Transcript Slide 1 - HIV Health and Human Services Planning Council

Delay, Drop-Out, and Connection
to Medical Care: Focus on
SRO Residents.
Angela Aidala and Sara Berk
Mailman School of Public Health, Columbia University
Needs Assessment Committee
May 27, 2008
INTRODUCTION

Persons living in SROs are disproportionately affected by HIV
but are often outside or marginal to systems of HIV care

What can the CHAIN Study tell us about SRO residents and
associations between SRO residency and access and
engagement with HIV primary care?

Three data sources:
- Ongoing cohort study of CHAIN agency recruited sample
- Separate effort to locate and interview PLWHA outside of care
- Interviews with providers from programs targeting PLWHA
outside of care
NYC C.H.A.I.N. STUDY
Community Health Advisory & Information Network
(CHAIN) Project
P Project of the NYC HIV Health and Human Services Planning
Council
P Multi-stage probability sampling:
Random selection of medical and soc svc agencies
Recruit random sample of clients within agencies
P Includes 1661 PLWHA recruited from clinics and agencies in 1994,
1998, 2002
P In-person comprehensive (2-3hr) interview every 6–12 mos
P Strong community support – 80-90% interview rate
DELAYERS/UNCONNECTED PLWHA
Delayers Study
 HIV positives who delay 4+ months to med care
- 1994-1995 CHAIN Cohort (n=247 delayers)
- 2001-2002 CHAIN Cohort (n=174 delayers)
 Quantitative and qualitative interview data
- demographics, health status, service need/use
- narrative descriptions: Why delay? Why drop out?
 Key informant interviews
- providers serving groups at risk for delay
 Focus groups with clients
DELAYERS/UNCONNECTED PLWHA
Unconnected Study
 Aware , no medical care, no case mgmt 6+months
- 1995 n=48 unconnected
- 1999 n=24 unconnected; 26 marginal
- 2003 n=25 unconnected; 35 marginal
n=36 high risk HIV - and out of care

Recruited through outreach in street and community
settings and referrals from agency recruited
participants

Quantitative and qualitative interview data
- demographics, health status, service need/use
- narrative descriptions: Why delay? Why drop out?
FINDING THE UNCONNECTED
# Sites
Visited
# Persons
Screened
HIV +
Unconnected
to Care
SRO
OUTREACH
5
12
6 (50%)
STREET
OUTREACH
15
171
13 ( 8%)
OTHER:
7
118
6 ( 5%)
27
301
25 (8%)
Needle Exchg
Soup Kitchen
Mobil Med Van
Drop-in Center
TOTALS
SRO RESIDENTS
 Classify CHAIN Participants based on coding
descriptions of current and recent housing
 Cross-check against addresses
 SRO Resident = PLWHA who currently or
in the 6
months prior to interview, lived in an SRO or ‘welfare
hotel’
 Focus on CHAIN Participants who were SRO residents at
the time of baseline interview, 2002
 10% of the entire sample had lived in an SRO; of these
one in four had multiple SRO stays
PATTERNS OF SRO RESIDENCY
 10% of the entire 2002 agency-recruited sample had
lived in an SRO during the study period
 Most PLWHA report one period of SRO residency
but one in four with any SRO experience had
multiple SRO stays
 Fewer PLWHA report living in an SRO than at earlier
periods of the CHAIN study
 Those found in SROs during most recent interview
periods are more likely to be PLWHA with multiple SRO
stays
COMPARING SRO RESIDENTS
Comparing SRO residents to PLWHA with no SRO
experience
 More likely to be male (74%); <35yrs old; have very
low incomes; jail experience
 Differences by borough of current residence: more
often Manhattan or Bronx
 Big difference transiency: 58% v. 7% changed
addresses 2+ time past 6 months
 No differences: race/ethnicity; hs grad; ever
worked; currently working
COMPARING SRO RESIDENTS
Comparing SRO residents to PLWHA with no SRO
experience
 No difference in risk exposure group, history of
problem drug use (heroin/ coke/ crack)
 Big differences in patterns of problem drug use:
- More likely to have a history of frequent use (17% v
5% used heroin/coke/crack weekly or more often)
- More often current users (57% v 23%)
 More likely to report unsafe sex
esp. men with hetero partners (14% v 5%)
Connection to Medical Care
SRO
Residents
No SRO
Experience
total sample n=
(69)
(624)
sig
No regular source of medical
care before HIV dx
73%
60%
**
Delayed entry into HIV care
32%
28%
Dropped out of care1
30%
20%
Dropped out because
dissatisfied with care2
32%
26%
Note: Most recent CHAIN cohort, 2002.
1.
Stopped going to the doctor for 6 months or longer
2.
Among those who ever dropped out of care
*
Connection to Medical Care
SRO
Residents
No SRO
Experience
total sample n=
(69)
(624)
No medical provider for HIV
7%
3%
#
Lacks comprehensive primary
care1
35%
23%
*
Care does not meet clinical
practice standards2
42%
25%
**
1+ Visit to ER past 6mo
39%
32%
Hospital inpatient past 6 mos
29%
18%
Note: Most recent CHAIN cohort, 2002.
1.
Care that coordinated, comprehensive, and provides 24hr access in case of
medical emergency
2.
Based on number of visits , diagnostic tests, and ARVs if needed
*
Health Outcomes and Perceived Service Need
SRO
Residents
No SRO
Experience
total sample n=
(69)
(624)
CD4 T-cell count <200
38%
22%
**
Viral load < 10K or ‘bad’
38%
19%
***
Health functioning indicates
impairment
64%
57%
**
Low mental health
functioning
55%
65%
#
Self-report need for mental
health services
20%
12%
*
Self-report need for AOD
treatment/ services
68%
51%
**
Note: Most recent CHAIN cohort, 2002.
PREDICTORS OF CONNECTION TO CARE
P SRO residence
 Co-morbidities: Low mental health functioning, current drug user
 Health status: T-cell count

Housing services: Assistance with housing needs past 6 months
or receipt of rental assistance
 Supportive services: Mental health services, drug treatment,
medical case management, social service case management,
transportation services
 Socio-demographics: Age, ethnicity, education, income
<$7500 yr, living in poverty neighborhood, risk exposure group
 Date of HIV diagnosis, date of cohort enrollment
ANALYSIS
 Logistical regression used to compare the odds of medical care
outcome associated with SRO residence v. other housing
 Also examine receipt of housing assistance vs. no assistance
 Adjusted odds ratios show odds of outcomes controlling for mental
health and substance use co-morbidities, receipt of supportive
services, socio-demographics, and time period
 Each interview with each participant provides opportunity to examine
which predictors are associated with medical care outcomes -1660
individuals interviewed 1-8 times for a total of over 5000 observation
points
 Models constructed using GEE procedures to adjust for dependency
among multiple observations contributed by the same individual
Access to Medical Care
Medical care
meets clinical practice
standards
SRO Residence
0.55 *
1.03
Low mental health functioning
0.82 #
0.84
Current problem drug use
0.95
0.92
Mental health services
1.48 **
1.47 **
Substance abuse treatment
1.13
1.08
Medical case management
1.44*
1.29
Social services case management
1.21
0.96
Housing assistance
N=571 individuals, 1650 observations, 2002 - 2008
1.81 ***
SUMMARY

PLWHA residing in SROs have multiple needs for clinical and
supportive services in addition to housing needs and appear to
recognize their need especially for mental health and drug treatment
services
 SRO residents significantly more likely than other PLWHA to be
marginally connected to HIV medical care – to out of care or not
receiving care that meets minimum clinical practice standards
P SRO residents have lower CD4 counts and higher viral loads than
other PLWHA and are more likely to be hospitalized.
P SRO residency remains a significant predictor of receiving
appropriate clinical care controlling for individual characteristics or
clinical status, or receipt of case management, mental health, drug
treatment, or other supportive services
P Receipt of housing assistance and mental health services are
significant predictors of receiving care that meets good clinical
practice standards and reduces the significance of SRO residency
CONCLUSIONS
 Findings provide strong and consistent evidence that:
-- housing needs are a significant barrier to receipt of
appropriate HIV medical care and continuity of care over time
-- receipt of housing assistance has a direct impact on improved
medical care outcomes for persons living with HIV/AIDS
 Any decrease in funding to provide housing assistance for
PLWHA would seem ill advised
 Improving access to housing will improve access to and
effectiveness of HIV medical care and treatment
 Housing 'expensive' but studies show more than offset by
savings associated with reduced emergency and inpatient
services, treatment failure, and decreased risk of HIV
transmission
ACKNOWLEDGEMENTS
This research was made possible by a series of grants from the US
Health Resources and Service Administration (HRSA) under Title I of
the Ryan White Comprehensive AIDS Resource Emergency (CARE) Act
and contracts with the New York City HIV Health and Human Services
Planning Council through the New York City Department of Health and
Medical and Health Research Association of New York City
Its contents are solely the responsibility of the Researchers and do not
necessarily represent the official views of the U.S. Health Resources
and Services Administration, the City of New York, or the Medical and
Health Research Association.
Special thanks is due to the 1661 persons living with HIV who have
participated in the CHAIN Project and shared their experiences with us.
Contact: [email protected]