Findings from Enhanced Surveillance
of Canadian Street Youth, 1999-2003
(November 2007)

Canadian Street Youth and Substance Use

Executive Summary Key Findings Introduction Purpose Methods Analysis Analysis Terms Laboratory Methods Limitations Demographic Profile of E-SYS Participants Substance Use Smoking Alcohol Use Recent Alcohol Intoxication General Drug Use Patterns Non-Injection Drug Use Injection Drug Use Injection Drug Use and Hepatitis C Sexual Behaviours and STIs Implications Conclusions References

2. Methods

Enhanced Surveillance of Canadian Street Youth is a multicentre sentinel surveillance system that monitors rates of STIs and BBIs, behaviours and risk determinants in the Canadian street youth population through repeated surveys accompanied by biological sampling (blood and/or urine testing). A pilot project launched in October 1998 (Phase I) investigated the feasibility of studying the street youth population. Since then, data collection has been conducted in large urban centres across Canada biannually, starting in 1999 (Phase II) and continuing in 2001 (Phase III) and 2003 (Phase IV).

In Phase II, the recruitment of youth involved informal snowball sampling methods, a method shown to be effective for hard-to-reach populations.¹² Participants were recruited through drop-in centres and outreach work. For the purposes of the behavioural surveys, the inclusion criteria were that respondents: (a) were between 15 and 24 years of age; (b) were able to speak either French or English; and (c) had, in the previous 6 months, been absent from their residence for at least 3 consecutive nights; run away from home (or another place of residence) for 3 days or more; been thrown out of their home for 3 days or more; or been without a fixed address for 3 days or more.

There were two interviewer-administered questionnaires. The first consisted of questions on demographics, lifestyle, sexual practices, attitudes about and knowledge of risk behaviours, and family history. In addition to completing the questionnaire, consenting youth were asked to provide urine and blood samples. If necessary, a second questionnaire about barriers to partner notification was administered as a follow-up for those whose urine or blood sample tested positive for STIs or BBIs. Phases III and IV of the study followed the same methodology as Phase II. Youth were permitted to participate in the survey only once during each data collection year. A total of 4728 street youth were recruited over the three phases of data collection: 1645 in 1999, 1427 in 2001 and 1656 in 2003.

2.1 Analysis

Data were analyzed using SAS Statistical Software (version 8, SAS Institute, Carey, NC). Chi-square statistics were used to compare distributions across demographic factors. A two-tailed p value of <0.05 was defined as statistically significant for univariate analysis. Further multivariate analyses using logistic regression models were conducted, with a p value of <0.10 defined as statistically significant, to select independent factors associated with specified outcomes.

For more detailed information, please refer to the methodology section of the 2006 Public Health Agency of Canada report Street Youth in Canada: Findings from Enhanced Surveillance of Canadian Street Youth 1999–2003.¹³ For the purposes of this report, younger youth are defined as those 15 to 19 years of age, while older youth are defined as those 20 to 24 years of age.

2.1.1 Analysis terms

Univariate analysis examines the relation of one independent variable to the outcome variable of interest, without taking other potential independent variables into account. In this report, univariate analysis was conducted using the chi-square test for heterogeneity.

Statistically significant relationships in univariate analyses are defined by a p value <0.05 and are denoted by ^§ in this report.

Multivariate analysis examines numerous factors or variables simultaneously: that is, it examines the relation of each independent variable to the outcome variable of interest while controlling for other variables. Logistic regression analysis was employed for multivariate analysis.

Independent associations were determined by multivariate analysis. They are denoted by ^¶ in this report. Variables found to be significantly associated with outcomes of interest in univariate analysis were included in multivariate analysis and were as follows: age; sex; previous STIs; age at initiation of sexual activity; number of sexual partners over lifetime; having been in jail, foster care, and/or group homes; having been assigned to a social worker; having had unwanted sex; having been expelled from school; having dropped out of school; and perceived risk of contracting STIs.

2.2 Laboratory Methods

A polymerase chain reaction (PCR) test was used to detect Chlamydia trachomatis and Neisseria gonorrhoeae (Roche Amplicor). Herpes simplex virus (HSV) antibody was screened using an HSV I/II enzyme immunoassay (EIA) (Meridian); repeatedly reactive and discordant results on EIA were confirmed by type-specific line immunoassay (MRL Diagnostics). Hepatitis C virus (HCV) antibody was screened using an EIA (Ortho HCV 3.0); repeatedly reactive results on EIA were confirmed by recombinant immunoblot assay (RIBA) (HCV 3.0 RIBA). PCR testing was used to detect recent seroconversion if the RIBA (v3) result was indeterminate. Testing was also done for hepatitis B virus (HBV) serology markers (antibodies to HBV surface antigen and core antigen). Syphilis testing was performed using serological testing (rapid plasma reagent [RPR]/Venereal Disease Research Laboratory [VDRL]) followed by confirmatory testing (fluorescent treponemal antibody absorption [FTA-ABS]/[microhemagglutination assay– Treponema pallidum [MHA-TP]).

2.3 Limitations

E-SYS has several limitations that should be noted.

• First, it is cross-sectional in design. As is the case for all cross-sectional research, it does not allow causality to be established.
• Second, the sample is limited to seven urban centres across Canada. This could be viewed as both a strength and limitation. It is a strength in that it documents the presence and the plight of street youth in cities, where they are most often situated. The limitation is that the findings may not be applicable to other geographic areas or to less populated areas.
• Third, the recruitment of youth involved informal snowball sampling methods; the study was well advertised and news of it passed verbally to potential participants without the use of specific geographic sampling frames or direct linked referrals. While this technique has been shown to be effective in hard-to-reach populations such as street youth, it could result in a selection bias.
• Finally, findings were based on self-reported data. As such, veracity of the information provided could not be established except for specific STIs and BBIs, for which blood or urine testing was performed. Also, street youth in this sample may have been reluctant to report socially undesirable behaviours such as unprotected sex, commercial sex trade work or injection drug use.