PrEP Persistence Among Key Populations in Ukraine: Analysis of Scaled PrEP Program Data from October 2020 through February 2022

While global HIV incidence has rapidly decreased during the last decade, Central and Eastern Europe have experienced a significant rise in new cases annually. Ukraine, located in Eastern Europe, has been one of the most ambitious in the region in implementing programs to combat HIV since 2017, including expanding access to PrEP among key populations starting in 2020. Little is known about persistence on PrEP in subpopulations accessing it in Ukraine. Methods Ukraine’s PrEP program was implemented in 40 health care facilities in 11 high-burden regions in Ukraine. We analyzed data from program initiation on October 1, 2020 through February 23, 2022. A Sankey diagram was constructed to visualize times between the first four PrEP visits. We conducted a Kaplan-Meier analysis to estimate persistence time for each key group, stratifying by men who have sex with men (MSM), people who inject drugs (PWID), sex workers (SW), and discordant couples and other people vulnerable to HIV acquisition (DC/Other). We also evaluated whether persistence varied by sex and age using the Kaplan-Meier method. We used a Cox regression, clustering by healthcare facility, to estimate hazard ratios for non-persistence by key group and sex, and adjust for age as a potential confounding factor. Results There were 2,033 clients newly starting PrEP with up to 510 days of observation. 65% of clients were males, with a median age of 35 years [IQR=29-42] and 35% were females with a median age of 37 years [IQR=31-43]. There were 51% DC/Other, 22% MSM, 22% PWID and 5% SW. The number of days between visits varied across clients and visit timepoints. Among clients returning to collect a 2nd PrEP prescription, 81% returned within 1-45 days after their previous visit, and 3% discontinued PrEP and then re-initiated after 106 days or more. Six-month persistence estimates were highest overall among PWID (28.9%; 95% CI: 23.4-35.7%), moderate in comparison among SW (18.5%; 95% CI: 11.1-30.8%) and DC/Other (12.7%; 95% CI: 10.0-16.0%), and dropped to nearly 0 among MSM (0.8%; 95% CI: 0.1-5.8%). Individuals aged 26 and above were slightly more likely to persist on PrEP longer than younger individuals and females were slightly more likely to persist on PrEP than males. PrEP persistence calculated using the Kaplan-Meier method differed significantly by key group and age (p<0.05), but not by sex. In the Cox regression model, compared with the female DC/Other reference group and adjusted for age, female PWID had the lowest risk of PrEP discontinuation (adjusted HR [aHR] 0.59; 95% CI: 0.31-1.11) while male SW had the highest discontinuation risk (aHR 1.87, 95% CI: 0.57-6.11), although findings were not statistically significant. Conclusion PrEP persistence at 6 months was low with the majority of individuals discontinuing PrEP by 6 months. Further research examining barriers and facilitators of persistence by groups can inform interventions to increase PrEP persistence.