ABSTRACT
Aims: This study examined the extent to which youth who live on the streets and in the slums of Kampala experience more HIV risky behaviors than representative national and urban school-attending youth. Place and Duration of Study: Analyses were based on three cross-sectional surveys: 1. the Kampala Youth Survey (service-seeking youth living in the slums; conducted in 2011; N=457); 2. The Global School-based Student Health Survey (GSHS; nationally representative sample of school-attending youth, conducted in 2003; N=3,215); 3. The Global School-based Student Health Survey (GSHS; urban representative sample of school-attending youth, conducted in 2003; N=1,709). Methodology: Analyses restricted to youth between ages 14-17 years assessed the differences in prevalence of sexual intercourse, age of sexual initiation, number of partners, and condom use among youth in the three surveys examined. A z-test was used to test the significance of difference in proportions. The prevalence reported in the Kampala Youth Survey was compared to the other two surveys of school-attending youth. Results: The lifetime prevalence of reporting sexual intercourse was statistically significantly higher (49%) among participants in the Kampala survey compared to the youth in the GSHS National (31%) and GSHS Urban (29%) surveys. Moreover, girls in the Kampala survey reported significantly higher prevalence of sexual intercourse (49%) compared to girls in the GSHS National (23%) and GSHS Urban (22%) surveys. In terms of condom use, the youth in the Kampala survey reported using condoms less in the past year (34%) compared to the youth in the GSHS National (64%) and the GSHS Urban (60%) who reported on condom use at the time of last sexual intercourse. Conclusion: The youth in the Kampala survey reported higher prevalence of sexual intercourse, having fewer sexual partners, and lower condom use compared to their nationally representative school-attending peers. Prevention strategies that seek to increase condom use specifically appear warranted.