Transcript The Aging of the Homeless Population: Fourteen

The Aging of the Homeless Population:
Fourteen-year Trends in San Francisco
Judy Hahn, Margot Kushel, David
Bangsberg, Elise Riley, Andrew Moss
Background
Changes in homelessness in the US
• 1930’s
– Young transient men in search of work
• 1950’s to 1970’s:
– “Skid row”, single older low income white men,
unstable housing (flophouses, SROs, missions)
• 1980’s onward:
– Loss of SRO hotels and affordable housing in urban
centers
– Shelter capacities nationwide increased
– More families and minorities, younger
– Poorer living conditions compared to Skid row
Rossi, American Psychologist 1990
Background
In San Francisco,
• Continued decline in the number of low cost
housing and SRO units in the 1990s
– Units lost due to earthquakes, fires and gentrification
• The response to homelessness
– Establishment of emergency shelters and soup kitchens
and large shelters with services (1980s)
– Policing programs (mid 1990s)
– Supportive housing (some late 1990s, most starting
2004), leveraging Federal $
Ilene Lelchuk, San Francisco Chronicle September 7, 2003.
Objectives
• We have studied HIV and TB in the homeless and
marginally housed in San Francisco from 1990 to
2003.
• In this analysis, we sought to examine changes in
the homeless population over time in:
–
–
–
–
–
Demographics – age, race, sex
Housing
Health status
Health service utilization
Drug use
Methods
• Four cross sectional studies (“waves”)
conducted at shelters and free meal programs
Wave 1:
Wave 2:
Wave 3:
Wave 4:
1990-1994
1996-1997
1998-2000
2003
• Over the entire study period we conducted sampling
at a total of 13 shelters and 8 free meal programs
Methods
• Inclusion criterion: Age 18 and older
• 45 minute interviewer-administered survey
• HIV antibody testing and counseling, TB
testing (waves 1 and 2)
• $10-$20 remuneration for participating
• Anonymous
Analysis
We included in this analysis:
• Shelters and meal programs that were sampled in
at least ¾ of the “waves”
– 4 shelters and 2 free meal programs (78% of those
sampled)
• Study participants who were “literally homeless”
in the prior year
– 87% of those sampled
RESULTS
Demographics (n=3534)
Male
Race
African American
Caucasian
Other, or mixed race
Veterans (of the men)
77%
52%
33%
15%
33%
Substance use and mental illness
Psychiatric hospitalization, ever
Crack use, ever
Injection drug use, ever
Heavy alcohol use, prior 30 days
At least one of the above
Two or more of the above
23%
63%
38%
35%
80%
49%
Age trends
Median age
50
46
45
41
40
42
37
35
30
1990
1995
Year
2000
2005
Age trends by group
100%
11%
18%
20%
32%
80%
60%
50%
60%
65%
52%
40%
20%
Age >=50
Age 35-49
Age <35
38%
22%
15%
15%
1998-2000
2003
0%
1990-1994
1996-1997
1% overall >= age 65
Demographics
60%
50%
56%
53%
49%
47%
40%
African American
36%
30%
31%
Veterans (% of
males)
28%
20%
Women
10%
0%
1990
1995
2000
2005
Housing
48
60%
42
40
50%
Months
36
30
30
24
30%
24
18
12
Median duration
ever homeless
Lived on street,
prior year
20%
12
10%
6
0
1990
40%
0%
1995
2000
2005
Lived in own apt,
prior year
Self-reported health
25%
21%
Hypertension
20%
Psychiatric
hospitalizations
15%
14%
15%
Diabetes
10%
10%
10%
Emphysema
8%
6%
5%
0%
1990
1993
1995
1998
2001
2004
Hospital utilization
60%
52%
50%
43%
42%
40%
31%
30%
21%
20%
Visited
emergency
department
Admitted to
hospital
21%
10%
0%
1995
2000
2005
Drug/alcohol use
60%
50%
Drank heavily
40%
40%
Used crack cocaine
30%
30%
31%
29%
Injected drugs
20%
12%
10%
9%
4%
0%
1990
1995
2000
2005
Used
methamphetamine
Aging in 6 cities
Median age
50
San Francisco
45
Los Angeles
St. Louis
40
Pittsburgh
35
Toronto
New York
City
30
1980
1985
1990
1995
Year
2000
2005
Conclusions and Implications
• The homeless population is getting older.
• This aging indicates that the homeless population
is static and not regenerating itself in time.
– A dynamic population would have as many new young
people joining the population as old people leaving the
population and would have a constant age over time.
• Good news: resources spent on housing the
homeless now may be finite.
Conclusions and Implications
• Of concern: the homeless will increasingly need
health care services -- either to control their
chronic disease or to treat the more serious
outcomes of unmanaged chronic disease.
• Control of chronic disease will be very difficult to
deliver to persons not in housing.
Recommendations
• Provide supportive housing with onsite medical
services for those age 50 and older in order to
intervene in the course of chronic disease early
• Base on the model of San Francisco Department
of Public Health’s Direct Access to Housing
–
–
–
–
–
Houses 1000 people in 12 buildings
3 buildings dedicated to seniors
Psychiatrists, nurses, physicians assistants
Case worker : resident ratio: 1:20
80% stay at least 1 year
Acknowledgments
REACH field staff and study participants
Grants: NIH 5R01DA004363, 1R01MH054907,
R01DA010164, and K08HS11415.
Contact info:
Judy Hahn, Ph.D.
Assistant Professor
EPI-Center, Department of Medicine
University of California, San Francisco
San Francisco, CA 94143-1372
[email protected]
Bonus data!
Younger vs. older homeless
2003 data wave
Age<50 (n=384) Age50 (n=140)
Median age (IQR)
40 (33-45)
53 (51-58)
Sex=Female*
25%
15%
Race/ethnicity*
African American
Caucasian
Mixed/other
43%
34%
23%
55%
28%
17%
*p<0.05
Younger vs. older homeless
Housing
Age<50 (n=384) Age50 (n=140)
Median total months homeless
(IQR)*
36 (9-76)
48 (18-108)
6 (1.5-12.5)
7 (3-16)
Lived on streets/outdoors, prior year
51%
50%
Lived in shelter, prior year
82%
87%
Lived in SRO, prior year
40%
46%
Lived in own apt, prior year*
23%
10%
Median years since first homeless
(IQR)*
*p<0.05
Younger vs. older homeless
Health
Health issue, prior year
Age<50 (n=384)
Age50 (n=140)
Visited ED
49%
44%
Admitted to hospital
29%
26%
Mental health admission
8%
5%
Any days ill (prior 30)
35%
36%
Heart disease*
3%
10%
Hypertension*
14%
34%
Diabetes
6%
6%
Emphysema*
3%
9%
Asthma
13%
15%
Chronic health problem
*p<0.05
Younger vs. older homeless
Substance use
30 day use
Age<50 (n=384) Age50 (n=140)
Heavy drinking
28%
23%
Injected drugs*
18%
9%
Crack cocaine
32%
29%
Powder cocaine
7%
4%
Heroin
13%
8%
Methamphetamine*
18%
8%
Drugs used:
*p<0.05
Summary of bonus data
• Many older homeless persons are using drugs or
drinking heavy amounts of alcohol, though
somewhat fewer than younger homeless persons
• Older homeless persons have the same rate of ED
visits and inpatient hospitalizations though higher
rates of chronic disease