TABLE 2—
Dependent Variable | Adjusted Odds Ratio | 95% Confidence Interval |
Personal characteristics | ||
Most money from selling sex, past 6 mos | 1.093 | 0.665, 1.797 |
Ever spent time in a correctional facility, jail, juvenile detention center or juvenile hall, prison, mental health ward of hospital, mental health facility, or other institution | 1.686 | 1.151, 2.470 |
Ever incarcerated | 1.525 | 1.074, 2.164 |
Ever in mental health ward or facility | 1.747 | 1.253, 2.435 |
Homeless, past 6 mos | 1.755 | 1.257, 2.451 |
Risk behaviors | ||
Engaged in receptive syringe sharing | 1.690 | 1.234, 2.316 |
Shared rinse water | 1.549 | 1.360, 2.112 |
Shared cotton | 1.309 | 0.962, 1.783 |
Mean no. of male sexual partners | β = 3.505 | −5.145, 12.154 |
Risk networks | ||
Ever injected with anyone who had hepatitis | 1.680 | 1.093, 2.580 |
Ever injected with another IDU who was at least 5 years older | 1.423 | 1.016, 1.994 |
Ever injected with other WSW | 4.923 | 3.466, 6.992 |
Ever had sex with an IDU | 1.726 | 1.152, 2.586 |
Ever injected with MSM IDU | 2.824 | 1.910, 4.177 |
Ever had unprotected sex with MSM IDU | 3.470 | 2.088, 5.767 |
Ever had sex with someone you knew or thought had HIV | 2.207 | 1.353, 3.601 |
Ever injected with someone who had HIV | 1.298 | 0.844, 1.996 |
Ever had sex with someone you knew or thought had AIDS | 1.479 | 0.679, 3.218 |
Ever had sex with someone you knew or thought had chlamydia | 1.713 | 0.727, 4.037 |
Infections overall | ||
Hepatitis B core antibody positive | 1.548 | 1.059, 2.263 |
HIV seropositivea | 1.607 | 0.857, 3.013 |
In the 2 sites with HIV prevalence >10% | ||
Ever injected drugs with a person who was HIV positive | 3.951 | 1.463, 10.672 |
HIV seropositive | 2.548 | 1.120, 5.797 |
In the 4 sites with HIV prevalence <10% | ||
Ever injected drugs with a person who was HIV positive | 1.018 | 0.632, 1.640 |
HIV seropositivea | 0.865 | 0.323, 2.318 |
Note. MSM = men who have sex with men.
aData for New Orleans, La, were omitted from the analyses of HIV prevalence because the lack of any infected subjects from that location led to quasi-complete separation in the estimates.