Declining mortality among street youth in Montral

1
Declining mortality amongstreet youth in
Montréal

• Élise Roy, MD, MSc Nancy Haley, MD,
  FRCPCJean-François Boudreau, MSc Pascale
  Leclerc, MSc Jean-François Boivin, MD, FRCPC
• Partners Solving Youth Homelessness
• National Conference
• November 19-21, 2008 - Toronto

2
Background I

• A follow-up prospective study (Cohort 1) with
  semi-annual interviews was carried out among
  Montréal street youth between 1995 and 2000
  (n1013)
• An analysis showed that a little more than 9 out
  of 1000 youth were dying every year
• Standardized mortality ratio was of 11.6 meaning
  that the mortality rate was 11.6 times higher
  than the mortality rate among youth of the same
  age and sex in Québec
• Roy É et al. JAMA, 2004292(5)569-74.

3
Background II

• Causes of death (n26) suicide (13), drug
  overdose (8), injury (2), fulminant hepatitis A
  (1), heart disease (1), unknown (1)
• New public services for street youth were
  implemented between 2000 and 2003 a new
  low-threshold primary care clinic, new social
  housing services, and training sessions on
  suicide prevention for outreach and community
  workers

4
Background III

• According to a second cohort study (Cohort 2)
  conducted between 2001 and 2005 (n 858),
  mortality had declined to 1.91/1000 person-years
• Causes of death (n5) suicide (1), drug overdose
  (2), injury (1), unknown (1)

5
Objective

• To explore the possible causes of the observed
  decrease of mortality among street youth

6
Hypotheses

• H1 Mortality declined similarly in the
• general population
• H2 Mortality risk factors were distributed
  differently in
• Cohort 1 and Cohort 2

7
Methods

• For H1
• Standardized mortality ratios (SMR) were computed
  for both cohorts using annual mortality rates of
  the Québec general population between 1995 and
  2006
• Comparison of the 2 SMR
• For H2
• Analyses of the risk factors conducted on a data
  base combining both cohorts
• Effect of participation in Cohort 1 on mortality,
  controlling for other independent risk factors

8
Results I - Description of participants

• Overall, 1687 subjects were included in the
  analyses
• 829 in Cohort 1 only
• 674 in Cohort 2 only
• 184 in both cohorts

9
Results II - Characteristics of participants
1 missing value 2 missing values 3 missing
values.
10
Results III Sexual history of participants
2 missing values 1 missing value 6 missing
values.
11
Results IV Substance use history of participants
1 missing value.
12
Results V - Hypothesis 1

• Standardized mortality ratios were different and
  this difference was statistically significant
• 11.6 in Cohort 1 versus 3.0 in Cohort 2
• The decrease was more important among street
  youth than in the general population
• SMR Cohort 1 11.6 ( 95 CI 7.6-17.0) SMR
  Cohort 2 3.0 ( 95 CI 1.0-6.9)

13
Results VI - Hypothesis 2
Analyses of Risk Factors of Mortality
Adjusted Hazard Ratio
14
Conclusion I

• Decreased mortality cannot be fully explained by
  our two hypotheses
• Another explanation is that in response to the
  mortality epidemic observed in Cohort 1, new
  public services for street youth were implemented
  between 2000 and 2003

15
Conclusion II

• Although our design does not allow us to claim
  that the implementation of these services led to
  the observed decrease, we think that this is
  highly plausible
• Mortality remains very high among street youth.
  We believe that the provision of services aiming
  at improving their health and decreasing their
  mortality continues to be essential

16
Acknowledgements

• The research staff
• The community based organisations
• The participants

17
Funding agencies

• Cohort 1
• Medical Research Council of Canada
• Ministère de la santé et des services sociaux du
  Québec
• Cohort 2
• Ministère de la santé et des services sociaux du
  Québec
• Health Canada
• Direction de santé publique de Montréal
• Human Resources Development Canada Communities
  Partnership Initiative