Background
For justice-involved people, the times during and immediately after incarceration can be high-risk periods for HIV acquisition since these periods afford limited access/adherence to effective HIV prevention. We report results from one of the first longitudinal studies to characterize PrEP use among justice-involved people in sub-Saharan Africa.
Methods
From August 2023—March 2024, we enrolled a cohort of PrEP users and non-users across four correctional facilities in Lusaka, Zambia. HIV-negative incarcerated persons were offered oral tenofovir (TFV)-based PrEP by the health system, and then approached by study staff for participation. Consenting participants were followed from incarceration through community release for HIV seroconversion and PrEP outcomes, including uptake, persistence, and adherence. Follow-up visits were scheduled at 1, 3, 6, 9 and 12 months after enrollment and a randomly selected sub-cohort of participants on PrEP underwent urine TFV screening for objective adherence monitoring. We used logistic regression to examine factors associated with PrEP outcomes.
Results
A total of 706 participants enrolled, with an average age of 28.6 years, majority (672; 95%) male and less than a third (200; 28%) with secondary education. Of 531 (75%) eligible to start PrEP, 326 (61%) initiated PrEP at enrollment. Participants were significantly more likely to initiate PrEP at enrollment if they were sentenced (OR: 1.8, 95% CI:1.3-2.5) vs awaiting trial, had heard of PrEP (OR: 2.3, 95% CI:1.6-3.3) or had greater worry about acquiring HIV in prison (OR 1.1, 95% CI:1.0-1.3). Of 101 persons initiating PrEP at enrollment with a 6-month follow-up visit, 54 (54%) were on PrEP. A total of 59 (29%) of 202 urine TFV screening tests at follow-up indicated adequate adherence. Self-reported PrEP adherence had a positive predictive value of 0.71 (95% CI:0.59-0.80) and negative predictive value of 0.97 (95% CI:0.92-0.99) compared to urine TFV. Two participants seroconverted (0.3%)—one in a PrEP user who had discontinued—and were linked to treatment. Women were 15.7 (95% CI:2.5-304) times as likely than men to persist on PrEP at 6 months.
Conclusions
Initial results suggest high PrEP uptake among incarcerated people in Zambia, particularly among those who had heard of PrEP before. PrEP adherence adjudicated objectively was low (29%) as was persistence, which was higher among women. Our preliminary findings point to the need for differentiated prevention services in this group, especially for men.