PO Box 25445
Albuquerque, New Mexico 87125-0445
Contact: Marsha McMurray-Avila
Phone: (505) 266-7683
Key words: transitional housing; substance
abuse; utilization of services
The purpose of this study was to evaluate
the effectiveness of Villa de Paz (VDP),
the Albuquerque Health Care for the
Homeless (AHCH) sober housing program,
using the following three indicators:
disease self-management; improved health
status and level of functioning; improved
quality of life; and involvement in
treatment.
METHODS
The VDP is a 23-unit apartment complex
operated by AHCH as a sober transitional
housing program since November 1992.
Single men or women who are homeless
and who have entered recovery from drugs
and/or alcohol are eligible for residence
for up to 14 months. The program is
focused on supporting participants in
remaining alcohol/drug-free and maintaining
housing through the transition period
following treatment (or individual recovery).
Since the project was initiated in November
1992, over 140 homeless people with
addiction problems have resided at VDP.
During that time, records have been
kept on each resident and aggregated
reports have been prepared for funders.
However, the data on individual residents
had never been analyzed to determine
if there are any predictors for successful
outcomes. The AHCH staff have needed
information regarding how well the program
is working, for whom, and areas where
changes might be needed. To that end,
this study undertook a retrospective
analysis of the data that already exist
for 110 VDP residents. Information was
collected on an additional 14 program
participants, but because there was
no indication of their alcohol/drug-free
status at time of exit, these were not
included in the analysis.
This study was not organized as experimental
research to prove cause and effect,
and therefore no control or comparison
groups were used. Rather, the focus
was on generation of useful information
in an accessible format for program
staff to be able to answer long-standing
questions and make decisions regarding
future program activities. For this
reason, methods were chosen that could
be easily replicated by HCH program
staff without extensive research background.
The basic approach used was to extract
percentages of successful outcomes overall
and use those as predictors for comparing
outcomes across several resident characteristics.
Where apparent differences were seen,
a Chi square test was used to determine
statistical significance. It is hoped
that this approach, while not statistically
sophisticated, will encourage further
outcome studies of VDP and other AHCH
services.
Because of the difficulties in tracking
residents after they leave the program
(and the numerous intervening variables
that are not related to effects of the
program), measures of successful outcome
were only applied to the period of time
that the resident was in the program.
Data that had been collected manually
in resident files, as well as relevant
data from the main AHCH database regarding
each residents utilization of other
AHCH services, were entered into a database
(using Microsoft Access) and analyzed
for outcomes in the following areas
recommended by the BPHC Outcomes Workgroup:
disease self-management (and, indirectly,
improved health status and level of
functioning); improved quality of life;
and involvement in treatment. To determine
if there were particular characteristics
that correlated with these outcomes,
comparisons were made across resident
characteristics of the following: age;
gender; ethnicity; amount of time clean
and sober before entry; nature and duration
of previous substance abuse; referral
source (specifically for those referred
directly from the AHCH recovery house,
Casa Los Arboles (CLA)); length of time
homeless; mental health; income/public
benefits; length of stay in the program;
and utilization of other AHCH services.
In the area of disease self-management,
the outcome question was Are program
participants maintaining a clean and
sober lifestyle? Maintenance of sobriety
was defined by whether or not the individual
relapsed into any substance use during
his or her residence in the program,
and the individuals alcohol/drug-free
status at time of exit from the program.
Improved health status and level
of functioning are important
outcomes to examine for this program,
but existing data were not available
in a form that could be used to accurately
measure changes in status. However,
maintenance of a clean/sober lifestyle
is a good indirect measure of improved
health status and level of functioning,
given the assumption that both health
and functioning in this population are
improved by abstinence from drugs and
alcohol.
To measure improved quality
of life, the primary outcome
addressed was housing status. How long
did the participants maintain stable
housing (length of stay) and what was
their housing status upon exit from
VDP? (Income status at entry was used
as a variable for comparing alcohol/drug-free
outcomes, however no analysis was done
for this study to compare changes in
income status from entry to exit.)
Examination of involvement in
treatment as a desirable outcome
includes both involvement in a recovery
plan and utilization of other AHCH services
(medical, dental, substance abuse counseling,
mental health, social work, case management,
outreach, and eye clinic). A very specific
question related to overall outcomes
was whether or not residents who had
previously lived at CLA, the AHCH recovery
house for men, had more successful outcomes
than those who were referred from other
sources or came directly from the street.
Results have been divided into three
areas: 1) description of characteristics
of program residents; 2) alcohol/drug-use
history; and 3) outcomes, including
comparison across variables.
Characteristics of program residents
Gender. Of the 110
residents studied, 79 (71.8 percent)
were male and 31 (28.2 percent) were
female.
Ethnicity. The majority
(70 percent) were either White or Hispanic,
with slightly over 39 percent White
and almost 31 percent Hispanic. Native
Americans made up 8.2 percent of the
total and Blacks were 7.3 percent. An
additional 9.1 percent classified themselves
as multiracial and 3.6 percent were
of unknown ethnicity.
Age. Residents of VDP
must be at least 18 years old. In the
age group 18-20, there were only three
participants (2.7 percent). Eleven percent
were between ages 21 and 30. The majority
(73.7 percent) were between 31 and 50
years of age, with 29.1 percent in the
31-40 group and 44.6 percent between
41 and 50. Eleven percent were between
51 and 60, and only two participants
(1.8 percent) were over age 60.
Family status. Over
two-thirds (67.3 percent) of the residents
were single, with males slightly higher
at 69.6 percent and females lower at
61.3 percent. A higher rate of males
(15.2 percent) than females (3.2 percent)
were separated. The reverse was true
for divorce only 1.3 percent of the
males were divorced, compared to 12.9
percent of the females. Small numbers
of both genders were married or living
with someone (1.8 percent total for
each) and only one female was widowed.
Thirteen (11.8 percent) had unknown
family status.
Veteran status. Over
one-fourth (26.4 percent) of the total
were veterans. Of these, two were female
(6.9 percent) and 27 were male (93.1
percent). Over half (55.2 percent) were
veterans of the Vietnam Era, one (3.4
percent) was a World War II vet, and
the remainder (41.4 percent) were veterans
during peacetime.
Years of education.
Information on years of education completed
was available for only about a third
of the participants. However, since
the question was part of a form that
was consistently completed during a
certain time period, it could be considered
a small sample of the total. Within
the group of 33 for whom this information
was available, 24.2 percent had some
college experience and one person (3
percent) had a Bachelors degree. Another
24.2 percent had not finished high school,
while 30.3 percent had either a high
school diploma or GED, and 18.2 percent
had vocational/technical education.
Legal issues. Almost 21 percent
of residents had previous or current
legal issues, indicated by either an
affirmative answer to the question Do
you have any current legal problems?
or listing of a probation or parole
officer. Females were more likely to
have had current or past legal issues
(25.8 percent) than males (19 percent).
Length of time homeless. To
be eligible for residence at VDP, a
person must either be homeless, at imminent
risk of becoming homeless, or have been
homeless before entering a treatment
or recovery program from which they
are being referred, such as CLA. Length
of time homeless before entering VDP
varied greatly from a matter of days
to several years. However, the majority
(74.5 percent) had been homeless for
less than 2 years. Further breakdown
of that number shows that almost twothirds
(62.7 percent) had been homeless for
1 year or less. Another 11.8 percent
had been homeless between 1 and 2 years.
The remaining one-fourth had been homeless
for over 2 years.
Length of time living in Albuquerque.
In contrast to the common stereotype
of homeless people as transients, it
was interesting to note that only 28.2
percent had been in Albuquerque for
less than a year. Males were less likely
(25.3 percent) than females (35.5 percent)
to have been in Albuquerque less than
a year. Over a quarter (26.6 percent)
of the males and 16.1 percent of the
females were lifetime residents_ of
over 20 years, for a combined percentage
of 23.6 percent. The remaining approximately
48 percent (both males and females)
had been residents of Albuquerque for
anywhere from one to 19 years. Hispanic
males had a lower than average rate
of new Albuquerque residents (21.4 percent
at less than one year) and the highest
rate of lifetime residents of over 20
years (39.1 percent). Income at time
of entry. Applicants for residence at
VDP are required to have some income,
either through employment, public benefits
or other sources. The majority during
the time of this study entered with
income from employment (53.6 percent),
part-time or full-time. Forty percent
were receiving public assistance as
follows: General Assistance (GA) at
26.4 percent; Supplemental Security
Income (SSI) at 8.2 percent; Social
Security Disability (SSD) at 4.5 percent;
and SSA/retirement at 0.9 percent (one
person). Another 8.2 percent were receiving
other income (e.g., workers compensation,
family trust).
Of those that were employed, 72.9 percent
were males and 27.1 percent were females,
consistent with the overall gender ratios
in the study. Residents with GA or SSD
were more likely to be male (82.8 percent
and 80 percent respectively), while
residents with SSI were more likely
to be female (55.5 percent).
Physical and mental health.
Forty-five (40.9 percent) of the residents
indicated a medical problem related
to physical health. Examples of some
of the diagnoses given by these 45 include:
16 (35.6 percent) with musculoskeletal
problems, especially of the back or
spine; 12 (26.7 percent) with heart
problems or hypertension; four (8.9
percent) with diabetes; five (11.1 percent)
with Hepatitis C; one (2.2 percent)
with HIV and Hepatitis B; and seven
(15.6 percent) with problems related
to other internal organs, including
kidneys, pancreas, liver, gall bladder
and stomach.
Three residents died while living at
VDP: one 51-year-old Hispanic female
who died of an overdose; one 57-year-old
Hispanic male who died of cirrhosis
of the liver; and one 46- year-old Hispanic
male who died of a heart attack.
Thirty-four residents (30.9 percent
of the total) either indicated a mental
health problem when applying for residency
or had mental health encounters through
AHCH. Of those 34, diagnoses were as
follows: depression (41.2 percent);
bipolar disorder (manic depression)
(14.7 percent); post traumatic stress
disorder (5.9 percent); issues from
childhood abuse (2.9 percent); and unknown
diagnoses with AHCH mental health encounters
documented (35.3 percent).
Length of time clean and sober.
The length of time residents
had been clean and sober before entering
VDP varied widely, with the majority
(60 percent) in early recovery for 1
to 6 months. Another 21.8 percent had
been in recovery for 7 to 12 months,
for a total of 81.8 percent being in
their first year of recovery, 10.9 percent
in recovery for 1 to 5 years and 4.5
percent for over 5 years.
There were differences between males
and females, with more of the males
(86.1 percent) in their first year of
recovery compared to 74.2 percent of
females. The ethnic group with the highest
rate of people in the first year of
recovery was Hispanic males with almost
93 percent in early recovery for 10
months or less. The maximum recovery
time for Hispanic males was 15 months,
in contrast to White males and females
who had two people each with over 5
years of recovery.
Alcohol/drug use preferences.
Less than half the residents reported
using both alcohol and drugs (42.7 percent).
Another 40.9 percent indicated that
alcohol was their primary drug of choice
and 13.6 percent indicated that drugs
other than alcohol were their preference.
More males (48 percent) than females
(29 percent) used both drugs and alcohol,
while females were more likely to use
primarily alcohol (48 percent) than
males (38 percent). Use of drugs only
was 16.1 percent of females and 12.7
percent of males.
Outcomes and Comparison Across
Variables
Disease self-management (and implied
improved health status and level of
functioning).
Are program participants maintaining
a clean and sober lifestyle?
Maintenance of sobriety was defined
by whether or not the individual relapsed
into any substance use during his or
her residence in the program, indicated
by alcohol/drug-free status at time
of exit from the program. (Any relapse
was grounds for eviction.) Sixty-five
of the 110 residents (59.1 percent)
were alcohol/drug-free during their
stay and at the point when they moved
out of VDP. Eleven of those (10 percent
of the total) were alcohol/drug-free,
but were evicted for non-payment of
rent or violation of rules (other than
use of alcohol or drugs). Of the remaining
54 that left alcohol/drug-free, 23 (20.9
percent) completed a full 14 months
or more at VDP, while 30 (27.3 percent)
left before 14 months, but on good terms.
(See results regarding length of stay
below.) One resident died of a heart
attack during his tenth month of residence,
but was alcohol/drug-free up until that
time. These percentages were then reviewed
for possible connections between the
outcome of alcohol/drug-free status
at exit and the following variables:
gender, age, ethnicity, family status,
veteran status, years of education,
past or current legal issues, time of
residence in Albuquerque, time homeless,
type of income at entry to program,
referral source into program, mental
health problems, alcohol/drug use history
(including age at first use of substances)
and amount of time clean and sober before
entering the program. Differences in
percentages (using the overall population
rate of 59.1 percent alcohol/drug-free
and 40.9 percent not alcohol/drug-free)
pointed to a few areas of possible correlation.
In general, those who seemed to do slightly
better were: females; residents over
40 years of age; veterans; residents
with no past or current legal issues;
residents who had been homeless less
than two years; those who had more than
one year of recovery; and those whose
drug of choice was only alcohol. Related
to ethnicity, Hispanics seemed to do
less well than all other groups. These
are all areas that program staff might
want to examine further, but none of
these differences was statistically
significant when a Chi square test was
applied. The only variable that proved
to be statistically significant was
the source of referral. A contingency
table with referral source data resulted
in a Chi square value of 3.97, a statistically
significant relationship (p<.05).
Again, one of the major questions asked
by AHCH staff since the inception of
VDP was whether or not those referred
from CLA, the AHCH social model recovery
house for males, did better at VDP than
those entering directly from the street
or referred from other sources. Although
a positive relationship was hypothesized,
this study showed the opposite to be
true. Residents who were previously
at CLA had a 45.5 percent rate of success,
compared to a 68.2 percent rate of success
for all other referrals (65.7 percent
when adjusted to include only males,
since CLA does not
house females).
Improved quality of life.
How long did participants maintain
stable housing (length of stay)?
What was their housing status
upon exit from VDP?
Length of stay ranged from
less than 1 month to 28 months. (Residents
are granted extensions beyond the 14
months based on employment, education
or health factors.) Average length of
stay was 8.9 months with a standard
deviation of 5.7, however, these statistics
varied depending on reason for exit
from the program and gender.
Of the 65 who were alcohol/drug-free
when they exited the program, 22 (20
percent of the total) left upon expiration
of their lease, with an average stay
of 16.1 months (standard deviation 4.9).
Nineteen percent of the males (15) and
22.6 percent of the females (7) fell
into this category.
Another 31 (28.2 percent of the total)
chose to leave before their lease expired
for reasons related to family issues
(e.g., moving in with boyfriend or girlfriend,
regaining custody of a child, getting
married, having a baby, etc.), finding
another place to live (on their own
or as part of another program), moving
out of state or _into the wilderness,
or leaving due to a job or training
elsewhere. Average length of stay for
this group was 7.6 months (standard
deviation 3.9), and included 22.8 percent
of the males and 41.9 percent of the
females.
The remaining 12 (18.5 percent) of those
who exited alcohol/drug-free were either
evicted due to non-payment of rent (7.3
percent) or other violations (2.7 percent),
or died from causes not related to alcohol
or drug use (one person at 0.9 percent).
Of the 45 residents (40.9 percent of
the total) who were not alcohol/drug-free
when they exited, 37 (33.6 percent of
the total) were evicted for use of alcohol
or drugs. The average length of stay
for this group was 6.4 months (standard
deviation 4.9) and was made up of 30
males (38 percent of the males) and
7 females (22.6 percent of females).
The remaining 8 were either evicted
for refusal to take a drug test (2.7
percent) or non-payment of rent (1.8
percent), chose to leave (.9 percent)
or died while in residence (1.8 percent).
Data on housing status at time
of exit from VDP is known for
only 60 percent of the residents. A
total of 26 (23.6 percent of the total
110) had acquired housing at the time
of exit. Of those, 11 were males who
were alcohol/drug-free (13.9 percent
of all males), 11 were females who were
alcohol/drug-free (35.5 percent of all
females), 3 were males who were not
alcohol/drug-free (2.7 percent of all
males), and 1 was a female who was not
alcohol/drug-free (3.2 percent of all
females).
Twenty-seven participants (24.5 percent)
either had plans to acquire housing
or moved in with friends or family.
Of those, 14 males were alcohol/drug-free
(17.7 percent of males), 6 females were
alcohol/drug-free (19.4 percent of females),
6 males were not alcohol/drug-free (7.6
percent of males), and 1 female was
not alcohol/drug-free (3.2 percent of
females). The remainder of the housing
outcomes for those who were alcohol/drug-free
included 2 males with other arrangements,
and 15 males and 5 females with unknown
disposition. The remainder of the housing
outcomes for those who were not alcohol/drugfree
included 1 female and 4 males who went
into treatment programs, 1 female who
went to jail, 3 males who returned to
the street, and an unknown disposition
for 4 females and 19 males.
Involvement in treatment.
What was the level of involvement
in a recovery plan?
What was the level of utilization of
other HCH services?
Although residents are required to attend
on-site community meetings, data on
attendance at those meetings was not
available at the time of this report.
In addition, because VDP is organized
as independent transitional housing
and is not an actual program in itself,
tracking each individuals recovery
plan is not done. Indications
at time of entry howed that all new
residents entered with some sort of
recovery plan, primarily because that
is a requirement for eligibility. The
majority of residents (86.4 percent)
listed involvement in a 12-step program
and 60.9 percent stated that they had
a sponsor or advocate as part of their
recovery plan. Other recovery plans
listed were: spirituality (26.4 percent);
counseling (22.7 percent) at AHCH, St.
Martins, the Center for Alcohol and
Substance Abuse Addictions, or another
program; fellowship with others who
are clean and sober (13.6 percent);
daily skills such as employment or school
(8.2 percent); attendance of other support
groups that are not 12-step (4.5 percent);
volunteer work (4.5 percent); and aftercare
at Turquoise Lodge (2.7 percent).
Although there is some verification
of the recovery plan done at the point
of entry (applicants provide name and
phone number of sponsor, and comments
from references are included in the
chart), there is no follow-up to track
compliance or changes in the plan. There
is also reason to believe that potential
residents say what they think the interviewer
wants to hear. For example, of the seven
people who listed counseling through
AHCH as part of their recovery plan,
only one had any substance abuse counseling
encounters during his stay at VDP. Two
others, in addition to the first one,
had mental health encounters with psychiatrists
during their stay which tend to be medication-related
more than counseling for substance abuse.
For these reasons, no further analysis
was done of this data.
Data on utilization of other
AHCH services was obtained
through the organization_s database,
and were divided for each resident by
number and type of encounters before
entering VDP, during VDP and after exiting
VDP. Of the total 110, 91 individuals
(82.7 percent) had at least one encounter
documented with an HCH service at some
point before, during or after residence.
Eighty individuals (72.7 percent) had
a total of 1,107 encounters before entering
VDP (an average of 13.8 encounters/person),
while 30 (27.2 percent) had no contact
previous to VDP residence. Sixty-four
individuals (not unduplicated from the
first group) had a total of 520 encounters
during their residence at VDP (an average
of 8.1 encounters/person), while 46
(41.8 percent) had no contact during
that time. Fifty-seven individuals (again
not unduplicated from previous groups)
had a total of 963 encounters after
leaving VDP (an average of 16.9 encounters/person),
while 53 (48.2 percent) had no further
contact.
Actual encounter rates varied tremendously
by type of service. Medical encounters
were the most common, with 70 individuals
having 662 encounters before VDP residence
(an average of 9.5 encounters/person,
but with the maximum for 1 person being
82 encounters); 41 individuals having
213 medical encounters during VDP residence
(an average of 5.2 encounters/person
with the maximum number for one person
of 54 encounters); and 45 individuals
having 448 medical encounters after
leaving VDP (an average of 10 encounters/person
with the maximum for 1 person of 87
encounters).
There are numerous issues
which merit discussion based on the
results of this study. However, only
four main points will be discussed here:
1) general observations on the meaning
of these outcomes; 2) consideration
of the utilization of services by residents;
3) discussion of rates of success connected
with amount of total recovery time;
and 4) potential explanations for the
difference in outcome for CLA referrals.
General observations.
A somewhat comparable population (homeless
alcoholics) in Albuquerque was studied
as part of an National Institute on
Alcohol, Abuse, and Alcholism-funded
research and demonstration project called
Project H&ART from 1990 to 1993.
(The AHCH project was a partner in that
study, and some study participants later
became residents of VDP.) The average
stay of participants in that controlled,
randomized study was 67 days, with about
one-quarter of the population graduating
(remaining sober through the full 4-month
period in one of the three groups, two
of which included housing in settings
similar to VDP and one of those which
included intensive case management).
(Lapham, Hall and Skipper, 1995)
Although only a cursory comparison has
been made with that study, it would
seem that the outcome of maintaining
a clean and sober lifestyle was seen
at a much higher rate than expected
in the present study, with 59 percent
of the 110 VDP residents staying alcohol/drug-free
during their stay in the program. The
outcome of housing stability was also
significant with an average length of
stay of 8.9 months for all residents,
10.6 months for those who stayed clean
and sober, and 16.1 months for those
who stayed clean and sober and left
when their lease expired (from 14 to
28 months).
It is important with results such as
these to examine the data to be sure
that the population being served was
not creamed, i.e., acceptance in the
program given only to those with the
greatest likelihood of success. Reviewing
the data on participants characteristics
and substance use history presented
above, it is clear that the population
was quite varied and included individuals
who had characteristics that would intuitively
be considered as predictors for failure,
including having been homeless for long
periods of time, a relatively short
recovery time, and physical and mental
health problems.
Although no follow-up was done to determine
what happened to residents after they
left VDP, the mere fact that they managed
to maintain sobriety and housing for
even that period of time could be considered
a success for many of the participants,
given their characteristics and substance
use history described above.
Utilization of AHCH services. Several
interesting results emerged related
to utilization of other AHCH services.
First, there was relatively less utilization
during the time of VDP residence. It
might be expected that with stable housing
and access to affordable health care
that participants would have increased
their utilization. However, there were
many cases of residents who had significant
utilization before and after, but relatively
little during their stay at VDP. More
examination of these residents AHCH
charts might provide some answers to
this question. Additionally, it would
be wise to ask the residents directly.
It is possible that simply the location
of VDP, which is outside of the downtown
area where AHCH services are located,
and/or the necessity to maintain their
employment, kept many from utilizing
services. Or it may be that having stable
housing actually improved their health
status and decreased their need for
services. A great deal more data would
be needed to accurately answer this
question.
Outcome, in terms of alcohol/drug-free
status, did not seem at first glance
to be affected at all by use of medical
services during residence at VDP. The
rate of success for those with more
than one medical encounter during their
stay was 63 percent, while the rate
for those with no medical encounters
was 64 percent. However, the success
rate for those who had more than one
AHCH medical encounter before entering
VDP was 47 percent and for those with
more than one medical encounter after
leaving VDP the rate was 40 percent.
If having more than one medical encounter
were interpreted to mean that the person
had medical problems, then it would
appear that having medical care during
their stay at VDP perhaps increased
their chances of staying alcohol/drug-free.
Much more study would need to be done
in this area to determine if there really
is a connection.
Another interesting finding was the
very low utilization of substance abuse
services. The AHCH substance abuse encounters
are basically for counseling and a much
higher utilization would be expected,
especially given the limited availability
of other counseling services for this
population. It is possible that the
data here is incomplete, due to inadequate
documentation by substance abuse staff
over time. Much more attention needs
to be paid to this issue.
A third observation concerns the already
know limitations of self-reporting.
Although the list of physical and mental
health problems that was provided by
residents was imprssive, it was clearly
not complete. Many of the residents
who listed no problems had significant
numbers of medical and/or mental health
encounters. Others who listed problems
had no utilization of AHCH services
at all. This raises another question
which should be included in future studies
Where do VDP residents receive health
care, if not from AHCH? Some possibilities
might be through the Veterans Administration
system, Indian Health Service or through
other facilities for those that have
Medicaid benefits related to disabilities.
One last observation related to the
encounter data is the number of residents
with heavy utilization of services.
It would be quite valuable to gather
the diagnoses of these residents and
review these utilization patterns. This
is not a recommendation to limit that
utilization, so much as to understand
which types of diagnoses lead to this
pattern. This information would particularly
be valuable for future cost studies.
Length of total recovery time.
An interesting observation when examining
the breakdown of length of stay during
the first 6 months is that the second
and third month seem to be a particularly
vulnerable time. Of the 15 residents
who exited in their second or third
month, 13 (86.7 percent) had relapsed
and only 2 (13.3 percent) were still
alcohol/drug-free. In order to accurately
determine whether or not this is a pattern,
it is necessary to look at total recovery
time, i.e., amount of recovery time
at entry plus the length of stay. From
that it appears that the first 9 months
of recovery are the crucial period with
a success rate of only 21.1 percent.
During the 10-16 month period the success
rate more than doubles to 46.5 percent,
and then increases dramatically to 85.4
percent for anything over 17 months.
Outcomes for Casa Los Arboles
referrals. The fourth discussion
point relates to the significant differences
between alcohol/drug-free status at
exit for those residents referred by
CLA and those from other sources. It
is important to note that CLA is the
single largest referral source into
VDP, and that CLA residents are given
priority in an attempt to develop a
continuum of care approach for recovering
substance abusers who are homeless.
Several factors were considered for
possible explanation of the difference
in success rates. First, the demographic
characteristics of CLA referrals were
compared to other referrals. When examining
differences in age, it was found that
in the category of 41-50 years old,
where the expected success rate for
males was 55 percent, the 41-50 year-old
males at CLA had a rate of 40.9 percent,
while the non-CLA referred males had
a rate of 72.2 percent. (Males 41-50
made up about 50 percent of each referral
group.) Also examined was the recovery
time participants had before entering
VDP. Overall, 81.8 percent of the study
population was in their first year of
recovery, with 60 percent in the first
6 months. Referrals from CLA were overwhelmingly
in their first year of recovery (95.5
percent) with 70.5 percent in their
first 6 months. The majority of referrals
from other sources were also in their
first year (74.2 percent) with 57.1
percent in their first 6 months. However,
in both time periods 1-6 months and
7-12 months non-CLA referrals did better
than CLA referrals (66.7 percent vs.
45.5 percent success at 7- 12 months
and 55 percent vs. 48.4 percent at 1-6
months). The conclusion here is that
the differences between CLA referrals
and non-CLA referrals hold up even within
specific demographic groups, and do
not give an explanation for that difference.
It may be that more information is needed
regarding the differences between the
program at CLA (which is more structured
and involved) and the environment at
VDP.
Perhaps CLA residents become accustomed
to a certain amount of support and interaction
that does not exist at the same level
of intensity in an independent living
situation such as VDP. Other factors
to be considered would be a more in-depth
look at when most relapses happen are
CLA residents moving into VDP during
their most vulnerable phase? More work
also needs to be done in looking at
trends over time. Both CLA and VDP are
relatively new programs, with changes
in both staffing and guidelines over
time. Looking at averages for the full
time period (over five years) may not
give a true picture. For example, when
comparing trends over time for CLA referrals
and non-CLA referrals, the rates vary
tremendously by year. This question
deserves a great deal of scrutiny from
all angles, both for decision-making
related to VDP and for indication of
possible changes needed at CLA.
Given the scarcity of appropriate/affordable
housing for homeless people in recovery,
and the impact of the potential stability
provided in their lives, this program
is definitely worth continuing.
Despite the limitations of the manual
data collection process that staff have
used over the years (there is a great
deal of missing data and what does exist
in the files is poorly organized), there
is a wealth of information here that
needs to be mined more deeply. Many
questions remain unanswered (and many
are still unasked). Continuation of
the data collection and analysis in
the format that has been developed is
highly recommended, with the addition
of the following:
- Improve systems for documentation
of VDP residents and link to the AHCH
client database.
- Do a similar study using the data
from CLA to compare dropout rates and
characteristics with VDP residents and
to continue searching for explanations
regarding the difference in success
rates for CLA referrals to VDP.
- Do client satisfaction surveys in
both VDP and CLA to round out the data
with more qualitative input and gain
valuable insight to some of the issues
in question. Include questions regarding
the elements of the program(s) that
support them, what doesn_t support them,
their suggestions for relapse prevention,
what they expected when they entered
the program (from themselves and from
the program) and how that compares to
what they got.
- Add a question to intake regarding
sources of health care other than AHCH.
- Review client records for other AHCH
services to better understand utilization
patterns.
- Verify self-report information on
physical and mental health problems
with the client record.
- Examine changes in resident income
from VDP entry to exit.
- Do a cost-effectiveness study of VDP.
