University of Southern California Dissertations and Theses

A longitudinal examination of gender and ethnicity across two models of community treatment for individuals with schizophrenia and related disorders

(USC Thesis Other)

A longitudinal examination of gender and ethnicity across two models of community treatment for individuals with schizophrenia and related disorders

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content A LONGITUDINAL EXAMINATION OF GENDER AND ETHNICITY
ACROSS TWO MODELS OF COMMUNITY TREATMENT FOR
INDIVIDUALS WITH SCHIZOPHRENIA AND RELATED DISORDERS
by
Elizabeth Stoll Phillips
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the Requirements for the Degree
DOCTOR OF PHILOSOPHY
(SOCIAL WORK)
May 2001
Copyright 2001 Elizabeth Stoll Phillips
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI Number: 3093420
UMI
UMI Microform 3093420
Copyright 2003 by ProQuest Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
ProQuest Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UNIVERSITY OF SOUTHERN CALIFORNIA
T h e G ra d u a te School
U n iv e rsity P ark
LOS ANGELES, CALIFORNIA 90089-1695
This dissertation, w ritten b y
E :li'Z & h eP k> , S^olI 'Phillips
Under th e direction o f h. D issertation
Com m ittee, and a pproved b y a ll its m embers,
has been p resen ted to an d accepted b y The
Graduate School, in p a rtia l fulfillm ent o f
requirem ents fo r th e degree o f
DOCTOR OF PHILOSOPHY
Dean o f Graduate Studies
August 7,2001
DISSER TA TION COMMITTEE
Chairperson
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Elizabeth Stoll Phillips John Brekke, Ph.D
Abstract
A Longitudinal Examination of Gender and Ethnicity Across Two Models of
Community Treatment for Individuals with Schizophrenia and Related Disorders
Objective: Gender and ethnicity were examined for their relationships to
prospective outcomes from two models of community-based treatment for
individuals with schizophrenia. Methods: Data on 164 individuals diagnosed with
schizophrenia or schizoaffective disorder was collected every six months for a
period of three years. The treatment models included psychosocial rehabilitation
(PSR) and case management (CM). Previous studies had established that certain
psychosocial interventions are effective in improving various outcomes. However,
the wide variation in the outcomes of these interventions suggests the presence of
individual factors which may account for these differences. A preliminary study
explored three ethnic groups (whites, African Americans, and Latinos) for their
relationship to prospective functional outcomes. Using data from the PSR
program, the study found ethnicity to be related to the outcomes “work” and
“socialization.” Whites appeared to benefit from this type of intervention, showing
gains on both measures, while Latinos, in the area of work, showed a dramatic
decline. The current study included gender, and expanded the analyses to include
subjective and clinical outcomes. Furthermore, because the data came from two
distinct treatment approaches, the study examined prospective program by gender
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
and ethnicity interactions. Results: The results using hierarchical linear modeling
revealed differences between males and females on subjective and clinical
measures, as well as gender by program interactions. Most notably, females in
CM become more symptomatic as measured by the BPRS, while those in PSR
improved. Males in PSR, on the other hand, became more symptomatic, but CM
males showed minimal change. With regard to ethnicity, minorities in CM showed
deterioration in work scores. Whites in PSR improved in self-esteem, while
Latinos declined. However, Latinos demonstrated improvement in subjective
levels of distress. Conclusions: Based on these findings, it is evident that there
are important variations among gender and ethnic groups in their responses to
different models of community care. Implications are discussed in terms of the
need for gender-specific and culturally-relevant services.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TABLE OF CONTENTS
ii
List of Tables iii
List of Figures iv
INTRODUCTION 1
Conceptual Definition of Psychosocial rehabilitation 2
Conceptual Definition of Case Management 3
LITERATURE REVIEW 7
Gender Studies 15
Ethnicity Studies 29
Preliminary Study 48
METHODS 54
PSR Programs 54
CM Program 55
Sample 56
Measures 58
Data Analysis 63
RESULTS 66
Gender Findings 75
Ethnicity Findings 96
Grand Summary of Findings by Program Type 103
DISCUSSION AND IMPLICATIONS. 109
REFERENCES 131
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
List of Tables
Table 1 PSR: Comparison of Gender Samples at Baseline 68
Table 2 PSR: Comparison of Ethnicity Samples at Baseline 68
Table 3 PSR: Comparison of Bi-Ethnicity Samples at Baseline 69
Table 4 CM: Comparison of Gender Samples at Baseline 69
Table 5 CM: Comparison of Ethnicity Samples at Baseline 70
Table 6 PSR: Comparison of Bi-Ethnicity Samples at Baseline 70
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
List of Figures
Figure 1 Work, PSR: Tri-ethnic Comparisons Over Time 52
Figure 2 Socialization, PSR: Tri-ethnic Comparisons 53
Figure 3 ISE, CM: Gender Comparisons Over Time 82
Figure 4 BSI, CM: Gender Comparisons Over Time 83
Figure 5 BPRS, CM: Gender Comparisons Over Time 84
Figure 6 BPRS, PSR: Gender Comparisons Over Time 85
Figure 7 BPRS (Disorganization) Gender Comparisons 86
Figure 8 BPRS (Anergia) Gender Comparisons 87
Figure 9 BSI, Gender by Program Interaction 88
Figure 10 BSI, Females Only Over time 89
Figure 11 BPRS, Gender by Program Interaction 90
Figure 12 BPRS, Males Only Over Time 91
Figure 13 BPRS Gender/ Program Interaction 92
Figure 14 BPRS (Disorganization), Males Only Over Time 93
Figure 15 Work, CM: Bi-ethnic Comparisons Over Time 100
Figure 16 BSI, PSR: Tri-ethnic Comparisons Over Time 101
Figure 17 ISE, PSR: Tri-ethnic Comparisons Over Time 102
Figure 18 Independent Living, Long-term Sample 108
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1
A longitudinal examination of gender and ethnicity across two models of
community treatment for individuals with schizophrenia and related disorders.
INTRODUCTION
While studies on community-based psychosocial rehabilitation programs
for persons with schizophrenia consistently show that certain psychosocial
interventions can improve various outcomes when compared to usual treatment,
there have been wide individual differences in the outcomes of these effective
interventions. Recent recommendations in the psychosocial rehabilitation
literature have been to make greater efforts to individualize treatment, match
consumers to treatment, improve our ability to predict consumer responsiveness to
specific interventions, and to develop new rehabilitative strategies based on
individual characteristics and needs (Attkisson, 1992; Mueser, 1998; Penn, 1996;
Scott, 1995). The broad purpose of this study is to contribute to an understanding
of individual characteristics that are related to outcomes over time from
community-based interventions. Furthermore, because the data come from two
modalities, psychosocial rehabilitation (PSR) and case-management (CM), it is
hoped that this study will begin to answer the question, “which type of
intervention works best for whom?”
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Specifically, this study will look at ethnicity (African American, Latino,
and white) and gender to see if there is a relationship between these variables and
functional, subjective, and clinical outcomes over time in two different models of
community care. The functional domains of interest are work, independent living,
hospitalizations, and socialization. The subjective variables include satisfaction
with life, self-esteem, and level of distress. The clinical outcome variables include
symptoms as measured by the Brief Psychiatric Symptom Inventory (Overall,
1962). The interventions were provided by three community-based programs in
Los Angeles (Brekke, 1997). Two provide intensive rehabilitative services; for the
purposes of this study, they will be grouped together under the psychosocial
rehabilitation model. The third is a case-management program operated by the
Los Angeles County Department of Mental Health.
Conceptual definition of psychosocial rehabilitation
According to Holland, Zipple, and Bat&cha (1993), “the mission of psychosocial
rehabilitation programs is to assist individuals disabled by mental illness to be
successful and satisfied in the environments of their choice with the least amount
of assistance from the helping professions” (p. 36). Psychosocial rehabilitation
centers typically offer a clubhouse milieu, which is open to members 5 days a
week, with regularly scheduled groups that focus on improving members’ level of
functioning. These treatments are aimed at enhancing interpersonal and role
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3
functioning, promoting reduced hospitalizations and independent living,
decreasing symptom severity, and improving illness management (Moore, 1990).
Conceptual definition of case-management
Central to case management is the function of linking clients with severe mental
illnesses with essential resources, and empowering clients to function as
independently as possible in securing the resources they need. In the wake of
deinstitutionalization, case management was developed to address patients’
problems in navigating the complex mental health system (Moore, 1990). The
traditional case-management model, or brokerage model, consists of specific
functions. These include assessment, linkage to services, monitoring, and
advocacy. The major emphasis of this approach is on assessing needs, referring to
appropriate resources, coordinating services, and monitoring treatment (Mueser,
1998). Typically, case managers have caseloads of thirty or more, their work is
office-based, and they visit their clients at their places of residence once every 1 to
3 months. There is little or no direct provision of services (e. g. therapy or social
skills training) provided by the case managers.
In the parent study (Brekke et al., 1997), the participants in the case
management group comprised the comparison sample. It was hypothesized that
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4
programs with more intense services, which targeted the specific functional
domains of hospitalization, work, independent living, and social skills, would be
associated with better outcomes. Not taking into account the individual factors of
gender and ethnicity, the psychosocial rehabilitation programs, in general, out
performed case management in all four areas. Of course not everyone who
entered the psychosocial rehabilitation programs succeeded, nor did all who
received case management do poorly. This study asks the question: are there
systematic differences among ethnic groups, or between males and females that
are associated with positive or negative outcomes over time? And if so, do these
individual factors interact with the type of treatment they receive, either
psychosocial rehabilitation or case management?
This study proceeds from a study by Phillips, Brekke, and Barrio (in press)
that looked at ethnicity in the functional domains of work, socialization,
hospitalization, and independent living. An examination of the individuals who
participated in the psychosocial rehabilitation model revealed striking differences
in the areas of work and socialization over a three-year period. In terms of work,
whites showed steady improvement over time, while African Americans showed
virtually no improvement. A dramatic change occurred among Latinos, who
entered psychosocial rehabilitation on par with their peers. They appeared to
decline in functioning over time as their work scores steadily dropped. In terms of
socialization, whites showed the most improvement, followed by Latinos. African
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5
Americans showed no improvement over time. It was suggested that the
minorities’ failure to thrive in the PSR milieu was due to Western mental health
professionals imposing individualistic values on people with a collectivist
orientation. This could be perceived as intrusive and stress-inducing, ultimately
undermining the healing influence of family and society. This study provides the
opportunity to compare PSR to the CM modality, which imposes minimal contact
in its effort to provide services.
In an effort to paint a richer view of the complexities of individual
differences, this project has expanded the outcome measures to include subjective
and clinical variables. It has been the assumption in psychosocial rehabilitation
programs that functional change is related to changes in one’s subjective
experience (e.g. living independently correlates with improved self-esteem).
However, in a recent investigation, Brekke and Long (2000) explored the
relationships among prospective functional, subjective and clinical outcomes over
three years of community-based treatment. The authors found that change in
subjective experience was only modestly associated with change in the other two
domains. And change in clinical outcomes was distinct from change in functional
outcomes. These findings imply that studies that measure one-dimensional
outcomes ignore essential components that comprise the human experience.
The current study also includes gender as an independent variable. This
follows years of schizophrenia research which have revealed important differences
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6
between men and women in a variety of areas that potentially impact their lives in
the community.
The specific aim of this research is to answer the following questions:
1) Is gender related to prospective functional, subjective, and clinical outcomes
from community-based psychosocial rehabilitation of persons with schizophrenia?
2) Is gender related to prospective functional, subjective, and clinical outcomes
from community-based case management of persons with schizophrenia?
3) Is ethnicity related to prospective functional, subjective, and clinical outcomes
from community-based psychosocial rehabilitation of persons with schizophrenia?
4) Is ethnicity related to prospective functional, subjective, and clinical outcomes
from community-based case management of persons with schizophrenia?
5) Does ethnicity interact with type of treatment model to influence prospective
functional, subjective, and clinical outcomes?
6) Does gender interact with type of treatment model to influence prospective
functional, subjective, and clinical outcomes?
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7
LITERATURE REVIEW
Community-based psychosocial treatment approaches have been
widely studied in recent years, although studies that have examined differences
with relation to ethnicity and gender are rare (Test, 1990; Barrio, 2000; Phillips,
Brekke, and Barrio, [in press]; Haas, 1990; Jerrell, 1997; Telles, 1995; Vandiver,
1998; Watkins, 1999). This is in spite of the fact that research has revealed
differences among these groups in a variety of domains directly related to the
disorder of schizophrenia (Hafner, 1993; Childers, 1990; Haas, Glick, Clarkin et
al., 1990; Mueser, Bellack, Randall, Morrison, & Wade, 1990; Alverson &
Vicente, 1998; Fabrega, Mezzich, & Ulrich, 1998; Valesquez & Callahan, 1990).
These include, but are not limited to, studies of symptomatology, help-seeking
behaviors, biological factors, and responses to hospital treatment. In this age of
deinstitutionalization, tremendous efforts are underway to develop the most
effective models to care for and rehabilitate persons with chronic mental illnesses.
As part of this process, it is critical to take into account the diversity of this
population, and not allow the diagnosis of schizophrenia to supersede ethnic and
gender identity. To date, while there are few studies which have looked at
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8
differences with relation to individual factors, there are no community
rehabilitation studies which have examined gender and ethnicity in combination
with different types of programs. In lieu of this research, a few salient studies
regarding community rehabilitation are highlighted, followed by a comprehensive
review of the literature on gender and ethnicity.
Reviews of case management intervention studies found them to be
successful in reducing time in the hospital, but their impact on psychosocial
functioning was variable (Mueser, 1998; Solomon, 1992; Mueser & Bond, 2000;
Baronet & Gerber, 1998; Bond et al., 1995; Gorey et al., 1998; Latimer, 1999;
Mueser etal., 1998; Rapp, 1998). Solomon (1992) reviewed twenty empirical
articles and found: i) most were effective in reducing the number of
hospitalizations, and, in the instance of hospitalization, reducing the length of stay;
ii) case management appeared to be cost effective and improved clients’ quality of
life; iii) treatment satisfaction was consistently high; and, iv) while case
management appeared to have negligible positive effects in clinical functioning, it
did not appear to produce negative effects, and was as effective as hospitalization
with subsequent aftercare. Similarly, other investigations have also found case
management to decrease hospitalization (Mueser, 1998; Homstra et al., 1993;
Santos et al., 1993; Bond et al., 1988), but to have only modest effects on clients’
quality of life (Bond, 1988; Franklin et al., 1987).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9
Recent case management reviews have focused on innovative and well
regarded case management approaches, which include the strengths model,
intensive case management, and assertive community treatment. The case
managers work in the field, have small caseloads (6-12 clients), and are often
assisted by teams. For the purpose of clarity, it is important to make the distinction
between these interventions, and the case management intervention that was used
in this study, which followed a more traditional style. Briefly, in traditional case
management, the case managers have thirty or more clients, their work is office-
based, and they visit their clients at their places of residence once every 1 to 3
months. The only consistent finding in the studies of traditional case management
has been that it is effective in reducing hospitalizations (Mueser et al., 1998;
Mueser & Bond, 2000).
Following the study of Mueser et al. (1998) which examined 75 models of
community care, the authors contended, “a more important goal than evaluating
which model is best, is to determine which model is best for whom...” (p. 64)
Indeed, this reflects a growing concern among several writers in the field of
community treatment who refer to the need to identify the unique characteristics
of individuals with severe mental disorders, and match them with appropriate
treatment (Brekke and Long, 2000; Barrio, 2000; Penn and Mueser, 1996;
Lehman, 1995; Blank, Jodi, & McCall, 1996; Watkins, Shaner, and Sullivan,
1999; Scott and Dixon, 1995). For example Attkinsson and colleagues (1992)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
performed a comprehensive assessment of the effectiveness of clinical
interventions in community settings. In their report they debated priorities for
future research on clinical services for people with severe mental illnesses. Central
to their deliberation was the issue of how to design treatment plans to match an
individual’s needs for service. They identified culture and gender as variables that
might impact not only the course of illness, but also the definition of what is
considered a valuable outcome by the individuals themselves, their families, and
the community.
The mission of intensive community-based psychosocial intervention
programs is “to assist individuals disabled by mental illness to be successful and
satisfied in the environments of their choice with the least amount of assistance
from the helping professions” (Holland, Zipple, and Batscha, 1993, p. 36). These
rehabilitation programs provide consumers with opportunities to develop the
various skills they need in order to achieve their personal goals.
Fountain House, which opened in 1948, pioneered community-based
psychosocial intervention, and was the site of the earliest research projects. Two
controlled studies were performed in order to measure the influence of services on
rehospitalization. Study I followed 252 experimental subjects discharged from the
hospital and 81 control subjects for 9 years. Study II, a virtual replication of study
I, followed 40 experimental and 34 controls subjects for 5 years. The results
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
11
indicated that after two years, a significantly fewer number of individuals in the
experimental groups needed to be rehospitalized.
One of the earliest studies to compare community-based psychosocial
intervention with traditional case management services was performed by Stein
and Test (1980). Subjects, (n=130) were randomly assigned to either community-
based psychosocial rehabilitation (the experimental group), or to short-term
hospitalization followed by case management (the control group). For the
experimental group, most of the treatment was done in vivo. Staff members went
to patients’ homes, communities, and places of work, assisting them in daily living
activities. They were also aided in the development of effective social skills. The
duration of this intervention was approximately 12 to 14 months. After that time
they received the same services as those in the control group.
The authors (Stein et al., 1980) looked at several outcome domains: living
situation, employment status, social relationships and leisure time activities,
satisfaction with life and self esteem, symptomatology, and medication
compliance. Their experiment revealed the following: In terms of the
participants’ living situations, the psychosocial rehabilitation group spent
significantly more time than those receiving case management in independent
living. They were less likely to be unemployed, and earned significantly more
income through competitive employment than the controls. A measure of leisure
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12
time activities showed no differences; however, those in the experimental group
were more likely to have contact with “trusted friends” (p. 394).
The experimental subjects were significantly more satisfied with their life
situations than the controls. At baseline, the experimental subjects reported higher
self-esteem scores, although these differences were not maintained at subsequent
data collection periods. Finally, with regard to symptoms, the experimental group
at various points during the first 12 months were less symptomatic in the areas of
depressed mood, suicidal ideology, anxiety, motor agitation, paranoia, and
hyperactivity.
Interestingly, the data from the latter part of the experiment during which
the participants of the experimental group were offered the same services as
members of the control group, showed that the subjects were unable to sustain
their progress. Within a short time they resembled statistically the group that
comprised the controls. There was one notable exception to this. In each data
collection period following the cessation of intensive rehabilitative services, there
was a gradual but definite increase in hospitalizations by the experimental group,
whereas the usage by members of the control group remained stable. This study
supported the authors’ contention that community programming should be
comprehensive and on-going.
A more recent study which compared community-based psychosocial
rehabilitation centers with case management was the parent study for this
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13
investigation. Brekke et al. (1997), in a quasi-experimental, follow-along design,
looked at 172 individuals diagnosed with schizophrenia and related disorders.
Two programs that provided rehabilitative services were compared to case
management only, which served as their comparison group. The authors
contended that models that had more intensive services would be associated with
higher rates of improvement over time. In addition, they hypothesized that
programs that targeted specific functional domains, e.g., the attainment of
independent living skills, would show greater gains in those areas. Finally, they
proposed that change scores would be in a positive direction as long as
participants remained in the programs, and that scores would decline following
their departure.
The authors looked at four functional domains: hospitalizations, work,
independent living, and socialization. The results supported their first hypothesis.
The intense rehabilitation programs had higher rates of improvement in all areas,
outperforming case management. The results supported their second hypothesis.
The program that emphasized vocational training revealed higher work scores,
while the program that stressed the acquisition of independent living skills showed
greater improvement in rates of independent living. And third, like the Stein and
Test study (1980), their results showed that once individuals exited the treatment
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
14
programs, some important gains were rapidly lost. In this instance, they were in
the areas of work and independent living.
To summarize, studies on community-based psychosocial rehabilitation
programs for persons with schizophrenia have consistently yielded the following:
(i) certain psychosocial interventions can improve functional outcomes when
compared with usual treatment (Brekke, Long, Nesbitt, and Sobel, 1997; Stein and
Test, 1980; Mueser, Bond, Drake, and Resnick, 1998); and (ii) there is wide
individual variation in the outcomes of these effective interventions (Test, Burke,
and Wallisch,1990). This individual variation suggests that in order to increase
the performance of these models of care, it is important to understand the
individual factors associated with the differential outcomes from these services
(NIMH, 1991; Attkinsson, 1992). Concurring, Brekke et al. (2000), emphasized
that the direction of future community rehabilitation research should be toward
“understanding the individual factors related to variation in treatment
responsiveness, since this is closely related to improving the effectiveness of our
treatment plans” (p. 678).
While individual factors have received scant attention in the psychosocial
rehabilitation literature for schizophrenia, the broader field of mental health has
contributed several decades o f research on client factors in psychotherapy outcome
studies (Garfield, 1994,1995). Investigators have studied the attributes of social
class, personality, diagnosis, gender, age, ethnicity, intelligence, and length of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
illness. Unfortunately, it has been difficult to garner definitive knowledge from
much of this work due, in part, to the likelihood that salient client characteristics
vary with the type of intervention. According to Garfield, “...it is apparent that
making predictions prior to therapy about continuation or outcome that are based
solely on client variables are not as successful as we would like” (p. 220). He
states further that predictions that pay attention to the type of client, and the
therapy process are more successful.
Gender
Individual factors that have repeatedly demonstrated significant variability
with regard to schizophrenia include gender and ethnicity. The former has been
shown to be related to schizophrenia in several ways including: age of onset
(Hambrecht, 1992), premorbid functioning (WHO, 1979; Lewis, 1989; Jablensky,
1992; Childers, 1990; Angermeyer, 1990), and treatment response (Angermeyer,
1990; Pfeiffer, 1996). It has been suggested that these differences, which favor
females, are determined by biological factors such as estrogen (Hafher, 1993;
Childers, 1990), or cerebral asymmetry (Lieberman, 1996). Women also have
been shown to have a more benign course of illness, at the same time showing
greater levels of self-esteem, and significantly better psychosocial adaptation than
men (Farmer, 1997).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
16
Most recently it has been proposed that women may be predisposed to an
affective subtype of schizophrenia that is genetically related to the affective
disorders, while men are more prone to a neurodevelopmental subtype (Salem,
1998). The former carries a hypothesized good prognosis, while the latter, a poor
prognosis.
Pronounced gender differences have been observed in the expression of
symptoms as well. For example, studies have found self-neglect and social
withdrawal to be more prevalent in males (Hafner, 1998; Mueser, 1990). A study
of homeless men and women with co-occurring schizophrenia and substance abuse
revealed that women had greater rates of comorbid anxiety and depression
(Brunette, 1998). They also had more children, were more socially connected, and
experienced higher rates of sexual and physical victimization than the men in the
study. The authors concluded that the women in the homeless, dually-diagnosed
population have distinct characteristics, vulnerabilities, and treatment needs when
compared to men.
A related study (Vandiver, 1998) addressed gender as a predictor of quality
of life among 102 community residents with schizophrenia in three countries,
Canada, Cuba, and the United States. The domains of interest were social
relationships, health, living situation, leisure, and finances. In all but one domain,
no differences were found. However, in the area of social relationships, gender
distinctions were discovered that may have implications for how to provide
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
17
effective community interventions. In Canada, women reported a higher quality of
life for social relations compared to the men in the sample. Qualitative comments
frequently referred to the accessibility of health care professionals that is afforded
by the Canadian health care system. This does not explain, however, why they
were more satisfied than males, who enjoy the same access. In contrast, Cuban
men reported higher social relationship scores when compared to women. Again,
qualitative comments were gathered which indicated that males used the out
patient hospitals as their main source of socialization. Women, on the other hand,
were constrained by the multiple roles of caregiver, housewife, and worker.
Unfortunately, there was no indication in the article as to what model of
community treatment was being received by the individuals being studied, nor was
it possible to assess whether these differences were sustained over time.
Another community study was concerned with the role of gender in
engaging individuals with coexisting mental disorders and substance abuse in
treatment (Watkins, Shaner, & Sullivan, 1999). This was a qualitative study
which used a convenience sample of 11 females and 10 males from three different
settings. These included a Veterans Affairs outpatient dual diagnosis program, a
county mental health clinic, and a homeless shelter. The study addressed the
needs o f the individuals as well as the obstacles they perceived toward obtaining
services. With respect to needs, they were similar across both gender groups.
Virtually everyone was concerned with getting housing, food, clothing, and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18
money. The obstacles to seeking treatment for women had a common theme: they
feared being harmed, abused, or victimized by people in the mental health system.
When they denied needing help, they frequently explained that they were more
comfortable doing things alone. The men in the study were more apt to say that
lack of motivation was what kept them out of treatment. This study, while very
limited in its scope and generalizability, addressed a problem not normally
discussed in the schizophrenia literature. This is that severely mentally ill females
with co-occurring substance abuse issues are more likely to have been exposed to
sexual, physical, or emotional abuse as children than women without a substance
abuse problem, or men (Test & Berlin, 1989).
To date, the literature addressing gender as a predictor of community-
based psychosocial rehabilitation of persons with schizophrenia is scarce (Test,
1990). The majority of treatment studies which include gender as an independent
variable are hospital-based, and review patients following discharge. For example,
Smith et al. (1997) studied a cohort of acutely-ill hospitalized patients who
participated in a manualized community reintegration skills training program.
Subjects (23 males and 21 females) were recruited from consecutive inpatient
admissions over a 10-month period. Eighty-five percent were diagnosed with
schizophrenia or schizoaffective disorder. At study entry, there were no gender
differences in skill level, or symptomatology.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19
The intervention was a daily group therapy program aimed at helping
patients learn how to identify symptoms, manage their medications, and be active
in the planning of their treatment. The authors hypothesized that females would
demonstrate greater post-training skill levels, and stronger associations between
post-training skill and post-treatment adherence. Their hypothesis was not
supported. Instead, they found that males showed greater associations between
post training skill level and post-discharge treatment adherence. The females in
the study showed a significantly higher rate of good post-discharge treatment
adherence, but no association between post-treatment skill levels and functional
outcome. In other words, males showed signs of clear benefit from the training,
yet females showed better post-treatment functioning, irrespective of training. The
authors concluded that factors other than the skills targeted by specific
psychosocial treatment programs are stronger determinants of social adjustment in
females.
Haas, Glick, Clarkin et al. (1990) studied gender differences in
symptomatology and role functioning in a sample of 49 male and 43 female
hospitalized schizophrenia patients. Role functioning was operationally defined
by the Role Activities Performance Scale and covered the domains of work, social,
family, and leisure activities. Symptomatology was measured using the Global
Assessment Scale and the Psychiatric Evaluation Form, which provides ratings for
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
19 psychiatric symptoms. Subjects were randomly assigned to one of two groups:
multi-modal hospital treatment with family intervention or multimodal treatment
without family intervention. Multimodal treatment consisted of neuroleptic drug
therapy, individual and group psychotherapy, and milieu therapy which included
learning limit-setting and social skills. The family intervention was
psychoeducational. It addressed the nature of schizophrenia, adjusting
expectations for patients, and identifying stressors. In addition, it educated both
family and patient as to the need for the patient’s continued treatment after
discharge.
The results revealed no gender differences or gender by treatment
interactions in rates of rehospitalization at 18-month follow-up. However, there
were treatment by gender interactions with respect to clinical and functional
outcomes. For females, the addition of family intervention was associated with
superior outcome in both symptomatology and functioning. At discharge, 6-
month, and 12-month follow-up, there was a progressive worsening for males
(relative to females). Their scores indicated that they were more impaired in
terms of grandiosity, as well as disorganized, and inappropriate behavior. There
was more severe alcohol use among males, and greater disability in maintaining
personal hygiene when compared to females (Haas et al., 1990).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
21
With regard to medication compliance, there was a main effect of gender,
with better compliance found among females. Also of interest was the finding that
the family intervention treatment was associated with less critical attitudes among
the families of female patients; conversely, among males, treatment that did not
include family intervention was associated with less critical attitudes (Haas et al.,
1990).
The authors offered that differences between genders may be attributable
to the social and occupational demands placed on individuals by their families and
by the greater society. For example, there may be greater expectations for males
living in the community, regardless of their disabilities, to be competent in the
area of work. With regard to traditional sex-role socialization practices, women
are encouraged to be dependent. Therefore, they may more readily accept and
respond to a psychosocial treatment in which families play a role. In contrast,
males are discouraged from dependency on families and the greater society (Haas
et al., 1990).
Another hospital study that, unfortunately, did not operationalize the
hospitalization treatment the subject received, is valuable to this discussion
nonetheless. The authors (Mueser, Bellack, Randall, Morrison, & Wade, 1990)
examined the relationship between gender and social skill in a sample of 57
schizophrenia patients, 33 affective disorder patients, and 20 non-patient controls.
The patients were admitted to the hospital for the treatment of symptom
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
22
exacerbation, and consented to participate in social skills assessment. The
investigators rated social skills among the three diagnostic groups (schizophrenia,
affective disorder, and controls). In addition, they performed longitudinal analyses
of social skills, symptoms, and social adjustment for the schizophrenia patients.
Social skill was evaluated using a role play test and included
measurements of nonverbal and verbal content skills: The nonverbal skills
included a) Gaze (the appropriateness of eye contact), b) Length (the duration of
speech), c) Meshing (the smoothness of turn-taking and conversational pauses),
and d) Affect (the subject’s tone of voice and facial expression). The verbal
content skills included a) Positive Assertion (the ability of subjects to express
praise and appreciation), b) Negative Assertion (the ability of subjects to
appropriately express dissatisfaction and resist hostility), and c) Overall Social
Skill. This is a rating of the subject’s overall effectiveness as a communicator.
The authors (Mueser et al., 1990) explored a diagnosis by gender
interaction by comparing the social skills of men and women in each group. They
found that female schizophrenia patients were more skilled than the male
schizophrenia patients; no gender differences were found among those with
schizoaffective disorder, or the controls. Within the sample of schizophrenia
patients, women were more skilled than men on length, meshing, the ability to
express praise and appreciation, and their overall social skills.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23
Analyses were conducted one year later using a repeated measures
ANOVA. There was a main effect for gender, with women having higher overall
social skills than men at both time periods. There was no change in skill over
time, and no gender by time interaction. There was also a significant difference in
affect, with women having better affective skills. In the area of verbal skills,
women were significantly more likely to possess the negative assertion skills
(Mueser et al., 1990).
The authors argue that gender differences cannot be accounted for solely
by social role expectations, given that both genders were assessed using the same
behavioral criteria (Mueser et al., 1990). They suggested that although women
with schizophrenia may benefit from less social stigma than men, and may enjoy
the advantage of greater social tolerance, these advantages may be, in part,
influenced by their own social skillfulness.
Gift, Harder, Ritzier, and Kokes (1985), noting that white males enjoy
several advantages in the general society, set out to determine if they were
similarly advantaged in a psychiatric population. This hospitalized sample
included 217 ethnically heterogeneous men and women. They were examined at
admission on eight variables, including quantity and quality of work, number of
social contacts, quality o f social relationships, and duration and level of psychotic
symptoms. After two years, they were re-examined, with the following outcome
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
24
variables added: time spent out of the hospital, ability to meet basic needs,
fullness of life, and overall level of function.
The authors (Gift et al., 1985) found few associations, suggesting that in
their population of patients hospitalized for functional psychiatric illness, sex and
race were, overall, not consistently related to psychiatric symptoms, disability, or
outcome. At baseline, females showed higher levels of neurotic symptoms than
males, but these differences disappeared at two-year follow-up. There was a sex
by race interaction when looking at psychotic symptoms showing Black males to
be more symptomatic than white males; these differences also were not sustained
over the two years. However, being female seemed to be advantageous in terms of
having fewer problems with employment two years after release from the hospital.
Unfortunately, the study did not inform the reader of the nature of the
hospital treatment. Nor did it determine what, if any, interventions ensued
following the hospitalizations. Furthermore, because the data were collected at
only two time points, baseline and two years, we have no way to determine if non
linear change occurred, and what the process of that change could have been.
Another flaw in the investigation was its failure to adequately delineate the ethnic
groupings. The categories consisted of whites, blacks, Orientals, and “other.” It
was into the latter category that Latinos were placed. Neither did the study reveal
cell sizes for any of the grouping variables. Given the size of the total sample, it is
possible that statistical power was not strong enough to reveal true differences.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
25
While the preceding hospitalization studies can be enlightening, they do
not address the fact that in response to recent fiscal, sociopolitical and
philosophical trends, more and more people with serious mental disorders are
receiving the greater part of their services as outpatients in conjunction with case
management, or are receiving long-term treatment as members in rehabilitative
programs (Pfeiffer, 1996).
Test et al. (1990) looked at 122 young adults (ages 18-30) with
schizophrenic disorders in various community life domains. The subjects were
randomly assigned to one of two treatment programs, each of which were
community-based with goals toward rehabilitation: the Training in Community
Living (TCL), or the usual system of care in Dane County. The TCL program is a
community support program that delivers biopsychosocial treatments to their
clients using assertive outreach and in vivo support and coaching. The study was
descriptive in nature, and focused on the first two years of a five year study.
Although they provided two treatment settings, treatment by gender interactions
have not been examined to date.
Two years into their five-year study, they revealed that more women than
men lived independently in apartments, had opposite-sex contact, and were
involved in child rearing. Men were more likely to live in rooming houses. They
also showed greater substance use, arrest rates, and spent more time in jail. There
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
26
were no significant differences in hospitalization rates, nor were there differences
in the number of friends the participants had contact with during the study period.
With regard to the gender differences in independent living, the authors
offered the following reasons for consideration: 1) Men are more often than
women without a residential partner, and thus may find rooms to be the only
source of affordable housing. 2) Men are more likely to be more gravely ill than
women, and therefore have less income. 3) Landlords may be more fearful of
renting to mentally-ill males. And finally, 4) staff members who assist clients in
finding housing may be operating from the assumption that “males don’t care as
much as women about where they live” (p. 342).
Summary of the Gender Literature
Gender has been shown to be related to schizophrenia in a variety of ways.
These differences extend from physiological and biological components, to
behavioral manifestations. In addition, gender has been a predictor of quality of
life, symptom expression, treatment engagement, and treatment response. This
cumulative evidence of differences validates the contention that gender, as an
individual factor, warrants attention in the field of community-based interventions.
The following summary is intended to highlight two main issues that have
treatment implications: 1.) Men and women have distinct characteristics that may
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
27
serve to impede or enhance their responses to treatment. 2.) Men and women have
distinct needs that require different treatment approaches.
In the study of dually diagnosed homeless men and women in the
community, it was revealed that women were more likely to suffer from affective
symptoms. They had more children, were more socially connected, and were
more often the victims of violent crimes. In terms of quality of life, Canadian
women reported that they were more satisfied than males in their social
relationships. In Cuba, males were more satisfied. The authors asserted that in
both instances, this difference could be related to the accessibility of their health
care facilities, which provided their main source of socialization. Cuban women,
constrained by the roles of caregiver, housewife, and worker, were less likely to
avail themselves of this opportunity.
With respect to the help-seeking behavior of dually-diagnosed individuals
in the community, Watkins et al.’s study (1999) showed that both genders were
concerned with obtaining food, clothing, housing and money. However, the men’s
involvement in treatment was impeded by lack of motivation, while the women’s
involvement was hindered by lack of trust.
Smith et al.’s study (1997) involved hospitalized patients who participated
in a community reintegration skills program. The results indicated that the men
showed a logical progression from learning the skills (medication management,
symptom identification, and participation in treatment planning), to actually
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
28
utilizing them. Women, on the other hand, were more effective in utilizing the
skills, but their skill level was not related to how well they did in the training
program. The authors concluded that factors other than the skills targeted by
specific psychosocial treatment programs are stronger determinants of community
adjustment in females.
Mueser et al. (1990) made the argument that women with schizophrenia
benefit from social role expectations. They are less often stigmatized for their
illness, and benefit from greater social tolerance. However, after evaluating men’s
and women’s verbal and non-verbal skills following hospital treatment, they also
asserted that women’s advantages in society are mitigated by their own social
skillfulness.
Another hospital study investigated the effectiveness of two treatment
groups: multimodal with family intervention, and multimodal without family
intervention (Haas et al., 1990). Irrespective of which treatment modality one was
enrolled in, the males in the study progressively became more symptomatic
following discharge. In addition, they were more likely to abuse alcohol, exhibit
anti-social behaviors, and be non-medication compliant. Treatment by gender
interactions revealed that the addition of family intervention was associated with
superior clinical outcomes for females, whereas males appeared worse off
following this type of intervention.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29
Finally, gender differences were found in the study of young men and
women in two community intervention settings (Test & Wallisch, 1990). Two
years into their participation in psychosocial rehabilitation, differences were found
in the types of housing they procured, involvement with the legal system, time
spent in child rearing practices, and amount of opposite-sex contact. The
differences between the two treatment conditions were distinct; unfortunately,
treatment by gender interactions were not examined.
Given the large number of gender differences found in the preceding
studies, it is clear that research on community interventions needs to include
gender as a key variable in its analyses. It is also clear that the analyses should not
stop there. Test and Wallisch stated, “...the fact that the lives of young men and
women in the community differ significantly holds implications for more basic
research investigating gender differences in the longer term outcome of
schizophrenia” (p. 342). It is the next step of basic research to abandon the one-
size-fits-all treatment approach and search for matchings of consumer-to-treatment
that effectively enhance the lives of all individuals seeking community care.
Ethnicity
Ethnicity and Culture Defined
The term “culture” was traditionally used by anthropologists in a restricted
tribal-traditional sense where mental illness is described in pre-literate societies
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30
(Al-Issa, 1995). It has been described as “a complex whole that includes
knowledge, belief, art, law, morals, and custom,” and “a set of rules that, when
acted upon by members of a society, produces behavior that is considered proper
and acceptable” (Searles, 1993, p. 21). In its most frequent use in the cross-
cultural study of mental illness, culture is equated with nation (Al-Issa, 1995). It is
assumed that each nation of the world has its own distinct culture, and that cultural
variations within a nation are less in scope than those that exist between nations.
In multicultural nations, such as the United States, the concept of culture is used in
the ethnic sense where the nation is divided into ethnic groups defined on the basis
of racial-physical characteristics, shared language, common historic origins, and
shared contemporary customs. Ethnic groups, within a nation, are “individuals
with a sense of belongingness and are thought by others and themselves to share a
common origin, as well as an important segment of a common culture” (p. 4).
Central to the discussion of ethnicity is the manner in which members of
different cultures perceive themselves in relation to each other and the larger
society. Scholars in the field of African American, Latino, and Asian American
studies frequently have asserted that ethnic minority groups tend to be higher in
collectivism and familism than are European Americans, who in turn tend to be
higher in individualism (Gaines, Bledsoe, Lyde, Rios, Garcia, & Page, 1997).
Collectivism refers to a tendency to place the needs of one’s in-group above one’s
own needs, while individualism is defined as a tendency to place one’s own needs
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
31
above the needs of one’s in-group. Familism can be subsumed under collectivism,
and is defined as “an orientation toward the welfare of one’s immediate and
extended family” (p. 1641).
According to Phinney (1996), African Americans reflect the continuing
African influence in contemporary Black lifestyles that results in characteristics
such as emotional vitality, collective survival, oral traditions, and interdependence,
particularly within the extended family. Latinos are characterized as being
interdependent, conformative, and willing to sacrifice for the welfare of in-group
members. They are also likely to show strong attachment and loyalty to their
extended family, and be obedient to those in authority.
These differences are reflected in the field of mental health, where many
variations exist across cultures. These can be found in family practices, beliefs
about the causes of mental illness, how mental illness is conceptualized, as well as
the stigma attached to mental illness (Barrio, 2000). To illustrate, relatively
unacculturated Mexican Americans often use the term “nervios” to refer to a wide
range of mental disorders. It is believed that preference for the use of this term is
linked to family efforts to reduce the stigma associated with schizophrenia
(Dassori, Miller, & Saldana, 1995). Another example can be found in how
different American cultural groups utilize the family as caregivers. Minority
families tend to have social networks that include more relatives than those of
European American families, and are more likely to go to those family members
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
32
for advice. European Americans, on the other hand, are more inclined to turn to
mental health professionals. Guamaccia and Parra (1996) contended that this was
related to differences in how the illness was conceptualized. European Americans
are likely to view illness as a medical problem, whereas African Americans and
Latinos attribute it to a wider range of causes and seek help in other areas, e.g.
religious advisors.
Barrio (2000) asserted that the strengths inherent in collectivist and
familistic societies are not cultivated in the community treatment of individuals
with mental illness. Rather, it appears that Western European values are being
imposed across cultural groups. She argues that the attempt to achieve
independence and self-sufficiency may conflict with cultural norms and be
countertherapeutic. On the other hand, she stated that if professionals would
“work from the perspective of the client’s cultural orientation, they (could)
coordinate services to support values of interdependence, validate sociocentric
behaviors, and instill hope, while providing individualized psychoeducational
services and respite care” (p. 882).
Ethnicity and Mental Health Research
As with the gender studies, there have been only a few published articles
on ethnicity and psychosocial rehabilitation (Brekke & Barrio, 1997; Telles et al.,
1995; Jerrell & Wilson, 1996), but differences have been found in other arenas
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
33
that are notable. For example, two major multinational studies sponsored by the
World Health Organization (WHO) including the International Pilot Study of
Schizophrenia (IPSS) and the Determinants of Outcome of Severe Mental
Disorders (DOSMD), have revealed that, while the incidence of a core
schizophrenia syndrome is similar across diverse cultural settings, there is a
distinct difference in the course of the illness. Regardless of the nature of onset
and presenting symptoms, there is a marked predominance of favorable outcomes
in the centers of developing counties as opposed to the developed nations of
Europe and North America.
Some problems exist with these studies, however. Murphy (1980) pointed
out that the IPSS study which had shown that the most technologically developed
nations which yielded the poorest results also included patients from these nations
who had a longer average history of schizophrenia than those patients in the
developing countries. In spite of this, and other methodological problems (Lin &
Kleinman, 1988), the consensus among most cross-cultural researchers is that
traditional cultures favor symptom recovery and the restoration of function.
(Brekke & Barrio, 1997; Kamo & Jenkins, 1993; Sartorius, Jablensky, Korten,
Emberg, Anker, Cooper, & Day, 1986). Furthermore, according to Dassori,
Miller, & Saldana, (1995) the powerful influence of environmental factors implies
the need for research into their nature, and their ability to be modified for the
purpose of therapeutic intervention.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
34
In the United States, cross-ethnic studies have focused primarily on the
expression of symptoms. Fabrega, Mezzich, and Ulrich (1988) examined
differences in psychopathology between African Americans and whites and found
that the latter were more impaired on items that rated emotional distance and flat
affect, as well as on items that rated depressed mood and low self-esteem. Chu,
Sallach, Zacheria, and Klein (1985) revealed that African Americans are more
likely to exhibit anger, disorientation, asocial behavior, and hallucinations, while
whites experience more symptoms of irrelevant speech and delusions.
Tri-ethnic studies comparing whites, Latinos and African Americans have
shown higher levels of paranoia and hysteria for African Americans (Valesquez &
Callahan, 1990), and higher levels of depression and hypochondriasis, or
somatization, for Hispanics (Escobar, 1987; Canino, Rubio Stipec, Canino, &
Escobar, 1992). Recently, Brekke and Barrio (1997) determined that whites had a
more symptomatic profile than African Americans and Latinos across the
following 7 variables: overall symptom severity, asociality, disorganization,
anhedonia, bizarre behavior, tension, and mannerisms. This finding supported
their hypothesis that ethnic minorities come from a more sociocentric culture,
which has a mediating effect on presenting symptoms.
Studies that address the treatment of patients with severe mental disorders
have uncovered noteworthy differences as well. For example, it was found that
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
35
minority patients were significantly more likely to be prescribed higher doses of
antipsychotic medications than their non-minority counterparts (Lehman &
Steinwachs, 1998), and were less likely to be offered psychotherapy as a treatment
option (Guamaccia & Parra, 1996). Correspondingly, a study that looked at
admission rates to various hospital settings found marked ethnic-related
differences (Snowden & Cheung, 1990). The African Americans in their sample
were more likely to overutilize hospital services, while those of Hispanic origin
tended to underutilize services when compared to whites. These findings, as they
relate to African Americans, are in sharp contrast to the findings of Farmer and
Pandurangi (1997).
The investigators (Farmer et al., 1997) placed 42 people with
schizophrenia into two groups, those with a larger ventricle-to-brain ratio or
cortico atrophy or both, and people without those conditions. The authors
expected to find that the former would have lower levels of psychosocial
adaptation, higher levels of stress, and lower levels of self-esteem than the latter.
Instead they found significant differences based on ethnic status. African
Americans were significantly older at age of illness onset, first hospitalization, and
first use of medication, which typically predicts a more benign course of the
illness. The authors proposed that the differences could be explained by culture
which, for African Americans, is family-oriented with an emphasis placed on
religion and spirituality. They asserted that African Americans, when compared to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
whites, are more likely to take care of their family members, and provide more
highly-developed social support networks.
The notion that culture affects psychopathology is certainly not new in the
research literature. Fabrega et al. (1968) wrote of a “cultural effect” wherein
cultural variables affect behavior structurally in a direct way through family
dynamics, socialization patterns, and values. They cautioned, however, that
reported clinical differences may reflect intervening variables, most problematic
being socioeconomic status, and the heterogeneity of diagnostic groups.
Expressed Emotions and Outcome
Many studies have confirmed the conclusion that measures of critical
comments of relatives and the degree of hostility and emotional involvement are
relevant to the relapse of patients with schizophrenia (Al-Issa, 1995; Kamo et al.,
1987). It has been proposed that cultural variations in the level of expressed
emotion (EE) in different households may be related to the degree of
Westernization. For example, while the EE level in households is found to be
67% in Los Angeles, (Vaughn, Snyder, Jones, Freeman, & Falloon, 1984), the
Mexican American level was 41% (Kamo et al., 1987). A couple of things may
account for this. First, it has been proposed tha'i cultural variation may be related
to the extent to which the patient is held responsible for his or her symptoms and
behavior (Ihsan, 1995). In traditional cultures, the cause of mental illness is often
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37
seen as external to the patient, or the workings of fate. The patient is thus relieved
from blame (Lefley, 1990). Second, Leff and Vaughn (1985) found that most
critical comments are directed toward negative symptoms such as lack of affect
and apathy rather than the positive symptoms of hallucinations and delusions.
Relatives in non-Westem households may be more accepting of these negative
symptoms as demonstrated among Mexican Americans (Kamo, Jenkins, De La
Selva & Santana, 1986).
With regard to ethnic minorities’ help-seeking behavior, several studies are
noteworthy. A participant-observation study followed nine Puerto Rican subjects.
Each participated in comprehensive vocational rehabilitation over 18 months in an
urban, New England city (Alverson & Vicente, 1998). The researchers found that
the most outstanding characteristic of the Hispanic clients with respect to the
mental health system is that they have few expectations for it. The clients, when
asked, reported that they were happy with their treatment, but would like more
than anything to be independent of mental health and welfare systems. The
authors further stated that Puerto Ricans do not utilize mental heatlth services as
fully as European Americans primarily because of lack of knowledge and lack of
proclivity to work the social services system.
Related to this is a series of studies that were performed in Washington
State. Sue and colleagues found that, compared to whites, ethnic minority patients
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
38
demonstrated a pattern of limited use and premature termination of outpatient
services in the Seattle-King County areas of Washington (Sue, McKinney, Allen,
& Hall, 1974; Sue, 1977, 1976). The authors called for research and program
development to integrate ethnic-sensitive services in an attempt to attract more
minorities to community mental health centers (CMHC). Their recommendations
included using more minority service providers, developing ethnic-specific
services within the minority communities, and prescribing treatment that was
culturally relevant. Ten years after the implementation of most of these
suggestions, O’Sullivan, Peterson, Cox, and Kirkeby (1989) went back to the
communities to replicate Sue and colleagues’ studies. They found that the ethnic
groups’ failure-to-retum rates had been greatly reduced, while their mean number
of contacts with the CMHCs had increased. Ethnic minorities that had been
previously underrepresented in the CMHCs were then represented in proportion to
the general population.
More recently, Alvidrez (1999) examined the predictors of mental health
service use among African American, Latina, and European American women.
While waiting for their clinic appointments, 187 low-income individuals were
interviewed about their attitudes toward mental illness and mental health services.
As the author predicted, Latinas more strongly endorsed the belief that problems
should be kept within the family. Furthermore, compared to European Americans,
Latina clients more strongly endorsed folk beliefs, such as the view that there are
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
39
imbalances in hot and cold forces which cause mental illness. African Americans
attributed causes to religious and supernatural forces, while European Americans
were more likely to state that the cause was related to imbalances in life style and
environmental factors. Given this understanding, it was not surprising that the
European American women in the study were significantly more likely to seek
professional mental health services.
Related findings were discovered in an earlier study of 90 family
caregivers (Guamaccia, 1996). The African American, Hispanic American, and
European American family members were asked to describe the kind of problem
their ill relative had. The responses were coded into four categories: “Medical”
responses included descriptions of the problem as a chemical imbalance.
“Emotional” responses focused on problems of nervousness. “Personality”
descriptors included adjectives like “selfish” and “aggressive.” The final category
was “Social”, which indicated the illness was conceptualized in terms of problems
of social interaction. The most frequent response from European American and
African American families was that their relative was suffering from an illness
that was medical in nature. The Hispanic families strongly felt that their ill
relative was suffering from an emotional problem.
Finally, three studies that revealed a relationship between ethnicity and
psychosocial interventions are highlighted. Barrio (1998) performed a cross
ethnic study on the impact of community-based psychosocial rehabilitation on
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40
cultural centricity. Borrowing data from the same parent study as this current
paper, she looked at 163 persons with schizophrenia and related disorders who
were participating in community-care programs. The author contended that
minorities (Latinos and African Americans) experience a less-debilitating course
of illness than whites due, in part, to their sociocentric culture. In response to
intensive psychosocial rehabilitation, however, the minority groups’ scores on
empathy, as defined by the Quality of Life Scale (QLS; Heinrichs et al., 1984)
declined, at the same time they remained stable for the non-minority group. In her
conclusion Barrio stated that cultural qualities and strengths can be negatively
impacted by the treatment culture, which can be experienced as intrusive and
stress-inducing.
A similar explanation was suggested following a study by Telles et al.
(1995). The investigators compared the effectiveness and cross-cultural
applicability of highly-structured and intensive behavioral family management
(BFM) with standard case management (CM) in preventing exacerbation of
symptoms and relapse among 40 Latino individuals with schizophrenia. BFM was
described as a highly structured behavioral intervention encompassing three
separate modules: patient-family education regarding schizophrenia; training in
communication skills; training in problem-solving skills. They operationalized
case management as goal-oriented, supportive psychotherapy with only the
identified patient, directed towards enhancing the patient’s functioning in the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
community. The subjects in both groups were seen in a clinic setting weekly for
the first six months, biweekly for the next three, and monthly for the final three
months of the study period.
The subjects comprised persons of Mexican, Guatemalan, and Salvadoran
descent. Fifty-eight percent were considered minimally acculturated to Euro-
American society; forty-two percent were highly acculturated. They were
randomly assigned to the two treatment groups.
The primary hypothesis of the study was that BFM, as compared to CM,
would have a beneficial effect on preventing relapse and reducing psychotic
symptoms. The Brief Psychiatric Rating Scale (BPRS) and the Global Assessment
Scale (GAS) were used to measure outcome. The effect of treatment was
significant. The direction of the effect was unexpected: the risk of exacerbation
of symptoms was significantly greater among patients who had received BFM.
After stratifying the sample into the dichotomous groups of high vs. low
acculturation, the authors made another discovery. Among the percentage of
subjects who were the most highly acculturated, the psychosocial treatment
approach made little difference. Among those who were less acculturated,
however, the pattern of treatment effects was significantly different. Those
patients who were treated with BFM had significantly poorer outcomes at 1-year
follow-up on BPRS items that rated thought disturbance, anergia, and total
severity, as well as the GAS when compared to case management.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
42
Telles et al. (1995) suggested that the type of highly structured family
intervention used in their study may have an adverse effect on patients who are
less acculturated, and that the benefits of this model of intervention are highly
questionable when applied to an ethnically diverse population. They stated that
the specific directives and exercises are culturally dystonic. For example, subjects
were asked to make eye contact and express negative feelings to authoritative
(paternal) figures, a practice that is considered disrespectful in traditional cultures.
Jerrell and Wilson (1997) compared white (n = 92) and ethnic minorities
(n = 40) in three types of community treatment. The minority group was
comprised of Hispanics (65%) African Americans (20%), Asians (7%), and Native
Americans (7%). Their purpose was to investigate differences in group means in
the areas of psychosocial functioning, psychiatric and substance abuse
symptomatology, as well as differences in service costs of the three programs. The
sample consisted of individuals suffering from a DSMIII-R Axis 1 diagnosis, as
well as a secondary
substance abuse disorder. Seventy-seven and one-half percent were diagnosed
with schizophrenia or schizoaffective disorder.
The types of treatment included the following: 1) Behavioral skills
training. This approach relied on cognitive-behavioral methods for teaching clients
self-management skills. The groups met once a week for repeated practice and
reinforcement of these skills. 2) Case management. This involved assistance by a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
43
clinician or a paraprofessional to provide housing for clients, and assistance with
daily living, money management, and legal problems. 3) A 12-step recovery
model. This intervention involved taking clients to in-house, as well as outside
12-step meetings, helping them to secure a sponsor, and providing ongoing 12-step
support (Jerrell & Wilson, 1997).
The results showed that in the area of psychosocial functioning, there was
a significant baseline difference between whites and ethnic minorities, with the
ethnic clients self-reporting lower functioning as measured by the Social
Adjustment Scale (Schooler, Weissman, & Hogarty, cited in Jerrell & Wilson,
1997). While both groups changed in a positive direction over six months, the
changes were not significant. Nor was there a change score by group interaction.
This same pattern held true for the Role Functioning Scale (RFS; Green &
Gracely, 1987). In terms of treatment by ethnicity interactions, significant
differences were found in the area of total drug and alcohol symptoms, with
symptoms being lower in the behavioral skills group, especially for the ethnic
minority clients.
An interesting aspect of the study was a gathering of clinical impressions
of the staff who were directly involved in treatment. The line staff reported that
families and friends of the ethnic clients often did not have basic information
about mental illness or substance abuse disorder, and experienced their loved ones
as willfully misbehaving and ignoring responsibilities. This led them to call the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
44
police, rather than mental health professionals, when their behavior was deemed
uncontrollable. Furthermore, the authors contended that it was not uncommon for
staff to view African-American male clients who entered the program with
criminal justice histories as malingering, antisocial, or non-compliant with
treatment, when compared to non-minority clients. Ethnic clients were also
viewed by clinical staff as suffering from more severe and persistent
symptomatology and as having lower psychosocial functioning. The authors
identified this discrepancy as being due to cross-cultural misperceptions, system
bias, and/or countertransference issues. They recommended that the treating staff
be engaged in advocating for culturally relevant services, particularly in the areas
of housing and work programs. They also recommended that an effort be made to
divert ethnic clients with mental illness and substance abuse out of the legal
system and into community rehabilitation.
Summary of Ethnicity Research
It is generally understood that there is a core schizophrenia syndrome that
is similar across diverse cultural settings. However, evidence tells that in
developing countries individuals enjoy a more favorable outcome than those in
Western industrialized nations. This has led to the consensus among cross-cultural
researchers that traditional cultures favor symptom recovery and the restoration of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
45
functioning. A suggested mediating influence has been the way individuals are
perceived, and how they are cared for within their collectivistic and familistic
societies.
The United States, as a multicultural nation, has been the site of a wide
range of studies that have attempted to discern and compare the unique
characteristics of ethnic groups. A heavy proportion of these studies have focused
on the expression of symptoms. Variations among whites, African Americans, and
Latinos extend from how they express anger to their experiences of hallucinations
and delusions. The majority of studies show no ethnic discrimination in the
severity of these symptoms. However, one study found whites to be more
symptomatic than Latinos and African Americans on several variables (Brekke &
Barrio, 1997). It was proposed that the buffering elements of sociocentric,
collectivist cultures were mediating factors in the experience of symptoms.
Studies that compared the treatment of individuals with severe mental
disorders revealed that minority patients were more likely to be prescribed high
doses of medication (Lehman & Steinwachs, 1998), and less likely to be offered
psychotherapy as a treatment option (Guamaccia & Parra, 1996) than non
minorities. With regard to help-seeking behavior, Latinos have been found to
underutilize mental health services when compared to whites, while African
Americans were more likely to overutilize services. The latter finding conflicts
with the assertion by Farmer et al. (1997) that African Americans, when compared
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
46
with whites, are more often cared for by family members. The former finding
could possibly be explained by a qualitative study which revealed that Latinos, for
a variety of reasons, would prefer to be independent of mental health and welfare
systems. However, when mental health systems provide ethnic-specific services,
and culturally relevant treatment, it has been shown that this underutilization by
ethnic minorities can be eliminated (O’Sullivan et al., 1989). Reasons for ethnic
differences in the utilization of mental health services might also be found in the
way mental illness is perceived. For example, Latinos have reported that mental
illness is caused by imbalances in hot and cold forces (Alvidrez, 1999) which
would not be remedied by a visit to a local mental health center.
In the literature on community-based psychosocial rehabilitation, three
studies revealed a relationship between ethnicity and psychosocial intervention.
Barrio (1998) observed that minorities became less empathic in response to
psychosocial rehabilitation, while non-minorities’ empathy scores went
unchanged. In her conclusion she stated that the cultural qualities and strengths
inherent in the Latino and African American groups could have been negatively
impacted by the treatment culture. A similar explanation was proposed by Telles
et al. (1995) following their study of behavioral family management vs. standard
case management. Among patients who were low in acculturation, there were a
number of significant treatment differences, all in favor of the less-structured and
less-directive case management. This was observed at one-year follow-up on the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
47
BPRS items that rated thought disturbance, anergia, and total severity. The
authors surmised that these unexpected findings were the result of specific
directives and exercises that were culturally dystonic. Not all directive
interventions, however, are unsuccessful with minority consumers.
Jerrell and Wilson’s (1997) study compared three types of community
treatment for individuals with mental illness and a co-occuring substance-abuse
disorder. These included behavioral skills training, case management, and a 12-
step recovery model. Significant treatment by ethnicity interactions were observed
in the area of total drug and alcohol symptoms. Behavioral skills training was the
most successful of the three, especially for the ethnic minority clients.
The preceding studies support the following contentions: 1.) Ethnicity is a
key variable in the search for individual factors that predict outcome from
community-based psychosocial interventions. 2.) Research on community
support programs needs to specifically examine differential treatment effects for
different ethnic groups. According to Barrio (2000), services should be
individualized, empower clients, incorporate natural supports, focus on strengths,
and be culturally appropriate. When the objective is to design and provide
services to a particular culture, this process must “appraise the fit between the
service system and the unique ethnocultural qualities, needs, and expectations of
the client to be served” (p.880).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
48
Preliminary Study
In preparation for this current project, a pilot study was performed to assess
ethnicity and its relationship over time to prospective functional outcomes from
community-based psychosocial rehabilitation (Phillips, Brekke, & Barrio, 2000, in
press). Using data from programs that had already been shown to result in
improved functional outcomes (Brekke et al., 1997; Brekke et al., 1999) the intent
of the study was to determine if there were systematic ethnic differences in either
the pattern or rate of longitudinal change. For example, using hierarchical linear
modeling, the authors were able to view change as a negative quadratic (an
improvement in group scores, followed by a decline) or a positive quadratic (an
initial drop in scores, followed by a rise). A significant quadratic slope-by-group
interaction meant that there were ethnic differences in the pattern of change.
Significant linear slope-by-group interactions indicated ethnic differences in the
rate of change.
The design, sample, and methods were largely identical to this study with
some important exceptions. The intervention was limited to psychosocial
rehabilitation only; therefore, no treatment by ethnicity interactions were
examined. The study focused only on the functional domains which included
work, socialization, hospitalization, and independent living. (For more information
on design and method issues, please refer to the appropriate sections in this paper.)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
49
Ninety-eight individuals diagnosed with schizophrenia or schizoaffective
disorder comprised the sample. Forty-two were white, 37 were African American,
and 19 were Latino. The results revealed significant differences among the three
groups in rehabilitative change in the areas of work and social functioning. In
terms of work (see figure 1), the linear slope was not significant. However, there
was a significant group by quadratic slope interaction (t = 2.470; df, 96; p < .02),
revealing critical differences in the pattern of change for the three ethnic groups.
Whites, at baseline, had the lowest scores (low scores indicate poor functioning),
followed by African Americans. Latinos demonstrated the highest scores,
although these differences were not statistically significant. Over the initial 18
months of the study period, whites improved. They then began to decline, but
their scores stayed above their baseline levels. African Americans experienced
only a slight change over the three-year period, and it was in the negative
direction. Latinos steadily dropped below their baseline levels for 30 months.
Afterward, they regained only marginally some of what they lost.
In terms of socialization (see figure 2), there were significant differences at
baseline among the three groups (t = 2.88; df, 96; p < .004), with Latinos having
the highest scores (high scores indicate better functioning), followed by African
Americans, then whites. The linear term was significant (t = 2.57; df, 96; p < .02)
indicating that, on the whole, there was improvement. There was a trend toward a
significant linear slope by group interaction (t = -1.63; df, 96; p < . 10). This
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
50
This showed that whites improved the most, followed by Latinos, followed by
African Americans. This trend was seen as important for the following reasons:
First, it replicated the findings on the work outcome. Second, it revealed that
while Latinos began the regime of psychosocial rehabilitation with the highest
socialization scores, their improvement was marginal compared to whites, and
African Americans showed no change.
In an effort to account for these striking variations among ethnic groups,
the authors explored the following possibilities (Phillips et al., in press). First,
there is some evidence that Latinos and African Americans have a more benign
symptom profile than their white counterparts, which is mediated by higher levels
of empathy and social competence (Brekke & Barrio, 1997). This is characteristic
of sociocentric cultures, which rely on a collectivist orientation for effective
problem solving. After exposure to psychosocial rehabilitation, it has been shown
that these qualities of empathy and social competence decline. It is possible,
therefore, that the more individualistic, Western model of behavioral change,
which glorifies independence, self-reliance, rationalism, and competitiveness may
be antithetical to the recovery and growth of members from collectivist cultures.
This would correspond to Telles et al.’s (1995) assertion that assertive behavioral
interventions are experienced as intrusive and stressful, and erode the buffering
structure of collectivist cultures.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
51
Second, it may be perceived by researchers, program directors, and other
mental health professionals that the experience of severe mental illness supersedes
any particular cultural or ethnic identification. However, given that the programs
in the study did not have any specific cultural programming, and given that
African Americans and Latinos did not fair well in the work and socialization
domains, there is a strong implication that the psychosocial interventions were not
relevant for the minority cultures.
The preliminary study set the foundation for the current dissertation in a
number of ways. First, it validated the argument that ethnicity is an important
contender in the search for individual factors that predict outcome from
community care. Second, it challenged the applicability of intensive interventions
in the rehabilitation of minority mental health consumers in terms of work and
social functioning. Third, by examining ethnic variation in relation to change
from one model of community care, intensive psychosocial intervention, it begged
the following question: would the same results hold true for whites and ethnic
minorities if they received less-intensive community intervention? Finally, it laid
the groundwork to assess the predictive value of other individual characteristics
which, for the purposes of this investigation, is gender.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Figure 1. Work Psychosocial Rehabilitation
Whites
African Americans
Latinos
0.4
6 5 3 4 2 1
Interview Period (six-month intervals)
KJ\
to
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Figure 2. Socialization Psychosocial Rehabilitation
4.5
3.5
s
o
o
V)
c
o
Whites
African Americans
Latinos
♦3 2.5
(0
N
m
o
o
tn
0.5
1 2 3 4 5 6 7
Interview Period (six-month intervals)
METHODS
54
This study used data from an earlier investigation that was conducted from
1989 to 1994. The parent study employed a quasi-experimental follow-along
design of patients admitted consecutively to three community-based programs in
Los Angeles. Two were high in service intensity and have been shown to be
effective in improving functional outcome (Brekke et al., 1997; Brekke & Long [in
press]). Offering their consumers a clubhouse milieu, they were open 5 days a
week with regularly scheduled groups aimed toward improving the members’ level
of functioning. They comprised the PSR model.
The third was a case-management program operated by the Los Angeles
County Department of Mental Health. Its primary functions were to link clients
with severe mental illnesses with essential resources, and empower them to
function as independently as possible in securing the resources they need. The
services had no particular rehabilitative focus.
PSR Programs
Portals is a comprehensive rehabilitation program located in urban Los
Angeles which provided its members with pre-employment training, job
development, and employment with support and follow-up. In addition it offered
educational and socialization programs, as well as linkages to other community
resources. Portals provided an average of 12 hr/month of staff-to-client contact
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
55
during the first 3 months that members were in the program. This dropped to
about 5 hr/month after that. Twenty-seven percent of the contacts were in the
vocational area, while 17% focused on independent living skills (Brekke, 1997).
The Community Living Program provided clients with 3 months of
intensive training in the areas of socialization and independent living, followed by
less intensive ongoing support and rehabilitation. In addition, it rendered crisis
management to prevent hospitalizations. CLP averaged over 35 hr/month of staff-
to-client contact during its 3-month training period, which dropped to about 7
hr/month after that. They provided virtually no vocational training (less than 1 %),
but did offer substantial training in daily and independent living skills (39%).
CM Program
The case management program, run by the Los Angeles County
Department of Mental Health, is connected to residential board and care homes.
Individuals are seen by their case managers a minimum of once every three
months, and are never dropped from their caseload. The case managers provide
assessment, planning, linkage to other services, and advocacy. While the program
offers no rehabilitation services, the intention is to help individuals with severe
mental disabilities to function optimally in the community.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
56
Sample
The sample (n = 164) is comprised of 41 females and 123 males diagnosed
with either schizophrenia or schizoaffective disorder. Eighty-seven are white, 50
are African American, and 27 are Latino. The average age is 33 years (SD = 7.3),
and the average length of illness was 11.2 years (SD = 7). Of the women, 80% had
never married, and 20% were divorced. Eighty-nine percent of the males were
single, 9% were divorced, and 1% was widowed.
Ninety-eight of the study participants were from the psychosocial
rehabilitation models (65 from Portals, 33 from CLP); 66 were in the case
management group. The psychosocial rehabilitation group had 42 (43%) whites,
37 (37%) African Americans, and 19 (19%) Latinos. The case management
sample consisted df 45 (68%) whites, 13 (20%) African Americans, and 8 (12%)
Latinos. In the psychosocial rehabilitation model there were 75 (75%) males and
25 (25%) females. There were 50 (76%) males in the case management group and
16 (24%) females.
Nearly all of the clients were receiving public assistance and had been
treated in the publicly-funded mental health system. No one was hospitalized at
the time they entered the study, and all were living in non-institutionalized
community settings.
There were four study admission criteria: (a) a diagnosis of schizophrenia
or schizoaffective disorder; (b) residence in Los Angeles for at least 3 months
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
before study admission; (c) 18-60 years of age; (d) neither a diagnosis of mental
retardation or organic brain syndrome, nor a primary diagnosis of substance
dependence. While not a criterion for admission, it was determined in an earlier
study (Brekke, 1997) that the Latino population was well-acculturated with 87%
having been bom in the US, and 100% having been exposed to the American
public education system. Social class was assessed using procedures outlined in
Hollingshead and Redlich (1958) and Meyers and Bean (1968). Based on a five-
point rating scale (1 representing the highest and 5 the lowest), the majority of
subjects were in class 3; there were no significant differences between ethnicity,
nor gender, and social class. Psychosocial functioning at baseline was also
assessed by comparing the ethnic and gender groups on four items: independent
living, hospitalization, socialization, and work. There were no differences across
gender groups.
Diagnostic procedures
Diagnoses were established through a two-step process. First, an initial
diagnostic interview was conducted by an admitting clinician at all program sites.
Following this initial screening, a face-to-face interview was conducted by a
licensed, doctorate-level clinician trained in the use of the Schedule for Affective
Disorders and Schizophrenia (SADS; Endicott, 1978). The SADS diagnosis was
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
58
determined through this structured interview as well as a review of clinical
records.
Measures
In order to comprehensively assess ethnic and gender variation, the
outcome model described in Brekke and Long (2000) was used. Using P-
technique factor analysis, the writers determined that functional, subjective, and
clinical variables form distinct, but linked clusters. In their study of 172
individuals with chronic mental illness, they found that change in subjective
experience was modestly associated with change in functional and clinical
domains. And change in clinical outcomes was distinct from change in functional
outcomes. Each domain, therefore, has unique value in determining outcome from
community-based interventions.
Functional Variables
In order to measure psychosocial functioning, it is important to take into
account the individual’s level of functioning in daily life. To achieve this, two
outcome measures were used in this study: the Role Functioning Scale (RFS;
Goodman, 1993; McPheeters, 1984), and the Strauss and Carpenter Outcome
Scale (SCOC; Strauss, 1972).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
59
The RFS rates patients with severe mental illness along several functional
domains. For the purposes of this study, the items which rate independent
living/self care and immediate social network relationships were selected. Each
domain consists of 7 anchor points, with “ 1” being the lowest, and “7” the highest
functioning. The lowest rating for independent living is, “Lacking self-care skills
approaching life-endangering threats.” Number 7 states, “Optimal care of
health/hygiene: independently manages to meet personal needs and household
tasks.” The measure of social relationships ranges from 1, “severely deviant
behaviors within immediate social networks often with imminent physical
aggression or abuse” to 7, “positive relationships with friends; assertively
contributes to these relationships.”
The items drawn from the SCOC were those that rated the frequency or
duration of hospitalization and employment. The highest rating, number 4 from
the hospitalization item reads, “not hospitalized in the past year.” Zero is
“hospitalized more than nine months in the past year.” The responses from the
“employment” item range from number 4, “employed” continuously in the past
year” to 0, “no useful work”. (It is important to note that “employment” covers
work as a volunteer, student, or homemaker, as well as competitive employment.)
The preceding scale ratings w et. extracted from a face-to-face interview
instrument called the Community Adjustment Form (CAF) which gathers data in
17 areas of community adjustment (Brekke et al., 1993; Test, 1991). A study of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60
interrater reliability using the intraclass correlation (ICC) was established during
intensive rater training, and during subsequent training assessments throughout the
study period (Brekke, 1992; Brekke, 1997). The ICC on the work, hospitalization,
social functioning, and independent living items ranged from .75 to .98, with an
average of .89.
Subjective Variables
Concepts of the self and identity have been central concerns in the study of
schizophrenia. The subjective experiences of individuals with schizophrenia and
related disorders provide a primary basis for understanding these disorders and the
individuals who have them (Strauss, 1989). This paper examined subjective
experience over time using the following variables: Hudson’s Index of Self-
Esteem (1SE; Hudson, 1982), the Satisfaction With Life Scale (SWL; Stein,
1980), and the Brief Symptom Inventory (BSI; Derogatis, 1983). These variables
have been recognized as “quantitative assessments of intrinsic and critical aspects
of the subjective life experience of individuals with schizophrenia” (Brekke, 1993
p. 601).
The ISE is a 25-item self-report measure that taps the subjective evaluation
o f self, as well as how one thinks others perceive him or her. For example item # 1
states, “I feel people would not like me if they knew me very well.” Participants
are asked to score each item with a number between 1 and 7, which represents
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
61
persistence of the feeling or attitude. Higher scores are indicative of greater levels
of self-esteem. Hudson (1982) reports excellent psychometric properties. The
alpha for the present study is .87.
The SWL (Stein, 1980) is a 21-item self-report measure intended to
quantify subjective satisfaction with one’s living situation, work, social contacts,
and psychological state. Each item is rated on a 5-point scale, and ranges from
“not at all,” to “a great deal.” Higher scores are indicative of a greater level of
satisfaction. The alpha for the present study is .89
The BSI (Derogatis, 1983) is used as a measure of subjective distress. It is
a 53-item self-report scale that measures the subject’s sense of discomfort caused
by a variety of symptomatic phenomena experienced in the previous 7 days. Each
item is rated on a 5-point scale of distress (0-4), ranging from “not at all” to
“extremely.” The BSI is conceived as measuring 9 primary symptom constructs.
These include the following: somatization, obsession-compulsion, interpersonal
sensitivity, depression, anxiety, hostility, phobic anxiety, paranoid ideation, and
psychoticism The alpha for the present study is .96.
Clinical Variables
The Brief Psychiatric Rating Scale (BPRS; Overall, 1962) provides an
efficient evaluation procedure that yields a comprehensive description of major
symptom characteristics. Since its inception in 1962, it has been used as not only
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
62
as a measure of symptom change, but as a tool to evaluate and examine psychiatric
diagnosis, to describe , measure, and empirically classify psychiatric patients and
their manifest psychopathology, and to develop clinical prediction models for
treatment response expectations (Hedlund, 1980).
Items 1 to 12 cover the following categories, and are rated primarily on the
basis of the patient’s self-report. Somatic Concern, Anxiety, Emotional
Withdrawal, Grandiosity, Suspiciousness, Hallucinatory Behavior, Unusual
Thought Content, Helplessness/Hopelessness, Guilt Feelings, Hostility, Depressive
Mood, and Disorientation. Items 13-22 are rated on the basis of observed behavior
and speech. They include Conceptual Disorganization, Tension, Mannerisms and
Posturing, Motor Retardation, Uncooperativeness, Blunted Affect, Elated Mood,
Excitement, Distractibility, and Motor Hyperactivity. The scale ranges from one
to seven with one being the absence of symptoms, and seven indicating severe
symptomatology. A total pathology score is obtained by summing the ratings on
all symptom constructs (Ventura, 1995). In an effort to reduce the risk of a type I
error, this study examined the symptoms in clusters, or factors. These include
Thought Disturbance, Anergia, Affect, and Disorganization. It was determined
through confirmatory factor analysis that these represent distinct symptom
constructs which endure over time (Long and Brekke, 1999).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
63
Data Analyses
Hierarchical Linear Modeling (HLM) offers a distinctive perspective to
modeling longitudinal data as a response to the developing interest in the
application of individual growth models in the social sciences and education
(Bryk, 1987; Francis, 1991). It provides a conceptual framework and a set of
analytic tools which focus the study of change at the individual client level. This
allows for the assessment of both intraindividual and interindividual differences in
change. Furthermore, HLM provides a process view of change; that is, a person’s
score reflects an ongoing process that underlies continuous change in the
expression of a variable (e. g., social functioning).
For the purpose of this study HLM involved a two-stage model of change.
At stage 1, each individual’s observed score is conceived of as a function of an
individual growth trajectory plus random error. This trajectory is determined by
the individual’s observed scores which are regressed on time, or a transformation
of time. This produces a growth curve. At stage 2 the individual parameters are
regressed on a measurable background variable, or covariate, (in this study the
level 2 covariates were gender, ethnicity, and program type) to determine whether
there are systematic differences in rate (i. e., slope) or type of change (linear or
quadratic) based on the covariates. This model also allows for the examination of
interactions between the covariates which, for this study, included treatment by
ethnicity, or treatment by gender effects. Final estimates of the growth curve
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
64
parameters are derived through empirical Bayes estimation, which provides a
composite procedure that uses both the information from each subjects’ data and
the information from the covariates in determining final parameter estimates. In
other words, each individual’s growth curve parameters are estimated with a
weighted combination of the level 1 and 2 estimates. This highlights the unique
strength of HLM in making individual predictions: if within-subject data is
precise, the model weights that heavily. If between-subject data is strong, that
data receives emphasis (Bryk, 1987). After the individual parameters are
estimated, hypothesis testing is used to evaluate the fit of the group linear and
quadratic curves in the study sample, as well as the significance of the covariate
interaction.
In this study the term “intercept” was used to describe the initial status, or
baseline functioning of the groups during the 6 months prior to the intervention.
The linear slope refers to a straight-line change, and can be either positive or
negative. Significant linear change can be viewed as capturing sustained rates of
change in the outcome variable. The quadratic slope can be either positive or
negative as well. A positive slope means that the groups’ score on a particular item
dropped from its initial baseline status, then rose again to some degree by the 36th
month. A negative slope indicates an initial rise in scores, followed by a decline.
A significant quadratic change can be viewed as capturing instability over time in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
65
the variable of interest. The course of an outcome variable can potentially follow
a linear or quadratic slope, or a combination of the two.
Effect size
In clinical research, where small samples or a few extreme values may
produce misleading results, effect size is a useful tool for assessing the practical
implications of rejecting the null hypothesis. For this study the product-moment
correlations, r (Rosenthal, 1993) were computed for significant HLM findings
using the following formula: r = [t2 /(t2 + df)]1/2. The criteria used to assess affect
size in clinical research is r = . 10 is small, r = .30 is medium, and r = .50 is large.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
66
RESULTS
Status of the Groups at Baseline
Assessing the equivalence of the groups at baseline is essential in quasi-
experimental outcome studies (Brekke et al., 1997). Tables 1 through 6 present
baseline data on the average age at study entry, prognosis, medication compliance,
as well as the functional, subjective, and clinical variables used in the study. In
analyses involving the ethnicity samples, they were divided tri-ethnically as well
as bi-ethnically. A tri-ethnic design allows for a richer picture of ethnic
differences. Furthermore, the inclusion of three comparison groups reduces the
problem of social class distinction. For example, if a difference exists between
Latinos and whites, but not African Americans and whites, the difference can not
be attributable to social class (Butcher, Braswell, and Raney, 1983). Due to small
sample sizes, the bi-ethnic design was also used to ensure greater statistical power.
In the psychosocial rehabilitation group a significant gender difference was
found in the BPRS score, with women presenting a more symptomatic profile than
men. Out of 22 items, women scored higher on four: self-neglect, conceptual
disorganization, emotional withdrawal, and excitement. In addition, Latinos and
African Americans demonstrated better socialization scores at baseline than
whites. This finding was supported in the bi-ethnic analysis.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
67
In the case management group women scored significantly higher on the
socialization scale. In both the tri-ethnic and bi-ethnic breakdowns, whites were
older than both African Americans and Latinos. In the bi-ethnic breakdown, there
was a trend toward significance on the work item, with minorities having better
work scores at baseline.
When controlling for gender and ethnicity, an examination across the two
treatment groups revealed that those in case management had some advantages
prior to the treatment interventions. Women, as well as non-minorities in the case
management program, had higher socialization scores than those in psychosocial
rehabilitation. And both men and women, as well as minorities and non-minorities
spent less time in the hospital in the six months prior to the study period than those
in psychosocial rehabilitation.
Attrition
Two types of attrition were examined, study attrition and treatment
attrition. The former concerned subjects who dropped out of the study. At 12 and
18 months 88% of the sample was retained. This dropped to 83% at 24 months,
80% at thirty months, and 70.7% at 36 months. There were no statistically
significant differences in dropout rates across gender or ethnicity variables.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
68
Table 1
Psychosocial rehabilitation: Comparison of Gender Samples at Baseline
Client Variable
Males
(n = 73)
Females
(n = 25) t (and dfs)
Age (and SD) 32.62 (8.36) 33.28 (7.56) /(96) = .35
Prognosis' 32.15(4.15) 32.64 (5.40) t(96) = .46
Medication Complianceb 168.78 (32.26) 153.75 (51.06) t(90) = 1.67
BPRS' score 41.31 (12.18) 48.16(15.28) /(96) = -2.26*
Satisfaction With Lifed score 38.95 (15.08) 39.33 (11.74) /(91) = .11
Index of Self -Esteem' score 84.49(15.60) 79.78 (13.70) t(9 2) = -1.29
Brief Symptom Inventory^ 63.23 (40.28) 65.78 (31.93) /(90) = .27
GAS 32.86 (9.69) 31.76 (9.48) t( 96) =-.49
RFS Independence 3.10(1.03) 2.72 (.73) t(96) = 1.25
SCOC Hospitalization 3.36(98) 3.32 (.95) t(96)= .16
SCOC Work 0.88(1.07) 0.64(1.15) t(96) = .94
Note. BPRS = Brief Psychiatric Rating Scale; 'Measured on the Strauss and Carpenter Outcome Scale;
b Number of days on medication in previous 6-month period (Measured on the Community Adjustment Form,
(Test et al., 1991); 'Brief Psychiatric Rating Scale (Overall and Gorham, 1962); d (Stein and Test, 1990);
Hudson (1982); fDerogatis and Melisaratos (1983); * p < .05.
Table 2
Psychosocial Rehabilitation: Comparison of Ethnicity Samples at Baseline
Client Variable
Whites
(n - 42)
Af-American
(/»= 37)
Latino
(// =19) F (and dfs)
Age (and SD) 32.76 (8.67) 34.49 (8.35) 29.53 (5.33) F(2, 97) = = 2.4
Prognosis* 31.42 (4.57) 32.48 (4.91) 33.68 (2.90) F(2, 97) = = 1.7
Medication Compliance1 1 157.49 (46.79) 171.53 (27.64) 167.65 (35.27) F(2, 89) = = 1.32
BPRS' score 45.45 (15.92) 39.56(10.48) 44.57 (10.75) F(2, 95)== 2.1
Satisfaction With Lifed score 38.53 (14.18) 35.82(13.85) 44.52 (14.11) F(2, 90) = = 2.2
Index of Self-Esteem' score 82.19(14.72) 83.69(13.60) 83.34(15.22) F(2, 91) = = 0.3
Brief Symptom Inventory 61.37(35.37) 61.47 (35.97) 75.50(49.12) F(2, 89) = = 0.9
GAS 30.86 (9.02) 34.41 (10.29) 32.58 (9.60) F(2, 95) = = 1.4
RFS Independence 2.86 (.84) 3.22(1.11) 3.21 (.85) F(2, 95) = = 1.68
RFS Social 2.50(1.73) 3.65 (2.04) 3.53 (1.74) F(2, 95) = = 4.29*
SCOC Hospitalization 3.33 (1.00) 3.27(1.07) 3.35 (.61) F(2, 95)== 0.45
SCOC Work 0.69(1.05) 0.84(1.14) 1.05(1.08) F(2, 92) = = 0.73
Note. BPRS = Brief Psychiatric Rating Scale; * Measured on the Strauss and Carpenter Outcome Scale;
Humber of days on medication in previous 6-month period (Measured on the Community Adjustment Form,
(Test et al., 1991); 'Brief Psychiatric Rating Scale (Overall and Gorham, 1962); d (Stein and Test, 1990);
"Hudson (1982); fDerogatis and Melisaratos (1983); * p < .05.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
69
Table 3
Psychosocial Rehabilitation: Comparison of Bi-Ethnicity Samples at Baseline
Client Variable
Whites
(n = 42)
Minorities
(n = 56) t (and dfs)
Age (and SD) 32.76 (8.67) 32.80 (7.78) t(96) = -.03
Prognosis' 31.45 (4.57) 32.89 (4.35) t(96) = -1.59
Medication CompIianceb 146.24 (60.95) 161.16 (48.46) t(96) = .18
BPRSC score 45.45 (15.92) 41.26 (10 74) t(96) = 1.55
Satisfaction With Lifed score 39.54 (14.18) 38.67 (14.40) t(91) = .77
Index of Self-Esteem' score 85.19(14.72) 84.27(15.70) t(92) = -.66
Brief Symptom Inventory^ 61.38 (35.37) 65.79 (40.51) t(90) = -.55
GAS 30.86 (9.02) 33.88 (9.90) t(96) = -1.55
RFS Independence 2.86 (.84) 3.21 (1.02) t(56) = -1.82
RFS Social 2.50(1.73) 3.61 (1.93) t(96) = .-2.94*
SCOC Hospitalization 3.33 (1.00) 3.36 (.94) t(96) = -.12
SCOC Work 0.69(1.05) .91 (1.12) t(96) = -.99
Note. BPRS = Brief Psychiatric Rating Scale; * Measured on the Strauss and Carpenter Outcome Scale;
'’ Number of days on medication in previous 6-month period (Measured on the Community Adjustment Form,
(Test et al., 1991); ^Brief Psychiatric Rating Scale (Overall and Gorham, 1962), d (Stein and Test, 1990);
“ Hudson (1982); fDerogatis and Melisaratos (1983). *p<.05.
Table 4
Case Management: Comparison of Gender Samples at Baseline
Males Females
Client Variable (n = 51) (n = 15) t (and dfs)
Age (and SD) 33.61 (5.57) 35.87 (5.21)
,(64) = -1.40
Prognosis' 30.71 (5.49) 34.47(4.16) t(64) = -2.45*
Medication CompIianceb 168.98 (43.60) 167.40 (33.29) t(62) = .13
BPRSC score 46.04(12.09) 42.60(10.62) <(64) = .87
Satisfaction With Lifed score 41.17(12.95) 41.00(16.68) <(61) = .04
Index of Self -Esteem' score 87.75 (19.94) 86.07(13.03) <(61) =.31
Brief Symptom Inventoryf 59.41 (45.66) 47.86 (34.25) <(58) = .87
GAS 34.88 (7.55) 32.07(5.09) <(64) =-1.35
RFS Independence 3.06 (.42) 3.00( 38)
<(64) = ..49
RFS Social 3.61 (2.06) 4.60(1.12)
<(64) = -1.78
SCOC Hospitalization 3.88 (.59) 3.87 (.35) <(64) = .10
SCOC Work 1.10(1.34) 1.27(1.10) <(63) = -.44
Note. BPRS = Brief Psychiatric Rating Scale; 'Measured on the Strauss and Carpenter Outcome Scale;
'’ Number of days on medication in previous 6-month period (Measured on the Community Adjustment Form,
(Test et al., 1991); 'Brief Psychiatric Rating Scale (Overall and Gorham, 1962); d (Stein and Test, 1990);
“ Hudson (1982); fDerogatis and Melisaratos (1983); * p < .05.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
70
Table 5
Case Management: Comparison of Ethnicity Samples at Baseline
Client Variable
White
(n = 45)
Af-American
("= 13)
Latino
(« =8) F(and dfs)
Age (and SD) 35.38 (5.26) 31.69(4.50) 31.00 (6.59) F(2, 63) = 4.03*
Prognosis* 31.24 (5.74) 31.88 (4 89) 32.46 (4.73) F(2, 63) = .26
Medication Compliance*1 157.49 (46.79) 171.53 (27.64) 167.65 (35.27) F(2, 61) = .36
BPRS“ score 46.24(13.66) 45.92 (15.22) 38.63 (6.12) F(2, 63) = 1.13
Satisfaction With Lifed score 40.91 (13.36) 38 45 (18.27) 46.71 (6.92) F(2, 60) = .78
Index of Self-Esteem' score 85.24 (17.93) 93.00 (20.73) 92.00(18.07) F(2, 60) = 1.03
Brief Symptom Inventor^ 55.54 (41.18) 67.00 (32.90) 45.75 (66.14) F(2, 57) = 0.63
GAS 33.58 (6.77) 35.38 (9.07) 36.13 (5.84) F(2, 63) = .63
RFS Independence 3.02 (.40) 3.15 (.55) 3.00(00) F(2, 63) = .57
RFS Socialization 3.84(1.87) 4.00 (2.24) 3.50(1.93) F(2, 63) = .17
SCOC Hospitalization 3.91 (.29) 3.69(1.11) 4.00 (.00) F(2, 63) = 1.06
SCOC Work 0.96(1.11) 1.77(1.69) 1.14(1.35) F(2, 60) = 0.78
Note. BPRS = Brief Psychiatric Rating Scale; * Measured on the Strauss and Carpenter Outcome Scale;
b Number of days on medication in previous 6-month period (Measured on the Community Adjustment Form,
(Test et al., 1991), “ Brief Psychiatric Rating Scale (Overall and Gorham, 1962), d (Stein and Test, 1990);
“ Hudson (1982); fDerogatis and Melisaratos (1983). *p < .05
Table 6
Case Management: Comparison of Ethnicity Samples at Baseline
Client Variable
Whites
(n = 45)
Minorities
(//=21) t (and dfs)
Age (and SD) 35.38 (5.26) 31.43 (5.24) t(64) = 2.84*
Prognosis* 31.24 (5.74) 32.24 (4.72) t(64) = -.69
Medication Compliance1 1 167.52 (42.00) 171.00(40.25) t(62) = -.31
BPRS“ score 46.24(13.66) 41.67(15.18) t(64) = .88
Satisfaction With Lifed score 40.91 (13.36) 38.45 (18.27) t(61) = -.19
Index of Self-Esteem' score 85.24(17.93) 93.00 (19.19) 1(61) =
-1.44
Brief Symptom Inventoryr 55.54 (41.18) 58.90 (47.88) t(58) = -.29
GAS 33.58 (6.77) 35.67 (7.84) 1(64) = -1.11
RFS Independence 3.02 ( 40) 3.10 (.44) t(64) = -.67
RFS Social 3.84(1.87) 3.81 (2.09) t(64) = ..07
SCOC Hospitalization 3.91 (.29) 3.81 (.87) t(64) = .71
SCOC Work 0.96(1.11) 1.55 (1.57) t(63) = -1.75
Note. BPRS = Brief Psychiatric Rating Scale; * Measured on the Strauss and Carpenter Outcome Scale;
'’ Number of days on medication in previous 6-month period (Measured on the Community Adjustment Form,
(Test et al., 1991); “ Brief Psychiatric Rating Scale (Overall and Gorham, 1962); d (Stein and Test, 1990);
“ Hudson (1982); r Derogatis and Melisaratos (1983). *p < .05
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
71
Treatment attrition concerned those who exited the PSR programs, but
remained in the study protocol. The exit rate for Portals was 19% at 6 months,
45% at 12 months, 76% at 18 months, 81% at 24 months, 90% at 30 months, and
91% at 36 months. For CLP the exit rate was 21% at 6 months, 56% at 12 months,
65% at 18 months, 71% at 24 months, 74% at 30 months, and 79% at 36 months.
Analysis of Variance was performed for the total PSR sample to determine if there
were any differences among the ethnicity groups in number of months they
remained in treatment. The differences were not significant (F = .141, df 2,
p. = .869). Nor did a T-test reveal significant differences between the genders
(t = .394, df 162, p = .694). Data was not collected for the CM sample on number
of months they remained in treatment. However, the overall attrition rate for this
group was relatively low. At 36 months, 88% of the sample continued to
participate in the study. Chi square analyses showed no differences between
genders, nor among ethnicity groups in terms of whether or not they completed the
study.
For the purpose of this investigation, analyses were done on those subjects
who stayed in the study protocol, regardless of whether or not they remained in
treatment. It was not determined whether these exits were planned or unplanned.
In either case, it was not assumed that the exits were premature; implicit in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
72
community-based psychiatric rehabilitation programs is that members are free to
stay for as little or as long as they like.
Medication Dosage and Use
Medication dosage and use was derived from self-report data. The sample
was medication compliant at baseline. With regard to neuroleptic dosage, no
significant differences existed between the three ethnic groups (F = .721, df 2,
p = .489). However, there were significant gender differences with women
reporting higher dosages than men (t = -2.37, df 96, p = .02). With regard to the
number of days on medication in the previous six months, there were no
statistically significant differences between the ethnicity groups (F = 1.53, df 2,
p = .22) nor gender groups (t = .76, df 96, p = .45). In order to assess the
potentially confounding impact of medication dosage, several correlations were
performed. It was found that dosage amount was not significantly correlated to
symptoms (BPRS score; -.071, p = .485). On the other hand, the correlation
between number of days on medication and BPRS score was significant (-.191,
p = .02). In an effort to further eliminate medication dosage as a confounding
variable, correlations were run on all outcome variables used in this study. No
relationships were found, making it unlikely that gender and ethnic differences
were confounded by medication dosage or number of days on medication.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
73
Results Using Hierarchical Linear Modeling
The study hypotheses are addressed below, and will follow the following
outline:
I. Gender
A. Gender findings in the CM group
1 . Functional outcomes
2. Subjective outcomes
3. Clinical outcomes
B. Gender findings in the PSR group
1. Functional outcomes
2. Subjective outcomes
3. Clinical outcomes
C. Gender by Program Interactions
1. Functional outcomes
2. Subjective outcomes
3. Clinical outcomes
II. Ethnicity
A. Ethnicity findings in the CM group
1. Functional outcomes
2. Subjective outcomes
3. Clinical outcomes
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
74
B. 1. Ethnicity findings in the PSR group
a. Functional outcomes
b. Subjective outcomes
c. Clinical outcomes
C. Ethnicity by Program Interactions
a. Functional outcomes
b. Subjective outcomes
c. Clinical outcomes
Using the two-level HLM data structure, the intercept, linear, and quadratic
coefficients were explored for statistical significance. A significant intercept term
revealed whether there were differences between or among the gender and ethnicity
groups at baseline. A significant linear slope illustrated sustained rates of change
over time, either positive or negative, in the outcome variables. A significant
quadratic slope illustrated a pattern of unstable change across time. Of particular
interest was the examination of group differences (interactive effects) in the linear
and quadratic trajectories with regard to the covariates, i.e. gender and ethnicity
over time. Finally, program type was added as a covariate in conjunction with
gender and ethnicity, and relative to the outcome variables. This enabled the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
75
study to answer whether there were gender by program, or ethnicity by program
interaction effects over time. Again, this was determined through linear and
quadratic trajectories.
I. Gender
It was hypothesized that gender would be related to the prospective
functional, clinical, and subjective outcomes from community-based case
management o f persons with schizophrenia over time.
A. Case Management Outcomes
1. Functional Domains: There were no statistically significant findings
with regard to the functional domains of work, socialization, independent living,
and hospitalization.
2. Subjective Domains: Figure 3 presents the predicted growth curves for
males and females in case management on the ISE. There was no group by
intercept interaction. The linear interaction term was significant t(61) = 1.88,
p = .05. The linear coefficient for males was 176, and for females it was .791.
The Quadratic Slope was significant, t(61) = 2.517, p < .01; The Group X
Quadratic Slope interaction was also significant t(61) = -2.43, p < .01. The
quadratic coefficients were .282 for males, and .922 for females. The average
effect size for the significant terms was .28. The results indicate that females’
scores deteriorated from their baseline levels for the initial 12 months of the study
period, tapered off, then began to show improvement. At 36 months, they showed
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
76
overall higher self-esteem scores. Males, on the other hand, showed virtually no
change from baseline levels over the three years.
Figure 4 presents the predicted growth curve for males and females on the
variable BSI. The differences at baseline were not significant, nor were the linear
terms. The Group X Quadratic Slot interaction was significant t(59) = 1.889,
p = .05, indicating that the process of change was different between gender groups.
The quadratic coefficients were .884 for males, and -1.190 for females. The effect
size for the quadratic term was .24. The results show that females became
progressively more symptomatic through the first half of the study period, then
showed steady improvement over the second half. Males’ scores, on the other
hand, dropped as the study period reached its half mark, then saw a return to near
baseline status by the study’s end.
There were no significant findings between men and women in case
management across the subjective variable SWL.
3. Clinical Domains: Figure 5 presents the predicted growth curves for
males and females in case management on the total BPRS score. There was no
group by intercept interaction. There was a trend toward significance in the linear
term t(94) = -1.848, p = .06. The linear interaction term was significant
t(64) = 2.188, p < .05. The linear coefficient for males was -.12, and for females it
was 1.134. The Quadratic Slope was not significant, nor was the Group X
Quadratic Slope interaction. The effect size for the significant linear term was .26.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
77
The results indicate that females were less symptomatic at the beginning of the
study period, but over the months of case management, they got progressively
more symptomatic. Males, on the other hand, showed neither improvement, nor
an exacerbation in their symptoms over 36 months.
It was hypothesized that gender would he related to the prospective
functional, clinical, and subjective outcomes from community-based psychosocial
rehabilitation o f persons with schizophrenia over time.
B. Psychosocial Rehabilitation Outcomes
1. Functional Domains: There were no statistically significant gender
differences with regard to the functional domains of work, socialization,
independent living, and hospitalization.
2. Subjective Domains: There were no statistically significant findings
with regard to the subjective variables ISE, SWL, and BSI.
3. Clinical Domains: Figure 6 presents the predicted growth curves for
males and females in community-based psychosocial rehabilitation on the total
BPRS score. There was a significant Group X Intercept interaction t(86) = 2.523,
p = .01. The intercept coefficients were 41.73 for males, and 49.25 for females.
The linear slope was significant t(86) = 2.18, p < .05. There was a weak trend
toward a significant linear interaction, t(86) = -1.51, p = .13. The linear
coefficients were .653 for males and -.041 for females. Neither the quadratic
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
78
slope, nor the quadratic interaction was significant. The average effect size for the
significant terms was .24. The results indicate that females were significantly
more symptomatic at the start of the study period, with their scores remaining
constant over the following 36 months. Males, on the other hand, showed less
symptomatology when the study began. They got progressively more symptomatic
over the next 12 months, but overall continued to maintain a less symptomatic
rating than their female counterparts.
Figure 7 illustrates the predicted growth curves for males and females on
the Disorganization subset of the BPRS score. There was a significant Group X
Intercept interaction t(86) = 2.996, p = .003. The intercept coefficients were 8.48
for males, and 11.13 for females. The linear slope was significant t(86) = 2.59,
p = .01. There was also a significant linear interaction, t(86) = -1.99, p = .04.
The linear coefficients were .283 for males and -.101 for females. Neither the
quadratic slope, nor the quadratic interaction was significant. The effect size for
the significant terms was .26. The results show an overall pattern that is similar to
that which is reflected in the total BPRS score. The difference between genders is
significant at baseline, with males being less symptomatic. However, over time,
females show slight improvement, while males show a worsening of symptoms.
Figure 8 illustrates the predicted growth curves for males and females on
the Anergia subset of the BPRS score. There was a significant Group X Intercept
interaction t(96) = -2.272, p = .02. The intercept coefficients were 8.28 for males,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
79
and 9.744 for females. The linear slope was significant t(96) = -2.911, p = .004.
There was also a significant linear interaction, t(96) = 2.44, p = .01. The linear
coefficients were .024 for males and -.27 for females. Neither the quadratic slope,
nor the quadratic interaction was significant. The effect size for the significant
terms was .26. The results show that the females in the study suffered more from
the anergia symptoms at baseline, but improved steadily over three years. Males
showed virtually no change, and by the end 36 months, their scores were
essentially equivalent.
It was hypothesized that gender will interact with type o f treatment model
to influence prospective functional, subjective, and clinical outcomes over time.
C. Gender by Program Outcomes
1. Functional Domains: There were no statistically significant Gender by
Program differences with regard to the functional domains of work, socialization,
independent living, and hospitalization.
2. Subjective Domains: Figure 9 represents the predicted Gender by
Program Interaction growth curves on the variable BS1. The intercept and linear
terms were not significant. The quadratic slope was significant t(143) = -2.36,
p = .01, showing a main effect for program. The quadratic Gender by Program
term was also significant t(143) = -1.92, p = .05. The quadratic coefficients were
as follows: males in PSR = 1.112; males in CM = .83; females in PSR = 1.534;
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
80
females in CM = -1.164. The average effect size for the significant terms was . 175.
Post hoc analysis revealed that the significant differences were found among the
females of the study. Figure 10 is presented to highlight the distinction. As can be
seen, females in the CM sample show a negative quadratic slope. Their subjective
level of distress increases over the first half of the study period, then declines.
Those in PSR show a positive quadratic curve. For the first year their scores on
the BSI reveal a lessening of distress. This is maintained over the next six months,
but is followed by a sharp rise over the remaining 18 months.
3. Clinical Domains: Figure 11 illustrates the predicted Gender by
Program Interaction growth curves on the total BPRS score. The intercept and
quadratic terms were not significant. The linear slope was significant
t( 121) = 2.171, p = .03. The linear Gender by Program interaction term was also
significant t( 121) = 2.05, p = .04. The linear coefficients were as follows: males
in PSR =1.191; males in CM = . 129; females in PSR = .205; females in
CM = 1.365. The average effect size for the linear terms was . 185. In this case,
post hoc analysis revealed that the significant differences rested among the males
in the sample. Figure 12 is presented to clarify the differences. As shown, males
in case management underwent virtually no change throughout the study period,
while the males in the PSR sample showed a steady increase in symptoms as
measured by the total BPRS score.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
81
Figure 13 illustrates the predicted Gender by Program interaction growth
curves on the Disorganization subset of the BPRS. There was a significant Gender
by Program interaction at baseline t( 149) = -2.97, p = .003. The linear interaction
term was significant t( 149) = 2.11, p = .04. The quadratic term was also
significant t(149) = -2.67, p = .008, while the quadratic interaction term was not.
The coefficients for the baseline interaction term were as follows: males in
PSR = 8.489; males in CM =11.311; females in PSR = 11.193; females in
CM = 9.32. The linear coefficients were males in PSR = .283, males in
CM = -.023, females in PSR = -.101, females in CM = .213. The quadratic
coefficient were males in
PSR = -.072, males in CM = -.06, females in PSR = -.038, females in CM = -.202.
The average effect size for the significant terms was .225. Post hoc analysis,
presented in figure 14, explains that males in case management were significantly
more impaired at the study’s inception, and followed a marginally negative slope
over time. Males in psychosocial rehabilitation were less impaired at baseline, but
became progressively more symptomatic as the study continued. Their scores
appeared to level off during the last 6 months.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 3 . IS E Case Management
82
CO
CN
O
GO
GO CO CN
GO
0 0
GO
O
0 )
aioos 3SI
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
interview Period ( 6 month intervals)
Figure 4 . BSi Case M anagement
83
u.
o in
in
o
c o
o in
c o
ajo3s isa
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
interview Period (six-month intervals)
Figure 5 . Case Management BPRS Gender
84
4 >
CD CM O 0 0
^r
CM
in
dJOOS SUdS
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Tim e Period ( 6 month intervals)
Figure 6 . BPRS Psychosocial Rehabilitation
85
p w
S u l
to G)
0 0
CD co
c o
9JOOS S ddS
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Interview Period (six-month intervals)
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Figure 7. BPRS (Disorganization) Psychosocial Rehabilitation
12 T
10 -
PSR Males
PSR Females
7 5 2 3 4 1
Interview Period (six-month intervals)
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Figure 8. BPRS (Anergia) Psychosocial Rehabilitation
10.5 -
10
9.5 -
Males
Females
o
o
to
.5 8.5
E?
0)
c
<
7.5
6.5
2 1 3 4 5 7 6
Interview Period (six-month intervals)
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Figure 9. BSI Gender by Program Interaction
£
o
o
< n
m
75
70
65
60
55
Males/PSR
Males/CM
Females/PSR
Females/CM
50
45
40
35
1 2 3 4 5 6 7
Interview Period (six-month intervals)
O O
0 0
Figure 10. BSI Females Only
89
O )
C L O
CO
o in o o o
aioos IS9
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Interview Period (six-month intervals)
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Figure 11. BPRS Gender by Program Interaction
2
o
o
< / >
(0
o'
a.
ca
Males/PSR
Males/CM
Females/PSR
Females/CM
40
2 5 7 3 4 6 1
Interview Period (six-month intervals)
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Figure 12. BPRS Males Only
49
g)
O
0
45
C O
c n
0 1
CD A 'x
Psychosocial Rehabilitation
Case Management
39
7 5 3 4 6 2 1
Interview Period (six-month intervals) ^ >
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Figure 13. BPRS (Disorganization) Gender by Program Interaction
13
12
11
Males/PSR
Males/CM
— a— Females/PSR
— Females/CM
10
8
7
6
7 6 5 4 3 2 1
Interview Period (six-month intervals)
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Figure 14. BPRS (Disorganization) Males Only
12
10
Psychosocial Rehabilitation
Case Management
1 2 3 4 5 6 7
Interview Period (six-month intervals)
94
Summary of Gender Findings
There was support for the hypothesis that gender would be associated with
the subjective variables ISE and BSI in the CM sample. Women’s self-esteem
scores went down over the first year of the study period, but after 18 months the
scores shifted direction and climbed upward. After 3 years their scores indicated
significantly higher levels than those of the men in the sample. In terms of the
BSI, which measures level of distress, a similar pattern followed for females.
They worsened over the first 18 months, then reverted back to baseline levels.
Males, on the other hand, initially moved in the direction of less impairment, then
went back to baseline levels. No differences were found for the variable which
measures satisfaction with life. With regard to clinical outcomes, there was
support for the hypothesis that gender would be associated with the total BPRS
score. Males remained fairly stable over time, while females showed a striking
increase in overall psychiatric symptoms.
Fewer differences were found among individuals in the PSR sample.
There was no support for the hypothesis that gender would be associated with the
functional variables which included work, hospitalization, independent living, and
socialization. Nor was gender associated with the subjective variables of self
esteem, satisfaction with life, and levels of distress. However, differences were
found on the total BPRS score, as well as the Disorganization subset of the BPRS.
In both analyses, the females maintained a higher, but more stable level of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
95
symptomatology across all time periods. Males, on the other hand, showed a
gradual exacerbation of symptoms over time.
Of major interest were gender by program interactions. These were
evident in the BSI scale, the total BPRS, and the Disorganization subset of the
BPRS. Regarding the BSI, females in psychosocial rehabilitation enjoyed reduced
levels of distress for the first 18 months, while those in case management reported
more intense levels over the same time period. These differences, however were
not sustained. The data illustrate that they actually reversed directions; for the
remaining half of the study period, the CM females showed significant
improvement. The females in the PSR group reported scores that reflected a
dramatic rise in distressful feelings which, by the study’s conclusion were
statistically worse than baseline levels.
The BPRS and the Disorganization subset revealed a difference between
males in PSR and males in CM. On the Disorganization subscale, males in CM
were more symptomatic at baseline, but the scores remained relatively constant
over three years. Their scores for the total BPRS scale also went unchanged over
the same time period. However, those in PSR showed an increase in symptoms at
nearly every time point starting from baseline. In the case of the total BPRS score,
their scores at 36 months actually surpassed their cohorts’ in CM.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
96
II. Ethnicity
This study proceeds from an earlier investigation which that found that
ethnicity was related to the prospective work and socialization outcomes following
community-based psychosocial rehabilitation (Phillips, Brekke, & Barrio et al.,
2000). This current study extends the effort to include prospective functional
outcomes from case management, as well as clinical, and subjective outcomes
from community-based case management and psychosocial rehabilitation. In all
the ethnicity analyses, tri-ethnic designs were employed initially. The groups were
then collapsed into minority/non-minority categories to bolster statistical power.
The hypotheses and results are as follows:
Ethnicity is related to the prospective functional, clinical, and subjective
outcomes from community-based case management ofpersons with schizophrenia
over time.
A. Ethnicity findings CM group
1. Functional Domains: Figure 15 presents the predicted growth curve for
minority/non-minority on the “work” item of the SCOC. There was no Group X
Intercept interaction. The linear term was not significant; the Group X Linear
Slope interaction showed a trend toward significance, t(64) = -1.64, p = . 10. The
quadratic slope indicated a trend, t(64) = -1.74, p = .07. The Group X Quadratic
Slope interaction was significant, t(64) = 2.029, p < .05. The quadratic
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
97
coefficients were -.011 for non-minorities, and .038 for minorities. The effect size
for the quadratic term was .24. The results reveal that, overall, whites showed
virtually no movement in the domain of work while receiving case management
services during the study period. Minorities, on the other hand, demonstrated
significant change, unfortunately in the direction of lower work functioning. Their
decline was steady over the first 24 months of the study period, at which time the
slope began to level off.
2. Subjective Domains: There were no statistically significant findings with
regard to the subjective variables ISE, SWL, and BSI.
3. Clinical Domains: There were no statistically significant findings with
regard to the clinical variable BPRS.
Ethnicity is related to the prospective clinical and subjective outcomes
from psychosocial rehabilitation o f persons with schizophrenia over time.
B. Ethnicity findings PSR Group
1. Functional Domains: (These results were reported in Phillips et al. [in
press]. They were also summarized on page 49 of this manuscript. See figures 1
and 2).
2. Subjective Domains: Figure 16 illustrates the predicted growth curve on
the BSI. The group X Intercept interaction was not significant; nor were the linear
slope and linear interaction terms. While the quadratic slope was not
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
98
significant, there was a significant Group X Quadratic Slope interaction
t(96) = 2.377, p = .01. The quadratic coefficients were .252 for whites, 1.024 for
African Americans, and 1.796 for Latinos. The effect size for the quadratic term
was .24. The results show that there was significant improvement in Latinos’
levels of distress, which continued for 24 months. Whites’ and African
Americans’ scores on the BSI remained virtually unchanged throughout the study
period.
Figure 17 illustrates the predicted growth curve on the ISE. The Group X
Intercept term revealed no significant differences in the baseline scores of the
three ethnicity groups. The linear slope was significant, however t(93) = 2.859, p
= .005, as was the Group X Linear interaction term t(93) = -2.663, p = .008. The
linear interaction terms were .804 for whites, .046 for African Americans, and
-.712 for Latinos. There was also a significant quadratic slope t(93) = -2.446,
p = .01. The quadratic coefficients were -.479 for whites, -.278 for African
Americans, and -.077 for Latinos. The average effect size for the significant terms
was .26.
The results indicate that whites showed a positive association between
psychosocial rehabilitation and self-esteem scores reaching their highest point at
24 months. After that there was some leveling off, but their scores at 36 months
were significantly better than when they entered the study. African Americans
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
99
showed virtually no movement. Latinos’ scores showed a significant pattern of
deterioration throughout the entire study period.
3. Clinical Domains: There were no statistically significant findings with
regard to the clinical variable BPRS.
Summary of Ethnicity Findings
There was support for the hypothesis that ethnicity would be associated
with the functional variable “work” in the CM sample. Differences did not show
up in the tri-ethnic breakdown, but when the groups were collapsed into minority/
non-minority categories an unstable pattern of change was revealed for the
minority group. They showed 12 months of improved work scores, which then
sharply declined. The non-minority group’s scores were stable but stagnant,
showing no improvement over the three years of the study. With regard to the
subjective and clinical variables, no differences were found in the CM sample
using either tri-ethnic or bi-ethnic analyses.
The hypothesis that ethnicity would be associated with the functional
variables work and socialization items for the PSR sample was also supported.
Particularly noteworthy was the scoring pattern from baseline to month 36 for the
three groups on the “work” and “socialization” items. Latinos consistently entered
into psychosocial rehabilitation with scores either equivalent to, or higher than
their African American and white associates. But they experienced less gain, and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 15. Case Management W ork Score
100
o o o
o
o o c o
aioos H J O m
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Interview Period (six-month intervals)
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Figure 16. BSi Psychosocial Rehabilitation
80
75
Whites
African Americans
Latinos
70
65
55
50
45
40
7 4 5 6 2 3 1
Interview Period (six-month intervals)
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Figure 17. ISE Psychosocial Rehabilitation
90
S»
o
u
< /)
E
0 )
2
(A
U J
0 )
< /)
Whites
African Americans
Latinos
80
78
7 6 5 4 3 2 1
Interview Period (six-month intervals)
103
in most instances where statistical significance or trends existed, actually
worsened over time. African Americans showed little change overall, and whites
appeared to respond positively to the interventions, although these improvements
attenuated over time.
The hypothesis that ethnicity would be related to the subjective variables
1SE and BSI in the PSR sample was supported. Latinos, as measured by the BSI,
showed the most dramatic improvement. This was sustained for two years.
Meanwhile, their self-esteem scores dropped. Conversely, Whites revealed
significant improvement in their self-esteem scores over time, but showed virtually
no change in the BSI. African Americans, on both measures, remained at near
baseline levels for the duration of the study period.
No differences were found with regard to the clinical variables, nor were
there ethnicity by program interactions.
Grand Summary of Findings bv Program Type
From the perspective of programs, case management was associated with
two positive outcomes: in the subjective domain, males’ BSI scores showed
improvement for the first 2 years of the study period. However, by the end of
three years they were back to near-baseline levels. The ISE for females showed an
overall improvement after three years, but this was followed by an initial
deterioration that extended over approximately 18 months. With regard to the total
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
104
BPRS scores, CM was associated with significant impairment over time for the
females in the study. Males, as a group, showed virtually no change in symptom
level.
In terms of work, there appeared to be no advantages to receiving CM
services. Whites, as a group, did not benefit at all, and the minorities in the study
showed a dramatic decline with this type of intervention. This last finding is of
particular interest. Recall that in the preliminary study (Phillips et al. [in press]),
Latinos who participated in PSR showed a similar pattern throughout the study
period. Whites, on the other hand, demonstrated a positive association between
PSR and work. It was postulated that PSR, due primarily to its intensity, was
detrimental to the buffering effects of the Latinos’ sociocentric culture. However,
the Latinos who received case management services, which is arguably less
demanding, also showed a decline in their work scores.
For those in the PSR sample, outcomes were somewhat more positive, and
include the following:
Scores on the BSI dramatically improved for Latinos for the first two years
of the study period. Some of these gains were lost during the last 12 months, but
they continued to reflect improvement over baseline levels.
Whites showed steady improvement in their self-esteem scores over two
years. During the last 12 months the gains attenuated, but they continued to stay
above baseline levels.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
With regard to the total BPRS score, as well as the Anergia subscale,
females began PSR more impaired than males. However, the overall scores
remained constant for the duration of the study period, and in terms of the Anergia
subscale, they reflected steady improvement.
Finally, in terms of the functional domains that were analyzed in the
preliminary study, Whites experienced considerable gains in the areas of work and
socialization.
PSR was also associated with several negative outcomes. In terms of the
functional variables work and socialization, Latinos appeared to worsen during
their participation in PSR even though they began with the highest levels of social
functioning. In addition, Latinos suffered a reduction in their self-esteem levels
over the three year period.
Finally, with regard to gender differences, the males who participated in
PSR became more symptomatic on both the total BPRS score, as well as the
Disorganization subscale
Short vs. Long Term Treatment
As stated earlier, two types of attrition were factors in this study. One,
treatment attrition, warrants consideration in the evaluation of outcomes over
time. For the purposes of this study those in PSR who exited treatment before the
18th month, the halfway mark, and those who stayed at least 18 months, were
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
106
dichotomized into long and short-term treatment groups. This was then entered
into the 2-level HLM as a covariate, to be examined over time in relation to the
subjective, and clinical outcome variables. Brekke et al. (1997) determined in the
parent study that those who remained in treatment showed sustained
improvements in the areas of work, while those who exited did not. With regard
to the subjective variables, self-esteem and satisfaction with life, no differences
were revealed between the two groups. However, with regard to the BPRS, a
significant quadratic interaction [t(76) = -2.03, p = .04] was found on the Anergia
subset. Anergia, according to Long et al.’s (1999) factor analysis study, includes
blunted affect, uncooperativeness, motor retardation, and emotional withdrawal.
Those who stayed n PSR fo- 18 months or longer were more likely to show an
unstable pattern of change which was reflected in a positive quadratic slope
(decline in symptom level followed by a rise). There were no gender, nor ethnicity
differences in rates of attrition.
For the purpose of this study, analyses were done on the 24 PSR subjects
who remained in treatment for at least 18 months. However, because the number
of subjects for this portion of the analysis was very small, results were included for
exploratory purposes only. Differences were found in the functional domain,
independent living only. Figure 18 presents the predicted growth curves for males
and females on the independent living score using the long-term treatment sample.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
107
There was a trend toward a Gender X Intercept interaction (initial status) (p < .06)
with females having better scores at baseline than males. The linear slope was
significant, t(22) = 2.36; the quadratic slope was a trend (p < . 10). Both slopes
were positive. There was also a significant quadratic Gender by Slope interaction
[t(22) = -2.56, p < .01], Women started out with higher scores, dropped down, and
then rose again to score significantly better than they had initially. Men improved
initially, but then experienced less independence than when they had started. The
average effect size for the significant terms was .46. This lends support to the
hypothesis that there is a relationship between gender and independent living.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Figure 18. Long-term Sample, Independent Living
5 T
4.5
£
o
o
(/)
o >
c
>
□
males
females
3.5
* *
c
0 )
■ o
c
a >
a
a »
xs
c
2.5
2 ----------------------------1 -----------------------------y ------------------------ 1 ------------------------------- (---
1 2 3 4 5
Interview Period (six-month intervals)
DISCUSSION AND IMPLICATIONS
109
This study tested the hypotheses that ethnicity and gender would be related
to a number of prospective outcomes from two models of community-based
interventions. Using data from PSR programs that had already been shown to
result in improved functional, subjective and clinical outcomes (Brekke et al.,
1993; Brekke et al, 1997; Brekke et al., 1999), as well as a CM program, this
study examined cross-ethnic and gender differences in the rates of change over a
three-year period. The study also examined the prospective interactions between
gender and ethnicity variables and program type. It represented the first study to
look at individual client factors in relation to program type within the past 20 years
of psychosocial rehabilitation research. In addition, the study employed HLM
which allows for analysis to begin at the client level, and provided a unique
opportunity to view change as a process, rather than simply as a linear projection.
Gender
The most notable gender differences were found in the clinical domain for
both the PSR and CM models. At baseline, women in PSR were more
symptomatic than the males in the study, but their scores remained stable. Males
in PSR, on the other hand, experienced a worsening of symptoms over three years.
In CM, the reverse was true: females became worse, while the males remained
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
110
stable. These findings were also reflected in a significant program-by-gender
interaction which showed that for males, participation in PSR was associated with
an increase in symptomatology in both the total BPRS score, and the
Disorganization subscale. The CM findings, as well as the baseline PSR findings
which found females at a clinical disadvantage are interesting given that the
preponderance of literature on gender differences in symptomatology has found
the course and outcome in schizophrenia to be worse for males (Lewis, 1992;
Angermeyer, 1989). It is also striking that men and women appeared to have
different responses to the two treatment models.
In terms of the subjective variables, significant differences were found as
well, and appear to be congruent with the clinical findings. The women in CM
showed a decline in their self-esteem scores which endured for the first year of the
study. At the same time, the CM females showed exacerbation in their levels of
distress, which continued for 18 months. These results, coupled with the increase
in overall BPRS symptom levels, call into question the appropriateness of this
treatment model for women. As for the males, their subjective experience was
more positive with this type of intervention, as measured by the BSI. This, too,
was congruent with the clinical findings.
It is helpful at this point to consider the source of the previous literature on
gender differences, and to reflect on the inherent qualities of the two interventions
and how they may have enhanced or impeded their clients’ progress toward
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Ill
recovery. Social role theory is presented as a framework for understanding the
complex relationship between men and women with chronic mental illness and
their different responses to community-based interventions.
With regard to the preponderance of literature favoring females, it should
be noted that studies have rarely looked at symptoms in relation to community-
based interventions, but rather in hospital or other institutionalized settings (see
Andia and Zisook 1991 for a review). This may indirectly explain the more
symptomatic ratings for the females in this study. First of all, recent
epidemiological studies report that the lifetime risk for schizophrenia appears to
be the same for men and women (Hafner & Heiden, 1997; Andia et al., 1991).
However, in the current study the number of men outweighs the number of women
by a 3 to 1 ratio. This suggests that there may have been a selection bias
threatening the composition of the two gender groups, i.e. the women may have
been more disabled than the general population of schizophrenia-disordered
females. Social role theory explains how pre-adult and adult factors can contribute
to very different lives for men and women which would impact their respective
needs for, and subsequent placement in, community treatment.
According to Cohler and Beeler (1996) as young girls women are taught to
be interdependent and relational, while boys are taught to be self-reliant and
independent. As adults, men have been expected to take the lead in courtship and
marriage. Unfortunately, men with chronic psychiatric disorders find it difficult to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
112
maintain social ties and carry on with these adult roles. Women, on the other hand,
even though impaired, are better able to meet society’s expectations in non-
assertive roles. They are rewarded with a stronger network of social and familial
ties, which reduces the need for long-term community treatment. Also, women
are several years older at the first onset of schizophrenia symptoms than men
(Watt and Szulecka, 1979). This difference makes it more likely that women, at
the time they first manifest symptoms, have both a spouse and children that can
serve as caregivers. It is arguable, therefore, that unmarried and more disabled
females are more likely to live in board-and-care facilities and seek case
management or psychosocial rehabilitation interventions.
To understand the nearly-opposite patterns of clinical and subjective
outcomes for men and women in the two treatment modalities requires a closer
look at the inherent nature of each treatment. As the name implies, psychosocial
rehabilitation is structured around social interactions. The majority of treatment
occurs in groups, and there is a strong emphasis placed on acquiring social skills.
Furthermore the activities that are available in the PSR milieu are rigorous and
demanding, with high expectations for achievement placed on its members.
On the other hand, case management takes place primarily between two
people: the client and his or her case manager. The focus is not on enhancing
skills through group interaction, but rather on acquiring and accessing resources in
the community. While the expectations for achievement depend on the unique
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
113
relationship between the two parties, the experience of CM is considerably less
intense than that of PSR.
The social role expectations of greater independence and assertiveness for
males, as well as the lower tolerance for deviant behavior might make the social
interactions required by the PSR milieu more difficult. According to Cohler et al.
(1996), although it may be assumed that participation in community life would
support more effective performance in social roles “community contact may only
enhance the sense of poor morale, burdening these adults with expectations that
they are unable to realize” (p. 749).
Andia et al. (1991) argued that the later age of onset in women enables
them to have better developed personality structures, and more mature defenses
and social skills than men. It is understandable, therefore, that women would
perform better in PSR, which empowers them to use the social skills which have
been developed and nurtured over time. Men, in turn, may experience less distress
and greater symptom stability from the less socially-taxing intervention, case
management.
Ethnicity
Strong cross-ethnic differences were found in both the functional and
subjective domains. The preliminary study revealed that whites showed
considerable gains during PSR in work and social functioning. African Americans
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
114
showed minimal change, and Latinos appeared to worsen during their participation
in this intervention. This same pattern was reflected in the self-esteem scores for
the PSR sample: whites improved significantly over time, African Americans
showed virtually no change, and Latinos self-esteem declined. CM also was
associated with deteriorating work functioning for minorities, while there was no
improvement over time for whites. Interestingly, Latinos in PSR were the only
ones to show improvement in their subjective levels of distress.
There is an assumption often made in the literature that work is associated
with increased self-esteem (Van Dongen, 1996). Certainly the convergent findings
for work and self-esteem support this assumption. As Scheid (1993) stated, “work
is central to the human existence; it provides a valued social position and identity
that affect self-concept and relationships with others” (p. 763). She further argued
that work is a primary source of self-esteem, providing an income and security,
and a structure from which we define ourselves.
In a first-person article in Schizophrenia Bulletin (Leete, 1989) this author
likely speaks for many who suffer from schizophrenia. Referring specifically to
work, she stated, “As I work, I become increasingly more self-confident, and my
self-image is bolstered. I feel important and grownup, which replaces my usual
sense of vulnerability, weakness, and incompetence” (p. 197).
It is interesting and unfortunate that the minorities in the study did not
enjoy the same benefits of improved work status and the accompanying
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
115
improvement in self-esteem. PSR on both outcome measures, as well as
socialization, was ineffectual for African Americans, and perhaps even deleterious
for Latinos. CM showed a similar pattern, although this model does not claim to
have a rehabilitative focus. This raises several critical issues concerning the
cultural relevancy of the community treatment programs which will be addressed
later.
Notwithstanding the decline in previously mentioned outcomes, Latinos
experienced a distinct and positive gain on the BSI, which measures subjective
levels of distress. Recall that the measure covers several symptoms commonly
associated with severe mental illness and the respondents report on how they are
directly experiencing the phenomena. For example, on the item “interpersonal
sensitivity” individuals are asked to report on their feelings of personal inadequacy
and inferiority, as well as discomfort during interpersonal interactions. While it is
perhaps counterintuitive that a decline in self-esteem can be juxtaposed with
improvement in levels of distress, it is possible that we have not been able to
adequately interpret subjective experiences in ethnic minority cultures to this
point. To illustrate, an article in American Psychologist (1996, author not listed)
stated that among cultures that stress interdependent concepts of self, self-esteem
is based more strongly on satisfactory relations with others, as opposed to personal
achievements. It was suggested that we reconceptualize our theories of self
esteem and how to enhance it, and approach it from a sociocultural perspective
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
116
with regard to people from ethnic minority backgrounds. Meanwhile, the dramatic
improvement in BSI scores indicate that the relationship between PSR and Latinos
has some advantages.
Findings from other studies may provide some insight into the differential
effects due to ethnicity in terms of the BSI outcomes. A possible explanation for
Latinos’ improvement can be found in the literature on the acculturation of ethnic
minorities. Briefly, acculturation refers to the psychosocial changes that occur
when individuals originating from one culture immigrate to a new host culture
(Bumam, Hough, Kamo, and Telles, 1987). This concept can be broadened to
include ethnic groups who move from a supportive Mexican American
community, for example, to the majority culture with its conflicting values and
expectations. Distress can occur when members of the minority culture are not
equipped with the knowledge and resources to obtain the goals valued by the new
culture. However, researchers have proposed that there is a curvilinear
relationship between acculturation and mental health (Birman, 1998; Burman et
al., 1987). According to this view, once individuals have endured the initial stress
of adapting to the host culture, there is th opportunity for them to become
bicultural, and successfully navigate the two worlds. Bumam et al. (1987) stated
that “individuals who have integrated both Anglo and Mexican cultural values and
behaviors are best able to adapt to the Anglo sociocultural environment while
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
117
simultaneously retaining the benefits of continuing involvement in the Mexican
American family and community” (p. 90).
Similarly, Flaskerud (1996) asserted that as levels of acculturation to the
Anglo culture increase, emotional and mental well-being improve. While the
process of acculturation may disrupt the emotional and social well-being of those
caught up in the process, it also improves the quality of life for many. The author
proposed that personal and social resources, such as knowledge and skill, as well
as a strong social network serve to mediate the impact of the stress of
acculturation, resulting in positive emotional adjustment. She further stated that
the characteristics of the receiving society may also affect adaptation, particularly
if it includes a subculture that protects the acculturating individual from the
stresses of adapting to the majority culture. Community-based psychosocial
rehabilitation programs, which proclaim to recognize diversity and build on
people’s strengths would qualify as meeting the criteria of a receiving society that
would mediate successful integration.
In this discussion of gender and ethnicity differences, it is also meaningful
to highlight areas where there was no evidence of differential treatment effects.
The important null findings are as follows:
First, no gender differences were found in the functional domains which
included work, socialization, independent living, and hospitalization. Recall that
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
118
in the parent study (Brekke et al., 1997) the investigators found that the
psychosocial rehabilitation programs, in general, out-performed case management
in all four areas. The null findings therefore indicate that there was no pervasive
differential effectiveness for the men or women in the study, and that both sexes
fared better in PSR with regard to these variables. However, this study did find
differences in independent living for those who stayed in PSR for 18 months and
beyond. It is unclear what to make of this due to the small number of women
available for analysis, but it is an important area for future research.
Second, in terms of ethnicity, there were no differences in independent
living and hospitalization outcomes for the PSR group. Neither were there
differences in independent living, hospitalization, and socialization for the CM
group. Again, this indicates that there was not a pervasive differential
effectiveness across ethnic groups on all functional outcomes. Based on the
findings of the parent study, PSR was equally effective for the three ethnicity
groups in terms of independent living and hospitalization.
Third, there were no gender or ethnic differences with regard to SWL.
This self-report measure is intended to quantify subjective satisfaction with one’s
living situation, work, social contacts, and psychological state. This indicates that
the subjects in the study were equally satisfied with life regardless of their levels
of functioning, experience of symptoms, self-esteem, or levels of distress.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
119
Finally, no ethnic differences were found in the total BPRS or its subscales.
This is inconsonant with much of the schizophrenia research which has found the
psychopathology of different cultures and ethnic groups to be expressed
differently. However, given that there was a high rate of medication compliance
across all ethnicity groups, it is not surprising that there were no differential
ethnicity outcomes.
Implications for Practice
For men, the symptoms of schizophrenia become manifest precisely at the
time when male role expectations demand that they succeed in work and build
meaningful sexual relationships. Mental health: viders have the opportunity to
help them rebuild their self-esteem and to view their futures with hope. This can
be achieved through providing role models, possibly via peer mentors or peer
counselors. Hiring people with schizophrenia to assist people more disabled than
they, can be immensely empowering for both parties. People with schizophrenia
are as likely as “normal” individuals to buy into the stereotype that schizophrenia
is an enduring, disabling disease from which people cannot recover. Working
closely with someone with the same diagnosis who is effectively meeting his or
her responsibilities and earning an income can help to alleviate the despairing
notion that schizophrenia is ultimately and permanently disabling.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
120
If indeed social role expectations place a burden on men in community-
based treatment, the role of practitioners would be to rethink the definition of what
it means to be a man (or a woman), and question the culturally accepted “macho”
image. Education groups, or staff-facilitated process groups could easily be
incorporated into the PSR milieu, and give men and women the opportunity to find
new ways of living and relating. In addition, practitioners could provide symptom-
management groups, which would explore the effects of work and socialization,
for example, on clients’ symtomatology and overall feelings of well-being.
It is also important for men, as well as women, to find a meaningful role in
their lives that transcends the role of “mental patient.” In our society, these are
most often found in work and relationships. Psychosocial rehabilitation models,
including the one in this study, are beginning to attend to the needs of people to
work and to develop social skills. However, it is inappropriate and ineffectual to
encourage the development of relationships without addressing the fact that people
with schizophrenia are sexual beings. If not attended to, sex can become “the
elephant in the room” that everyone tiptoes around, hoping not to awaken. Lifting
the taboo of sex can free men to talk about and find solutions to problems such as
impotency, a common side effect of psych tropic medication. A forum is
recommended where men can share with each other the difficulties inherent in
establishing intimate relationships. American society has come far toward
equalizing the responsibilities between the two sexes. However, men more often
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
121
than not, continue to bear the burden of being the one who pursues, who takes the
initiative, and who finances the dinner and movie.
It was stated earlier that women were underrepresented in the study, which
could be due to a number of undiscovered factors. But keeping in line with what is
known about the benefits that social role expectations incur on females in our
society, it is likely that they were either being cared for by family and friends, or
they themselves were at home caring for children. In either case, the community of
caregivers has an obligation to discover what the unique needs are of women, and
to make readily available their services as requested. This will likely call for
innovative outreach interventions, and perhaps incentives such as on-site child
care for the clients’ offspring.
Social and sexual relationships, pregnancy, childbirth, and parenting are
major issues for most women. In a study by Test et al. (1990), 38% of the female
participants had children, and most were parenting them. The writers’
observations suggested that the stresses and burdens on the mothers were
enormous. They were single mothers for the most part, poor, and suffering the
symptoms of severe mental illness. It is arguable that they view community-based
interventions, if they are aware of them at all, as an indulgence they do not have
time for. Mental health professionals can help engage these individuals in
treatment by embracing the spectrum of needs with which they present. This may
involve providing respite care for a few hours each week from parenting duties, or
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
122
holding women’s groups where they can gain support from each other and reduce
their feelings of isolation. In addition, providers could develop a forum in which to
discuss relevant issues such as contraception, and the risks of contracting sexually
transmitted diseases.
PSR lends itself to multiple solutions to the problems women face. By its
very nature it already appeals to many of the social and relationship needs of its
female consumers. It also can take advantage of its staff and consumer resources
to broaden its menu of therapeutic tools, including child care, for example.
Traditional case management, on the other hand, lacks the structure and the
supportive community that women respond to. However, case management is
associated with reduced levels of distress for men, and symptom stability. Perhaps
the ultimate therapeutic benefit can be had by linking the two services. By way of
illustration, case managers can refer male clients to PSR programs that will stretch
their tolerance for social interactions and address their very human need for
intimate partnerships. In the field, they could assess the needs of single mothers,
and include a PSR socialization group in their list of referrals.
In terms of ethnicity, several practice implications are apparent. The fact
that the Latinos in the PSR model showed a decrease in their levels of distress
implies this model of treatment is providing some benefit for this group. It was
suggested that in this milieu there is the opportunity for mental health
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
123
professionals to help individuals to integrate both Anglo and Latino cultural
values. In this way, they are able to be proficient in both cultural systems and
switch between them as the situation requires. According to psychosocial
rehabilitation principles, this is accomplished through an attitude of acceptance
and valuation of the clients’ original culture, and an openness and willingness to
share on the part of members from the majority culture. This, in turn, can lead to
subjective feelings of well-being.
On the other hand, the Latinos in PSR and CM were unsuccessful in terms
of work. It was discussed in the preliminary study that the ideology of the PSR
intervention was based on Western behavioral models that apply the values of
individualism— independence, self-reliance, rationalism, competitiveness, mastery
over one’s destiny, autonomous action, and emotional detachment (Phillips et al.,
in press). The implication was that community-based interventions subscribed to
in this model could be noxious to participants’ sociocentric cultural orientation. It
is also possible that ethnic minorities have different attitudes and orientations
toward work, which should be explored in the context of vocational rehabilitation,
and before mental health workers impose their expectations for work performance.
Scheid (1993) presented the argument that there is a contradictory nature
of work in that, on one hand, it is fundamental to self-identity. On the other hand
it can be alienating, and be perceived as an instrument of harm and oppression.
She asserted that psychiatric clients are often employed in low-pay, menial jobs
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
124
which can undermine the value of the work itself. It might be that minorities who
are struggling with the hardships incumbent upon interfacing with the majority
culture, as well as dealing with a disabling mental illness, are making an adaptive
choice to withdraw from work that is experienced as dehumanizing. In any event,
PSR vocational counselors have an obligation to discuss with their clients what it
means for them to work, and help them to use their talents in ways that are
productive, as well as meaningful.
It is notable that neither CM nor PSR was effective for the African
Americans in the study. On all cross-ethnic analyses, the outcomes that showed
differential effectiveness for the three groups showed that African American
clients appeared unaffected either positively or negatively for the duration of the
study. At best, the community interventions employed in this study were helpful in
maintaining their African American clients at baseline levels, but it does not
appear that PSR achieved its mission of enhancing interpersonal and role
functioning, decreasing symptom severity, and improving illness management
(Moore, 1990). However, recall that on several of the measures, including
hospitalization, independent living, satisfaction with life, and the BPRS, there
were no ethnic differences.
Whatever the contributing factors to the ethnicity differences related here,
the fact that they exist, and that, by and large they favor members of the majority
culture, is indicative of the need to develop culturally responsive treatment
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
125
programming. Several suggestions for improving the delivery of services to
minority clients have been suggested. Barrio (2000) offered the following:
Services should be culturally appropriate and effectiveley coordinated to ensure
continuity of care. They should be individualized and flexible, incorporate natural
supports, and focus on strengths. Service providers should address the cultural
orientations of individuals by honoring the attitudes, values, and behaviors of
individuals, and recognize howr these influence the use and outcome of services.
Interventions for individuals from sociocentric cultures should incorporate family
networks and use culturally congruent group modalities.
Other recommendations have been to provide more ethnic and bilingual
staff; and to educate service providers with the goal of eliminating biases,
stereotypes, and prejudices which result in culturally destructive or ineffective
practices and structures (Kopelowicz, 1998; Sims et al., 1996).
Speaking specifically for African Americans, Plummer (1996) stated that
service providers must begin treatment with an accurate assessment. She suggested
that cultural differences in the use of language, mannerisms, communication
styles, and expressions of distress often lead to misdiagnoses. “Because one’s
culture influences his or her behavior and thinking, this cultural world view
translates into one’s manner o f expression and language” (p. 38). For example,
confusion, impulsiveness, and disorientation placed in a cultural context may not
represent symptoms of schizophrenia, but could be more appropriately interpreted
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
126
as depression. Also, including the aspect of culture in diagnostic measures raises
the awareness of differences, and sets the stage for culturally responsive treatment
planning.
Plummer (1996) also endorsed the idea of involving local churches, as well
as the local police departments in the psychosocial rehabilitation milieu.
Regarding the latter, the writer emphasized the importance of building trust on
both sides. Educating members of the justice system about the symptoms and
management of chronic mental illness reduces the likelihood that policemen will
be abusive to psychiatrically disabled citizens.
Finally, she recommended that service providers understand their own
culture and the influence it has on their behavior and interpretations of reality.
This will assist them in understanding how racial identity functions in other
individuals (Plummer, 1996).
Policy Implications
At the policy level, it is important that we advocate for the funding of
culture and gender-specific services for clients, as well as comprehensive training
programs for the service providers. In addition, it is essential to maintain and
develop new funding sources that will support research in gender and ethnic
differences in how mental illness is experienced and it can be optimally treated.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
127
This will require several strategies, including the construction of grant proposals,
and persuasion of local and federal legislators.
Strengths and Limitations
Although schizophrenia is among the most heavily studied disorders in
cross cultural and gender research, this was the first to look at these variables
longitudinally with such multifaceted assessment measures, and in combination
with two distinct community treatment models. The number of significant
findings validated that this is indeed an important direction for schizophrenia
research if we are to improve the delivery of services to men and women from
various ethnic groups.
This study used a tri-ethnic design which allowed for a richer picture of
ethnic differences than the minority/non-minority delineation, and reduced the
problem of social class distinction. Still, a limitation of the study was the
relatively small Latino representation.
The differences in clinical outcomes cannot be explained by medication
compliance, because it was established that there were no gender or ethnicity
differences in number of days on medication. Also, people who had a diagnosis of
substance abuse disorder were excluded from the study. This can be viewed as a
strength, as well as a limitation. Persistent use of alcohol and other illicit drugs can
make psychotropic medications ineffectual, which would have confounded the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
128
data on medication compliance. Given that, however, it is important to note that
these results cannot be generalized to dually diagnosed individuals, who comprise
a significant portion of the schizophrenia population.
Before concluding, a few methodological caveats need to be considered.
First, the study did not use a random sample of subjects from the community;
therefore, the study sample cannot be said to represent the population at large. To
illustrate, Latinos, according to the 1990 US census comprise 38% of Los Angeles
County. They represented only 18% of the study sample. It is possible that the
Latino clients in community rehabilitation programs are systematically different
from those who are treated at home or by private physicians. Females, as well,
were underrepresented. African Americans, who comprise 11% of the general
population in Los Angeles were over-represented.
Second, the equivalence of the different ethnic and gender groups was
established at baseline on numerous variables that have been related to outcomes
in schizophrenia (Brekke et al., 1997). However, there is always the possibility
that the groups are different in ways that have affected the findings to some
degree.
Third, the treatment attrition rate was high. It is therefore difficult to
interpret the quadratic slopes. For example, after many months of improvement on
the BSI, women in PSR started to show a return to baseline distress levels. This
raises the question of whether this was a consequence of their leaving the program
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
129
or, could it have been related to an aspect of treatment? Similarly, after two years
of improvement, the self-esteem scores for Latinos changed directions. Perhaps
this was due to attrition. And if so, perhaps the scores would have continued to
deteriorate if they had stayed in treatment. These questions remain unanswered.
The sample sizes for the Latino and African American groups were small,
particularly for the case management intervention. This lack of statistical power
could have resulted in type II errors. To increase power, the groups were
collapsed into minority/non-minority groups. However, this eliminated the
possibility of detecting critical differences between the distinct ethnic minority
groups.
Finally, the PSR model was treated as one variable under the heading of
community-based psychosocial rehabilitation. In fact, PSR consisted of two
separate rehabilitation facilities, each with its own strengths and limitations. To
illustrate, CLP had virtually no vocational rehabilitation component^ as opposed to
Portals, whose vocational program was well-developed. On the other hand CLP
prioritized training that addressed the acquisition of independent living skills.
Merging the two programs could have diluted pronounced changes that may have
been revealed had the study looked at specific services in relation to outcome.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
130
Future Directions
This study represented only an initial phase in the search for individual
factors that predict outcome from community-based psychosocial rehabilitation.
Future research should continue to identify specific client characteristics, and
examine them in relationship to different models of community care. It would also
be of value to examine culture-specific and gender-specific services in relation to
outcome. These might include models that offer innovative cultural diversity
training to its employees, as well as models that employ more family involvement,
offer child care and parenting classes, or provide groups that address gender-
specific problems in establishing intimate relationships. Matching subjects on
background variables, such as marital status, or level of acculturation would be
useful in determining true differences in future quasi-experimental designs.
Finally, research is needed using larger sample sizes to replicate these results.
Replications across program models as well, are essential to establish the
generalizability of these findings.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
131
REFERENCES
Al-Issa, I. (1995). Handbook o f Culture and Mental Illness: An International
Perspective. Madison: International Universities Press.
Alverson, h., & Vicente, E. (1998). An ethnographic study of vocational
rehabilitation for Puerto Rican Americans with severe mental illness. Psychiatric
Rehabilitation Journal, 22(1), 69-72.
Alvidrez, J. (1999). Ethnic variations in mental health attitudes and service use
among low-income African America, Latina, and European American young
women. Community Mental Health Journal, 36(6), 515-530.
Andia, A. M., & Zisook, S. (1991). Gender differences in schizophrenia. Annals o f
Clinical Psychiatry, 5(4), 333-340.
Angermeyer, M. C., Goldstein, J. M., & Kuhn, L. (1989). Gender differences in
schizophrenia: Rehospitalization and community survival. Psychological
Medicine, 19, 365-382.
Angermeyer, M. C., Kuhn, L., & Goldstein, J. M. (1990). Gender and the course of
schizophrenia: Differences in treated outcomes. Schizophrenia Bulletin, 16, 293-
308.
Aquila, R., Santos, G., Malamud, T. J., & McCrory, D. (1999). The rehabilitation
alliance in practice: The clubhouse connection. Psychiatric Rehabilitation
Journal, 25(1), 19-23.
Attkinsson, C. (1992). Clinical services research. Schizophrenia Bulletin, 18(A),
561-626.
Baker, F. M., Stokes-Thompson, J., Davis, O. A., Gonzo, R., & Hishinuma, E. S.
(1999). Two-year outcomes of psychosocial rehabilitation of black patients with
chronic mental illness. Psychiatric Services, 50(4), 535-539.
Baronet, A., & Gerber, G. J. (1998). Psychiatric rehabilitation: Efficacy of four
models. Clinical P sychology Review, 18, 189-228.
Barrio, C. (1998). A longitudinal cross-ethnic study o f sociocentricity in
schizophrenia. Unpublished Dissertation, University of Southern California, Los
Angeles.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
132
Barrio, C. (2000). The cultural relevance of community support programs.
Psychiatric Services, 51(7), 879-883.
Beard, J. H., Propst, R. N., & Malamud, T. J. (1982). The Fountain House model
of rehabilitation. Psychosocial Rehabilitation Journal, 5 ,47-53.
Bell, P. H., Lysaker, P. H., & Milstein, R. M. (1996). Clinical benefits of paid
work activity in schizophrenia. Schizophrenia Bulletin, 22(1), 51-67.
Bergin, A. E., & Garfield, S. L. (Eds.). (1995). Handbook o f Psychotherapy and
Behavior Change (4 ed.). New York: John Wiley and Sons, Inc.
Birman, D. (1996). Biculturalism and perceived competence of Latino immigrant
adolescents. American Journal o f Community Psychology, 26(3), 335-354.
Blank, M. B., Jodi, K. M., & McCall, B. R. (1996). Psychosocial rehabilitation
program characteristics in urban and rural areas. Psychiatric Rehabilitation
Journal, 20(1), 3-10.
Bond, G. R., McGrew, J. H., & Fekete, D. M. (1995). Assertive outreach for
frequent users of psychiatric hospitals: A meta-analysis. Journal o f Mental Health
Administration, 22, 4-16.
Bond, G. R., Miller, L. D., Krumwied, R. D., & Ward, R. S. (1988). Assertive case
management in in three CMHCs: A controlled study. Hospital and Community
Psychiatry, 39, 411-418.
Brekke, J., Levin, S., Wolkon, G., Sobel, G., & Slade, E. (1993). Psychosocial
functioning and subjective experience in schizophrenia. Schizophrenia Bulletin,
19, 599-608.
Brekke, J., & Test, M. A. (1992). A model for measuring the implementation of
Community Support Programs: Results from three sites. Community Mental
Health Journal, 2 8 ,227-247.
Brekke, J. S., Ansel, M., Long, J., Slade, E., & Weinstein, M. (1999). Intensity and
continuity of services and functional outcomes in the rehablitation of persons with
schizophrenia. Psychiatric Services, 50, 248-252.
Brekke, J. S., & Barrio, C. (1997). Cross-ethnic symptom differences in
schizophrenia: The influence of culture and minority status. Schizophrenia
Bulletin, 23(2), 305-316.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
133
Brekke, J. S., & Long, J. D. (in press). Community-based psychosocial
rehabilitation and prospective change in functional, clinical, and subjective
experience variables in schizophrenia. Schizophrenia Bulletin.
Brekke, J. S., Long, J. D., Nesbitt, N., & Sobel, E. (1997). The impact of service
characteristics on functional outcomes from community support programs for
persons with schizophrenia: A growth curve analysis. Journal o f Consulting and
Clinical Psychology, (55(3), 464-475.
Brunette, M., & Drake, R. E. (1998). Gender differences in homeless persons with
schizophrenia and substance abuse. Community Mental Health Journal, 4(6), 627-
642.
Bryk, A. S., & Raudenbush, S. W. (1987). Application of hierarchical linear
models to assessing change. Psychological Bulletin, 101, 147-158.
Bumam, M. A., Hough, R. L., Kamo, M., Escobar, J. I., & Telles, C. A. (1987).
Acculturation and lifetime prevalence of psychiatric disorders among Mexican
Americans in Los Angeles. Journal o f Health and Social Behavior, 28, 89-102.
Butcher, J. N., Braswell, L., & Raney, D. (1983). A cross-cultural comparison of
American Indian, Black, and White inpatients on the MMPI and Presenting
Symptoms. Journal o f Consulting and Clinical Psychology, 5/(4), 587-594.
Canino, I. A., Rubio-Stipec, M., Canino, G., & Escobar, J. I. (1992). Functional
Somatic Symptoms: A cross-ethnic comparison. American Journal o f
Orthopsychiatry, 62(4), 605-612.
Childers, S. E., & Harding, C. M. (1990). Gender, premorbid social functioning,
and long-term outcome in DSM III schizophrenia. Schizophrenia Bulletin, 16, 309-
318.
Chu, C„ Sallach, H. S., Zakeria, S. A., & Klein, H. E. (1985). Differences in
psychopathology between African-American and white schizophrenics.
International Journal o f Social Psychiatry, 31, 252-257.
Cohler, B. J., & Beeler, J. A. (1996). Schizophrenia and the life course:
Implications for family relations and caregiving. Psychiatric Annals, 26(12), 745-
756.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
134
Dassori, A. M., Miller, A. L., & Saldana, D. (1995). Schizophrenia among
Hispanics: Epidemiology, phenomenology, course, and outcome. Schizophrenia
Bulletin, 2/(2), 303-312.
Derogatis, L. R., & Melisaratos, N. (1983). The Brief Symptom Inventory: An
introductory report. Psychological Medicine, 13, 595-605.
Dongen, V. (1996). Quality of life and self-esteem in working and non-working
persons with mental illness. Community Mental Health Journal, 32(6), 535-548.
Drake, R. E., & Mueser, K. T. (2000). Psychosocial approaches to dual diagnosis.
Schizophrenia Bulletin, 26(1), 105-118.
Endicott, J., & Spitzer, R. (1978). A diagnostic interview: The schedule for
affective disorders and schizophrenia. Archives o f General Psychiatry, 5, 837-844.
Escobar, J. I. (1987). Cross-cultural aspects of the somatization trait. Hospital and
Community Psychiatry, 38(2), 189-196.
Fabrega, H., Mezzich, J., & Ulrich, R. F. (1998). African-American-White
differences in psychopathology in an urban psychiatric population. Comprehensive
Psychiatry,, 29(3), 285-297.
Farmer, R. L., & Pandurangi, A. K. (1997). Diversity in schizophrenia: Toward a
richer biopsychosocial understandingfor social work practice. Health and Social
Work, 22(2), 109-116.
Flaskerud, J. H., & Uman, G. (1996). Acculturation and its effect on self-esteem
among immigrant Latina women. Behavioral Medicine, 22(3), 123-133.
Francis, D. J., Fletcher, J. M., Stuebing, K. K., Davidson, K. C., & Thompson, N.
M. (1991). Analysis of change: Modeling individual growth. Journal o f
Consulting and Clinical Psychology, 59(1), 27-37.
Franklin, J. L., Solovitz, B., Mason, M., & Miller, G. (1987). An evaluation of case
management. American Journal o f Public Health, 77, 674-678.
Gaines, S. O., Marelich, W. D., Bledsloe, K. L., Steers, W. N., Henderson, M. C.,
& Granrose, C. S. (1997). Links between race/ethnicity and cultural values as
mediated by racial/ethnic identity and moderated by gender. Journal o f
Personality and Social Psychology, 72(6), 1460-1473.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
135
Garfield, S. L. (1995). Handbook o f Psychotherapy and Behavior Change (4th
ed.). New York: John Wiley and Sons, Inc.
Gift, T. E., Harder, D. W., Ritzier, B. A., & Kokes, R. F. (1985). Sex and race of
patients admitted for their first psychiatric hospitalization: Correlates and
prognostic power. American Journal o f Psychiatry, 742(12), 1447-1449.
Goodman, S. H., Seawell, D. R., Cooley, E. L., & Leavitt, N. (1993). Assessing
levels of adaptive functioning: The role functioning scale. Community Mental
Health Journal, 29, 119-131.
Gorey, K. M., Lesley, D. R., Morris, T., Carruthers, W. V., John, L., & Chacko, J.
(1998). Effectiveness of case management with severely and persistently mentally
ill people. Community Mental Health Journal, 34, 241-250.
Green, R. S., & Gracely, E. J. (1992). Selecting a rating scale for evaluating
services to the chronically mentally ill. Community Mental Health Journal, 28(3),
163-180.
Guamaccia, P., & Parra, P. (1996). Ethnicity, social status, and families'
experiences of caring for mentally ill family members. Community Mental Health
Journal, 32(3), 243-260.
Haas, G. L., Glick, I. D., Clarkin, J. F., Spencer, J. H., & Lewis, A. B. (1990).
Gender and schizophrenia outcome: A clinical trial of an inpatient family
intervention. Schizophrenia Bulletin, 16(2), 277-292.
Hafiner, H., Heiden, W., Behrens, S., Gattaz, W. F., Hambrecht, M., Loffier, W.,
Maurer, K., Munk-Jorgensen, P., Nowotny, B., Reicher-Rossler, A., & Stein, A.
(1998). Causes and consequences of the gender difference in age of onset of
schizophrenia. Schizophrenia Bulletin, 24(1), 99-113.
Hafner, H., Maurer, K., Loffier, W., & Riecher-Rossler, A. (1993). The influence
of age and sex on the onset and early course of schizophrenia. British Journal o f
Psychiatry,, 162, 80-86.
Hambrecht, H., Maurer, K., Hafher, H., & Sartorius, N. (1992). Transitional
stability of gender differences in schizophrenia? European Archives o f psychiatry
and Clinical Neuroscience, 242, 6-12.
Hedlund, J. L., & Vieweg, B. W. (1980). The Brief Psychiatric Rating Scale
(BPRS): A comprehensive review. Journal o f Operational Psychiatry, 11, 48-65.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
136
Heinrichs, D. W., Hanlon, T E., & Carpenter, W. T. J. (1984). The Quality of Life
Scale: An instrument for rating the schizophrenia deficit syndrome. Schizophrenia
Bulletin, 10, 388-398.
Holland, P. M., Zipple, A. M., & Batscha, C. L. (1993). Borderline personality
disorder in psychosocial rehabilitation programs. Innovations and Research, 2(4),
35-40.
Hollingshead, A., & Redlich, F. (1958). Social class and mental illness. New
York: Wiley.
Homstra, R. K., Bruce-Wolfe, V., Sagduyu, K., & Riffle, D. W. (1993). The effect
of intensive case management on hospitalization of patients with schizophrenia.
Hospital and Community Psychiatry, 44(9), 844-847.
Hudson, W. W. (1982). The Clinical Measurement Package. Pacific Grove, CA.
Pacific Grove, CA: Brooks/Cole.
Jablensky, A., Sartorius, N., & Emberg, G. (1992). Schizophrenia: Manifestations,
incidence and course in different cultures. A World Health organization ten-
country study. Psychological Medicine, Monograph Suppl. 20.
Jerrell, J. M., & Wilson, J. L. (1997). Ethnic differences in the treatment of dual
mental and substance abuse disorders: A preliminary analysis. Journal o f
Substance Abuse Treatment, 14(2), 133-140.
Kamo, m., Jenkins, J., De La Selva, A., Santana, F., Telles, C., Lopez, S., & Mintz,
J. (1987a). Expressed emotion and schizophrenic outcome among Mexican-
American families. Journal o f Nervous and Mental Disease, 175(3), 143-151.
Kamo, M., & Jenkins, J. H. (1993). Cross-cultural issues in the course and
treatment of schizophrenia. Psychiatric Clinics o f North America, 16(2), 339-350.
Kamo, M., Jenkins, J. H., De La Selva, A., Santana, F., Telles, C., Lopez, S., &
Mintz, J. (1987b). Expressed emotion and schizophrenic outcome in Mexican-
American families. Journal o f Nervous and M ental Disease, 175, 143-151.
Kopelowitz, A. (1998). Adapting social skills training for Latinos with
schizophrenia. International Review o f Psychiatry, 70(1), 47-50.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
137
Latimer, E. (1999). Economic impacts of assertive community treatment: A
review of the literature. Canadian Journal o f Psychiatry, 44, 443-454.
Leete, E. (1989). How I perceive and manage my illness. Schizophrenia Bulletin,
18(2), 197-200.
Leff, J., & Vaughn, C. (1985). Expressed Emotions in Families. New York:
Guilford Press.
Lefley, H. P. (1985). Culture and chronic mental illness. Hospital and Community
Psychiatry, 41(3), 277-286.
Lehman, A. F., & Steinwachs, D. M. (1998). Patterns of usual care for
schizophrenia: Initial results from the schizophrenia patient outcomes research
team (PORT) client survey. Schizophrenia Bulletin, 24(11-20).
Lewis, M. S. (1989). Age incidence and schizophrenia, Part I. The season of birth
controversy. Part II. Beyond age incidence. Schizophrenia Bulletin, 89, 59-80.
Lewis, S. (1992). Sex and schizophrenia: Viva la difference. British Journal o f
Psychiatry, 161,445-450.
Lieberman, J. A., Alvir, J. M., Koreen, A., Geisler, S., Chakos, M„ Sheitman, B.,
& Woemer, M. (1996). Psychobiologic correlates of treatment response in
schizophrenia. Neuropsychopharmacology, 14, 13S-21S.
Lin, K. M., & Kleinman, A. M. (1988). Psychopathology and clinical coursr of
schizophrenia: A cross-cultural perspective. Schizophrenia Bulletin, 14(4), 555-
567.
McPheeters, H. L. (1984). Statewide mental health outcome evaluation: A
perspective of two southern states. Community Mental health Journal, 20 ,44-55.
Meyers, J., & Bean, L. L. (1968). A Decade Later: A Follow-up o f Social Class
and Mental Illness. New York: Wiley.
Moore, S. T. (1990). A social work practice model of case management. Social
work, 35(5), 444-448.
Mueser, K. T., Bellack, A. S., Morrison, R. C., & Wade, J. H. (1990). Gender,
social competence, and symptomatology in schizophrenia: A longitudinal
analysis. Journal o f Abnormal Psychology, 99(2), 18-147.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
138
Mueser, K. T., & Bond, G. R. (2000). Psychosocial treatment approaches for
schizoophrenia. Current Opinion in Psychiatry, 13(1), 27-35.
Mueser, K. T., Bond, G. R., Drake, R. E., & Resnick, S. G. (1998). Models of
community care for severe mental illness: A review of research on case
management. Schizophrenia Bulletin, 24(1), 37-74.
NIMH. (1991). Caring for People With Severe Mental Disorders: A National Plan
of Research to Improve Services. Superintendent of Documents . Washington,
D.C.: U. S. Government Printing Office.
Norman, R. M. J., & Malla, A. K. (1991). Subjective stress in schizophrenia.
Social Psychiatry and Psychiatric Epidemiology, 26, 212-216.
O'Sullivan, M. J., Peterson, P. D., Cox, G. B., & Kirkeby, J. (1989). Ethnic
populations: Community mental health services ten years later. American Journal
o f Community Psychology, 17, 17-30.
Overall, J. E., & Gorham, D. R. (1962). The Brief Psychiatric Rating Scale.
Psychology Report, 10, 977-812.
Penn, D. L., Mueser, K. T. (1996). Research update on the psychosocial treatment
of schizophrenia. American Journal o f Psychiatry,, 15(5), 607-617.
Penn, D. L., Mueser, K. T., & Spaulding, W. (1996). Information processing,
social skills, and gender in schizophrenia. Psychiatry Research, 59(3), 213-220.
Pfeiffer, S., O'Malley, D. S., & Shott, S. (1996). Factors associated with the
outcome of adults treated in psychiatric hospitals: A synthesis of findings.
Psychiatric Services, 47(3), 263-269.
Phinney, J. S. (1996). When we talk about American ethnic groups, what do we
mean? American Psychologist, 51(9), 918-927.
Plummer, D. L. (1996). Developing culturally responsive psychosocial
rehabilitative programs for African Americans. Psychiatric Rehabilitation
Journal, 19(4), 37-43.
Rabkin, J. G. (1980). Stressful life events and schizophrenia: A review of the
literature. Psychological Bulletin, 87,408-425.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
139
Rapp, C. T. (1998). The active ingredients of effective case management: A
research synthesis. Community Mental Health Journal, 34, 363-380.
Ruiz, P. (1985). Clinical care update: The minority patient. Community Mental
Health Journal, 21( ., 208-2 6.
Salem, J. E., & Kring, A. M. (1998). The role of gender differences in the
reduction of etiologic heterogeneity in schizophrenia. Clinical Psychology Review,
18, 795-819.
Santos, A. B., Hawkins, G. D., Julius, B., Deci, P. A., Hiers, T. H., & Bums, B. J.
(1993). A pilot study of assertive community treatment for patients with chronic
psychotic disorders. American Journal o f Psychiatry, 150(3), 501-504.
Sartorius, N., Jablensky, A., Korten, A., Emberg, G., Anker, M., Cooper, J. E., &
Day, R. (1986). Early manifestations and first contact incidence of schizophrenia
in different cultures. Psychological Medicine, 16, 909-928.
Scheid, T. L. (1993). An investigation of work and unemployment among
psychiatric clients. International Journal o f Health Services, 23(4), 763-781.
Scott, J. E., & Dixon, B. (1995). Psychological interventions for schizophrenia.
Schizophrenia Bulletin, 4, 621-660.
Searles, R. T. (1993). Faces o f Culture (4th ed.). Fort Worth: Harcourt Brace.
Sims, E. M. (1998). Principles of multicultural psychiatric rehabilitation services.
Psychiatric Rehabilitation Journal, 21(3), 219-223.
Smith, T. E., Hull, J. W., Anthony, D. T., Goodman, M., Hedayat-Harris, A.,
Felger, T„ Kentros, M. K„ MacKain, S. J., & Romanelli, S. (1997). Post
hospitalization treatment adherence of schizophrenic patients: Gender differences
in skill aquisition. Psychiatry Research, 6> } , 123-129.
Snowden, L. R., & Cheung, F. K. (1990). Use of mental health services by
members of ethnic minority groups. American Psychologist,, 45(3), 347-355.
Solomon, P. (1992). The efficacy of case management services for severely
mentally disabled clients. Community Mental Health Journal, 28(3), 163-180.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
140
Stein, L. 1., & Test, M. A. (1980). Alternatives to mental hospital treatment:
Conceptual model treatment program and clinical evaluation. Archives o f General
Psychiatry, 37, 392-397.
Strauss, J. S., & Carpenter, W. T. (1972). The prediction of outcome in
schizophrenia: I. Characteristics of outcome. Archives o f General Psychiatry, 27,
739-746.
Strauss, J. S., & Estroff, S. E. (1989). Subjective experience of schizophrenia and
related disorders: Implications for understanding and treatment. Schizophrenia
Bulletin, 15, 177-178.
Sue, S. (1976). Client's demographic and therapeutic treatment: Differences that
make a difference. Journal o f Consulting and Clinical Psychology,, 44, 864.
Sue, S. (1977). Community mental health services to minority groups: Some
optimism, some pessimism. American Psychologist, 2 ,616-624.
Sue, S., McKinney, H., Allen, D., & Hall, J. (1974). Delivery of community
mental health services to black and white clients. Journal o f Consulting and
Clinical Psychology, 42, 794-801.
Telles, C., Kamo, M., Mintz, J., Paz, G., Arias, M., Tucker, D., & Lopez, S.
(1995). Immigrant families coping with schizophrenia: Behavioral family
intervention with a low-income Spanish-speaking population. British Journal o f
Psychiatry, 167(4), 473-479.
Tennant, C. (1985). Stress and schizophrenia: A review. Integrative Psychiatry, 3,
248-261.
Test, M. A., & Berlin, S. B. (1981). Issues of special concern to chronically
mentally ill women. Professional Psychology, 12, 136-145.
Test, M. A., Burke, S. S., & Wallisch, L. S. (1990). Gender differences of young
adults with schizophrenic disorders in community care. Schizophrenia Bulletin,
16, 331-344.
Test, M. A., Knoedler, W. H., Allness, D. J., & Burke, S. S. (1985). Characteristics
of young adults with schizophrenic disorders treated in the community. Hospital
and Community Psychiatry, 36(8), 853-858.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
141
Test, M. A., Knoedler, W. H., Allness, D. J., Burke, S. S., Brown, R. L., &
Wallisch, L. S. (1991). Long-term community care through an assertive continuous
treatment team. In C. Tamminga & S. Schultz (Eds.), Advances in
Neuropsychiatry and Psychopharmacology, Volume I: Schizophrenia Research .
New York: Raven Press.
Thomson, J., Davis, O. A., Gonzo, R., & Hishinuma, E. S. (1999). Two-year
outcomes of psychosocial rehabilitation of black patients with chronic mental
illness. Psychiatric Services, 50(4), 535-539.
Valesquez, R. J., & Callahan, W. J. (1990). MMPI's of Hispanic, Black, & White
DSM-III schizophrenics. Psychological Reports, 66(3), 819-822.
Vandiver, V. L. (1998). Quality of life, gender, and schizophrenia: A cross
national survey in Canada, Cuba, and U.S.A. Community Mental Health Journal,
34(5), 501-511.
Vaughn, S., Snyder, K. S., Jones, S., Freeman, W. B., & Falloon, I. R. H. (1984).
Family factors in schizophrenic relapse: A California replication of the British
research on expressed emotion. Archives o f General Psychiatry, 41, 1169-1177.
Ventura, J., Nuechterlein, K. H., Subotnik, K., & Gilbert, E. (1995). Symptom
dimensions in recent-onset schizophrenia. Paper presented at the International
Congress on Schizophrenia Research, Hot Springs, VA.
Watkins, K. E., Shaner, A., & Sullivan, G. (1999). The role of gender in engaging
the dually diagnosed in treatment. Community Mental Health Journal, 55(2), 115-
126.
WHO. (1979). Schizophrenia. An International Follow-up Study . Wiley,
Chichester.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Impact of housing arrangements on social support and health status among Chinese American elderly

PDF

Childhood videotaped neuromotor and social precursors of schizophrenia: A prospective investigation

PDF

Client characteristics, service characteristics, and psychosocial outcomes in the community treatment of schizophrenia

PDF

Gay men with eating disorders and food, body image, and exercise concerns: A group treatment approach

PDF

A longitudinal cross-ethnic study of sociocentricity in schizophrenia

PDF

Family functioning, intergenerational conflict, and psychological symptomology of Asian American college students

PDF

Associations and mechanisms among attention deficit hyperactivity disorder symptoms, cognitive functioning, and drinking habits

PDF

A methodology to identify high -risk patients with diabetes in the California Medicaid populations (Medi -Cal)

PDF

Control and dynamics during horizontal impulse generation

PDF

A Positive Psychology goal intervention using strengths and values to enhance goals and increase well -being

PDF

Education and intelligence test scores: Predictors of dementia?

PDF

A cluster analysis of the differences between expert and novice counselors based on real time training

PDF

Atypicality: Benefit and bane. Provocation, trigger, typicality, and the expression of aggression and generalization

PDF

Assessment of fatigue as a late effect of therapy among survivors of childhood leukemia

PDF

Gender and the bump: An investigation of the reminiscence effect in the Long Beach Longitudinal Study

PDF

A foundation for general-purpose natural language generation: Sentence realization using probabilistic models of language

PDF

Intergenerational conflict, family functioning, and acculturation experienced by Asian American community college students

PDF

Heterosexuals' attitudes toward lesbians and gay men and the attitude functions they serve: Correlates, stability and gender differences

PDF

Criminal justice response to domestic violence: A closer look at Los Angeles County, 1995--1998

PDF

Health-related quality of life, occupation and prosthesis use in elderly people with lower extremity amputation in Taiwan

Asset Metadata

Creator Phillips, Elizabeth Stoll (author)

Core Title A longitudinal examination of gender and ethnicity across two models of community treatment for individuals with schizophrenia and related disorders

School Graduate School

Degree Doctor of Philosophy

Degree Program Social Work

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag health sciences, mental health,OAI-PMH Harvest,Social Work

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Brekke, John (committee chair), Chung, Ruth H. (committee member), Ell, Kathleen (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-240455

Unique identifier UC11339332

Identifier 3093420.pdf (filename),usctheses-c16-240455 (legacy record id)

Legacy Identifier 3093420.pdf

Dmrecord 240455

Document Type Dissertation

Rights Phillips, Elizabeth Stoll

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA