Skip to main content
. 2021 Sep 16;4(9):e2125365. doi: 10.1001/jamanetworkopen.2021.25365

Table 3. Unadjusted and Adjusted HIV Incidence, Death, and Time to Incident HIV or Deatha.

Participant exposure HIV incidence (n = 2551)b Death (n = 2474)c Time to incident HIV or death (n = 2474)
Unadjusted HR (95% CI) P value Adjusted HR (95% CI) P value Unadjusted HR (95% CI) P value Adjusted HR (95% CI) P value Unadjusted HR (95% CI) P value Adjusted HR (95% CI) P value
Institutional vs family-based setting 1.33 (0.41-4.35) .64 1.49 (0.46-4.83) .50 0.29 (0.09-0.99) .047 0.26 (0.07-1.02) .05 0.70 (0.28-1.75) .44 0.73 (0.29-1.86) .51
Street vs family-based setting 13.33 (5.42-32.79) <.001 17.31 (5.85-51.25) <.001 7.76 (3.34-18.04) <.001 5.46 (2.30-12.94) <.001 8.84 (4.46-17.53) <.001 7.82 (3.48-17.55) <.001
Male vs female NA NA 0.28 (0.11-0.69) .006 NA NA 1.81 (0.81-4.01) .15 NA NA 0.69 (0.37-1.30) .25
Baseline age ≥12 y vs <12 y NA NA 1.83 (0.74-4.51) .19 NA NA 2.02 (0.83-4.89) .12 NA NA 2.03 (0.99-4.16) .05
Baseline HIV-positive vs HIV-negative status NA NA NA NA NA NA 6.84 (1.88-24.88) .004 NA NA NA NA

Abbreviations: HR, hazard ratio; NA, not applicable.

a

All models (adjusted and unadjusted) for HIV incidence and time to incident HIV or death omitted participants who had HIV-positive status at enrollment. The adjusted models were adjusted for age and sex.

b

Estimated using competing risks analysis, in which death was a competing risk for HIV infection.

c

The adjusted model for death was adjusted for baseline HIV status, age, and sex.