational rehabilitation services should be studied to determine the types and combinations of services that are beneficial for various types of persons; the comparative effectiveness of ongoing versus time-limited services; the effect of labor market features and employment and hiring trends on different types of vocational programs; the degree to which State vocational rehabilitation systems serve people with severe mental disorders; and ways to enhance access and effectiveness in such State systems. • Research is needed on the role of habilitation, particularly the further development of its application to people with severe mental disorders; the identification of factors associated with need for kabilitation services and training; and the special application of habilitation services to the needs of older people with severe mental disorders. • Residential and independent living services, particularly the types of residential services desired and required for different subpopulations should be studied, as well as the following: the impact of housing costs, availability, and access on residential needs; the impact of programmatic factors within residential care settings that affect outcome; the need for longitudinal research designs; the housing needs of people who are unable to live independently in the community where safety and asylum are important; the integration of vocational rehabilitation programs with residential programs; and the residential issues in the rehabilitation of people who live with their families or are excluded from residential programs because of acquired immunodeficiency syndrome (AIDS) or forensic and other considerations. Research is also needed on the organization and financing of residential care facilities, especially compari-

VOL 18, NO.4, 1992

sons between for-profit housing models and their nonprofit counterparts. • Cross-cultural issues and their impact on rehabilitation outcomes should be researched. • The role of the family in the rehabilitation process, particularly the study of ways familial expertise can be integrated with rehabilitation interventions, should be studied, along with ways to determine the value of family and professional collaboration in rehabilitation interventions, and the role of consumers, families, and consumer advocates in policy, service delivery, and evaluation. • The consumer perspective and its incorporation into all aspects of rehabilitation must be considered. Outcome Research. • Clinical services research must include multiple domains of outcome, not just clinical measures. Systematic reviews of psychiatric diagnostic and psychosocial scales are necessary, as is development of instruments in areas that are incompletely measured. • Multiple perspectives of outcome should be taken into account, as should different utilities. • Consensus on a basic set of measures to be used as test batteries in clinical and community settings is also needed. A balance must be sought between the cost of any battery proposed and the value of the information desired. • The validity and reliability of instruments used in general health care settings to measure health, mental health, and functioning should be tested in mental health care settings and among clients in outpatient treatment and rehabilitation. • Testing the performance of instruments used in longitudinal studies, which include repeated measures, is

609

critical. Because many of the aspects of outcome that we wish to measure are not stable over time, we must take this inherent instability into account when determining the reliability of measures, and when it is most appropriate to measure outcomes. • Analytic techniques that include consideration of the difficulties encountered in repeated measures are critical (e.g., missing data, irregularly spaced observations, correlated errors, and covariates that do and do not change over time).

References Abramowitz, S.I., and Murray, J. Race effects in psychotherapy. In: Murray, J., and Abramson, P., eds. Bias in Psychotherapy. New York: Praeger, 1983. pp. 215-255. Acosta, F.X.; Yamamota, J.; Evans, L.A.; and Skilbeck, W.M. Preparing low-income Hispanic, black, and white patients for psychotherapy: Evaluation of a new orientation program. Journal of Clinical Psychology, 39:872-877, 1983. American Psychiatric Association. DSM-III: Diagnostic and Statistical Manual of Mental Disorders. 3rd ed.

Washington, DC: The Association, 1980. American Psychiatric Association. Manual of Psychiatric Peer Review.

2nd ed. Washington, DC: American Psychiatric Press, Inc., 1981. American Psychiatric Association. DSM-III-R: Diagnostic and Statistical Manual of Mental Disorders. 3rd

ed., revised. Washington, DC: The Association, 1987. American Psychiatric Association. The Practice of ECT: Recommendations for Treatment, Training, and Privileging. Washington, DC: Amer-

ican Psychiatric Press, Inc., 1990.

Anderson, C.M.; Reiss, D.J.; and Hogarty, G.E. Schizophrenia and the Family. New York: Guilford Press,

1986. Andreasen, N.C. The Scale for the Assessment of Negative Symptoms

(SANS). Iowa City, IA: The University of Iowa, 1984g. Andreasen, N.C. The Scale for the Assessment of Positive Symptoms

(SAPS). Iowa City, IA: The University of Iowa, 1984b. Andreasen, N.C. Positive vs. negative schizophrenia: A critical evaluation. Schizophrenia Bulletin, 11:380-389, 1985. Angel, R., and Thoits, P. The impact of culture on the cognitive structure of illness. Culture, Medicine, and Psychiatry, 11:465494, 1987. Angst, J.; Felder, W.; and Frey, R. The course of unipolar and bipolar affective disorders. In: Schou, M., and Stromgren, E., eds. Origin, Prevention and Treatment of Affective

-

Disorders. London: Academic Press,

1979. Anthony, J.C.; Folstein, M.; Romenoski, A.J.; Von Korif, M.R.; Nestadt, G.R.; Chahal, R.; Merchant, A.; Brown, C.H.; Shapiro, S.; Kramer, M.; and Gruenberg, E.M. Comparison of the lay Diagnostic Interview Schedule and a standardized psychiatric diagnosis. Archives of General Psychiatry, 42:667-675, 1985. Anthony, W.A., and Blanch, A. Research on community support services: What have we learned. Psychosocial Rehabilitation Journal,

12:55-81, 1988. Anthony, W.A.; Buell, G.J.; Sharratt, S.; and Althoff, M.E. Efficacy of psychiatric rehabilitation. Psychological Bulletin, 78:447-456, 1972. Anthony, W.A.; Cohen, M.R.; and Danley, K.S. The psychiatric rehabili-

-

-

610

tation model as applied to vocational rehabilitation. In: Ciardiello, J.A., and Bell, M.D., eds. Vocational Rehabilitation of Persons With Prolonged Psychiatric Disorders. Baltimore, MD: Johns Hopkins University Press, 1988. pp. 59-79.

Anthony, W.A.; Cohen, M.R.; and Vitalo, R. The measurement of rehabilitation outcome. Schizophrenia Bulletin, 4:365-383, 1978.

Anthony, W.A., and Parkas, M. A client outcome planning model for assessing psychiatric rehabilitation interventions. Schizophrenia Bulletin, 8:13-38, 1982.

SCHIZOPHRENIA BULLETIN

Aylett, M.J. Seeing the same doctor. Journal of Royal College of General Practitioners, 26:47-52, 1976.

Azrin, N.H., and Besalel, V.B. Job Club Counselor's Manual: A Behavioral Approach to Vocational Counseling. Baltimore, MD: University

Park Press, 1980. Azrin, N.H., and Philip, R.A. The Job Club method for the job handicapped: A comparative outcome study. Rehabilitation Counseling Bulletin, 23:

tal and Community Psychiatry, 35:1023-1028, 1984.

Berkowitz, R.; Kuipers, L.; EberleinVries, R.; and Leff, J. Lowering expressed emotion in relatives of schizophrenics. In: Goldstein, M.J., ed. New Development in Interoentions

With Families of Schizophrenics. San

144-155, 1979.

Francisco: Jossey-Bass, 1981. pp.

Bachrach, L.L. Continuity of care for chronic mental patients: A conceptual analysis. American Journal of Psychi-

27-48.

atry, 138:1449-1456, 1981.

Anthony, W.A., and Jansen, M.A. Predicting the vocational capacity of the chronically mentally ill: Research and policy implications. American

Baker, F., and Intagliata, J. Quality of life in the evaluation of community support systems. Evaluation and Pro-

Psychologist, 39:537-544, 1984.

gram Planning, 5:69-79, 1982.

Anthony, W.A., and Liberman, R.P. The practice of psychiatric rehabilitation: Historical, conceptual, and research base. Schizophrenia Bulletin,

Bass, R.D., and Windle, C. A prelixninary attempt to measure continuity of care in a community mental health center. Community Mental Health

12:542-559, 1986.

Journal, 9:53-62, 1973.

Appleby, L., and Desai, P. Residential instability: A perspective of system imbalance. American Journal of Or-

Beiser, M. Commentary on culturebound syndromes and international disease classifications. Culture, Medi-

thopsychiatry, 57:515-524, 1987.

cine and Psychiatry, 11:29-33, 1987.

Ashcraft, M.L.F.; Fries, B.F.; Nerenz, D.R.; Falcon, S.Z.P.; Srivastava, S.V.; Lee, C.Z.; Berki, S.E.; and Errera, P. A psychiatric patient classification system: An alternative to diagnosis-related groups. Medical

Beiser, M. Influences of time, ethnicity, and attachment on depression in Southeast Asian refugees. American Journal of Psychiatry, 145:46-51, 1988.

Atkins, B.J., and Wright, G.N. The vocational rehabilitation of blacks.

Bellack, A.; Hersen, M.; and Himznelhoch, I. A comparison of social skills training, pharmacotherapy, and psychotherapy of depression. Behavior

Journal of Rehabilitation, 46:42-46,

Research and Therapy, 21:101-107,

1980.

1983.

Care, 27:543-557, 1989.

Bellack, A.; Turner, S.; Hersen, M.; and Luber, R. An examination of the efficacy of social skills training for chronic schizophrenic patients. Hospi-

Bernheim, KY., and Lehman, A.F. Working With Families of the Mentally Ill. New York: Guilford Press, 1985.

Berwick, D.M. Health service research and quality of care: Assignments for the 1990s. Medical Care, 27:763-771, 1989.

Beiwick, D.M., and Weinstein, M.C. What do patients value? Willingness to pay for ultrasound in normal pregnancy. Medical Care, 23:881-893, 1985.

Bigelow, D.A.; Brodsky, G.; Stewart, L.; and Olson, M. The concept and measurement of quality of life as' a dependent variable in evaluation of mental health services. In: Stahler, G.J., and Tash, W.R., eds. Innovative Approaches to Mental Health Evaluation. New York: Academic Press, 1982. pp. 345-366.

Bilder, R.M.; Mukherjee, S.; Rieder, R.O.; and Pandurangi, A.K. Symptomatic and neuropsychological components of defect states. Schizo-

Avison, W.R., and Speechely , K.N. The discharged psychiatric patient: A review of social, social-psychological, and psychiatric correlates of outcome.

Bellack, A.; Hersen, M.; and Turner, S. Generalization effects of social skills training with chronic schizophrenics: An experimental analysis.

phrenia Bulletin, 11:409-419, 1985.

American Journal of Psychiatry,

Behaviour Research and Therapy,

American Medical Association,

144:10-18, 1987.

14:391-398, 1976.

253:58-61, 1985.

Black, D.W.; Warrack, C.; and Winokur, G. Excess mortality among psychiatric patients. Journal of the

611

VOL 18, NO. 4, 1992

Blanch, A.K.; Caning, P.J.; and Ridgway, P. Normal housing with special-

ized supports: A psychiatric rehabilitation approach to living in the community. Rehabilitation Psycho! ogy, 33:47-55, 1988.

Bolton, B.. and Cooper, P.C. Comments on vocational rehabilitation of blacks. Journal of Rehabilitation, 46:47-49, 1980.

Bond, G.R. 'Psychiatric Vocational Programs: A Meta-Analysis." Presented at the Annual Conference of the International Association of Psychosocial Rehabilitation Services, Cleveland, OH, June 1986. Bond, G.R., and Boyer, S.L. Rehabilitation programs and outcomes. In: Ciardiello, J.A., and Bell, M., eds. Vocational Rehabilitation of Persons

With Prolonged Psychiatric Disorders. Baltimore, MD: Johns Hopkins University Press, 1988. pp. 231-263.

Bond, G.R., and Fniedmeyer, M.H. Predictive validity of situational assessment at a psychiatric rehabilitation center. Rehabilitation Psychology, 32:99-112, 1987.

Bond, G.R.; Miller, L.D.; Krumwied, R.D.; and Ward, R.S. Assertive case management in three CMHCs: A controlled study. Hospital and Community Psychiatry, 39:411-418, 1988.

Borland, A.; McRae, J.; and Lycan, C. Outcomes of five years of continuous intensive case management. Hospital and Community Psychiatry, 40:369-376, 1989.

Boyer, C.A. Obstades in urban housing policy for the chronically mentally ill. In: Mechanic, D., ed. improving Mental Health Services: What the Social Sciences Tell Us: New Directions for Mental Health Services, 36. San Francisco: Jossey-Bass, 1987. pp. 71-81.

Boyle, R.M., and Rockhold, F.W. An analysis of returning patients in family practice. Journal of Family Practice,

and Johnston, S.A. Overview of adult health status measures fielded in RAND's health insurance study. Med-

8:1029-1036, 1979.

ical Care, 17(Suppl.):1-131, 1979.

Brady, J.P. Social skills training for psychiatric patients, II: Clinical outcome studies. American Journal of

Brown, G.W. Experiences of discharged chronic schizophrenic mental hospital patients in various types of living groups. Milbank Memorial

Psychiatry, 141:491-498, 1984.

Braun, P.; Kochansky, C.; Shapiro, R.; Greenberg, S.; Gudeman, J.E.; Johnson, S.; and Shore, M.F. Overview: Deinstitutionalization of psychiatric patients, a critical review of outcome studies. American Journal of Psychiatry, 138:736-749, 1981.

Bravo, M.; Canino, G.J.; and Bird, H. El DIS en Espaf*ol: Su traduccion y adaptacion en Puerto Rico. Aria Psi-

Fund Quarterly, 37:105-131, 1959.

Brown, G.W.; Birley, J.L.; and Wing, J.K. Post-hospital adjustment of chronic mental patients. Lancet, 11:685-698, 1958.

Brown, G.W.; Binley, J.L.; and Wing, J.K. Influence of family life on the course of schizophrenic illness: A replication. British Journal of Psychiatry, 121:241-258, 1972.

quiatrica y Psicologica de America L.atina, 33:27-42, 1987.

Burke, J.D. Diagnostic assessment. In: Skodol, A., and Spitzer, R.L.,

Breakey, W.R. "Recent Empirical Research on the Homeless Mentally Presented at the ADAMHA Meeting on Research Methodologies Concerning Homeless Persons With Serious Mental illness and/or Substance Abuse Disorders, Bethesda, MD, July

eds. Structured Clinical interview for DSM-ii-R. New York: Biometrics Research Department, New York State Psychiatric Institute, 1988. pp. 13-22.

1987.

Brenner, H.D.; Boker, W.; Hodel, B.; and Wyss, H. Cognitive treatment of basic pervasive dysfunctions in schizophrenia. In: Schulz, S.C., and Tamminga, C.A., eds. Schizophrenia: Scientific Progress. New York: Oxford University Press, 1989. pp. 358-367.

Breslau, N., and Reeb, K.G. Continuity of care in a university-based prac-

Burkhauser, R.V.; Haveman, R.H.; and Wolfe, B.L. "The Economic Status of the Disabled, 1967-1987: Leveli, Changes, andTh eir Policy Implications." Unpublished manuscript, Syracuse University, 1991. Burnam, M.A.; Karno, M.; Hough, R.L.; Escobar, J.I.; and Forsythe, A.B. The Spanish Diagnostic Interview Schedule: Reliability and comparison with clinical diagnoses. Archives of General Psychiatry, 40:1189-1196,

tice. Journal of Medical Education,

1983.

50:965-969, 1975. Brook, B.D. An alternative to psychiatric hospitalization for emergency patients. Hospital and Community Psychiatry, 24:621-624, 1973.

Burnarn, M.A.; Wells, K.B.; Leake, B.; and Landsverk, J. Development of a brief screening instrument for detecting depressive disorders. Medical

Brook, R.H.; Ware, J.E., Jr.; DaviesAvery, A.; Stewart, A.L.; Donald, C.A.; Rogers, W.H.; Williams, K.N.;

Burns, B.J., and Lehman, A.F. Quality of life measurement for the chronically mentally ill. Quality of Life and

Care, 26:775-789, 1988.

612

Cardiovascular Disease, 4:142-144, 1988. Calabrese, J.R., and Delucchi, C. Spectrum of efficacy of vaiproate in the treatment of rapid cycling manic depression. American Journal of Psychiatry, 147:431-434, 1990. Canino, C., and Canino, I.A. Culturally syntonic family therapy for migrant Puerto Ricans. Hospital and Community Psychiatry, 33:299-303, 1982. Carling, P.J.; Randolph, F.L.; Blanch, A.K.; and Ridgway, P.A. Review of the research on housing and community integration for people with psydiiatric disabilities. National Rehabilitation Information Center Quarterly,

1:6-18, 1988. Carpenter, M.D. Residential placement for the chronic psychiatric patient: A review and evaluation of the literature. Schizophrenia Bulletin, 4:384-398, 1978. Carpenter, W.T., Jr. Thoughts on the treatment of schizophrenia. Schizophrenia Bulletin, 12:527-539, 1986. Carpenter, W.T., Jr.; Heinrichs, D.W.; and Alphs, L.D. Treatment of negative symptoms. Schizophrenia Bulletin, 11:440-452, 1985. Carpenter, W.T., Jr.; Mcclashan, T.H.; and Strauss, J.S. The treatment of acute schizophrenia without drugs: An investigation of some current assumptions. American Journal of Psychiatry, 134:14-20, 1977. Cavanaugh, S.V. The prevalence of emotional and cognitive dysfunction in a general medical population: Using the MMSE, GHQ, and BDI. General Hospital Psychiatry, 5:15-24, 1983. Cervantes, R.C.; Salgado de Snyder, V.N.; and Padilla, A.M. Posttraumatic stress in immigrants from Central America and Mexico. Hospi-

SCHIZOPHRENIA BULLETIN

tal and Community Psychiatry,

Journal of Orthopsychiatry, 51:636-

40:615-619, 1989. Chamberlin, J. On Our Own: Patient-

645, 1981. Cometa, M.S.; Morrison, J.K.; and Ziskoven, M. Halfway to where: A critique of research on halfway

Controlled Alternatives to the Mental Health System. New York: McGraw-

Hill, 1979. Chamberlin, J.; Rogers, J.A.; and Sneed, C.S. Consumers, families, and community support systems. Special Issue: The Community Support System Concept. Psychosocial Rehabilitation Journal, 12:93-106, 1989. Ciardiello, J.A. Job placement success of schizophrenic clients in sheltered workshop programs. Vocational Eval uation and Work Adjustment Bulletin, 14:125-128, 1981.

Cleary, P.D. The need and demand for mental health services. In: Taube, C.A.; Mechanic, D.; and Hohmann, A., eds. The Future of Mental Health Services Research. Washington, DC: National Institute of Mental Health. DHHS Publication No. (ADM)891600, 1989. pp. 161-184. Coates, D.B.; Kendall, L.M.; MaCurdy, E.A.; and Goodacre, R.H. Evaluating hospital and home treatment for psychiatric patients. Canada's Mental Health, 24:28-32, 1976. Cohen, B.F., and Anthony, W.A. Functional assessment in psychiatric rehabilitation. In: Flalpern, A., and Fuhrer, M., eds. Functional Assessment in Rehabilitation. Baltimore, MD: Paul Brooks Publishing Co., 1984. pp. 79-100. Cohler, B.; Pickett, S.; and Cook, J. The psychiatric patient grows older: Issues in family care. In: Light, E., and Lebowitz, B., eds. The Elderly With Chronic Mental Illness. New York: Springer Publishing Co., 1991. pp. 82-110. Comas-Diaz, L. Puerto Rican espiritismo and psychotherapy. American

houses. Journal of Community Psychology, 7:23-27, 1979.

Comrey, A.L.; Backer, T.E.; and Glaser, E.M. A Sourcebook for Mental Health Measures. Los Angeles:

Human Interaction Research Institute, 1973. Conger, J.C., and Conger, A.J. Assessment of social skills. In: Ciminero, A.R.; Calhoun, K.S.; and Adams, H.E., eds. Handbook of Behavioral Asses5ment. 2nd ed. New York: John Wiley & Sons, Inc., 1986. pp. 526-560. Conte, L. Sheltered employment and disabled citizens: An analysis of the work stations in the industry model. Dissertation Abstracts International,

43:2976A, 1983. Cook, J.A. Who "mothers" the mentally ill? Family Relations, 37:42-49, 1988. Cook, J.A. Job ending among youth and adults with severe mental illness. Journal of Mental Health Administration, in press. Cook, J.A., and Cohler, B. Reciprocal socialization and the care of offspring with cancer and schizophrenia. In: Datan, N.; Greene, A.; and Reese, H., eds. Life-Span Developmental Psychology. I-Iillsdale, NJ: Erlbaum, 1986. pp. 223-243. Cook, J.A.; Jonikas, J.; and Solomon, M. Models of vocational rehabilitation for youth and adults with severe mental illness. American Rehabilitation, in press. Cook, J.A., and Pickett, S. Feelings of burden and criticalness among parents residing with chronically mentally ill

VOL 18, NO.4, 1992

offspring. Journal of Applied Social Sciences, 12:79-107, 1987. Cook, J.A., and Razzano, L. Natural vocational supports for persons with severe mental illness: Thesholds Supported Competitive Employment Program. In: Stein, L., ed., Innovations in Mental Health Services. San Francisco: Jossey-Bass, in press. Cook, J.A.; Solomon, M.; and Mock, L. What happens after the first job placement? Programming for Adolescents With Behavior Disorders, 4:71-93, 1989.

Cook, l.A.; Wessell, M.; and Dincin, J. Predicting educational achievement levels of the severely mentally ill: hnplications for psychosocial program administrators. Psychosocial Rehabili-

tation Journal, 10:23-37, 1987.

Cook, T.D., and Campbell, D.T. Quasi-Experimentation: Design & Analysis Issues for Field Settings. Chi-

cago: Rand McNally, 1979. Cornblatt, B.A.; Lenzenweger, M.F.; Dworkin, R.H.; and ErlenmeyerKimling, L. Positive and negative schizophrenic symptoms, attention, and information processing. Schizophrenia Bulletin, 11:397-408, 1983.

Cournos, F. The impact of environmental factors on outcome in residential programs. Hospital and Community Psychiatry, 38:848-852, 1987.

Creer, C., and Wing, J.K. Schizophrenia at Home. London: National Schizophrenia Fellowship, 1974. Creer, C., and Wing, J. Living with a schizophrenic patient. British Journal of Hospital Medicine, 14:73-82, 1975.

Crow, T.J. The two-syndrome concept: Origins and current status. Schizophrenia Bulletin, 11:471-486, 1985.

Cuellar, I., and Arnold, B. Cultural considerations and rehabilitation of

613

Mexican-Americans. Journal of Rehabilitation, 54:35-41, 1988. Danek, M.M., and Lawrence, R. Women in rehabilitation: An analysis of State agency services to disabled women. Journal of Applied Rehabilitation Counseling, 16:16-18, 1985.

Daniels, L.V., and Carling, P.J. "Community Residential Rehabilitation Services for Psychiatrically Disabled Persons in Kitsap County, Washington." Unpublished manuscript, Center for Psychiatric Rehabilitation, Boston University, 1986. Delano, D., and Anthony, W. On the relative effectiveness of institutional and alternative placements of the psychiatrically disabled. Journal

of Social Issues, 37:21-33, 1981. Derogatis, L.R., and Cleary, P.A. Confirmation of the dimensional structure of the SCL-90: A study in construct validation. Journal of Clini-

Doll, W. Family coping with the mentally ill: An unanticipated problem of deinstitutionalization. Hospital and Community Psychiatry, 27:183-186, 1976.

Donabedian, A. Explorations in Quality Assessment and Monitoring: The Definition of Quality and Approaches to Its Assessment. Ann Arbor, Ml:

Health Administration Press, 1980. Douzinas, N., and Carpenter, M. Predicting the community performance of vocational rehabilitation clients. Hospital and Community Psychiatry, 32:409-413, 1981.

Draguns, J.G. Psychological disorders of clinical severity. In: Triandis, H.D., and Draguns, J.G., eds. Handbook of Cross-Cultural Psychology and Psychopathology. Boston, MA: Allyn and Bacon, 1980. pp. 99-174.

Diamond, R. Drugs and the quality of life: The patient's point of view. Jour-

Drake, R.E.; Wallach, M.A.; and Hoffman, J.S. Housing instability and homelessness among aftercare patients of an urban State hospital. Hospital

nal of Clinical Psychiatry, 46:29-35, 1985.

and Community Psychiatry, 40:46-51, 1989.

Dicta, P.E., and Dvoskin, J.A. Quality of life for the mentally disabled.

Dunner, D.L., and Fieve, R.R. Clinical factors in lithium prophylaxis failure. Archives of General Psychiatry,

cal Psychology, 33:981-989, 1977.

Journal of Forensic Sciences, 25: 926-931, 1980.

Dincin, J. Psychiatric rehabilitation. Schizophrenia Bulletin, 1(Experimental Issue No. 14):131-148, 1975. Dincin, J.; Selleck, V.; and Streicker, S. Restructuring parental attitudes: Working with parents of the mentally ill. Schizophrenia Bulletin, 4:597-608, 1978.

Dion, G.L., and Anthony, W.A. Research in psychiatric rehabilitation: A review of experimental and quasiexperimental studies. Rehabilitation Counseling Bulletin, 30:177-203, 1987.

30:229-233, 1974.

Dworkin, R.J., and Adams, G.L. Retention of Hispanics in public sector mental health services. Community Mental Health Journal, 23:204-216, 1987.

Elkin, I.; Shea, M.; Watkins, J.; hnber, S.; Sotsky, S.; Collins, J.; Glass, D.; Pilkonis, P.; Leber, W.; Docherty, J.; Fiester, S.; and Panloff, M. NIMH treatment of depression collaborative research program: General effectiveness of treatments. Archives of General Psychiatry, 46:971-982, 1989.

-

614

Ellsworth, R.B.; Casey, N.A.; Hickey, R.H.; Twemlow, S.W.; Collins, J.F.; Schoonover, R.A.; Hyer, L.; and Nesseiroade, J.R. Some characteristics of effective psychiatric treatment programs. Journal of Consulting and Clinical Psychology, 47:799-817, 1979. Endicott, J., and Spitzer, R.L. A diagnostic interview: Schedule for Affective Disorders and Schizophrenia. Archives of General Psychiatry, 35:837844, 1978. Endicott, J.; Spitzer, R.L.; Fleiss, J.L.; and Cohen, J. The Global Assessment Scale—A procedure for measuring overall severity of psychiatric disturbance. Archives of General Psychiatry, 33:766-771, 1976. Engelhardt, D.M., and Rosen, B. Implications of drug treatment for the social rehabilitation of schizophrenic patients. Schizophrenia Bulletin, 2:454-462, 1976. Escobar, 1.; Randolph, E.T.; and Hill, M. Symptoms of schizophrenia in Hispanic and Anglo-veterans. Culture, Medicine & Psychiatry, 10:259-276, 1986. Essock-Vitale, S.M. Patient characteristics predictive of treatment costs in inpatient psychiatric wards. Hospital and Community Psychiatry, 38:263269, 1987. Fabrega, H.R., Jr. An ethnomedical perspective of Anglo-American psychiatry. American Journal of Psychia-

146:588-596, 1989. Fairweather, G.W. The prototype Lodge Society: Instituting process principles. In: Fairweather, G.W., ed. New Directions for Mental Health by,

Services—The Fairweat her Lodge: A Twenty-Five Year Retrospective.

San Francisco: Jossey-Bass, 1980. pp. 13-32.

SCHIZOPHRENIA BULLETIN

Fairweather, G.W.; Sanders, D.H.; Maynard, H.; Cressler, D.L.; and Bleck, D.S. Community Life for the Mentally M. Chicago: Aldine, 1969. Falloon, I.R.H.; Boyd, 1.; McGill, C.; Razani, J.; Moss, FL; and Gilderman, A. Family management in the prevention of exacerbations of schizophrenia: A controlled study. New England Journal of Medicine,

306:1437-1440, 1982. Falloon, LR.H.; Liberman, R.P.; Liilie, F.; and Vaughn, C.E. Family therapy of schizophrenics with high risk of relapse. Family Process, 20:211221, 1981. Farkas, M.; Rogers, S.; and Thurer, S. Rehabilitation Outcome for the Recently Deinstitutionalized Client: The Ones We Left Behind. Boston,

MA: Center for Psychiatric Rehabilitation, Boston University, 1986. Feighner, J.P.; Robins, E.; Guze, S.B.; Woodruff, R.A., Jr.; Winokur, C.; and Munoz, R. Diagnostic criteria for use in psychiatric research. Archives of General Psychiatry, 26:57-63, 1972. Fischer, P.J.; Shapiro, S.; Breakey, W.; Anthony, J.; and Dramer, M. 'Mental Health and Social Characteristics of the Homeless: A Survey of Mission Users." Presented at the 112th Annual Meeting of the American Public Health Association, Anaheim, CA, November 1984. Flaskerud, J.H., and Anh, N.T. Mental health needs of Vietnamese refugees. Hospital and Community Psychiatry, 39:435-437, 1988.

Fountain House. Evaluation of Clubhouse Model Community-Based Psychiatric Rehabilitation: Final Report to the National Institute of Handicapped Research. Washington, DC:

National Institute of Handicapped Research, Contract No. 300-84-0124, 1985.

Foy, D.; Eisler, R.; and Pinkston, S. MOdeled assertion in a case of explosive rages. Journal of Behavior The,apy and Experimental Psychiatry,

6:135-137, 1975. Frankenburg, P.R.; Tohen, M.O.; Cohen, B.M.; and Lipinski, J.F. Longterm response to carbamazepine: A retrospective study. Journal of Clinical Psychopharmacology, 8:130-132, 1988. Franklin, J.L.; Kittredge, L.D.; and Thrasher, J.H. A survey of factors related to mental hospital readmissions. Hospital and Community Psychiatry, 26:749-751, 1975.

Franklin, J.L.; Solovitz, B.; Mason, M.; Clemons, J.; and Miller, G. An evaluation of case management. American Journal

of Public Health,

77:674-678, 1987. Freedman, R.I., and Moran, A. Wanderers in a promised land: The chronically mentally ill and deinstitufionalization. Medical Care, 22(Suppl.): S1-S60, 1984. Fujiwara, T., and Christian, A. The Asian/Pacific Island community—A diverse cultural and ethnic people. Community Support Network News,

4:5-6, 1988. Fyer, M.R.; Frances, A.J.; Sullivan, T.; Hurt, S.W.; and Larkin, J. Comorbidity of borderline personality disorders. Archives of General Psychiatry, 45:348-352, 1988. Gary, L. Attitudes of black adults toward community mental health cen-

ters. Hospital and Community Psychiatry, 38:1100-1105, 1987.

Gibbons, R.D.; Hedeker, D.R.; Waternaux, C.; and Davis, J.M. Random regression models: A comprehensive approach to the analysis of longitudinal psychiatric data. Psychopharmacology Bulletin, 24:438-443, 1988.

-

VOL. 18, NO. 4, 1992

Gibbons, R.D.; Lewine, R.R.J.; Davis, J.M.; Schooler, N.R.; and Cole, J.O. An empirical test of a Kraepelinian vs. a Bleulerian view of negafive symptoms. Schizophrenia Bulktin, 11:390-396, 1985. Gibson, R.L.; Marone, J.; Coutu, C.; and King, E.W. A rehabilitation program for chronically hospitalized pafients. Hospital and Community Psychiat,,j, 23:381-383, 1972. Gift, T.E.; Strauss, J.; and RitzIer, B. The failure to detect low IQ in psychiatric assessment. American Journal of Psychiatry, 135:345-349, 1978. Goering, P.; Wasylenki, D.; Farkas, M.; Lancee, W.; and Ballantyne, R. What difference does case management make? Hospital and Community Psychiatry, 39:272-276, 1988. Goldberg, D.; Steele, J.J.; Johnson, A.; and Smith, C. Ability of primary care physicians to make accurate ratings of psychiatric symptoms. Archives of General Psychiatry,

39:829-833, 1982. Goldberg, S.C. Negative and deficit symptoms in schizophrenia do respond to neurolepfics. Schizophrenia Bulletin, 11:453-456, 1985. Goldman, H.H. Mental illness and family burden: A public health perspective. Hospital and Community Psychiatry, 33:557-560, 1982.

Goldman, H.H.; Gattozzi, A.A.; and Taube, C. Defining and counting the chronically mentally ill. Hospital and Community Psychiatry, 32:21-27, 1981. Goldman, H.H., and Manderscheid, R.W. Chronic mental disorders in the United States. In: Manderscheid, R.W., and Barrett, S.A., eds. Mental Health, United States, 1987. Washington, DC: Superintendent of Documents, U.S. Government Printing Office, DHHS Publication No.

615

(ADM)87-1518, 1987. pp. 1-13. Goldsmith, J., and McFall, R. Development and evaluation of an interpersonal skill-training program for psychiatric inpatients. Journal of Abnormal Psychology, 84:51-58, 1975. Goldstein, M.J., and Doane, J.A. Family factors in the onset, course, and treatment of schizophrenic spectrum disorders. Journal of Nervous and Mental Disease, 170:692-700, 1982. Goldstein, M.J., and Kopeikin, H.S. Short and long-term effects of combining drug and family therapy. In: Goldstein, M.J., ed. New Devel opments in Interventions With Fami-

lies of Schizophrenics. San Francisco: Jossey-Bass, 1981. pp. 5-26. Goldstein, M.J., and Rodnick, E.H. The family's contribution to the etiology of schizophrenia: Current status. Schizophrenia Bulletin, 1(Experimental Issue No. 14):48-63, 1975. Goldstrom, I.D., and Manderscheid, R.W. The chronically mentally ill: A descriptive analysis from the Uniform Client Data Instrument. Community Support Service Journal, 2:4-9, 1982. Gordis, L., and Markowitz, M. Evaluation of the effectiveness of comprehensive and continuous pediatric care. Pediatrics, 48:766-776, 1971. Grad, J., and Sainsbury, P. The effects that patients have on their families in a community care and a control psychiatric service—A two-year followup. British Journal of Psychiatry,

114:265-278, 1968. Greenfield, T.K., and Attkisson, C.C. Steps toward a multifactorial satisfaction scale for primary care and mental health services. Evaluation and Program Planning, 12:271-278, 1989.

Greist, J.H.; Mathisen, K.; Klein, M.; Benjamin, L.; Erdman, H.; and Evans,

F. Psychiatric diagnosis: What role for the computer? Hospital and Community Psychiatry, 35:1089-1090, 1093,

1984. Griffith, J. A community survey of psychological impairment among Anglo and Mexican Americans and its relationship to service utilization. Community Mental Health Journal,

21:2841, 1985. Griffiths, R.D. Rehabilitation of chronic psychotic patients. Psychological Medicine, 4:316-325, 1974. Hamilton, M. Development of a rating scale for primary depressive illness. British Journal of Social and Clinical Psychology, 6:278-296, 1967.

Hammer, M. Social supports, social networks, and schizophrenia. Schizophrenia Bulletin, 7:45-57, 1981. Hankin, J., and Oktay, J.S. Mental Disorders and Primary Medical Care: An Analytical Review of the Literature. Washington, DC: Superintend-

ent of Documents, U.S. Government Printing Office, DHEW Publication No. (ADM)78-661, 1979. Hargreaves, W.A. "Case Management Measurement Methods Under Development." Presented at the Conference on Homelessness and Mental illness: Toward the Next Generation of Research Studies. National Institute of Mental Health, Bethesda, MD, February 1989. Hargreaves, W.A.; Attkisson, C.C.; and Sorensen, J.E., eds. Resource Ma-

terials for Community Mental Health Program Evaluation. Washington,

DC: Superintendent of Documents, U.S. Government Printing Office, DHEW Publication No. (ADM) 77-328, 1977. Hargreaves, W.A.; Shaw, R.; Shadoan, R.; Walker, E.; Surber, R.; and Gaynor, J. Measuring case management activity. Journal of Nervous and Mental Disease, 172:296-300, 1984.

SCHIZOPHRENIA BULLETIN

616

Hargreaves, W.A., and Shumway, M. Effectiveness of mental health services for the severely mentally ill. In: Taube, C.A.; Mechanic, D.; and Hohmann, A.A., eds. The Future of Mental Health Services Research. Washington, DC: Superintendent of Documents, U.S. Government Printing Office, DMHS Publication No. (ADM)89-1600, 1989. pp. 253-283. Harris, M., and Bergman, H.C. Networking with young adult chronic patients. Psychosocial Rehabilitation Journal, 8:28-35, 1985. Harris, M., and Bergman, H.C. Clinical case management for the chronically mentally ill: A conceptual analysis. In: Harris, M., and Bachrach, L., eds. New Directions in Mental Health Services, San Francisco:

Jossey-Bass, 1988. pp. 5-13. Hatfield, A.B. Psychological costs of schizophrenia to the family. Social Work, 23:355-359, 1978. Hatfield, A.B. 'Emerging Issues of the Family Self-Help Movement." Presented at International Symposium, Hebrew University, Jerusalem, Israel, October 1983. Hatfield, A.B., ed. Families of the Mentally Ill: Meeting the Challenges.

San Francisco: Jossey-Bass, 1987. Hatfield, A.B. Serving the unserved in community rehabilitation programs. Psychosocial Rehabilitation Journal,

13:71-82, 1989. Hatfield, A.B.; Fierstein, R.; and Johnson, D. Meeting the needs of families of the psychiatrically disabled. Psychosocial Rehabilitation Journal,

6:17-40, 1982. Hathaway, S.R., and McKinley, J.C. Construction of the schedule. In: Welsh, G.S., and Dahlstrom, W.G., eds. Basic Readings in the MMPI.

Minneapolis, MN: University of Minnesota Press, 1956. pp. 60-63.

Heinrichs, D.W.; Hanlon, T.E.; and Carpenter, W.T., Jr. The Quality of Life Scale: An instrument for rating the schizophrenic deficit syndrome. Schizophrenia Bulletin, 10:388-398, 1984. Henn, F.A., and Nasrallah, H.A.

Bednarski, P. The increased risk for specific psychiatric disorders among persons of low socioeconomic status. The American Journal of Social Psy-

New York: Oxford University Press, 1982. Hennen, B.K. Continuity of care in family practice: Dimensions of con-

chiatry, 6:259-271, 1986. Horn, S.D.; Chambers, A.F.; Sharkey, P.D.; and Horn, R.A. Psychiatric severity of illness: A case mix study. Medical Care, 27:69484, 1989. Hoult, J.; Reynolds, 1.; and Charbonneau-Powis, M. Psychiatric hospital versus community treatment. Aus-

tinuity. Journal of Family Practice,

tralian and New Zealand Journal

2:371-372, 1975. Hersen, M., and Bellack, A. Social skills training for chronic psychiatric patients: Rationale, research findings and future directions. Comprehensive Psychiatry, 17:559-580, 1976. Hersen, M.; Turner, S.; Edeistein, B.; and Pinkston, S. Effects of phenothiazine and social skills training in a withdrawn schizophrenic. Journal of Clinical Psychology, 31:588-594, 1975. Hoeper, E.W.; Kessler, L.G.; Nycz, G.R.; Burke, J.D.; and Pierce, W.E. The usefulness of screening for mental illness. Lancet, 1:33-35, 1984. Hoffman, R., and Koran, L. Detecting physical illness in patients with mental disorders. Psychosomatics, 25:654660, 1984. Hogarty, G.E.; Anderson, C.M.; Reiss, D.J.; Kornblith, S.J.; Greenwald, D.P.; Javna, C.D.; and Madoma, M.J. Family psychoeducation, social skills training, and maintenance chemotherapy in the aftercare treatment of schizophrenia. Archives of General Psychiatry, 43:633-642, 1986. Hollingshead, A.B., and Redlich, F.C. Social Class and Mental Illness. New York: John Wiley & Sons, 1958. Holzer, C.E. Ill; Shea, B.M.; Swanson, J.W.; Leaf, P.J.; Myers, J.K.; George, L.; Weissman, M.M.; and

Psychiatry, 17:160-167, 1983. Hull, J.T., and Thompson, J.C. Factors which contribute to normalization in residential facilities for the mentally ill. Community Mental Health Journal, 17:107-113, 1981. Institute of Medicine. Homelessness, Health, and Human Needs. Washington, DC: National Academy Press, 1988. Intagliata, J. Improving the quality of community care for the chronically mentally disabled: The role of case management. Schizophrenia Bulletin, 8:655-674, 1982. Jacobs, H.; Kardashian, S.; Krienbring, R.; Ponder, R.; and Simpson, A. A skills-oriented model for facilitating employment among psychiatrically disabled persons. Rehabilitation Counseling Bulletin, 28:87-96, 1984. Jampala, V.C.; Sierles, F.S.; and Taylor, M.A. Consumers' views of DSM-Ill: Attitudes and practices of U.S. psychiatrists and 1984 graduating psychiatric residents. American Journal of Psychiatry, 143:148-153, 1986. Jampala, V.C.; Sierles, F.S.; and Taylor, M.A. The use of DSM-III in the United States: A case of not going by the book. Comprehensive Psychiatry, 29:39-47, 1988.

Schizophrenia as a Brain Disease.

of

617

VOL. 18, NO.4, 1992

Jenkins, J.H. Conceptions of schizophrenia as a problem of nerves: A cross-cultural comparison of MexicanAmericans and Anglo-Americans.

Keefe, S.E. Help-seeking behavior among foreign-born and native-born Mexican Americans. Social Science and Medicine, 16:1467-1472, 1982.

Klein, H.E.; Person, T.M.; Cetingok, M.; and Itil, T.M. Family and community variables in adjustment of Turkish and Missouri schizophrenics.

Social Science and Medicine, 26:1233-1246, 1988a.

Keith, S.J., and Matthews, S. Group therapy. In: Bellack, A.S., ed. Schizo-

Comprehensive Psychiatry, 19:233240, 1978.

Jenkins, J.H. Ethnopsychiatric interpretations of schizophrenic illness: The problem of nervios within Mexican-American families. Culture, Med -

phrenia: Treatment, Management, and Rehabilitation. Orlando, FL: Grune & Stratton, Inc., 1984. pp. 337-345.

Kleinman, A. Rethinking Psychiatry: From Cultural Category to Personal Experience. New York: The Free Press, 1987.

icine and Psychiatry, 12:301-329, 1988b.

Kelly, I.; Urey, J.; and Patterson, J. Improving heterosocial conversational skills of male psychiatric patients through a small group training procedure. Behavior Therapy, 11:79-

Kleinman, A. The Illness Narratives:

Johnstone, A., and Goldberg. D. Psychiatric screening in general practice. Lancet, 11:605-608, 1976.

Jones, B.E., and Gray, B.A. Problems in diagnosing schizophrenia and affective disorders among blacks. Hospital and Community Psychiatry, 37:6165, 1986.

Kane, J.; Honigfeld, G.; Singer, J.; Meltzer, H.Y.; and the Cloraril Collaborative Study Group. Clozapine for the treatment-resistant schizophrenic: A double-blind comparison with chiorpromazine. Archives of General Psychiatry, 45:789-796, 1988.

Kanter, J. Clinical case management: Definition, principles, components. Hospital and Community Psychiatry, 40:361-368, 1989.

Karno, M.; Golding, J.M.; Sorenson, S.B.; and Burnam, M.A. The epidemiology of obsessive-compulsive disorders in five U.S. communities. Archives

of General Psychiatry, 45:

1094-1099, 1988.

83, 1980.

Kessler, L.G.; Cleary, P.D.; and Burke, J.D. Psychiatric disorders in primary care: Results of a follow-up

of General Psychiatry, 42:583-587, 1985.

study. Archives

Kiesler, C.A. Mental hospitals and alternative care: Noninstitutionalization as potential public policy for mental patients. American Psychologist, 37:349-360, 1982.

Kiloh, L.G.; Andrews, G.; and Neilson, M. The long-term outcome of depressive illness. British Journal of Psychiatry, 153:752-757, 1988.

Kinzie, J.D.; Frederickson, R.H.; Rath, B.; Fleck, J.; and Karls, W. Posttraumatic stress disorders among survivors of Cambodian concentration camps. American Journal of Psychiatry, 141:645-650, 1984.

Katerndahl, D.A. Nonpsychiatric disorders associated with depression.

Kinzie, J.D., and Manson, S.M. The use of self-rating scales in crosscultural psychiatry. Hospital and

Journal of Family Practice, 13:619624, 1981.

Community Psychiatry, 38:190-195, 1987.

Katz, S. Summary. In: Jones, B.E., ed. Treating the Homeless: Urban Psychiatry's Challenge. Washington, DC: American Psychiatric Press,

Klein, D.F.; Gittelman, R.; and Quitkin, F. Diagnosis and Drug Treatment

1986. pp. 109-114.

Suffering, Healing and the Human Condition. New York: Basic Books, 1988.

Koegel, P.; Burnam, M.A.; and Parr, R.K. The prevalence of specific psychiatric disorders among homeless individuals in the inner city of Los Angeles. Archives of General Psychiatry, 45:1085-1092, 1988.

Kopeikin, H.S.; Marshall, V.; and Goldstein, M. Stages and impact of crisis-oriented family therapy in the aftercare of acute schizophrenia. In: McFarlane, W.R., ed. Family Therapy in Schizophrenia. New York: Guilford Press, 1983. pp. 69-98. Koran, L.M.; Sox, H.C., Jr.; Marton, K.!.; Moltzen, S.; Sox, C.H.;. Kraemer, H.C.; Imai, K.; Kelsey, T.G.; Rose, T.G., Jr.; Levin, L.C.; and Chandra, S. Medical evaluation of psychiatric patients. I: Results in a state mental health system. Archives

of General Psychiatry, 46:733-740, 1989.

Kraepelin, E. Lectures on Clinical Psychiatry. Translated and edited by Thomas Johnstone. New York: Hafner, 1904.

of Psychiatric Disorders: Adults and

Kramer, L.W., and Beidel, D.C. Jobseeking skill groups: Review and application to a chronic psychiatric pop-

Children. 2nd ed. Baltimore, MD: Williams & Wilkins Company, 1980.

ulation. Occupational Therapy in Mental Health, 2:37-44, 1982.

618

Kravitz, P. Left homeless, mentally ill struggle with life on the street. Delaware County (PAl Daily Times, October5, 1987, p.1. Kreisman, D.E.; Sirnmens, S.J.; and Joy, V.D. 'Deinstjtutionaljzation and the Family's Well-being." Presented at the Annual Meeting of the American Psychological Association, New York State Psychiatric Institute, August 1979. Laird, N.M., and Ware, J.H. Random effects models for longitudinal data. Biometrics, 38:963-974, 1982. Lamb, H.R. Therapist-case managers: More than brokers of services. Hospital and Community Psychiatry, 31: 762-764, 1980. Lamb, H.R. Treating the Long-Term Mentally Ill: Deinstitutionalization.

San Francisco: Jossey-Bass, 1982. Lamb, H.R., and Goertzel, V. High expectations of long-term ex-State hospital patients. American Journal of Psychiatry, 129:131-135, 1972. Lamb, R., and Oliphant, E. Schizophrenia through the eyes of families. Hospital and Community Psychiatry,

9:803-806, 1978. Langsley, D.C.; Machotka, P.; and Flomenhatt, K. Avoiding mental hospital admission: A follow-up study. American Journal of Psychiatry,

127:1391-1394, 1971. Lawson, W.B. Racial and ethnic factors in psychiatric research. Hospital and Community Psychiatry, 37:5054, 1986. Lee, A.S., and Murray, R.M. The long-term outcome of Maudsley depressives. British Journal of Psychiatry, 153:741-751, 1988. LeBruzza, A. Physical illness presenting in psychiatric disorders: Guidelines for differential diagnoses. Journal of Operational Psychiatry, 12:2431, 1981.

SCHIZOPHRENIA BULLETIN

Leete, E. The role of the consumer movement and persons with mental illness. In: Perlman, L.G., and Hansen. C.E., eds. Rehabilitation of Persons With Long-Term Mental Illness in the 1990s: A Switzer Monograph.

Alexandria, VA: National Rehabilitation Association, 1988. pp. 65-73. Leff, J. Schizophrenia and sensitivity to the family environment. Schizophrenia Bulletin, 2:568-574, 1976. Leff, J.; Kuipers, L.; Berkowitz, R.; Eberlein-Vries, R.; and Sturgeon, D. A controlled trial of social intervention in the families of schizophrenic patients. British Journal of Psychiatry, 141:121.-134, 1982. Lefley, H. Families of the mentally ill in cross-cultural perspective. Psychosocial Rehabilitation Journal, 8:57-75, 1985. Lefley, H. Aging parents as caregivers of mentally ill adult children: An emerging social problem. Hospital and Community Psychiatry, 10:10631070, 1987a. Lefley, H. Culture and mental illness: The family role. In: Hatfield, A.B., and Lefley, H., eds. Families of the Mentally Ill. New York: Guilford Press, 1987b, pp. 30-59. Lehman, A.F. A quality of life interview for the chronically mentally ill. Program Evaluation and Planning,

-

Lehman, H.E.; Fenton, F.R.; Deutsch, M.; Feldman, S.; and Engelsmann, F. An 11-year follow-up study of 110 depressed patients. Acta Psychiatrica Scandinavica, 78:57-65, 1988. Levine, I.S., and Kennedy, C. The homeless mentally ill: A consultation challenge. Consultation: An International Journal, 4:52-63, 1985. Lewine, R.R.J. Negative symptoms in schizophrenia: Editor's introduction. Schizophrenia Bulletin, 11:361-363, 1985. Lewis, M.A.; Leake, B.; Clark, V.; and Leal-Sotelo, M. Case mix and outcomes of nursing home patients: The importance of prior nursing home care and admission from home versus hospital. Medical Care, 27:376-385, 1989. Liberman, R.P. Assessment of social skills. Schizophrenia Bulletin, 8:6283, 1982. Liberman, R.P. Psychiatric Rehabilitation of Psychiatric Patients. Washington, DC: American Psychiatric Press, 1988. Liberman, R.P.; Lillie, F.; Falloon, I.R.H.; 1-larpin, E.; Hutchinson, W.; and Stoute, B. Social skills training f or relapsing schizophrenics: An experimental analysis. Behavior Modification, 8:155-179, 1984. Liberman, R.P.; Mueser, K.; and Wallace, C.J. Social skills training for schizophrenic individuals at risk for

11:51-62, 1988. Lehman, A.F.; Corty, E.; and Myers, C. Assessment and classification of patients with psychiatric and substance abuse syndromes. Hospital and Community Psychiatry, 40:10191025, 1989. Lehman, A.., and Burns, B.J. Severe mental illness in the COmmunity: Quality of life assessment. In: Spilker,

and competence. Schizophrenia Bulletin, 12:631-647, 1986b.

Press, 1990.

Lin, K.M., and Finder, E. Neurolepfic dosage for Asians. American Journal of Psychiatry, 140:490-491, 1983.

B., ed. Quality of Lift Assessments in Clinical Trials. New York: Raven

relapse. American Journal try, 143:523-526, 1986a.

of Psychia-

Liberman, R.P.; Mueser, K.; Wallace, C.J.; Jacobs, H.; Eckman, T.; and Massel, K. Training skills in the psychiatrically disabled: Learning coping

VOL. 18, NO.4, 1992

Lin, K.M., and Kleinman, A. Psychopathology and clinical course of schizophrenia: A cross-cultural perspective. Schizophrenia Bulletin, 14:555567, 1988. Linn, L., and Yager, J. The effect of screening sensitization and feedback on notation of depression by physicians. Journal of Medical Education, 55:929-942, 1980. Linn, L., and Yager, J. Screening of depression in relationship to subsequent patient and physician behavior. Medical Care, 20:1233-1240, 1982. Linn, M.W.; Caffey, E.M., Jr.; Klett, C.J.; Hogarty, G.E.; and Lamb, H.R. Day treatment and psychotropic drugs in the aftercare of schizophrenic patients. A Veterans Administration cooperative study. Archives of General Psychiatry, 36:1055-1066, 1979. Linn, M.W.; Klett, C.J.; and Caffey, E.M., Jr. Foster home characteristics and psychiatric patient outcome: The wisdom of Gheel confirmed. Archives of General Psychiatry, 37:129-132, 1980. Lipkowitz, M.H., and Idupuganti, S. Diagnosing schizophrenia in 1982: The effect of DSM-III. American Journal of Psychiatry, 142:634-637, 1985. Upton, A.A., and Simon, F.S. Psychiatric diagnosis in a State hospital: Manhattan State revisited. Hospital and Community Psychiatry, 36:368373, 1985. Upton, F.; Sabatini, A.; and Micheels, P. Characteristics and service needs of the homeless mentally ill. In: Jones, B.E., ed. Treating the Homeless: Urban Psychiatry's Challenge. Washington, DC: Ainerican Psychiatric Press, 1986. pp. 31-46. Lopez, S., and Hernandez, P. When culture is considered in the evaluation

619

and treatment of Hispanic patients. Psychotherapy, 24:120-126, 1987. Lyerly, S.B. Handbook of Psychiatric Rating Scales. 2nd ed. Washington, DC: Superintendent of Documents, U.S. Government Printing Office, DHEW Publication No. (HSM)739061; 1973. Lynch, M.M., and Kruzich,J.M. Needs assessment of the chronically mentally ill: Practitioner and client perspectives. Administration in Mental Health, 13:237-248, 1986. Magaro, P.A.; Talbott, J.; and Click, I. The inpatient care of chronic schizophrenia. In: Bellack, A.S., ed Schizophrenia: Treatment, Management, and Rehabilitation. Orlando, FL: Grune & Stratton, Inc., 1984. pp. 193-218. Malgady, R.G.; Rogler, L.H.; and Costantino, C. Ethnocultural and linguistic bias in mental health evaluation of Hispanics. American Psychol ogist, 42:228-234, 1987. Mancuso, L. Reasonable accommodation for workers with psychiatric disabilities. Psychosocial Rehabilitation Journal, 14:3-19, 1991. Manschreck, T.C. Delusional (paranoid) disorders. In: Kaplan, Hi., and Sadock, B.J., eds. Comprehensive Textbook of Psychiatry. 5th ed. Baltimore, MD: Williams & Wilkins Company, 1988. pp. 816-829. Manson, S.M.; Walker, R.D.; and Kivlahan, D.R. Psychiatric assessment and treatment of American Indians and Alaska Natives. Hospital and Community Psychiatry, 38:165-173, 1987. Marcos, L.R.; Alpert, M.; Urguyo, L.; and Kesselman, M. The effect of interview language on the evaluation of psychopathology in Spanish-American schizophrenic patients. American

Journal of Psychiatry, 130:549-553,

1973. Marks, J.; Goldberg, D.; and Hillier, V.G. Determinants of the ability of general practitioners to detect psychiatric illness. Psychological Medicine, 9:337-354, 1979. Marsh, G.N. The family doctor's responsibility to the individual. Journal of the Royal College of General Practitioners, 22:149-160, 1972. Martinez-Diaz, J.; Massel, H.; Wong, S.; Wiegand, W.; Bowen, L.; Edelstein, B.; Marshall, R.; and Liberman, R. "Training of Generalization of Conversational Skills in Chronic Schizophrenics." Presented at the World Congress on Behavior Therapy, Washington, DC, December 1983. Matson, J., and Stephens, R. Increasing appropriate behavior of explosive chronic psychiatric patients with a social skills training package. Behavior Modification, 2:61-77, 1978. Mayer, M.; Alpert, M.; Stastny, P.; Perlick, D.; and Empfield, M. Multiple contributions to clinical presentation of flat affect in schizophrenia. Schizophrenia Bulletin, 11:420-426, 1985. McDowell, I., and Newell, C. Measwing Health: A Guide to Rating Scales and Questionnaires. New York: Oxford University Press, 1987. McFarlane, W.R., ed. Family Therapy in Schizophrenia. New York: Guilford Press, 1983. McGlashan, T.H. The Chestnut Lodge follow-up study. U: Long-term outcome of schizophrenia and the affective disorders. Archives of General Psychiatry, 41:586-601, 1984. McClashan, T.H. The prediction of outcome in chronic schizophrenia. N: The Chestnut Lodge follow-up study. Archives of General Psychiatry, 43:167-176, 1986a.

620

SCHIZOPHRENIA BULLETIN

McGlashan, T.H. Schizophrenia: Psychosocial treatments and the role of psychosocial factors in its etiology and pathogenesis. In: Frances, Aj., and Hales, R.E., ed. American Psychi-

Mezzich, J.E., and Sharfstein, S.S. Severity of illness and diagnostic £ ormulation: Classifying patients for prospective payment systems. Hospi-

Mosher, L.R., and Menn, A.Z. Community residential treatment for schizophrenia: Two-year follow-up.

tal and Community Psychiatry,

atric Association Annual Review.

36:770-772, 1985. Miller, G.H., and Wilier, B. Predictors of return to a psychiatric hospi-

29:715-723, 1978. Mulkern, V.M., and Manderscheid, R.W. Characteristics of community support program dients in 1980 and 1984. Hospital and Community Psychiatry, 40:165-172, 1989. Munoz, J.A. Difficulties of a Hispanic-American psychotherapist in the treatment of Hispanic-American patients. American Journal of Orthopsychiatry, 51:646-653, 1981. Myerson, A.T., and Fine, T., eds.

Vol.5. Washington, DC: American Psychiatric Press, 1986b. pp. 96-111. McGlashan, T.H. Borderline personality disorders and unipolar affective disorders: Long-term effects of comorbidity. Journal of Nervous and Mental Disease, 175:467-473, 1987. McGlashan, T.H. A selective review of recent North American long-term followup studies of schizophrenia. Schizophrenia Bulletin, 14:515-542, 1988. McGlynn, E.A.; Norquist, G.S.; Wells, K.B.; Sullivan, G.; and Liberman, R.P. Quality of care research in mental health: Responding to the challenge. Inquiry, 25:157-170, 1988. McKnight, D.; Nelson, R.; Hayes, S.; and Jarrett, R. Importance of treating individually assessed response classes in the amelioration of depression. Behavior Therapy, 15:315-335, 1984. McNeil, B.J., and Pauker, S.C. Optimizing patient and societal decisionmaking by the incorporation of individual values. In: Disorders, R.L., and Disorders, R.A., eds. Values and Long-Term Care, Lexington, MA: D.C. Heath, 1982. pp. 215-230. McWhinney, I.R. Continuity of care in family practice: Implications of continuity. Journal of Family Practice, 2:373-374, 1975. Meltzer, H.Y.; Bastani, B.; Kwon, K.Y.; Ramirez, L.; Burnett, S.; and Sharpe, I. A prospective study of dozapine in treatment resistant schizophrenic patients. I: Preliminary report. Psychopharmacology, 99: 568-S72, 1989.

tal. Journal of Consulting and Clinical Psychology, 44:898-900, 1976.

Minkoff, K. A map of the chronic mental patient. In: Talbott, J.A., ed. The Chronic Mental Patient: Problems, Solutions, and Recommendations for a Public Policy. Washington,

DC: American Psychiatric Press, 1978. pp. 11-37. Moos, R., and Otto, J. The Community-Oriented Programs Environment Scale: A methodology for the facilitation and evaluation of social change. Community Mental Health Journal,

8:28-37, 1972. Moriatry, J.B.; Walss, R.T.; and McLaughlin, D.E. Employability of clients served in State vocational rehabilitation agencies: A national census. Rehabilitation Counseling Bulletin,

32:108-121, 1988. Morrison, R.L., and Bellack, A.S. Social skills training. In: Bellack, A.S., ed. Schizophrenia: Treatment, Management, and Rehabilitation. Or-

lando, FL: Crone & Stratton, Inc., 1984. pp. 247-279. Morrison, R.L., and Bellack, A.S. Social functioning of schizophrenic patients: Clinical and research issues. Schizophrenia Bulletin, 13:715-725, 1987. Mocicki, E.K.; Locke, B.Z.; Rae, D.S.; and Boyd, J.H. Depressive symptoms among Mexican Americans: The Hispanic Health and Nuffition Examination Survey. American Journal of Epidemiology, 130:348360, 1989.

Hospital and Community Psychiatry,

Psychiatric Disability: Clinical, Legal, and Administrative Dimensions.

Washington, DC: American Psychiatric Press, 1987. Naranjo, D. Differences form the basis for inclusion—It is time to consider the needs of ethnic and racial minority persons. Community Support Network News, 4:4, 1988.

National Center for Health Statistics. Plan and Operation of the Hispanic Health and Nutrition Eza,nination Survey, 1982-84. Series 1, No. 19.

Washington, DC: U.S. Department of Health and Human Services, 1985. National Institute of Mental Health. The Assessment of Psychopathology and Behavioral Problems in Children: A Review of Scales Suitable for Epidemiological and Clinical Research (1967-1979). Series AN, No.1. Wash-

ington, DC: Superintendent of Documents, U.S. Government Printing Office, DHHS Publication No. (ADM)80-1037, 1980. National Institute of Mental Health. Measuring Social Functioning in Mental Health Studies: Concepts and Instruments. Series DN No.5.

Washington, DC: Superintendent of Documents, U.S. Government

VOL. 18, NO.4, 1992

Printing Office, DHHS Publication No. (ADM)85-1384, 1985. National Institute of Mental Health. Guidelines for Meeting the Housing Needs of People With Psychiatric Disabilities. Rockville, MD: National

Institute of Mental Health, 1987a. National Institute of Mental Health. "Request for Applications: Mental Health Services Demonstration Grants," Community Support Program for Adults. Rockville, MD: National Institute of Mental Health, 1987b. National Institute of Mental Health. Caring for People With Severe Mental Disorders: A National Plan of Research to Improve Seroices. Wash-

621

Orleans, C.T.; George, L.K.; Houpt, J.L.; and Brodie, H. How primary care physicians treat psychiatric disorders: A national survey of family practitioners. American Journal of Psychiatry, 142:52-56, 1985.

Platt, J.; Siegel, J.; and Spivack, C. Do psychiatric patients and normals see the same solutions as effective in solving interpersonal problems? Jour-

O'Sullivan, M.J.; Peterson, P.D.; Cox, C.; and Kirkeby, J. Ethnic populations: Community mental health services ten years later. American Journal of Community Psychology, 17:17-30, 1989.

Plait, J., and Spivack, C. Problemsolving thinking of psychiatric pa-

Overall, i.E., and Gorhain, D.R. The Brief Psychiatric Rating Scale. Psychological Reports, 10:799-812, 1962. Pasamanick, B.; Scarpitti, F.; and Dinitz, S. Schizophrenics in the Community. New York: Appleton-Century-

ington, DC: Superintendent of Documents, U.S. Government Printing Office, Dl-IHS Publication No. (ADM) 91-1762, 1991. Nielsen, A.C., and Williams, T.A. Depression in ambulatory medical patients: Prevalence by self-report questionnaire and recognition by nonpsychiatric physicians. Archives of General Psychiatry, 37:999-1004, 1980. Olshansky, S. A five-year follow-up of psychiatrically disabled clients. Rehabilitation Literature, 34:15-16, 1973. Olshansky, S., and Beach, D. A fiveyear follow-up of mentally retarded clients. Rehabilitation Literature, 35:4849, 1974. Olshansky, S., and Beach, D. A fiveyear follow-up of physically disabled

Pincus, H.A.; Pardes, H.; and Rosenfeld, A. Consensus development conferences: Assessing and communicating advances in mental health.

clients. Rehabilitation Literature,

American Journal

36:251-252, 258, 1975. Opjordsmoen, S. Long-term course and outcome in unipolar affective and schizoaffective psychoses. Ada Psychiatrica Scandinavica, 79:317-326, 1989.

140:1329-1331, 1983.

Crofts, 1967. Paul, G.L. Assessment instruinentation for service and research. In: Paul, G.L., ed. Assessment in Residential Treatment Settings. Part II: The TimeSample Behavioral Checklist. Cham..

paign, IL: Research Press, 1987. Paul, G.L., and Lentz, R.J. Psychosocial Treatment of Chronic Mental Patients: Milieu Versus Social Learning

Program. Cambridge, MA: Harvard University Press, 1977. Pettengill, I., and Vertrees, J.C. Reliability and validity in hospital casemix measurement. Health Care Financing Review, 4:101-130, 1982.

of Psychiatry,

Plait, I., and Siegel, J. MMPI characteristics of good and poor social problem-solvers among psychiatric patients. Journal of Psychology, 94:245-251, 1976.

nal of Consulting and Clinical Psychology, 43:279, 1975.

tients. Journal

of Consulting and Clin-

ical Psychology, 39:148-154, 1972. Platt, J., and Spivack, C. Means of solving real life problems. 1: Psychiatric patients versus controls, and cross-cultural comparisons of normal females. Journal of Community Psychology, 2:45-48, 1974. Pogue-Geile, M.F., and Harrow, M. Negative symptoms in schizophrenia: Their longitudinal course and prognostic importance. Schizophrenia Bullatin, 11427-439, 1985. Polak, P., and Kirby, M. A model to replace psychiatric hospitals. Journal

of Nervous and Mental Diseases,

162:13-22, 1976. Post, R.M.; Uhde, T.W.; Ballenger, J.C.; and Squillace, K.M. Prophylactic efficacy of carbamazepine in manic-depressive illness. American Journal of Psychiatry, 140:1602-1604, 1983. Prien, R.F., and Gelengery, A.J.Alternatives to lithium for preventive treatment of bipolar disorder. American Journal of Psychiatry, 146:840848, 1989. Prince, R., and TchengLarouche, F. Culture-bound syndromes and international disease classifications. Culture, Medicine and Psychiatry,

11:3-19, 1987. Putnam, J.F.; Cohen, N.L.; and Sullivan, A.M. Innovative outreach services for the homeless mentally ill. International Journal of Mental Health,

14:112-124, 1986.

622

Radloff, L.S. The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement,

1:385-401, 1977.

Rapp, R.E. A normalization approach to the vocational training of mentally retarded adults. Dissertation Abstracts International, 40:1410A, 1979.

Regier, D.A.; Burke, J.D.; Manderscheid, R.W.; and Burns, B.J. The chronically mentally ill in primary care. Psychological Medicine, 15:265-

273, 1985.

Regier, D.A.; Goldberg, I.D.; Burns, B.; Hankin, J.; Hoeper, E.; and Nycz, G. Specialist/generalist division of responsibility for patients with mental disorders. Archives of General Psychiatry, 39:219-224, 1982.

Regier, D.A.; Goldberg, I.D.; and Taube, C.A. The de facto U.S. mental health services system. Archives of General Psychiatry, 35:685-693, 1978.

Regier, D.A.; Myers, J.K.; Kramer, M.; Robins, L.N.; Blazer, D.C.; Hough, R.L.; Eaton, W.W.; and Locke, B.Z. The NIMH Epidemiologic Catchment Area Program. Archives of General Psychiatry, 41:934-941,

1984.

Rittenhouse, J.D. Without Hospitalization: An Experimental Study of Psychiatric Care in the Home. Denver: Swallow Press, 1970. Robins, L.N.; Helzer, J.E.; Croughan, J.; and Ratcliff, K.S. National Institute of Mental Health Diagnostic Interview Schedule: Its history, characteristics, and validity. Archives of General Psychiatry, 38:381-389, 1981.

Robins, L.N.; Helzer, i.E.; Weissman, M.M.; Orvaschel, H.; Gruenberg, E.; Burke, J.D., Jr.; and Regier, D.A. Lifetime prevalence of specific psychiatric disorders in three sites. Archives

SCHIZOPHRENIA BULLETIN

of General Psychiatry, 41:949-958,

1984.

Robins, L.N.; Locke, B.Z.; and Regier, D.A. The overview: Psychiatric disorders in America. In: Robins, L.N., and Regier, D.A., eds. Psychiatric Disorders in America. New York: The Free Press, 1991. pp. 328-366. Rogers, J., and Curtis, P. The concept and measurement of continuity in primary care. American Journal of Public Health, 70:122-127, 1980.

Rogler, L.H.; Malgady, R.G.; Costantino, C.; and Blumenthal, R. What do culturally sensitive mental health services mean? The case of Hispanics. American Psychologist, 42:565-570,

1987.

Roussel, A.E., and Cook, J.A. The role of work in psychiatric rehabilitation: A therapeutic alternative to competitive employment. Sociological Practice, 6:149-170, 1987.

Rutman, I. 'Preventing Chronicity: A Study of Three Alternatives." Unpublished manuscript, Matrix Research Institute, Philadelphia, PA, 1971. Rutman, I., and Armstrong, K. "A Comprehensive Evaluation of Transitional Employment Programs in the Rehabilitation of Chronically Mentally Disabled Clients." Unpublished manuscript, Matrix Research Institute, Philadelphia, PA, 1983-84, 1985. Samson, J.A.; Simpson, J.C.; and Tsuang, M.T. Outcome studies of schizoaffective disorders. Schizophrenia Bulletin, 14:543-554, 1988.

Santiago, J.M.; Bachrach, L.L.; Berren, M.R.; and Hannah, M.T. Defining the homeless mentally ill: A methodological note. Hospital and Community Psychiatry, 39:1100-1102,

1988.

Sattler, J.J. The effects of therapistclient racial similarity. In: Gurman, A.S., and Razin, A.M., eds. Effective

Psychotherapy: A Handbook of Research. New York: Pergamon Press,

1977. pp. 252-290.

Schottstaedt, M.J.; Deckert, G.H.; and Schneider, J.M. Depression in a university outpatient medical clinic. Journal of the Oklahoma State Medical Association, 64:311-316, 1971.

Schulberg, H.C., and Bromet, E. Strategies for evaluating the outcome of community services for the chronically mentally ill. American Journal of Psychiatry, 138:930-935, 1981.

Schulberg, H.C.; Saul, M.; McCleIland, M.; Ganguli, M.; Christy, W.; and Frank, R. Assessing depression in primary medical and psychiatric practices. Archives of General Psychiatry,

42:1164-1170, 1985.

Schurman, R.A.; Kramer, P.; and Mitchell, J. The hidden mental health network: Treatment of mental illness by non-psychiatrist physicians. Archives of General Psychiatry,

42:89-92, 1985.

Schwab, J.J.; Bialow, M.; Brown, J.M.; and Holzer, C.E. Diagnosing depression in medical inpatients. Annals of Internal Medicine, 67:695-707,

1967.

Segal, S.; Baumohl, J.; and Johnson, E. Falling through the cracks: Mental disorders and social margin in a young vagrant population. Social Problems, 24:387-400, 1977.

Segal, S.P., and Aviram, U. The Mentally Ill in Community-Based Shel tered Care: A Study of Community Care and Social Integration. New York: John Wiley & Sons, 1978.

Seller, R.H.; Blascovich, J.; and Lenkei, E. Influence of stereotypes in the diagnosis of depression by family practice residents. Journal of Family Practice, 12:849-854, 1981.

Shepard, M.; Cooper, B.; Brown, A.C.; Kalton, G.; and Clare, A.

VOL 18, NO.4, 1992

Psychiatric Illness in General Practice. London: Oxford University Press, 1966.

Sherman, P.S., and Porter, R. Mental health consumers as case management aides. Hospital and Community Psychiatry, 42:494-498, 1991. Solomon, P. Racial factors in mental health service utilization. Psychosocial Rehabilitation Journal, 11:3-12, 1988a. Solomon, P. Services to severely mentally disabled and homeless persons, and to emergency food and shelter providers. Psychosocial Rehabilitation Journal, 12:3-13, 1988b. Solomon, P.; Baird, R.; Everstine, L.; and Escobar, A.J. Assessing the community care of chronic psychotic patients. Hospital and Community Psychiatry, 31:113-116, 1980. Solomon, P.; Beck, S.; and Gordon, B. A comparison of perspectives on discharge of extended-care facility clients: Views of families, hospital staff, community mental health workers, and clients. Administration in Mental Health, 15:166-174, 1988. Sommers, A.A. Negative symptoms: Conceptual and methodological problems. Schizophrenia Bulletin, 11:364379, 1985. Spaniol, L.; Jung, H.; Zipple, A.M.; andPitzgerald, S. Familjesasaresource in the rehabilitation of the severely psychiatrically disabled. In: Hatfield, A.B., and Lefley, H.P., eds. Families of the Mentally Ill: Coping and Adaptation. New York: Guilford Press, 1987. pp. 167-190. Sparks, D. Migrant farm workers in South Texas. OSERS New in Print, 3:16-19, 1991. Spaulding, W.D.; Storms, L.; Goodrich, V.; and Sullivan, M. Applications of experimental psychopathol-

623

ogyin psychiatric rehabilitation. Schizophrenia Bulletin, 12:560-577, 1986. Spitzer, R.L.; Endicott, J.; and Robins, E. Research Diagnostic Criteria: Rationale and reliability. Archives of General Psychiatry, 35:773-782, 1978 Spitzer, R.L.; Williams, J.B.W.; Gibbon, M.; and First, M.B. Structured Clinical Interoieu., for DSM-III-R (SCID). New York: Biometrics Research Department, New York State Psychiatric Institute, 1989. Stanton, A.H.; Gunderson, J.; Knapp, P.; Frank, A.; Vannicelli, M.; Schnitzer, R.; and Rosenthal, R. Effects of psychotherapy in schizophrenia. I: Design and implementation of a controlled study. Schizophrenia Bulletin, 10:520-563, 1984. Stein, L.I., and Test, M.A. Alternative to mental hospital treatment. I: Conceptual model, treatment program, and clinical evaluation. Archives of General Psychiatry, 37:392-397, 1980. Stein, L.I.; Test, M.A.; and Marx, A.J. Alternative to the hospital: A controlled study. American Journal of Psychiatry, 132:517-522, 1975. Steinwachs, D.M.; Cullum, H.M.; Dorwart, R.A.; Flynn, L.; Frank, R.; Friedman, M.B.; Herz, M.I.; Mulvey, E.P.; Snowden, L.; Test, M.A.; Iremaine, L.S.; and Windle, C.D. Service systems research. Schizophrenia Bulletin, 18:627-668, 1992. Stewart, A.L.; Hays, R.D.; and Ware, J.E. The MOS short-form general health survey. Medical Care, 26:724735, 1988. Stine, O.C.; Fischer, P.J.; and Breakey, W.R. "Co-morbid Physical Problems of Homeless Mentally ill Persons in Baltimore." Presented at the Annuai Meeting of the American Public -

Health Association, Boston, MA, November 1988. Strauss, J.S. Negative symptoms: Future developments of the concept. Schizophrenia Bulletin, 11:457-460, 1985. Strauss, J.S. Subjective experiences of schizophrenia: Toward a new dynamic psychiatry—U. Schizophrenia Bulletin, 15:179-187, 1989. Strauss, J.S., and Carpenter, W.T., Jr. The prediction of outcome in schizophrenia. I: Characteristics of outcome. Archives of General Psychiatry, 27:739-746, 1972. Stroul, B. Toward Community Support Systems for the Mentally Disabled: The NIMH Community Support Program. Boston, MA: Center for Psychiatric Rehabilitation, Sargent College of Allied Health Professions, 1984. Stroul, B. 'Models of Community Support Services: Approaches to Helping Persons With Long-Term Mental illness." Unpublished manuscript prepared for the National Institute of Mental Health Community Support Program, Boston, MA, 1986. Stroul, B. Community Support Systems for Persons with Long-Term Mental Illness. Rockville, MD: National Institute of Mental Health, Community Support Program, 1988. Struening, E.L. 'Measures of Impact on Caretakers of the Mentally ill." National Institute of Mental Health Grant (RO1-44690), Rockville, MD, 1991. Sue, D., and Sue, S. Culture factors in the clinical assessment of Asian Americans. Journal of Consulting and Clinical Psychology, 55:479-487, 1987. Sue, S. Psychotherapeutic services for ethnic minorities: Two decades of research findings. American Psychologist, 43:301-308, 1988. -

-

624

Szapocznik, J.; Santiesteban, D.; Hervis, 0.; Spencer, F.; and Kurtines, W.M. Treatment of depression among Cuban American elders: Some validation evidence for a life enhancement counseling approach. Journal of Consulting and Clinical Psychology, 49:752-754, 1981.

Talbott, J.A., and Glick, I.D. The inpatient care of the chronically mentally ill. Schizophrenia Bulletin, 12:129-140, 1986.

Taube, C.A.; Thompson, J.; Burns, B.; Widem, P.; and Prevost, C. Prospective payment and psychiatric discharges from general hospitals with and without psychiatric units. Hospital and Community Psychiatry, 7:754-760, 1985.

Tessler, R.C. Continuity of care and client outcome. Psychosocial Rehabilitation Journal, 10:39-53, 1987.

Tessler, R.C. What have we learned to date? Assessing the first generation of NIMH-supported research studies. In: Morrissey, J.P., and Dennis, D.L., eds. Homelessness and Mental Illness: Toward the Next Generation of Research Studies. Rockville, MD:

National Institute of Mental Health, Office of Programs for the Homeless Mentally Ill, 1990. pp. 9-23. Tessler, R.C.; Gamache, G.M.; and Fisher, G.A. Patterns of contact of patient's families with mental health professionals and attitudes toward professionals. Hospital and Community Psychiatry, 42:929-935, 1991.

Tessler, R.C., and Goldman, H.H. The Chronically Mentally Ill: Assessing Community Support Programs. Cambridge, MA: Ballinger, 1982.

Tessler, R.C.; Willis, C.; and Cubman, G.D. Defining and measuring continuity of care. Psychosocial Rehabilitation Journal, 10:27-38, 1986.

SCHIZOPHRENIA BULLETIN

Test, M.A., and Stein, L.I. Community treatment of the chronic patient: Research overview. Schizophrenia Bulletin, 4:350-364, 1978.

Test, M.A., and Stein, L.I. An alternative to mental hospital treatment. 111: Social cost. Archives of General Psychiatry, 37:409-412, 1980.

Tischler, G.L., and Reidel, D.C. A criterion approach to patient care evaluation. American Journal of Psychiatry, 130:913-916, 1973.

Torrey, E.F. Continuous care teams in the care of the chronic mentally ill. Hospital and Community Psychiatry, 37:1243-1247, 1986.

Torrey, E.F. Nowhere to Go: The Tragic Odyssey of the Homeless Mentally M. New York: Harper & Row,

Publishers, 1988. Torrey, E.F., and Wolfe, S.M. Care of the Seriously Mentally Ill: A Rating

of State Programs. Washington, DC: Public Citizen Health Research Group, 1986. Trevino, F.M., ed. Hispanic Health and Nutrition Examination Survey, 1982-1984: Findings on Health Status and Health Care Needs. American Journal

of Public Health, 80(Suppl.):

1-72, 1990.

Unger, K.V., and Anthony, W.A. Are families satisfied with services to young adult chronic patients? In: Pepper, B., and Pyglewicz, H., eds. New Directions for Mental Health Services Sourcebook. San Francisco: Jossey-

Bass, 1983. pp. 91-97. Unger, K.V.; Danley, K.S.; Kohn, L.; and Hutchinson, D. Rehabilitation through education: A universitybased continuing education program for young adults with psychiatric disabilities on a university campus. Psychosocial Rehabilitation Journal, 10:35-49, 1987.

Van Putten, T., and May, P.R. A milieu therapy of the schizophrenias. In: West, L.J., and Flinn, D., eds. Treatment of Schizophrenia: Progress and Prospects. New York: Grune &

Stratton, 1976. pp. 217-243. Van Tosh, L. The Training and Employment Project: Overview, Curriculum, and Summary Report. Philadel-

phia, PA: Project SHARE and Mental Health Association of Southeastern Pennsylvania, 1989. Vaughn, C.E., and Leff, J.P. The influence of family and social factors of the course of psychiatric illness: A comparison of schizophrenic and depressed neurotic patients. British Journal of Psychiatry, 129:125-137, 1976.

Vaughn, C.E.; Sorenson, K.; Jones, S.; Freeman, W.; and Falloon, I.R.H. Family factors in schizophrenic relapse: Replication in California of British research on expressed emotion. Archives of General Psychiatry, 41:1169-1177, 1984.

Verbrugge, L.M. How physicians treat mentally distressed men and women. Social Science and Medicine, 18:1-9, 1984.

Vergare, M.J., and Acre, A. Mental illness in the homeless, in: Jones, B.E., ed. Treating the Homeless: Urban Psychiatry's Challenge. Washington, DC: American Psychiatric Press, 1986. pp. 17-30.

Vernez, G.; Burnam, M.A.; McGlynn, E.A.; Trade, S.E.; and Mittman, B.S. Review of California's Program for the Mentally Disabled.

Santa Monica, CA: The RAND Corporation, 1988. Vernon, S., and Roberts, R.E. Prevalence of treated and untreated psychiatric disorders in three ethnic groups. Social Science and Medicine, 16:1575-1582, 1982.

VOL 18, NO. 4, 1992

625

Vertrees, J.C.; Manton, K.G.; and Mitchell, K.C. Case-mix adjusted analyses of service utilization for a Medicaid health insuring organization in Philadelphia. Medical Care, 27:397-411, 1989. Vitalo, R. An application in an aftercare setting. In: Anthony, W.A., ed. Principles of Psychiatric Rehabilita-

tion. Baltimore, MD: University Park Press, 1979. pp. 193-202. Von Korif, M.; Shapiro, S.; Burke, J.D.; Teitelbauxn, E.; Skinner, E.; Gern'an, P.; Turner, R.; Klein, L.; and Burns, B. Anxiety and depression in a primary care clinic. Archives of General Psychiatry, 44:152-156, 1987. Wallace, Cj., and Liberman, R.P. Social skills training for patients with schizophrenia: A controlled clinical trial. Psychiatry Research, 15:239247, 1985. Wallace, C.J.; Nelson, C.J.; Uberman, R.P.; Aitchinson, R.H.; Lukoff, D.; Elder, J.P.; and Ferris, C. A review and critique of social skills training with schizophrenic patients. Schizophrenia Bulletin, 6:42-63, 1980. Ware, J.E., Jr. Measuring health and functional status in mental health services research. In: Taube, C.A.; Mechanic, D.; and Hohmann, A.A., eds. -

The Future of Mental Health Services Research. Washington, DC: Superin-

tendent of Documents, U.S. Government Printing Office, DHHS Publication No. (ADM)89-1600, 1989. pp. 289-302. Waskow, I.E., and Parloff, M.B., eds. Psychotherapy Change Measures.

Washington, DC: Superintendent of Documents, U.S. Government Printing Office, DHEW Publication No. (ADM)74-120, 1975. Wasylenki, D.; Goeririg, P.; Lancee, W.J.; Fischer, L.; and Freeman, S.J.J. Psychiatric aftercare: Identified needs versus referral patterns. American

Journal of Psychiatry, 138:1228-1231,

1981. Waternaux, C.; Laird, N.M.; and Ware, J.H. Methods for analysis of longitudinal data: Blood lead concentrations and cognitive development. Journal of the American Statistical Association, 84:33-41, 1989.

Weinberg, J.L., and Lustig, P. A workshop experience for posthospitalized schizophrenics. In: Wright, G.N., and Trotter, A.B., eds. Rehabilitation Research. Madison, WI: University of Wisconsin, 1968. pp. 72-78. Weinman, B.; Sanders, R.; Kleiner, R.; and Wilson, S. Community-based treatment of the chronic psychotic. Community Mental Health Journal,

6:13-21, 1970. Weinstein, M.C., and Stason, W.B. Foundation for cost-effectiveness analysis for health and medical practices. New England Journal of Medicine,

296:716-721, 1977. Weisbrod, B.A. A guide to benefitcost analysis, as seen through a controlled experiment in treating the mentally ill. Journal of Health Politics, Policy, and Law, 7:805-845, 1983. Weisbrod, B.A.; Test, M.A.; and Stein, L.I. Alternative to mental hospital treatment. U: Economic costbenefit analysis. Archives of General Psychiatry, 37:400-404, 1980. Weissman, M.M., and Bothwell, S. Assessment of social adjustment by patient self-report. Archives of General Psychiatry, 33:1111-1115, 1976. Weissman, M.M.; Bruce, M.L.; Leaf, P.J.; Florio, L.P.; and Holzer, C.E. M. Affective disorders. In: Robins, L., and Regier, D.A., eds. Psychiatric Disorders in America. New York: The Free Press, 1991. pp. 53-80. Wells, K.B. Major Depression as a Tracer Condition for the National

of Medical Care Outcomes: Background Review. Santa Monica, Study

CA: The RAND Corporation, 1985. Wells, K.B., and Brook, R.H. Historical trends in quality assurance for mental health services. In: Stricker, C., and Rodriguez, A., eds. Hand-

book of Quality Assurance in Mental Health: A Comprehensive Overview.

New York: Plenum Press, 1988. pp. 39-63. Wells, K.B., and Brook, R.H. The quality of mental health services: Past, present, and future. In: Taube, C.A.; Mechanic, D.; and Hohmann, A.A., eds. The Future of Mental Health Services Research. Washington, DC: Superintendent of Documents, U.S. Government Printing Office, DHHS Publication No. (ADM)89-1600, 1989. pp. 203-224. Wells, K.B.; Goldberg, G.; Brook, R.H.; and Leake, B. Management of patients on psychotropic drugs in primary care clinics. Medical Care, 26:645-656, 1988a. Wells, K.B.; Golding, J.M.; and Burnarn, M.A. Psychiatric disorder in a sample of the general population with and without chronic medical conditions. American Journal of Psychiatry, 145:976-981. 1988b.

Wells, K.B.; Hays, R.; Burnam, M.A.; Rogers, W.H.; Greenfield, S.H.; and Ware, J.E., Jr. Detection of depressive disorders in prepaid and fee-for-service health care settings: Results from the Medical Outcomes Study. Journal of the American Medical Association, 262:298-302, 1989a.

Wells, K.B.; Hough, R.L.; Golding, J.M.; Burnam, M.A.; and Karno, M. Which Mexican-Americans underutilize health services? American Journal of Psychiatry, 144:918-922, 1987. Wells, K.B.; Manning, W.G.; Duan, N.; Newhouse, J.P.; and Ware, J.E., Jr. Use of outpatient mental health

626

SCHIZOPHRENIA BULLETIN

services by a general population with insurance coverage. Hospital and Community Psychiatry, 37:11191125, 1986. Wells, K.B.; Stewart, A.; Hays, R.D.; Burnam, M.A.; Rogers, W.; Daniels, M.; Berry, S.; Greenfield, S.; and Ware, J.E., Jr. The functioning and well-being of depressed patients: Results from the Medical Outcomes Study. Journal of the American Medical Association, 262:914-919, 1989b. Westermeyer, J. Psychiatric diagnosis aeross cultural boundaries. American Journal of Psychiatry, 142:798-805, 1985. Westermeyer, J. DSM -III psychiatric disorders among refugees in the United States: A point prevalence study. American Journal of Psychiatry, 145:197-202, 1988. Whitehead, C. Sheltered workshops in the decade ahead: Work and wages or welfare? In: Bellamy, G.T.;

O'Connor, C.; and Karen, O.C., eds. Vocational Rehabilitation

of Severely

Handicapped Individuals. Baltimore, MD: University Park Press, 1979. pp. 71-84.

Yess, J. What families of the mentally ill want. Community Support Services Journal, 2:1-3, 1981.

Wilkinson, C. Mental Health Prac-

Zarin, D.A., and Pass, T.M. Lithium and the single episode: When to begin long-term prophylaxis for bipolar disorders. Medical Care, 25(Suppl.): S76-S84, 1987.

tices in Primary Care Settings: An Annotated Bibliography. London: Tavistock Publications, 1985. Wilier, B., and Biggin, P. Comparison of rehospitalized and non-rehospitalized psychiatric patients on community adjustment: Self-assessment guide. Psychiatry, 39:239-244, 1976. Williams, D.H. The epidemiology of mental illness in Afro-Americans. Hospital and Community Psychiatry, 37:42-49, 1986.

Zastowny, T.R., and Lehman, A.F. Patient satisfaction with mental health services: New thoughts and links with satisfaction with medical care. Quality Review Bulletin, 14:284-289, 1988. Zimmerman, M.; Coryell, W.; Corenthai, C.; and Wilson, S. A self-report scale to diagnose major depressive disorders. Archives of General Psychiatry, 43:1076-1081, 1986.

Williams, P.; Williams, W.; Soinmer, R.; and Sommer, B. A survey of the California Alliance for the Mentally Ill. Hospital and Community Psychiatry, 7:253-256, 1986.

Zubin, J. Negative symptoms: Are they indigenous to schizophrenia? Schizophrenia Bulletin, 11:461-470, 1985.

Yates, A. Current status and future directions of research on the American Indian child. American Journal of Psychiatry, 144:1135-1142, 1987.

Zubin, 7., and Spring, B. Vulnerability: A new view of schizophrenia. Journal of Abnormal Psychology, 86:103-126, 1977.