This webinar has been made possible with support from the
Canadian Institutes of Health Research
Welcome!
Social Determinants
of Health
Program Planning:
What’s the evidence?
You will be placed on hold until the webinar begins.
The webinar will begin shortly, please remain on the line.

What’s the evidence?
 Fitzpatrick-Lewis, D., Ganann, R., Krishnaratne, S.,
Ciliska, D., Kouyoumdjian, F., Hwang, S.W. (2011).
Effectiveness of interventions to improve the health
and housing status of homeless people: A rapid
systematic review. BMC Public Health, 11: 638.
 http://www.health-evidence.ca/articles/show/21957

Housekeeping
 Connection issues
 Recommend you use a wired Internet
connection rather than wireless, to help prevent
connection challenges
 Please contact the WebEx 24/7 help line:
1-866-229-3239

Side Panel in WebEx
Housekeeping Feedback icon
 How to post comments/questions Hand icon
during the webinar Mic request
 To write in CHAT or Q&A
• Address questions to all Chat
panelists
• Raise hand using the ‘hand’ icon
(indicated on the right)
• Respond to general comments
using the feedback icon
Q&A
 To talk:
• If have a head set, you can ask to
be passed the mic (mic request
icon on right)

This webinar has been made possible with support from the
Canadian Institutes of Health Research
Welcome!
Social Determinants
of Health
Program Planning:
What’s the evidence?
You will be placed on hold until the webinar begins.
The webinar will begin shortly, please remain on the line.

The Health Evidence Team
Maureen Dobbins Kara DeCorby Daiva Tirilis
Scientific Director Administrative Director Research Coordinator
Tel: 905 525-9140 ext 22481 Tel: (905) 525-9140 ext. 20461 Tel: (905) 525-9140 ext. 20460
E-mail: dobbinsm@mcmaster.ca E-mail: kdecorby@health-evidence.ca E-mail: dtirilis@health-evidence.ca
Lori Greco Heather Husson Robyn Traynor Lyndsey McRae
Knowledge Broker Project Manager Research Coordinator Research Assistant

What is www.health-evidence.ca?
Evidence
inform
Decision Making

Why use www.health-evidence.ca?
1. Saves you time
2. Relevant & current evidence
3. Transparent process
4. Supports for EIDM available
5. Easy to use

Meetings, Planning &
Dissemination Project
CIHR Funded MOP-238541

CIHR-Funded Reviews
 Hwang, S. W., Tolomiczenko, G., Kouyoumdjian, F. G., &
Garner, R. E. (2005). Interventions to improve the health
of the homeless: A systematic review. American Journal of
Preventive Medicine, 29(4), 311-319.
 UPDATE: Fitzpatrick-Lewis, D., Ganann, R., Krishnaratne,
S., Ciliska, D., Kouyoumdjian, F., Hwang, S.W. (2011).
Effectiveness of interventions to improve the health and
housing status of homeless people: A rapid systematic
review. BMC Public Health, 11: 638.

Rapid Reviews
 Rapid reviews are literature reviews that use
methods to accelerate or streamline
traditional systematic review processes.
(Ganann, 2010)

Questions?

Summary Statement:
Fitzpatrick-Lewis (2011)

Overall Considerations
Considerations for Public Health Practice
Conclusions from Health Evidence General Implications
Provision of housing for homeless or marginally
housed populations leads to:
• increased housing stability
• small, but significant, decreases in
substance/alcohol use
• longer durations of abstinence
• reduced emergency department or
psychiatric inpatient use
• improved quality of life

Overall Considerations
Considerations for Public Health Practice
Conclusions from Health Evidence General Implications
Provision of housing for homeless or marginally
housed populations leads to:
• increased housing stability
• small, but significant, decreases in
substance/alcohol use
• longer durations of abstinence
• reduced emergency department or
psychiatric inpatient use
• improved quality of life
Adding case management and/or day treatment
services to housing provision for varying
homeless populations leads to:
• improved housing stability
• less need for substance abuse treatment
• improved antiretroviral adherence

Overall Considerations
Considerations for Public Health Practice
Conclusions from Health Evidence General Implications
Public health programs should include and/or
Provision of housing for homeless or marginally
housed populations leads to: support:
• increased housing stability • provision of housing with rent subsidy for
• small, but significant, decreases in homeless people with mental illness
substance/alcohol use • housing, preferably abstinent contingent, for
• longer durations of abstinence homeless people with substance abuse
• reduced emergency department or • individual counseling to reduce risk among
psychiatric inpatient use homeless people with HIV/AIDS
• improved quality of life • weekly educational sessions for homeless
or runaway youth
Adding case management and/or day treatment • the provision of housing and/or moderate-
services to housing provision for varying consistent case management for homeless
homeless populations leads to: people with HIV/AIDS
• improved housing stability
• less need for substance abuse treatment
• improved antiretroviral adherence

Overall Considerations
Considerations for Public Health Practice
Conclusions from Health Evidence General Implications
Public health programs should include and/or
Provision of housing for homeless or marginally
housed populations leads to: support:
• increased housing stability • provision of housing with rent subsidy for
• small, but significant, decreases in homeless people with mental illness
substance/alcohol use • housing, preferably abstinent contingent, for
• longer durations of abstinence homeless people with substance abuse
• reduced emergency department or • individual counseling to reduce risk among
psychiatric inpatient use homeless people with HIV/AIDS
• improved quality of life • weekly educational sessions for homeless
or runaway youth
Adding case management and/or day treatment • the provision of housing and/or moderate-
services to housing provision for varying consistent case management for homeless
homeless populations leads to: people with HIV/AIDS
• improved housing stability Non-abstinent contingent housing with case
• less need for substance abuse treatment management is not recommended for
• improved antiretroviral adherence homeless people with concurrent disorders
to decrease psychiatric symptoms & substance,
but is recommended to stabilize housing.

Homeless people with
mental illness
Studies Homelessness Definitions
Forchuk et al., 2008 Patients being discharges from psychiatric wards to
shelters and no fixed address.

Interpreting the Evidence
Interventions for homeless people with mental illness (1 RCT)
What’s the evidence? Implications for practice &
policy
• The provision of discharge support (i.e. assistance
with finding housing and payment of first/last
month’s rent) led to significant improvements in
housing status up to 6 months post-discharge,
compared to individuals receiving usual care (i.e.
social work referral) (p < 0.001).

Interpreting the Evidence
Interventions for homeless people with mental illness (1 RCT)
What’s the evidence? Implications for practice &
policy
• The provision of discharge support (i.e. assistance • Public health decision makers should
with finding housing and payment of first/last consider the positive impact
month’s rent) led to significant improvements in supportive housing and rental
housing status up to 6 months post-discharge, assistance has on housing status of
compared to individuals receiving usual care (i.e. those discharged from psychiatric
social work referral) (p < 0.001). care, while acknowledging the effect is
based on a single study.

Homeless people with
substance abuse issues
Studies Homelessness Definitions
Larimer et al., 2009 ‘Chronically homeless’ list of individuals with high
local crisis services utilization patterns. Chronic
homelessness is not further defined
Milby et al., 2004; Milby Lacking a fixed overnight residence, including shelters
et al., 2003; Milby et al., or temporary accommodations, or were at
2000; Milby et al., 2005; immediate risk of being homeless.
Kertesz et al., 2007

Interpreting the Evidence
Interventions for homeless people with substance abuse issues
(3 studies)
What’s the evidence? Implications for practice & policy
• Housing + on-site case management minimally
decreased the risk of alcohol consumption up to
12 months (RR 0.98; 95% CI 0.96 – 0.99)
What is relative risk?
Ratio of the risk of an event among an exposed
population (intervention group) to the risk among
the unexposed (control group).
(DiCenso, Guyatt, & Ciliska, 2005)

Interpreting the Evidence
Interventions for homeless people with substance abuse issues
(3 studies)
What’s the evidence? Implications for practice & policy
• Housing + on-site case management minimally
decreased the risk of alcohol consumption up to
12 months (RR 0.98; 95% CI 0.96 – 0.99)
• Behavioural day treatment + abstinence-
contingent housing and therapy (DT+)
increased proportion of days abstinent at 2
months (71% vs. 41%) and 6 months (41% vs.
15%) vs. day treatment alone (DT). Relapse was
lower in the DT+ vs DT group (55% vs. 81%).

Interpreting the Evidence
Interventions for homeless people with substance abuse issues
(3 studies)
What’s the evidence? Implications for practice & policy
• Housing + on-site case management minimally
decreased the risk of alcohol consumption up to
12 months (RR 0.98; 95% CI 0.96 – 0.99)
• Behavioural day treatment + abstinence-
contingent housing and therapy (DT+)
increased proportion of days abstinent at 2
months (71% vs. 41%) and 6 months (41% vs.
15%) vs. day treatment alone (DT). Relapse was
lower in the DT+ vs DT group (55% vs. 81%).
• Abstinence-contingent housing increased mean
consecutive weeks of abstinence (7.32) vs. the
no-housing group (5.28) (p = 0.024), and vs. the
non-abstinent-contingent group (4.68) (p =
0.0031) with days housed increased for all
groups (p < 0.0001).

Interpreting the Evidence
Interventions for homeless people with substance abuse issues
(3 studies)
What’s the evidence? Implications for practice & policy
• Housing + on-site case management minimally • Public health decision makers should
decreased alcohol consumption up to 12 months promote and support the provision of
(RR 0.98; 95% CI 0.96 – 0.99) housing, preferably abstinence-
• Behavioural day treatment + abstinence- contingent with on-site case
contingent housing and therapy (DT+) management, to reduce substance use
increased proportion of days abstinent at 2 among homeless people with
months (71% vs. 41%) and 6 months (41% vs. substance abuse issues. DT+ can also
15%) vs. day treatment alone (DT). Relapse was be used to increase days abstinent and
lower in the DT+ vs DT group (55% vs. 81%). also reduce relapse among homeless
• Abstinence-contingent housing decreased mean people with substance abuse issues.
consecutive weeks of abstinence (7.32) vs. the
no-housing group (5.28) (p = 0.024), and vs. the
non-abstinent-contingent group (4.68) (p =
0.0031) with days housed increased for all
groups (p < 0.0001).

Interpreting the Evidence
Interventions with no impact for homeless people with
substance abuse issues
What’s the evidence? Implications for practice & policy
• No impact on days housed in those • Public health should not promote or use
receiving DT+ compared to those receiving DT+ if the sole purpose is to increase the
day treatment only; or, weeks of abstinence number of days housed.
for the non-abstinent contingent housing
group compared to the no housing
group (p = 0.51)

Homeless people with
concurrent disorders
Studies Homelessness Definitions
Gulcur et al., 2003; Spent 15 out of the last 30 days on the street (not
Tsemberis et al., 2004; including shelters) and experienced period of ‘housing
Tsemberis et al., 2003; instability’ (not defined) within last six months.
Padgett et al., 2006;
Greenwood et al., 2005;
Stefancic et al., 2004

Interpreting the Evidence
Interventions for homeless people with concurrent disorders
(1 RCT)
What’s the evidence? Implications for practice & policy
• Study participants given independent apartments
without requirement of abstinence/psychiatric
care spent 66% fewer days homeless (p < 0.001),
and had less need for substance abuse treatment
at 36 months compared to participants receiving
outreach/drop-in services plus group living (p =
0.05). The independent-living, group, however
utilized health care services slightly more, at 48
months post-intervention (p = 0.025).

Interpreting the Evidence
Interventions for homeless people with concurrent disorders
(1 RCT)
What’s the evidence? Implications for practice & policy
• Study participants given independent apartments • Public health decision makers may
without requirement of abstinence/psychiatric advocate non abstinence-contingent,
care spent 66% fewer days homeless (p < 0.001), independent housing as a way to
and had less need for substance abuse treatment improve housing stability and decrease
at 36 months compared to participants receiving need for substance abuse treatment
outreach/drop-in services plus group living (p = for homeless individuals with
0.05). The independent-living, group, however concurrent disorders, while
utilized health care services slightly more, at 48 acknowledging that positive findings
months post-intervention (p = 0.025). are limited to a single study.

Interpreting the Evidence
Interventions with no impact for homeless people with
concurrent disorders
What’s the evidence? Implications for practice & policy
• No difference between groups in • Non-abstinent contingent housing should
psychiatric symptoms and substance use. not be used to improve psychiatric
symptoms, decrease substance use rates, or
decrease healthcare utilization.

Homeless people with HIV
Studies Homelessness Definitions
Woliski et al., HIV+ individuals living in the following housing contexts: having one’s
2009; Kidder et al., own place to live, being unstably housed (staying temporarily with
2007 others/living in a transitional setting and had not been homeless), or
being homeless ≥ one night (e.g., sleeping in shelters or locations not
suitable for human habitation) in the last 90 days.
Rotheram-Borus HIV+ marginally housed individuals including reports of currently
et al., 2009 being homeless, living in a shelter/welfare hotel, or having lived in
other condition within the 12 months prior to each assessment.
Kushel et al., 2006 HIV+ homeless and marginally housed individuals. Homeless was
defined as ≥ one night on street or in shelter in the last quarter,
whereas marginally housed was defined as ≥ 90% of nights in single-
room occupancy dwelling in past quarter with no nights spent on
street or in shelter.
Schwarcz et al., HIV+ individuals. Cases were defined as homeless if medical records
2009 documented individuals were homeless or if addresses listed in chart
were for shelters, health care clinics, or a general delivery address
not connected to an address.

Interpreting the Evidence
Interventions for homeless people with HIV (4 studies)
What’s the evidence? Implications
• Rental assistance + case management improved self-
reported physical/mental health, and housing status (88%
intervention group housed at 18 months vs. 51% with no
intervention, p ≤ 0.0001).

Interpreting the Evidence
Interventions for homeless people with HIV (4 studies)
What’s the evidence? Implications
• Rental assistance + case management improved self-
reported physical/mental health, and housing status (88%
intervention group housed at 18 months vs. 51% with no
intervention, p ≤ 0.0001).
• Three individual counselling modules (five 90-minute
sessions), led to significant decreases in: # days
alcohol/marijuana use (35.77 to 27.54, p = 0.002); # of
risky sexual acts (5.03 to 1.75, p = 0.037); and # days of
hard drug use (27.76 to 24.00, p = 0.042), vs. no
intervention.

Interpreting the Evidence
Interventions for homeless people with HIV (4 studies)
What’s the evidence? Implications
• Rental assistance + case management improved self-
reported physical/mental health, and housing status (88%
intervention group housed at 18 months vs. 51% with no
intervention, p ≤ 0.0001).
• Three individual counselling modules (five 90-minute
sessions), led to significant decreases in: # days
alcohol/marijuana use (35.77 to 27.54, p = 0.002); # of
risky sexual acts (5.03 to 1.75, p = 0.037); and # days of
hard drug use (27.76 to 24.00, p = 0.042), vs. no
intervention.
• Moderate case management improved antiretroviral
adherence (β = 0.13; 95% CI, 0.02-0.25) vs. none or
minimal case management. Consistent and moderate case
management led to 50% improvement in CD4+ cell count.

Interpreting the Evidence
Interventions for homeless people with HIV (4 studies)
What’s the evidence? Implications
• Rental assistance + case management improved self-
reported physical/mental health, and housing status (88%
intervention group housed at 18 months vs. 51% with no
intervention, p ≤ 0.0001).
• Three individual counselling modules (five 90-minute
sessions), led to significant decreases in: # days
alcohol/marijuana use (35.77 to 27.54, p = 0.002); # of
risky sexual acts (5.03 to 1.75, p = 0.037); and # days of
hard drug use (27.76 to 24.00, p = 0.042), vs. no
intervention.
• Moderate case management improved antiretroviral
adherence (β = 0.13; 95% CI, 0.02-0.25) vs. none or
minimal case management. Consistent and moderate case
management led to 50% improvement in CD4+ cell count.
• Risk of death was 20% higher for those not receiving
supportive housing post-HIV diagnosis (Relative Hazard
1.20; 95% CI 1.03, 1.41).

Interpreting the Evidence
Interventions for homeless people with HIV (4 studies)
What’s the evidence? Implications
• Rental assistance + case management improved self- • Public health should support
reported physical/mental health, and housing status (88% housing provision programs,
intervention group housed at 18 months vs. 51% with no preferably with moderate to
intervention, p ≤ 0.0001). consistent case management
• Three individual counselling modules (five 90-minute to improve the health and
sessions), led to significant decreases in: # days housing status of homeless
alcohol/marijuana use (35.77 to 27.54, p = 0.002); # of people with HIV, and to
risky sexual acts (5.03 to 1.75, p = 0.037); and # days of reduce HIV-risk behaviours
hard drug use (27.76 to 24.00, p = 0.042), vs. no and risk of death.
intervention.
• Moderate case management improved antiretroviral
adherence (β = 0.13; 95% CI, 0.02-0.25) vs. no or minimal
case management. Consistent and moderate case
management led to 50% improvement in CD4+ cell count.
• Risk of death was 20% higher for those not receiving
supportive housing post-HIV diagnosis (Relative Hazard
1.20; 95% CI 1.03, 1.41).

Interpreting the Evidence
Interventions for homeless people with HIV (4 studies)
What’s the evidence? Implications
• Rental assistance + case management improved self- • Public health should support
reported physical/mental health, and housing status (88% housing provision programs,
intervention group housed at 18 months vs. 51% with no preferably with moderate to
intervention, p ≤ 0.0001). consistent case management
• Three individual counselling modules (five 90-minute to improve the health and
sessions), led to significant decreases in: # days housing status of homeless
alcohol/marijuana use (35.77 to 27.54, p = 0.002); # of people with HIV, and to
risky sexual acts (5.03 to 1.75, p = 0.037); and # days of reduce HIV-risk behaviours
hard drug use (27.76 to 24.00, p = 0.042), vs. no and risk of death.
intervention. • Decision makers should
• Moderate case management improved antiretroviral advocate for case
adherence (β = 0.13; 95% CI, 0.02-0.25) vs. no or minimal management support for
case management. Consistent and moderate case homeless people with HIV, to
management led to 50% improvement in CD4+ cell count. promote adherence to anti-
• Risk of death was 20% higher for those not receiving retroviral therapy and
supportive housing post-HIV diagnosis (Relative Hazard improve CD4+ cell counts.
1.20; 95% CI 1.03, 1.41).

Interpreting the Evidence
Interventions with no impact for homeless people with HIV
What’s the evidence? Implications for practice & policy
• No impact of multiple individual counselling • Public health should not rely on multiple
sessions compared to no intervention on individual counselling sessions to achieve
abstinence, or provision of rental assistance substance use abstinence.
with case management on # of sexual • Public health should not use rental subsidy
partners, condom use, or sex trading. with case management to impact # of
sexual partners, condom use or sex trading

Homeless or runaway
youth
Studies Homelessness Definitions
Gulcur et al., 2003; Spent 15 out of the last 30 days on the street (not
Tsemberis et al., 2004; including shelters) and experienced period of ‘housing
Tsemberis et al., 2003; instability’ (not defined) within last six months.
Padgett et al., 2006;
Greenwood et al., 2005;
Stefancic et al., 2004

Interpreting the Evidence
Interventions for homeless or runaway youth (1 RCT)
What’s the evidence? Implications for practice & policy
• 12 weekly sessions covering life skills and
psychiatric issues plus HIV/AIDS education
led to a greater reduction in substance use
from baseline (37%) compared with usual
care (17%) (time effect p < 0.001).

Interpreting the Evidence
Interventions for homeless or runaway youth (1 RCT)
What’s the evidence? Implications for practice & policy
• 12 weekly sessions covering life skills and • Public health decision makers should
psychiatric issues plus HIV/AIDS education consider weekly education sessions
led to a greater reduction in substance use covering life skills, mental health and
from baseline (37%) compared with usual HIV/AIDS education to reduce substance
care (17%) (time effect p < 0.001). use among homeless youth, while
acknowledging that positive findings are
limited to a single study.

Overall Considerations
Considerations for Public Health Practice
Conclusions from Health Evidence General Implications
Public health programs should include and/or
Provision of housing for homeless or marginally
housed populations leads to: support:
• increased housing stability • provision of housing with rent subsidy for
• small, but significant, decreases in homeless people with mental illness
substance/alcohol use • housing, preferably abstinent contingent, for
• longer durations of abstinence homeless people with substance abuse
• reduced emergency department or • individual counseling to reduce risk among
psychiatric inpatient use homeless people with HIV/AIDS
• improved quality of life • weekly educational sessions for homeless
or runaway youth
Adding case management and/or day treatment • the provision of housing and/or moderate-
services to housing provision for varying consistent case management for homeless
homeless populations leads to: people with HIV/AIDS
• improved housing stability Non-abstinent contingent housing with case
• less need for substance abuse treatment management is not recommended for
• improved antiretroviral adherence homeless people with concurrent disorders
to decrease psychiatric symptoms & substance,
but is recommended for stable housing.

Questions?

What is the most
important thing you
learned today?

Discussion Forum
Please continue to discuss this topic and other
topics on our discussion forum.
www.health-evidence.ca/forum/
Login with your health-evidence username and password or
register if you aren’t a member yet.
Join us for a LIVE on Monday, December 12 at 1:00 pm
EST to have your questions answered in real time!

Evaluation
Please check your email for the evaluation
link. It take 5 minutes to complete!
If you’ve been watching with someone else and did
not personally register for the webinar, please
e-mail Jennifer McGugan at
mcgugj@mcmaster.ca to be sent the survey.
Thank you for your participation!