Introduction
Among patients with severe mental illness (SMI), there is a group—the
“Revolving Door” patients (Hiday & Scheid-Cook, 1991, p. 83)—who
are non-adherent to medications and treatment and high users of
psychiatric services (Hafemeister & Banks, 1996; Cuffel, 1997).
They also have frequent contact with other community services and
agencies such as the police, courts, jails, prisons and emergency
medical facilities around episodes of dangerousness and
victimization (Swartz & Monahan, 2001). In the past,
these patients might have been referred for involuntary admission to a
psychiatric hospital but now they are likely to be referred for
involuntary outpatient treatment (Torrey & Kaplan, 1995).
Although some mental health professionals consider involuntary
outpatient commitment to be helpful in engaging in treatment persons
with SMI who are at risk for danger to themselves or others but who are
unable or unwilling to accept on-going treatment (Muentz, Grande,
Kleist & Peterson, 1996; Swartz, et al., 2001; Torrey &
Zdanowicz, 2001), others believe that court ordered treatment is the
extreme end of a continuum of coercion (Diamond, 1995). Family
members of the National Alliance for the Mentally Ill (NAMI) see
involuntary outpatient treatment as a way of getting help for family
members before they become so ill that they are a danger to themselves
or others (“AOT Update”, 2000). Consumers of mental
health services have a wide range of opinions with one consumer, Frese,
(1997) saying that involuntary outpatient treatment
helped him when he was unable to realize that he was ill, but with
others expressing concern that involuntary treatment will create fear,
destroy trust between consumer and clinician, and cause people to avoid
seeking help (Mad Nation, 1999). Geller (1995, p. 233) views
involuntary treatment as a compassionate measure, a compensatory
therapeutic intervention for individuals who have biological,
psychological and social deficits due to their mental illness or to the
residual effects of the illness--a “form of prosthesis for a
biopsychosocial deficit”. Policy makers view involuntary
outpatient commitment as a solution for many problems associated with
severe mental illness-- homelessness, frequent use of emergency health
services, contact with the police/arrests, and violent behaviors
(Swartz, et al., 1999; Applebaum, 2001).
According to Muentz, Grande, Kleist and Peterson (1996) and Appelbaum
(2001),
involuntary outpatient commitment is associated with a reduction of
hospitalizations, increase in time in treatment, increase in ability to
remain in the community and a reduction of episodes of violence
(Torrey & Zdanowicz, 2001). On the other hand, there
are those who believe that availability and provision of appropriate
mental health care services would make involuntary outpatient treatment
unnecessary (Diamond, 1995). There have been two large studies
that have examined the effect of outpatient commitment on outcome-- the
Duke North Carolina Study (Swartz, et al., 2001) and the Bellevue pilot
study (Steadman, et al., 2001). The Duke study did not find any
difference in hospitalization between the patients committed for
outpatient treatment and the controls at one year but the authors
suggest that outpatient commitment is effective in reducing hospital
stays and violence as long as the intervention is sustained and
combined with a high intensity of community services. The
Bellevue study compared outcomes of patients mandated to receive
community treatment and a control group. Both groups received enhanced
community services. There was no significant difference in
hospitalization and arrests and other outcomes .
With the passage of Kendra’s Law, Assisted Outpatient Treatment (AOT),
in 1999, New York joined the 40 states and the District of Columbia
that have some form of involuntary outpatient commitment laws
(Applebaum, 2001).The New York State Office of Mental Health assumed
responsibility for AOT and delegated implementation to the
counties. Across the state, from the cosmopolitan New York City
area to the dormant steel towns of Western New York, counties with
varying resources faced the task of providing treatment with limited
information about the effectiveness of different treatment modalities
under usual practice conditions to guide them (Holloway, Szmukler &
Sullivan, 2000). Counties often used existing programs such as
Assertive Community Treatment (ACT) (Swanson, et al., 1997).
Extensive studies of ACT for voluntary patients have shown that ACT
reduces hospital readmissions, use of emergency services, legal
problems, and homelessness and improves global functioning and clinical
symptoms (Scott & Dixon, 1995; Lehman, et al., 1999). There is less
evidence that ACT improves social functioning and employment (Mueser et
al., 1998). ACT programs are widely accepted as the “gold-standard” of
treatment for persons with SMI who are frequent users of inpatient
psychiatric care and emergency psychiatric services (Burns &
Santos, 1995; Lehman, et al., 1999; Phillips et al., 2001) but there is
little information about effectiveness of ACT in providing care for
patients who have been court-ordered to receive treatment or as a “jail
diversion program” (Phillips, et al., 2001, p.772). Arboleda-Florez
(1999, p. 4) writes that ACT “is not better than standard community
care on imprisonment, arrests or police contacts”.
Mays (1995) questions the effectiveness of ACT for those high users of
emergency psychiatric services who deny their need for on-going care
and reject efforts of ACT staff to develop a trusting, cooperative
relationship with them. Mueser, et al. (1998) raise the
question of the need to modify ACT to meet the needs of patients with
SMI who have high rates of illegal behaviors.
This study examines the effectiveness of ACT in providing care for
patients
referred for treatment under Kendra’s Law. Because severe mental
illness affects many aspects of a person’s life, Lehman (1999, p. 30)
recommends that measures of treatment effectiveness include a broad
range of outcomes—“ clinical, rehabilitative, humanitarian and public
welfare domains”; and, that both proximal and distal outcomes be
considered. Based on Lehman’s suggestions, the following research
questions were posed: 1. Was there improvement of proximal
outcomes –basic needs, psychiatric symptoms, psychological distress,
stress, and dangerousness to self and to others---after six months of
ACT? 2. Was there improvement of distal outcomes –health,
functioning, social support, life satisfaction, employment,
homelessness, police contacts, arrests/incarcerations and use of
psychiatric and non-psychiatric emergency rooms and hospitals-- after
one year of ACT? 3. How satisfied were the patients with
treatment? 4. Was there a relationship between patients’
status—voluntary or referred for involuntary outpatient treatment-- on
admission to ACT and outcomes and satisfaction with treatment?
Proximal and Distal Outcomes
The terms proximal and distal outcomes carry the meaning of cause and
time relationships to the treatment intervention. In comparison
to distal outcomes that are broad and tend to be evidenced at a greater
distance of time, the effect of the intervention is more likely to be
stronger or more direct on proximal outcomes and to occur close in time
to the intervention (Lehman, 1996). Because of the course of severe
mental illness, psychiatric symptoms and residual disabilities,
persons with SMI often have impaired functioning, inability to tolerate
stress, limited social skills and compromised judgment (National
Institute of Mental Health, 1991) that result in unmet needs
(Arvidsson, 2003), poor health (Massaro, 1992), poverty, homelessness
and victimization (Steinwachs, Kaspser & Skinner, 1992), increased
contact with the police, arrests and incarcerations (Schellenberg,
Wasylenki, Webster & Goering, 1992), high rates of unemployment
(Anthony & Blanch, 1987; Mueser, Drake & Bond, 1997), lack of
social support (Cohen & Farkas, 1986; Estroff, Zimmer, Lachicotte
& Benoit, 1994; Shankar & Collyer, 2002) and diminished
quality of life, (Mueser, Drake & Bond, 1997). In addition to
having increased rates of medical problems, persons with SMI often lack
the ability to access health care, including preventive care, and the
ability to modify unhealthy life-style practices (NIMH, 1991; Felker,
1996; Hutchinson, 1996; Getty, Perese & Knab, 1998; Berren, 1999;
Brown, 1999). Evaluation of effectiveness of treatment for
persons with SMI must consider the intervention’s ability to bring
about change for these multiple problems
Proximal outcomes
In this study, proximal outcomes include perceived basic needs,
psychiatric symptoms, psychological distress, stress, and dangerousness
to self or others. Basic needs. Malm, May and Dencken
(1981) identified the needs of persons with SMI as: medical care, human
relationships, material supplies (housing, clothing food), work and
employment opportunities, safety, leisure activities, communication and
transportation, and inner experiences such as a sense of purpose. In
Buffalo, New York, located in the western part of New York State and
the setting of this study, when 73 members of a chapter of (NAMI) who
had SMI were surveyed about basic needs, the needs identified as unmet
by more than half of the respondents were: a friend (62%), a
role-in-life [a job] (60%), a group to belong to (56%), a sense of
self-identity (55%), and information about mental illness for others
[experienced stigma] (53%) (Perese, 1997).
In another Buffalo study, among 34 patients with SMI who were
psychosocial club members, the most frequently identified unmet needs
were: role in life (job), self-identity, financial resources, a friend
and a group to belong to (Perese, Getty & Wooldridge, unpublished).
A Swedish five-year study of 377 persons with SMI receiving treatment
in the community found similar unmet needs. The overall unmet
needs improved over a five year period but there was no change in unmet
needs for work and social contact with friends and relatives
(Arvidsson, 2003). Psychiatric symptoms and Psychological
distress. Even when receiving treatment, many persons with SMI
continue to experience psychiatric symptoms (Clinton, Lunney, Edwards,
Weir & Barr, 1998; Mueser, Bond, Drake & Resnick, 1998).
In studying subjective distress of patients with SMI, Bradshaw and
Brekke (1999) found that distress was influenced not only by high
levels of psychiatric symptoms but also by patients’ perception
that people were angry with them. Stress.
In a study of 34 persons with SMI, the stress level was found to be
6.67 on a scale of 1 to 10 with nearly half (42%) reporting high levels
(8, 9 or 10). The subjects identified daily hassles as: being
criticized, problems getting along with others, family relations,
physical health problems and financial worries; and major events
as: acute episodes of physical illness, surgery, death of a
family member or friend, homelessness, being robbed, and problems with
housing (Perese, Getty & Wooldridge, 2003).
Dangerousness. In comparison to the general population, there is
an increased risk of violent behavior among a subgroup of persons with
SMI who have a prior history of violence and substance abuse disorder
or antisocial personality disorder (Eronen, Angermeyer & Schulze,
1998). The risk is further increased when these patients are
non-compliant with medication (Swartz, et al., 1998). Hiday and
Wales (2003) say that, if violence is defined broadly to include being
in a fight, the proportion of persons with SMI who are violent is about
50%.
Distal outcomes
Health - Over all,
persons with SMI have poorer health than the general population
(Lehman, 1996). The effects of their illness and the residual
disabilities further compromise their health by limiting their ability
to maintain health and prevent health problems from occurring
(Hutchinson, 1996). In addition, high rates of unhealthy
lifestyle practices, poverty, victimization and stigma negatively
affect their health (Felker, Yazel & Short, 1996; Berren, Hill,
Merikle, Gonzalez & Santiago, 1999; Brown, Birtwistle, Roe &
Thompson, 1999). Although about 70% of people with SMI who report
health problems receive treatment, only half of those with dental
problems receive treatment (Dixon, Postrado, Delahanty, Fischer &
Lehman, 1999). Use of health care services by persons with
SMI has been found to be inefficient and costly (Felker, 1996; Berren,
et al., 1999), with high rates of non-urgent use of emergency services
linked to perceived unavailability of other sources of care and
psychosocial factors such as lack of social support (Padgett &
Brodskky, 1992) and limited knowledge and capacity for self-care
(Getty, Perese & Knab, 1998).
Functioning - Among
persons with SMI, impaired functioning often results in problems in
carrying out a role in life such as a worker, homemaker or student;
problems with relationships; problems with leisure or social activities
and problems with caring for oneself including managing mental and
physical health problems (Pratt & Mueser, 2002).
Social Support - In 1988,
House, Landis and Umberson described a link between lack of social
support and higher mortality rates and higher rates of physical and
psychiatric illnesses in the general population; and, in 1994, Berren
et al. reported similar findings among persons with SMI. In
addition to social support’s influence on health, Clinton et al., 1998,
found that social support plays an important role in helping persons
with SMI adjust to living in the community. Among the
general population, social support was 5.4 when measured with the
Social Support Index that has a range of 0-8 (Bell, LeRoy &
Stephenson, 1982). In Buffalo, among 34 persons with SMI,
social support was 4.6 (Perese, et al., 2003). Lack of treatment
effect on social support among patients with SMI was found by Walsh
(1996) and Getty, Perese and Wooldridge (unpublished).
Life satisfaction - In the
general population, the mean life satisfaction score as measured with
Lehman’s (1988) General Life Satisfaction Scale (GLSS) was 5.4.
That is between mostly satisfied and pleased (Sullivan, Wells &
Leake, 1991). Among persons with SMI living in the community,
life satisfaction was 4.4 (Lehman, Ward & Linn, 1982). A
similar rate (4.2) was found among persons with SMI who were NAMI
members (Perese, 1997) but a lower rate (3.8) was found among
psychosocial club members, (Perese, et al., 2003).
Homelessness -
Approximately one-third of the homeless population has severe mental
illness; and, among this population there is a high rate of co-existing
medical problems, substance abuse, relationship problems, contact with
law enforcement agencies and limited education (Shern, et al., (2000).
In Buffalo, the two year rate of homelessness or “on-the –streets”
varied dramatically from 17.8% among NAMI members (Perese, 1997) to
26.5% among psychosocial club members (Perese, et al., 2003).
Psychiatric emergency room visits -
Psychiatric emergency rooms are the first line of care for many
persons with severe mental illness in crisis; and for some, the only
source of mental health care (Arfken, Zeman, Yeagaer, Mischel &
Amirsadri, 2002). About one-third of the visits are accounted for
by repeat visitors (Ellison, Blum & Barsky, 1986). In an early
study, Munves (1983) reported that the rate of repeat emergency room
visits among persons with chronic mental illness was 20% in one
year. In Buffalo, the two-year rates were 45% (Perese, 1997) and
39% (Perese, et al., 2003).
Readmission to psychiatric hospital
- Hafemeister and Banks (1996), who examined re-hospitalization
rates in six New York State hospitals, predicted that 32-57% would be
readmitted at one year and 43-64% at two years. In Buffalo,
two-year readmission rate was 64% (Perese, 1997) and one-year
rate was 32% (Perese, et al., 2003). Interestingly, reduction of
hospitalizations is believed by clinicians to be related to decreased
substance use and, by consumers, to be related to improvement in social
support and finances (Lang, et al., 1999). Police contact,
arrests and incarceration. Because of diminished ability to
manage crises and limited skills for surviving in the community,
persons with SMI are likely to have more encounters with the police
than the general population (Lamb, Weinberger & DeCuir, 2002). With
a duty to protect both the welfare of the community and individuals
with disabilities who cannot manage on their own, police often try to
settle crises, for non-mentally ill as well as for mentally ill, by
defusing the situation. If they are not able to resolve the
crisis, they decide whether the person will be arrested or taken to a
psychiatric evaluation setting (Hiday & Wales, 2003).
According to Schellenberg et al., (1992), between 33% and 50% of
psychiatric patients have a history of being arrested; and, they say
that, even after engagement in treatment, one in thirteen patients will
be arrested. Factors that predict increased risk of arrest are:
being young, male, nonwhite, homeless, substance abusing and having a
history of prior arrests. Persons with SMI who are arrested tend to be
charged with nuisance crimes such as trespassing, failure to pay for
meals, vandalism, loitering and disorderly conduct; or with shoplifting
and theft. They are more likely to threaten harm than to use weapons or
to cause injury.
Employment - Anthony and
Blanch (1987) report that less than 15% of patients with severe mental
illness are competitively employed. Barriers to employment include the
residual disabilities of mental illness (Mueser, et al., 1997; McGurk
& Mueser, 2003), lack of economic incentives, such as fear of
losing social insurance program benefits and treatment programs that
emphasize prevocational goals rather than competitive employment
(Drake, McHugo, Becker, Anthony & Clark, 1996).
Satisfaction with Treatment
- Howard, et al., (2001) emphasize the need for consumers’
evaluation of their satisfaction with service, quality of care and
effectiveness of treatment as a vital part of evaluating the
effectiveness of a treatment option. In measuring client
satisfaction with Assertive Community Treatment, Gerber and Prince
(1999) found that 80% of the clients indicated overall satisfaction
with the treatment. They were less satisfied in the areas of
medications and their ability to influence treatment choices.
Method
Setting and Subjects
The setting was an ACT program provided by a not-for-profit Behavioral
Health Services Organization. After obtaining approval from
the University at Buffalo’s Institutional Review Board and the agency’s
clinical director, the 71 patients in the ACT program were invited to
join the study. Fifty-six (79%) agreed to participate and signed
consent forms. There was no significant difference in age, race,
marital status, living situation, employment, dangerousness,
arrests/incarcerations, hospital readmissions, and functioning between
patients who agreed to participate and those who refused. They differed
significantly on three items. Women were more likely to refuse, as were
patients with higher levels of educational achievement and patients
with lower rates of alcohol and substance abuse.
Study Design
This comparative study was conducted between November 2000 and March
2002. Data were obtained using record review, face-to-face
interviews and staff observations. Record review carried out by
trained research assistants provided information about demographic
characteristics and about hospital admissions, arrests, incarcerations,
employment and homelessness for the year prior to admission to the ACT
program and for the year following admission to the program.
Proximal outcomes and distal outcomes were obtained in face-to-face
interviews conducted by the PI and, for three patients, by a Spanish
speaking agency member. ACT staff completed standard scales that
related to severity of psychiatric symptoms and level of functioning
based on their observations. At the end of the study, patients
completed a questionnaire that measured their degree of satisfaction
with the program in an interview with research assistants or on their
own if they preferred.
Intervention
The ACT program provides 24-hour, 7 day a week services by a
multidisciplinary team that includes a psychiatrist, a nurse, case
managers, substance abuse and vocational specialists and peer
counselors. It collaborates with other agencies such as housing,
transportation and social services and has an agreement with a primary
care clinician to provide medical care. In a study that used The
Dartmouth Asssertive Community Treatment Scale (DACTS) (Winter &
Calsyn, 2000) to evaluate the fidelity to the ACT model of 28 ACT
programs in New York State, the Buffalo ACT program had the highest
rating of fidelity, a score of 3.81 out of a possible 5 (Laughlin, C.,
personal communication, March 21, 2002). Priority for admission
to ACT is given to patients who are non-responsive to treatment,
homeless, recently incarcerated, at risk of danger to themselves or
others, frequent users of psychiatric emergency services and to
patients with repeated hospital admissions. Patients are admitted
to ACT voluntarily, often upon discharge from the hospital; or, as
referrals from the AOT committee for enhanced outpatient
treatment. If patients are found by the AOT committee to be in
need of AOT, they are given the choice of agreeing to a voluntary
enhanced outpatient treatment plan, the AOT diversion agreement, or
facing a court order to follow a recommended treatment plan. In
Erie County, the proportion of people who accepted the diversion
agreement from passage of Kendra’ Law in the fall of 1999 through
February 2003 is much higher (83%) than the rate for all of New York
State (39%) (“Western New York Assisted Outpatient Treatment Report”,
2003). In Erie County, during the time of the study, 90% of
individuals referred to AOT chose the diversion agreement (The
Challenger, 2000).
Instrumentation
The Buffalo Client Assessment Inventory (BCAI), developed by the
authors and used to obtain data, includes structured questions and
research instruments—10 brief self-report instruments and 3 instruments
that used observation- based ratings. For this study, psychiatric
diagnosis was the diagnosis entered in the patient’s record by the ACT
team psychiatrist on enrollment in ACT and was based on longitudinal
data, an extensive psychiatric evaluation and data available from
collateral sources. Data relating to readmissions to a psychiatric
hospital, police contact, arrests, and incarcerations during the year
preceding admission to ACT were obtained from the record. Legal
status was obtained from the record review and, for the purpose of this
study, is represented by AOT status and includes AOT diversion
agreement and court petitions, the Group I patients, and
voluntary status, the Group II patients. Among proximal outcomes,
basic needs (food, clothing, transportation, medical care, dental care,
housing and finances) were measured using the Meeting Basic Needs
scale, a subscale of The Colorado Client Assessment Record (Ellis,
Wilson & Foster, 1984), that was modified to include medical and
dental needs. The scale has a range of 7 to 28 with higher scores
indicating more unmet needs. Psychiatric symptoms were assessed
with the Brief Psychiatric Rating Scale (BPRS) (Overall & Gorham,
1962), score range of 18 to 126 with higher scores indicating greater
severity of symptoms. Psychological distress was measured with
the Six-Item Indexes of Psychological Distress (Rosen, Drescher, Moos,
Finney & Murphy, 2000), score range of 6 to 30 with higher scores
indicating more psychological distress. The scale asks the
patients to indicate on a scale of 1 (not at all) to 5 (extremely
distressed) how distressed they are by certain feelings or ideas.
Stress was measured with the Modified Stress Ladder (Duffy et al.,
1992), score range of 1 to 10 with higher scores indicating more
stress. Dangerousness was measured with the Violent Behavior
Scale (Neale & Rosenheck, 2000) that elicits patients’ history of
dangerous behaviors. The scale consists of eight questions about
thoughts, discussions, threats or actions, four relating to self and
four to others. The patient responds with a yes or no. Each yes
response counts as one. Score range is 0 to 8 for the total scale
or 0 to 4 for danger to self and 0-4 for danger to others with higher
scores indicating greater risk for dangerous behavior.
Distal outcomes--health,
social support and life satisfaction--were measured respectively with
the Self-rated Health Scale (Stewart, Hays & Ware, 1992), an eight
item scale that has a score range 1-5, lower score indicates better
health (not reported in this article) and the Self-reported
Health scale (Idler & Angel, 1990), a one-item scale that has a
score range of 1-5 with lower score indicating better health; the
Social Support Index (SSI) (Bell, et al., 1982) that has a score range
of 0-8 with higher scores indicating greater support and the
General Life Satisfaction Scale (GLSS) (Lehman, 1988), that has a score
range of 1, terrible, to 7, delighted. Functioning was
measured with two instruments, the Role Functioning Scale (RFS)
(Goodman, Sewell, Cooley & Leavitt, 1993) and the Global Assessment
of Functioning (GAF) (DSM-IV, 1994). The Role Functioning Scale (RFS)
was used to measure overall level of functioning and functioning in
four domains-- working/productivity; independent living/self-care;
immediate social relationships and extended social relationships. The
RFS is considered to be a reliable scale for measuring functioning
among persons with SMI (Green & Gracely, 1987). The RFS score range
for individual domains is 1 to 7 and for overall functioning, 7 to 28.
Higher score indicates better functioning. The GAF (DSM-IV, 1994)
has a score range from 0 to 100. Higher scores reflect better
functioning, with scores of 50 to 41 indicating serious distress and
dysfunction, 40 to 31, severe distress and dysfunction, 30-21,
inability to function in most areas, and 20-11, danger to self or
others (Rosse & Deutsch, 2000). All members of the ACT team
received training in use of the GAF before the research study was
initiated. Satisfaction with treatment was measured with the
Client Evaluation of Services (CSQ-8) (Nguyen, Attkisson & Stegner,
1983), an eight-item scale with a score range of 8 to 32. Higher
scores indicated greater satisfaction.
Analysis of data
A repeated measures analytical approach was used to answer our research
questions. Descriptive statistics were used to present demographic
characteristics. Chi-square analysis was used to examine demographic
characteristics between patients referred by AOT and voluntary
patients. Paired t-tests were used to compare the effectiveness
of the intervention between prior year or baseline and 6 months for
proximal outcomes. For distal outcomes, paired
t-tests were used to compare the effectiveness of the intervention
between prior year or baseline and 12 months. To manage missing
data, existing time one data was used to create a regression
equation. Based on the parameters estimated by these regression
equations, we substituted the missing values. Chi-square and
independent t-tests were also used to compare outcomes of Group I and
Group II. Statistical significance level was set at .05 level and
1-tailed test was used.
Results
Demographic characteristics
As shown in Table 1, the majority of the 56 participants were male
(77%) and African-American (59%). Twenty-four (44%) had not completed
high school. Most common clinical diagnosis was
schizophrenia. According to admission data in the record, 30
(54%) had a diagnosis of alcohol abuse and 32 (57%), drug abuse; 51
(91%) had been hospitalized at least once in the prior year; and 13
(31%) of the 41 participants with available data had exhibited
dangerousness to others. Nineteen (34%) lived in unstable situations
(emergency housing, shelters or “staying with friends”) and nineteen
(34%) had been homeless at least once during the prior year.
Thirty-seven had been referred by AOT. Nineteen were voluntary
admissions to ACT.
|
Table 1.
Demographic, Clinical and Community Adaptation
Characteristics of
|
| Patients on Admission to ACT |
N=56
|
|
|
|
Demographic
Characteristics
|
|
n
%
|
|
Gender
|
|
|
|
Male
43 77.0
|
|
|
|
Female
13 23.0
|
|
Age
|
|
|
|
18-29
9
16.1
|
|
|
|
30-49
39 69.6
|
|
|
|
50-69
8 14.3
|
|
Race
|
|
Caucasian
19 34.0
|
|
African American
33 59.0
|
|
Latino
3
5.0
|
|
Pacific Rim
1
2.0
|
|
Marital
Status
|
|
Divorced
11
20.0
|
|
Single
45
80.0
|
|
Education
|
|
Less than high school
24
44.0
|
|
High school
22
41.0
|
|
Some college
5
9.0
|
|
College degree
3
6.0
|
|
Employment
|
|
Unemployed
18
32.0
|
|
Employed
5 9.0
|
|
Total disability
33
60.0
|
|
Living
situation (n=50)
|
|
Stable housing
31
55.4
|
|
(house,
apartment,
|
|
room,
group home,
|
|
board and care)
|
|
Unstable
housing
25
44.6
|
|
(shelter,
motel, friend, homeless)
|
|
|
|
Legal
status
|
|
AOT
37
66.1
|
|
Non-AOT
19
33.9
|
|
Clinical Characteristics
|
Psychiatric
diagnosis
|
|
Schizophrenia
40
71.4
|
|
Schizoaffective
7
12.5
|
|
Unipolar disorder
2
3.6
|
|
Bipolar disorder
3
5.4
|
|
Anxiety
Disorder
1
1.8
|
|
Alcohol abuse alone
7
12.5
|
|
Drug abuse alone
8
14.2
|
|
Alcohol & drug abuse
22
39.2
|
|
Psychosis NOS
2
3.6
|
|
Borderline Personality Disorder
2
3.6
|
|
Antisocial Personality Disorder
7
12.5
|
|
Personality Disorder NOS
3
5.4
|
|
|
Danger to self
(n=43)
|
|
(year
prior to ACT enrollment)
|
|
0
39
91
|
|
1-3 times
4
9
|
Danger to
others (n=41)
|
|
(year
prior to ACT enrollment)
|
|
0
29
69
|
|
1-3 times
9
21
|
|
more than 3 times
4
10
|
|
|
Hospital
admissions
|
|
(year prior to ACT enrollment)
|
|
0
5
9
|
|
1-3 times
36
65
|
|
more than 3 times
14
26
|
|
|
|
Community Adaptation
Characteristics
|
|
(year
prior to ACT enrollment)
|
|
Contact with police (n=34)
|
|
0
14
41
|
|
1-3 times
14
41
|
|
more than 3 times
6
18
|
|
Arrests (n=48)
|
|
0
30
63
|
|
1-3 times
13
27
|
|
more than 3 times
5
10
|
|
Incarcerations (n=40)
|
|
0
29
73
|
|
1-3 times
11
27
|
|
more than 3 times
0
|
Chi square analysis indicated that there was no significant difference
in gender, race, marital status, education, housing, diagnosis, alcohol
and substance abuse, arrest/incarcerations, dangerousness or hospital
admissions between Group I and Group II on admission. There was a
difference in age (2=6.28, df=2, p=.043). Among 37 patients in
Group I, 22 (59.5%) were 30 to 49 years old and there were nearly equal
numbers of younger patients (7 or 18.9%) and older patients (8 or
21.6%). Among 19 patients in Group II, 17 (89.5%) were in 30 to
49 years old and two younger than 30. The statistical difference in
employment status (2=7.73, df=2, p=.02) was of little clinical
importance; e.g. four (10.8%) Group I patients employed and one (5.3%)
in Group II.. Social support (score range 0-8) among Group I
patients (3.35, s.d.=2.1) was significantly lower than among Group II
patients, (4.77 , s.d.=2.1, df=2, p=.021). (Test result is t
(54)=1.71, p=.046.)
Research question 1: Was there improvement of proximal
outcomes?
Table 2 summarizes the pre-post comparisons of proximal outcomes. Basic
needs.
The overall rating of needs at the initial assessment was 12.68; e.g.,
the patients perceived their needs as met between “most of the time”
and “some of the time”. (Data obtained by asking patients about
their unmet needs for the year prior to admission to ACT are not
included in the measurement of the effect of treatment from baseline to
6 months; however, the total rating of needs for the prior year was
14.0, significantly more unmet needs than at the initial
assessment.) At initial assessment, the three needs most
frequently rated as unmet were transportation, housing and dental
care. There was no significant change in patients’ perception of
needs met between initial assessment and six months. Patients’ comments
included: “If I want to go somewhere, I still have to walk”; “ I have a
place to stay but it is not what I want; and, “I am afraid to go to the
dentist”.
|
Table 2.
Proximal Outcomes of Patients in ACT
|
|
|
|
Initial Assessment
6 month Assessment
|
|
M
sd n
M
sd
n
|
|
Total Basic
Needsa
12.68
5.39 50
11.74 3.21 50
|
|
|
|
Psychiatric Symptoms
|
|
BPRS b
40.80 16.35 50
32.7*
10.47 50
|
|
Psychological
Distress 12.21
5.54 56
10.86*
5.27 56
|
|
SCL-6c
|
|
|
|
Stress
4.98 2.37 53
3.94* 2.27
53
|
|
Stress
Ladderd
|
|
|
Dangerousness
|
|
Violent Behavior Scalee
|
|
Toward self
.25
.76 43
.07* .46
43
|
|
(score range 4-8)
|
|
|
|
Toward others
.21 .66 39
.21 .52
3
|
(score range 4-8 *significant at .05 (one-tail)
a
Meeting Basic Needs
scale (Ellis, Wilson &
Foster, 1984) score range 7-28, higher score indicates more unmet needs
b
Brief Psychiatric Rating
Scale (Overall & Gorham, 1962), score range 18-126, higher
score
indicates greater severity
c
Six-Item Indexes of
Psychological Distress (Rosen, Drescher, Moos, Finney & Murphy,
2000),
score range 6-30, higher score indicates greater psychological distress
d
Modified Stress Ladder,
(Duffy et al. 1992), score range 1-10, higher score indicates greater
stress
e
Violent Behavior Scale (Neale & Rosenheck, 2000), score
range 8-24,
higher score indicates greater violence.
Subscale
danger to self, score 4-12; danger to others, score 4-.
Psychiatric symptoms,
Psychological distress, Stress and Dangerousness.
There was a significant improvement of psychiatric symptoms that were
rated by the ACT team and also a significant improvement of
psychological distress that was rated by the patients. On initial
assessment of distress, feelings such as depression, fear,
anxiety, hopelessness and thoughts such as there is something
wrong with your mind were rated in the moderately distressful range and
at one year they were rated in the range of little
distress. Perceived stress showed significant
improvement, decreasing one point. There was a dramatic decrease
in dangerousness toward self; but there was no significant change in
dangerousness toward others. In responding affirmatively to the
questions of whether they talked about hurting or striking someone or
threatened to do so, patients often added comments indicating that
those actions were required to protect themselves in the
community. Of the six proximal outcomes, five (86%) showed
significant improvement.
Research question 2: Was there improvement of distal outcomes?
Table 3 summarizes the results from pre-post comparisons of distal
outcomes. Health. There was a significant positive change in
perception of general health; e.g., patients perceived their health as
improved from time of admission. Functioning. There was
significant improvement of one measure of functioning, the GAF that
includes psychiatric symptoms but no improvement in functioning as
measured by the RFS that does not include psychiatric
symptoms. Social support. There was no significant
change in perceived social support. Data obtained by asking the
patients about their perceived social support for the year prior to
admission to ACT showed that social support was 3.69 s.d. 2.1,
considerably lower than the score obtained on the initial assessment,
with patients referred for Assisted Outpatient Treatment having
significantly lower social support, (3.4, s.d. 2.1) than voluntary
patients (4.4, s.d. 2.1). Life satisfaction. There
was significant improvement in life satisfaction.
Employment. There was no significant increase in
employment. Arrests/incarcerations. There was significant
decrease in police contacts and arrests but no significant decrease in
incarcerations. Use of medical and psychiatric emergency rooms and
hospitals. There was significant decrease in medical emergency
room visits, medical hospital admissions and psychiatric emergency room
visits. The change was striking with the rates cut nearly in
half. There was no decrease in psychiatric hospital admissions. Of the
13 distal outcomes examined, 8 (62%) showed improvement.
|
Table 3. One-year distal outcomes of patients in ACT
|
|
1 year prior
Initial
Interview 12 month
Interview
|
|
M sd n
M
sd
n
M
sd
n
|
Health
|
Self-rated Healtha
2.98 1.12
40
2.63* 1.21
40
|
Functioning
|
|
1)
RFSb |