HIV seroprevalence among orphaned and homeless youth: no place like home.

OBJECTIVES We evaluated the combined influences of orphaned status and homelessness on HIV seroprevalence and risk among street-involved Ukrainian youth in 2008. DESIGN Systematic, multicity, community-based, cross-sectional assessment. METHODS Time-location sampling was used to identify eligible youth aged 15-24 after city-wide mapping of 91 sites where street-involved youth gathered in Odessa, Kiev, and Donetsk. Universal sampling identified 961 youth in 74 randomly selected sites; 97% consented. Youth reporting one or both parents dead were classified as orphaned; those without a stable residence or sleeping outside their residence at least two nights per week were classified as homeless. Trained staff provided HIV counseling and rapid testing via mobile vans. Adjusted odds ratios (AORs) were calculated using logistic regression, accounting for intracluster homogeneity. RESULTS We found 32% (300 of 929) were both orphaned and homeless; 48% either (but not both) homeless [37% (343 of 929)] or orphaned [11% (104 of 929)]; and [20% (182 of 929)] neither orphaned nor homeless. HIV seroprevalences were 7% for neither orphaned/homeless; 16 and 17%, respectively, for either orphaned/homeless; 28% for both orphaned/homeless (P for trend <0.0001). AORs for HIV infection were 1 for neither; 2.3 and 2.4 for either homeless [95% confidence interval (CI) 1.7-2.9] or orphaned (CI 1.8-3.3); 3.3 for both orphaned/homeless (CI 2.3-4.4). Ever-use of injection drugs increased from 15 to 32 to 48% for those who neither, either, or both orphaned and homeless, respectively (P for trend <0.0001). CONCLUSIONS One of four youths who were both homeless and orphaned was HIV-infected; these youths were significantly more likely to be HIV infected and to report injection drug use than those with adequate housing and living parents.