Abstract
Russia’s invasion of Ukraine and the ongoing war have had devastating effects on health programs. This observational study assessed uptake and persistent use of HIV pre-exposure prophylaxis (PrEP) during wartime. The study included newly-initiating daily PrEP users from 94 clinics who enrolled on PrEP from October 2022 through September 2023. PrEP persistence was defined as returning for a PrEP refill no later than 7 days after the expected refill date. We used Kaplan–Meier curves and sex-stratified Cox regression to explore PrEP discontinuation by sex, key population group, age group, and wartime location. 4537 clients initiated daily PrEP, with 31.9% being men who have sex with men (MSM), 29.7% based in a frontline location, and 16.7% aged 18–25. Six-month persistence was 37.3% overall. It was lower among MSM (28.6%), young adults aged 18–25 (17.2%), and frontline residents (15.6%), and higher among sex workers (SW) (81.5%). In adjusted analysis among women, SW had a lower risk of discontinuing PrEP compared with discordant couples (adjusted hazard ratio [aHR]: 0.22, 95% confidence interval [CI] 0.14–0.34). In both sexes, clients aged 18–25 had a higher risk of discontinuing PrEP (aHR: 1.43, 95% CI 1.25–1.65 for men; aHR: 1.60, 95% CI 1.16–2.20 for women), as did those based in frontline locations (aHR: 2.19, 95% CI 1.99–2.41 for men; aHR: 1.26, 95% CI 1.04–1.53 for women). The findings highlight the resilience of Ukraine’s PrEP program and the need to support PrEP persistence among younger and frontline populations.
Keywords: HIV pre-exposure prophylaxis (PrEP), Ukraine, Key populations, War and health, Delivery of health care, Routinely collected health data
Introduction
On February 24, 2022, the Russian Federation invaded Ukraine, launching a brutal attack on the country, leading to more than 30,000 civilian casualties, displacement of millions of citizens, and destruction of homes, infrastructure, and healthcare facilities [1, 2]. Prior to the war, Ukraine had 245,000 people with HIV (PWH), the second largest population of PWH in Eastern Europe [3–5]. Wartime disruption to HIV testing, prevention, care and treatment services has been profound; 31 of 250 (12.4%) medical facilities providing HIV-related services closed because of the war, and many others experienced direct attacks, personnel shortages, and supply chain disruptions [5]. These disruptions combined with heightened rates of sexual violence, human trafficking, and intimate partner violence are expected to increase the transmission of HIV and other infectious diseases [6–8]. Key population (KP) groups at risk of acquiring HIV, such as sex workers (SW) and men who have sex with men (MSM), have become more vulnerable throughout the war [9, 10].
The Ukrainian Public Health Center estimated there were 11,658 new cases of HIV and 157,435 PWH who remained in unoccupied areas of Ukraine in 2023 [11]. Prior to the war, there were 16,357 new HIV diagnoses in 2019 and 15,658 in 2020 [12]. Ukraine continues to invest in HIV epidemic control despite the war [4, 5, 11]. In keeping with World Health Organization (WHO) guidelines for HIV pre-exposure prophylaxis (PrEP) as a key HIV prevention strategy [9], Ukraine had introduced PrEP services in 2017 via a two-year pilot project [13]. The number of people receiving PrEP in Ukraine grew from 2258 in 2020 to 13,147 in 2023 [11].
The effectiveness of oral PrEP depends on persistent, daily use [14]. It is unclear how wartime conditions have impacted the ability of at-risk populations to access and sustain PrEP use in Ukraine. Our team previously published an analysis of Ukraine’s PrEP program outcomes before the full-scale Russian invasion, which found 6-month persistence of 42.4% under a definition that allowed a 15-day buffer period for PrEP refills [15]. The aims of the present study were to describe PrEP uptake and persistence within wartime Ukraine. We specifically studied how persistence levels varied by KP groups, sex, age, and location of residence relative to the frontlines of the war.
Methods
Study Design and Setting
This observational study was led by the International Training and Education Center for Health (I-TECH), with a team based at the University of Washington and in Kyiv, Ukraine. I-TECH has provided technical assistance to Ukraine’s Ministry of Health (MOH) in strengthening HIV-related clinical training and service delivery, with funding from the US Health Resources and Services Administration (HRSA) within the U.S. President’s Emergency Plan for AIDS Relief (PEPFAR). With the onset of the full-scale war, PEPFAR provided funding to support HIV services and provision of antiretroviral therapy (ART) and PrEP medication throughout the country [5, 16]. As of September 2023, I-TECH supported MOH and private healthcare providers to deliver PrEP services within 107 hospitals and primary care clinics offering HIV-related services across 12 Ukrainian regions (oblasts). The study used routinely collected, deidentified data from Ukraine’s PrEP program.
Ukraine’s PrEP program follows international guidelines, with availability of an oral PrEP regimen of emtricitabine and tenofovir disoproxil fumarate (TDF/FTC 300/200 mg per day) for use in daily PrEP as well as on-demand or event-driven PrEP. PrEP medication is provided cost-free to clients through the national program. Clinical guidelines for PrEP dispensation recommend a 30-day PrEP supply at the first visit, a 60-day supply at the second visit, and a 90-day supply for subsequent visits.
Study Participants and Data Source
Participants included clients who initiated daily PrEP between October 1, 2022 and September 30, 2023, at 94 clinical sites within 12 oblasts. Given our interest in PrEP persistence, we excluded on-demand/event-driven PrEP clients, since they had unpredictable refill schedules (Fig. 1).
Fig. 1.
Flow diagram of study population inclusion
The study used routinely collected clinical data recorded within the Information System for Socially Significant Diseases (IS-SSD), an electronic health record system created in 2016 to capture data on HIV-related client care. IS-SSD contains client demographic data, clinical observations, and data on pharmacy prescriptions and laboratory testing. Demographic data abstracted from clinic records included client sex, age, KP group, and treatment facility. Clients were categorized into four age groups (18–25, 26–35, 36–45, and 46 +) and one of six KP groups (men who have sex with men [MSM], people who engage in transactional sex work [SW], transgender people [TG], people who inject drugs [PWID], discordant couples [DC)], and individuals with risky heterosexual behavior [RHB], i.e., heterosexual contact with PWH or unprotected sexual contact with casual sexual partners), as described elsewhere [15]. Clients with more than one KP group noted were reclassified to the KP group with the highest risk for HIV, based on HIV prevalence estimates in the following order (from highest to lowest risk): PWID, MSM, SW, TG, DC, and RHB. Sex was classified as male or female. Gender identity was not recorded, but providers could record transgender clients as a distinct KP group.
We also grouped clients by wartime location at the frontlines of the war, based on the location of the health facility where they initiated PrEP. The government of Ukraine maintains a list of sub-regional areas (called rayons) grouped in three wartime categories: (1) temporarily occupied by Russian forces; (2) experiencing active hostilities; and (3) facing possible hostilities (meaning they neighbor areas with active hostilities) [17]. We classified clients as living in a “frontline” location if their health facility was in a rayon that experienced occupation, active hostilities, or possible hostilities at any point during the study observation period. All others were classified as “non-frontline.”
Data Analysis
We compared the demographic profile of new daily PrEP clients by KP group. We conducted a time-to-event analysis to estimate the proportion persisting on PrEP during the first 6 months after initiation by KP group, sex, age, and wartime location. We used Kaplan–Meier curves to compare persistence time for each group and a log-rank test to determine statistical significance of any differences. We performed Cox regression using sex-stratified models adjusted for KP group, age group, and wartime location, to investigate the independent effect of each characteristic on PrEP persistence, censoring participants after the first instance of PrEP discontinuation.
Our primary outcome, PrEP persistence at 6 months (180 days), was defined as a client returning for a PrEP medication dispensation no later than 7 days after their expected refill date (termed “Buffer7”). Clients who returned for a refill before the expected refill date were counted as persisting on PrEP. Expected refill dates were recorded by providers after adding the number of pills dispensed to the client’s most recent dispensation date. There were a total of 11,343 PrEP dispensations across daily PrEP clients, and among these 3657 (32.2%) lacked an expected refill date. To handle missing expected refill dates, we considered PrEP dispensation guidelines as well as the median intervals observed in non-missing data. On this basis, we imputed values assuming 30-day supply of medication for 1st dispensations, 60-day supply for 2nd dispensations, 90-day supply for 3rd through 6th dispensations, 60-day supply for 7th dispensations, and 30-day supply for 8–11th dispensations. Only 71 clients had 7 or more dispensations during the 12-month period. For clients who persisted on PrEP while always returning no later than 7 days after their expected refill dates, time on PrEP was calculated as the number of days between their initial dispensation and the administrative censor date of September 30, 2023. If clients received a positive HIV test, they were censored at the test date.
We also considered two alternative PrEP persistence outcome definitions, one more strict, requiring a client to return on or before their expected refill date (termed “Buffer0”), and one less strict, returning no later than 14 days after their expected refill date (termed “Buffer14”). The Buffer7 definition was our primary outcome, since it provided some flexibility for clients to receive their refill if their calculated refill date fell on a weekend or if they were unable to make it to the health center on time.
All analyses were completed in R version 4.2.3 [18].
Ethical Approval
The Ethics Committee of the Ukraine Ministry of Health Center for Public Health, the University of Washington Human Subjects Division, and the US Centers for Disease Control and Prevention (CDC) reviewed the study protocol. All three approved the study as a program evaluation with minimal risk to human subjects and waived the requirement for client consent to participate based on the secondary use of routinely collected, deidentified client data.
Results
Uptake of Daily PrEP and Client Characteristics
We identified a total of 4,537 daily PrEP users who initiated PrEP during the study period, for an average of 383 new clients per month (Fig. 1). Our sample population was classified as 31.9% MSM, 27.7% DC, 22.0% PWID, 16.4% RHB, and 2.0% SW (Table 1). Only one daily PrEP client was classified as TG, and was excluded from further analysis. Among included clients, 74.1% were male and 25.9% were female. They ranged in age from 18 to 78 years; males had a median age of 35 years (interquartile range [IQR]: 27–42) while females had a median age of 39 years (IQR: 34–46). Most MSM were 35 years or younger (82.7%), while most DC (67.6%), PWID (66.9%), SW (57.1%), and RHB (71.1%) clients were older than 35. A total of 1,346 (29.7%) of clients were from frontline locations, with a greater share of males (33.4%) than females (18.9%) in frontline locations.
Table 1.
Demographic characteristics of clients > = 18 years who initiated PrEP from 10/1/22–9/30/23
| Population groups (n, col %) | ||||||
|---|---|---|---|---|---|---|
| All N = 4537 |
MSM N = 1447 |
DC N = 1257 |
PWID N = 997 |
RHB N = 744 |
SW N = 91 |
|
| Males | 3363 (74.1%) | 1447 (100.0%) | 567 (45.1%) | 844 (84.7%) | 505 (67.9%) | 1 (1.1%) |
| Females | 1173 (25.9%) | 0 (0.0%) | 691 (54.9%) | 153 (15.3%) | 239 (32.1%) | 90 (98.9%) |
| Age in years, median (IQR) | 36 (29–43) | 28 (22–33) | 40 (34–47) | 38 (34–44) | 41 (43–49) | 37 (32–47) |
| Age 18–25 | 758 (16.7%) | 574 (39.7%) | 72 (5.7%) | 40 (4.0%) | 66 (8.9%) | 5 (5.5%) |
| Age 26–35 | 1431 (31.5%) | 623 (43.1%) | 335 (26.7%) | 290 (29.1%) | 149 (20.0%) | 34 (37.4%) |
| Age 36–45 | 1457 (32.1%) | 201 (13.9%) | 494 (39.3%) | 478 (47.9%) | 259 (34.8%) | 25 (27.5%) |
| Age 46 + | 891 (19.6%) | 49 (3.4%) | 356 (28.3%) | 189 (19.0%) | 270 (36.3%) | 27 (29.7%) |
| Frontline | 1346 (29.7%) | 719 (49.7%) | 249 (19.8%) | 246 (24.7%) | 127 (17.1%) | 5 (5.5%) |
| Non-frontline | 3191 (70.3%) | 728 (50.3%) | 1008 (80.2%) | 751 (75.3%) | 617 (82.9%) | 86 (94.5%) |
| Males | ||||||
| Age in years, median (IQR) | 35 (27–42) | 28 (22–33) | 40 (34–48) | 38 (34–44) | 41 (34–48) | 57 |
| Age 18–25 | 687 (20.4%) | 573 (39.7%) | 32 (5.6%) | 31 (3.7%) | 50 (9.9%) | 0 (0.0%) |
| Age 26–35 | 1096 (32.6%) | 623 (43.1%) | 141 (24.9%) | 242 (28.7%) | 90 (17.8%) | 0 (0.0%) |
| Age 36–45 | 1013 (30.1%) | 201 (13.9%) | 219 (38.6%) | 408 (48.3%) | 185 (36.6%) | 0 (0.0%) |
| Age 46 + | 568 (16.9%) | 49 (3.4%) | 175 (30.9%) | 163 (19.3%) | 180 (35.6%) | 1 (100.0%) |
| Frontline | 1124 (33.4%) | 719 (49.7%) | 117 (20.6%) | 210 (24.9%) | 78 (15.4%) | 0 (0.0%) |
| Non-frontline | 2239 (66.6%) | 728 (50.3%) | 450 (79.4%) | 634 (75.1%) | 427 (84.6%) | 0 (0.0%) |
| Females | ||||||
| Age in years, median (IQR) | 39 (34–46) | NA | 39 (34–46) | 38 (34–43) | 42 (34–50) | 37 (32–46) |
| Age 18–25 | 71 (6.1%) | NA | 41 (5.8%) | 9 (5.9%) | 16 (6.7%) | 5 (5.6%) |
| Age 26–35 | 335 (28.6%) | NA | 194 (28.1%) | 48 (31.4%) | 59 (24.7%) | 34 (37.8%) |
| Age 36–45 | 444 (37.9%) | NA | 275 (39.9%) | 70 (45.8%) | 74 (31.0%) | 25 (27.8%) |
| Age 46 + | 323 (27.5%) | NA | 181 (26.2%) | 26 (17.0%) | 90 (37.7%) | 26 (28.9%) |
| Frontline | 222 (18.9%) | NA | 132 (19.1%) | 36 (23.5%) | 49 (20.5%) | 5 (5.6%) |
| Non-frontline | 949 (81.1%) | NA | 556 (80.9%) | 117 (76.5%) | 190 (79.5%) | 85 (94.4%) |
KPs key populations, PWID people who inject drugs, MSM men who have sex with men, SW sex workers, DC discordant couples, RHB risky heterosexual behavior, NA not applicable (no/few individuals in this group)
PrEP Persistence
Overall daily-PrEP persistence was 86.9% at 1 month, 59.5% at 3 months, and 37.4% at 6 months among daily PrEP users, based on the Buffer7 outcome definition (Table 2). Estimates were slightly higher, 90.6% at 1 month, 64.1% at 3 months, and 41.5% at 6 months, based on the Buffer14 definition (Table 2). The strictest definition, Buffer0, requiring participants to return on or before the expected refill date, produced lower estimates of 62.9% at 1 month, 26.0% at 3 months, and 13.4% at 6 months (Table 2).
Table 2.
PrEP persistence estimates, by outcome buffer period, timepoint after PrEP initiation, KP group, and frontline location (n = 4536)
| % Persisting (Buffer7) | % Persisting (Buffer0) | % Persisting (Buffer14) | |
|---|---|---|---|
| Proportion Persisting (n at risk for failure outcome at each time point) | |||
| Total n = 4536 | |||
| 1-month | 86.9% (3,975) | 62.9% (3,299) | 90.6% (4,047) |
| 3-months | 59.5% (2,086) | 26.0% (928) | 64.1% (2,230) |
| 6-months | 37.4% (839) | 13.4% (348) | 41.5% (913) |
| MSM total n = 1447 | |||
| 1-month | 90.1% (1257) | 68.7% (1,019) | 94.7% (1,290) |
| 3-months | 52.7% (596) | 18.5% (201) | 57.6% (646) |
| 6-months | 28.6% (155) | 11.0% (88) | 29.7% (163) |
| DC total n = 1257 | |||
| 1-month | 83.5% (1084) | 61.6% (933) | 86.0% (1,096) |
| 3-months | 61.2% (566) | 31.7% (300) | 65.0% (596) |
| 6-months | 38.5% (236) | 17.4% (105) | 42.2% (256) |
| PWID total n = 997 | |||
| 1-month | 83.7% (897) | 52.3% (714) | 88.9% (917) |
| 3-months | 59.2% (485) | 24.8% (215) | 65.9% (531) |
| 6-months | 37.8% (213) | 15.3% (96) | 46.0% (237) |
| RHB total n = 744 | |||
| 1-month | 89.4% (649) | 67.0% (560) | 91.2% (656) |
| 3-months | 67.0% (369) | 29.9% (169) | 69.6% (383) |
| 6-months | 47.7% (186) | 7.0% (33) | 52.8% (206) |
| SW total n = 91 | |||
| 1-month | 96.6% (88) | 71.6% (73) | 98.9% (88) |
| 3-months | 88.2% (70) | 49.7% (43) | 93.0% (74) |
| 6-months | 81.5% (49) | 33.2% (26) | 84.9% (51) |
| Age 18–25 total n = 757 | |||
| 1-month | 88.1% (668) | 61.0% (499) | 93.0% (686) |
| 3-months | 46.5% (304) | 11.9% (76) | 52.7% (337) |
| 6-months | 17.2% (55) | 4.0% (21) | 19.6% (65) |
| Age 26–35 total n = 1431 | |||
| 1-month | 86.8% (1,255) | 64.9% (1,080) | 90.9% (1,276) |
| 3-months | 60.2% (624) | 29.2% (304) | 64.5% (664) |
| 6-months | 40.0% (266) | 17.1% (133) | 43.9% (286) |
| Age 36–45 total n = 1457 | |||
| 1-month | 86.2% (1,274) | 60.6% (1,053) | 89.0% (1,293) |
| 3-months | 63.4% (720) | 29.9% (347) | 67.8% (764) |
| 6-months | 42.2% (327) | 15.5% (132) | 47.6% (358) |
| Age 46 + total n = 891 | |||
| 1-month | 87.3% (778) | 64.9% (667) | 90.3% (792) |
| 3-months | 63.4% (438) | 27.4% (201) | 67.4% (465) |
| 6-months | 43.9% (191) | 12.6% (62) | 48.1% (204) |
| Frontline total n = 1346 | |||
| 1-month | 85.3% (1,107) | 53.7% (728) | 90.2% (1,167) |
| 3-months | 43.3% (511) | 12.9% (153) | 49.8% (573) |
| 6-months | 15.7% (139) | 5.2% (47) | 19.6% (169) |
| Non-frontline total n = 3190 | |||
| 1-month | 87.6% (2,868) | 66.8% (2,571) | 90.7% (2,880) |
| 3-months | 66.9% (,1575) | 32.0% (775) | 70.6% (1,657) |
| 6-months | 47.8% (700) | 17.2% (301) | 52.0% (744) |
Outcome definitions: Time intervals: 1 month (30 days); 3 months (90 days); 6 months (180 days). Buffer7 = returned for PrEP refill no later than 7 days after the expected next dispensation date; Buffer0 = returned for PrEP refill on or before the expected next dispensation date; Buffer14 = returned for PrEP refill no later than 14 days after the expected next dispensation date. Clients who returned for a PrEP refill before the expected refill date were counted as persisting on PrEP
KPs key populations, PWID people who inject drugs, MSM men who have sex with men, SW sex workers, DC discordant couples, RHB risky heterosexual behavior, NA not applicable. The single transgender client was excluded from the analysis
PrEP persistence varied significantly across KP groups (p-value < 0.0001) (Fig. 2). Six-month persistence was 81.4% (95% CI 73.4–90.5) among SW and but only 28.6% (95% CI 26.1–31.5) among MSM. When stratified by age group, 6-month persistence was lowest among those aged 18–24, but similar in those aged 25–35, 36–45, and 46 + years (p-value < 0.0001). It was higher among females than males (43.5% versus 35.3%, p-value < 0.0001). It was significantly lower among clients in frontline locations than in non-frontline locations (15.6% versus 47.8%, p-value < 0.0001) (Fig. 3).
Fig. 2.
Probability of PrEP persistence for clients > = 18 years who initiated PrEP from 10/1/22–9/30/23, by key population group (Kaplan Meier curve) (n = 4536). KPs key populations, MSM men who have sex with men, DC discordant couples, PWID people who inject drugs, RHB risky heterosexual behavior, SW sex workers. The single transgender client was excluded from the analysis
Fig. 3.
Probability of PrEP persistence for clients > = 18 years who initiated PrEP from 10/1/22–9/30/23, by wartime location (Kaplan Meier curve) (n = 4536). A single transgender client was excluded from the analysis
In the sex-stratified adjusted Cox model for males, MSM, PWID, and RHB groups all had lower risk of PrEP discontinuation compared to the DC reference group, after adjustment for age and wartime location (adjusted hazard ratios [aHR]: 0.85, 0.84 and 0.84 respectively, all p < 0.05, see Table 3). In the adjusted Cox model for females, SW had a lower risk of discontinuing PrEP compared with the DC reference group (aHR: 0.23, 95% CI 0.15–0.36) (Table 3). For both males and females, the 18–25 year age group had significantly higher risk of PrEP discontinuation (among males, aHR: 1.43, 95% CI 1.25–1.65, p < 0.001; among females, aHR: 1.60, 95% CI 1.16–2.20, p < 0.01). For both males and females, clients in frontline locations had a significantly higher risk of discontinuing PrEP (among males, aHR: 2.19, 95% CI 1.99–2.41, p < 0.001; among females, aHR: 1.26, 95% CI 1.04–1.53, p = 0.02).
Table 3.
Adjusted and unadjusted risk of PrEP discontinuation, by key population group, sex, age group, and wartime location (Cox Proportional Hazard model)
| Characteristic | Unadjusted hazard ratio | 95% CI | p-value | Adjusted hazard ratio ± | 95% CI | p-value |
|---|---|---|---|---|---|---|
| Males (n = 3362) | ||||||
| Key Population (ref = DC) | ||||||
| MSM | 1.20 | (1.06–1.36) | 0.005 | 0.85 | (0.74–.98) | 0.03 |
| PWID | 0.88 | (0.77–1.01) | 0.07 | 0.84 | (0.73–0.96) | 0.01 |
| RHB | 0.82 | (0.70–0.96) | 0.02 | 0.84 | (0.72–0.99) | 0.04 |
| Age Group (ref = 36–45 years) | ||||||
| 18–25 | 1.78 | (1.58–2.00) | < 0.001 | 1.43 | (1.25–1.65) | < 0.001 |
| 26–35 | 1.10 | (0.98–1.23) | 0.10 | 1.12 | (0.99–1.26) | 0.07 |
| 46 + | 0.98 | (0.85–1.12) | 0.7 | 0.97 | (0.84–1.11) | 0.60 |
| Wartime Location (ref = Non-frontline) | ||||||
| Frontline | 2.33 | (2.13–2.54) | < 0.001 | 2.19 | (1.99–2.41) | < 0.001 |
| Females (n = 1173) | ||||||
| Key Population (ref = DC) | ||||||
| PWID | 1.23 | (0.98–1.55) | 0.07 | 1.25 | (1.0–1.57) | 0.06 |
| RHB | 0.92 | (074–1.13) | 0.40 | 0.94 | (0.76–1.17) | 0.60 |
| SW | 0.23 | (0.15–0.36) | < 0.001 | 0.23 | (0.15–0.36) | < 0.001 |
| Age Group (ref = 36–45 years) | ||||||
| 18–25 | 1.52 | (1.11–2.09) | 0.01 | 1.60 | (1.16–2.20) | < 0.004 |
| 26–35 | 0.99 | (0.81–1.21) | > 0.9 | 1.12 | (0.91–1.27) | 0.30 |
| 46 + | 0.92 | (0.75–1.13) | 0.4 | 0.97 | (0.79–1.19) | 0.80 |
| Wartime Location (ref = Non-frontline) | ||||||
| Frontline | 1.40 | (1.15–1.69) | < 0.001 | 1.26 | (1.04–1.53) | 0.02 |
A single transgender client was excluded from the analysis, and the male model excluded a single client in the SW category. ± Adjusted for all characteristics shown in the table. Results with statistical significance of p < 0.05 are shown in bold
Discussion
Among more than 4,500 clients served in 94 Ukrainian health facilities from October 2022-September 2023, following Russia’s full-scale invasion, the largest key groups initiating daily PrEP were MSM, discordant couples, and PWID, who together made up more than 80% of all PrEP clients. Although HIV prevalence among men and women is similar in Ukraine [19], about three-quarters of new PrEP clients in the wartime setting were men. Overall PrEP persistence was moderate, with more than a third of clients continuing PrEP 6 months after initiation. SW demonstrated the highest persistence, while MSM experienced the lowest persistence. In adjusted analyses, factors associated with lower persistence included younger age and location in frontline war areas, while SW showed higher persistence relative to other KPs.
Our analysis allows for a comparison of Ukraine’s PrEP program during wartime with an analysis of PrEP services in 40 health facilities from before the full-scale war [15]. The age distribution of clients was similar, but the participation of KP groups shifted somewhat. During wartime, there was a lower proportion of female PrEP initiators (25% vs. 35% pre-war), a higher proportion of MSM (32% vs. 22% pre-war), and a lower proportion of SW (2% vs. 5% pre-war). SW accounted for only 7.6% of the women in our study and most were clustered in a small number of health facilities. It is possible that SW remain underserved, and/or that they were misclassified into other risk groups because of stigmatization and criminalization of sex work in Ukraine. The low numbers of women and SW initiating PrEP in the wartime context is concerning, though the displacement of more than 4.7 million adult female refugees outside of Ukraine’s borders may have contributed to the low proportion of females enrolling in PrEP services [20, 21].
It is difficult to compare PrEP persistence outcomes during pre-war and wartime periods in Ukraine, or to compare outcomes in Ukraine with outcomes from real-world observational studies in other settings, due to different data sources and definitions of PrEP persistence [22]. In our pre-war study of PrEP persistence in Ukraine, which lacked information on clients’ expected PrEP refill dates, an outcome definition that assumed 90-day refill intervals across all dispensations plus a 15-day buffer period [15] was somewhat similar to the estimate in the present study that used the 14-day buffer period. Using these respective definitions, we estimated overall 6-month PrEP persistence of 41.5% during wartime, remarkably similar to the pre-war estimate of 42.4% [15]. In comparison, a meta-analysis of PrEP persistence based on 59 published studies from around the globe found a pooled estimate of PrEP persistence of 59.0% at 6 months, with substantial heterogeneity based on global region, sex and gender group, and whether the estimates arose from observational or interventional studies [22]. Large observational studies in the US, Belgium, and France have noted PrEP discontinuation rates of 18–44% within the first year of PrEP use, but with widely variable definitions of persistence [23–25]. In terms of evidence from HIV programs in other wartime settings, we noted evidence on elevated HIV transmission risk [26, 27] and disruption of HIV antiretroviral therapy during war [4, 28], but no studies specifically on PrEP persistence in wartime.
Our results suggest that absolute levels of PrEP persistence among SW in Ukraine may have improved during wartime compared to pre-war. Again comparing the most similar 6-month outcome definitions, the pre-war estimate for SW was 52.1%, while during wartime, the estimate was notably higher, 84.9% [15]. This contrast might reflect a change in the phenomenon of sex work during wartime. A recent study of HIV risk in wartime Ukraine found that SW saw a decrease in business from their normal clients due to financial instability or relocation, but an increase in military clients, therefore potentially increasing their perceived risk of HIV acquisition due to the influx of new sexual contacts [29]. Although there is no direct evidence of how changes in sex work may have affected PrEP uptake or persistence, the changes could have motivated SW to adhere to PrEP regimens to lower their risk of HIV. An alternative explanation is that the handful of health facilities who disproportionately served SW during wartime typically provided complementary social services through partnerships with non-governmental organizations, which may have helped SW to maintain persistent use of PrEP. It is also possible that measurement issues related to small sample sizes and misclassification of SW differentially biased our estimates between the two time periods.
Before the war, age was not a significant predictor of PrEP discontinuation [15], yet in wartime we found higher risk of discontinuation among clients aged 25 and younger, with a 60% excess risk among females and a 40% excess risk among males, after adjusting for KP group and frontline location. Young adults face significant challenges in accessing and staying engaged in healthcare services for chronic conditions like HIV due to difficulties navigating adult health systems and lack of service awareness [30, 31]. This may have been exacerbated by wartime conditions. The deployment of young Ukrainian men and women into combat could have decreased their ability to access PrEP and maintain consistent use, though further study would be needed to verify this hypothesis. The elevated risk of PrEP discontinuation in frontline wartime locations, especially among males who faced a more than two-fold excess risk of discontinuation in frontline locations, underscores the vulnerability of clients at risk of HIV in the areas most heavily impacted by the war.
Our results showed low levels of PrEP persistence among MSM, consistent with a trend observed before the war [15]. MSM in our study were disproportionately young and based in frontline locations, with nearly 40% falling in the 18–25 year age category and almost half located at the frontlines. These co-occurring factors—male sex, youth, and frontline location—appeared to largely explain the low absolute PrEP persistence levels among MSM, since we found reduced risk of PrEP discontinuation among MSM compared to the DC reference group, after stratification by sex and adjustment for age and frontline location.
Our study offers insights on the remarkable efforts to sustain HIV prevention services during the war. Wartime disruptions to the health care system, internal displacement of Ukrainians, and economic hardship have increased risk factors for HIV and impacted the ability of vulnerable populations to access PrEP services [8, 32]. For example, in the months after Russia’s full-scale invasion, PWIDs experienced shortages in sterile needles, which increased syringe sharing, and many were internally displaced and lacked the necessary information to find areas to safely use drugs in their new environments [8]. An influx of newly displaced SW relocating to areas in Western and South-western Ukraine like Odesa may have increased competition among SW and resulted in riskier sexual behavior [8].
Our study demonstrates how the Ukrainian healthcare system has proved resilient and adaptable despite wide destruction and forced closure of many healthcare facilities. Although the war has created obstacles to providing and receiving PrEP services, the response of the Ukrainian MOH, non-governmental organizations (NGOs), and community workers has been creative and strong [33, 34]. Wartime service adaptations include meeting with patients online when possible and distributing medications within communities via courier services or mobile vans [5, 6, 34]. Our results underscore an ongoing need to identify additional strategies for raising awareness of PrEP services among women at risk of HIV, especially women involved in sex work, and for promoting age-tailored interventions that support PrEP persistence among young people.
This study benefited from several strengths that enhanced the reliability and applicability of its findings. Firstly, the large sample size allowed for a comprehensive analysis and provided statistical power to detect significant differences in PrEP persistence across sub-groups. Additionally, the availability of data recorded at health facilities on various KPs, age groups, and wartime locations in Ukraine, provided insight into specific dynamics within these groups.
There were also several limitations to this evaluation. Routinely-collected PrEP dispensation and refill data represents an imperfect proxy for PrEP adherence. For example, a prior study among PWID PrEP users in Ukraine found evidence of “white coat compliance” where 79% of clients returned for PrEP refill visits but only about 4% of those who returned had biomarker evidence showing consistent daily use of PrEP [35, 36]. A second limitation of our study was data inaccuracies and missing data. Despite the robust quality assurance protocols implemented by I-TECH Ukraine, some PrEP providers were not directly involved in I-TECH’s technical assistance activities, which may have led to variable knowledge of data collection procedures, lower motivation to document services with high completeness and accuracy, and less communication to resolve misalignments in data. The relatively high degree of missing data for expected PrEP refill dates, in nearly one third of all PrEP refills, necessitated the imputation and our assumptions may have introduced bias due to misclassification of PrEP persistence. Focusing on enhancing data quality control measures and collecting comprehensive records will continue to strengthen the validity of future program evaluations.
There are several areas for future research on the implementation and outcomes of Ukraine’s PrEP program. Further study is needed of the unique barriers, stigmas, and facilitators of access for at-risk groups, especially women and young people, so that those at risk of HIV exposure are aware of PrEP and can initiate PrEP use within the wartime setting [10, 37]. Further study of the reasons for discontinuation of PrEP in different geographies and among groups vulnerable to HIV is needed to determine the degree to which PrEP discontinuation represents low adherence despite high risk of HIV exposure versus transition to low risk of HIV exposure. Further study of the distinctive barriers to PrEP adherence and program models to overcome these barriers among clients with ongoing risk of HIV exposure is also needed, in order to ensure that PrEP programs are aligned to the needs of the most vulnerable.
Conclusion
In conclusion, we observed moderate PrEP persistence in wartime Ukraine, consistent with levels observed pre-war. PrEP services are critical to controlling HIV transmission amidst wartime conditions that favor a surging epidemic. The accomplishments of Ukraine’s national PrEP program reflect the resilience of Ukrainian healthcare providers and clients, as well as continued support from international partners. Further work is needed to enhance PrEP uptake and sustainted PrEP use by KPs and other populations vulnerable to HIV, especially women, young adults, and those residing in frontline areas.
Acknowledgements
On behalf of the International Training and Education Center for Health (I-TECH) at the University of Washington, the authors would like to thank the US Health Resources and Services Administration (HRSA) and the US Centers for Disease Control and Prevention (CDC) in Ukraine for a decade-long strategic partnership in-country and support of Ukraine’s national HIV pre-exposure prophylaxis (PrEP) program.
Author Contributions
E.G.M, A.P.I. and N.H.P. performed the research. A.P.I., L.I.H., A.P.K., and O.V.D. facilitated data access. A.P.I., A.P.S., M.R.M. and N.H.P. designed the research study. A.P.I., L.I.H., A.P.K., and O.V.D. supported interpretation of results. E.G.M, A.P.I., and N.H.P. analysed the data. E.G.M, L.E.M, and N.H.P. wrote the paper. All authors have read and approved the final manuscript.
Funding
This manuscript was developed by I-TECH with funding from HRSA of the U.S. Department of Health and Human Services (HHS) under Cooperative Agreement No. U91HA06801. The information and conclusions in this document are those of the author(s) and should not be construed as the official position or policy of, nor should any endorsements be inferred by HRSA, HHS or the U.S. Government.
Declarations
Conflict of interest
L.E.M. as received research materials and funding from Hologic, Inc. and Nabriva Therapeutics, Ltd. and conference support from Hologic, Inc., unrelated to this work. This publication was made possible by a grant to the International Training and Education Center for Health (number U91HA06801) from the U. S. Department of Health and Human Services, Health Resources and Services Administration (HRSA), Office of Global Health. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the government.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.United Nations (UN). Ukraine: Report reveals war’s long-term impact which will be felt ‘for generations’. United Nations News. https://news.un.org/en/story/2024/02/1146842 (accessed 21 October, 2024).
- 2.World Health Organization (WHO). WHO records more than 1000 attacks on health care in Ukraine over the past 15 months of full-scale war. WHO/Europe Press Office. https://www.who.int/europe/news/item/30-05-2023-who-records-1-000th-attack-on-health-care-in-ukraine-over-the-past-15-months-of-full-scale-war (accessed 21 October, 2024).
- 3.Gokengin D, et al. PrEP scale-up and pep in central and eastern Europe: changes in time and the challenges we face with no expected HIV vaccine in the near future. Vaccines (Basel). 2023. 10.3390/vaccines11010122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Vasylyev M, et al. Unified European support framework to sustain the HIV cascade of care for people living with HIV including in displaced populations of war-struck Ukraine. Lancet HIV. 2022;9(6):e438–48. 10.1016/S2352-3018(22)00125-4. [DOI] [PubMed] [Google Scholar]
- 5.Public Health Center of the Ministry of Health of Ukraine (PHC), Annual Report of the Public Health Center of the MOH of Ukraine: National Response of HIV, TB, Viral Hepatitis, and SMT Programmes in the Context of Full-Scale Russian Invasion, Ministry of Health, Kyiv, Ukraine, 2023. Accessed: 10 October 2024. [Online]. Available: https://phc.org.ua/sites/default/files/users/user90/National_response_HIV_TB_VH_SMT_war_2023_ENG.pdf
- 6.Barsukova A, et al. Impact of war on HIV-related healthcare services and health workers in eastern Ukraine: a qualitative study. AIDS Care. 2024. 10.1080/09540121.2024.2332445. [DOI] [PubMed] [Google Scholar]
- 7.Obrizan M, Iavorskyi P. Health consequences of the war in Eastern Ukraine: comparing 2015–16 to 2012–13. Eur J Comp Econom. 2023;20(2):239–63. 10.25428/1824-2979/024. [Google Scholar]
- 8.Friedman SR, Smyrnov P, Vasylyeva TI. Will the Russian war in Ukraine unleash larger epidemics of HIV, TB and associated conditions and diseases in Ukraine? Harm Reduct J. 2023;20(1):119. 10.1186/s12954-023-00855-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.World Health Organization (WHO), Differentiated and simplified pre-exposure prophylaxis for HIV prevention: update to WHO implementation guidance, World Health Organization, Geneva, 2022. Accessed: 23 October 2024. [Online]. Available: https://iris.who.int/bitstream/handle/10665/360861/9789240053694-eng.pdf?sequence=1
- 10.Rousseau E, Julies RF, Madubela N, Kassim S. Novel platforms for biomedical HIV prevention delivery to key populations—community mobile clinics, peer-supported, pharmacy-led PrEP delivery, and the use of telemedicine. Curr HIV/AIDS Rep. 2021;18(6):500–7. 10.1007/s11904-021-00578-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Korshunova K, Antonenko Z, Zubko M et al., newsletter of the public health center of the ministry of health of Ukraine: HIV Infection in Ukraine. Public Health Center, Kyiv, 2024
- 12.Antonenko Z, Kovalchuk A, Martsynovska V, Pohorielova O, Zubko M, HIV Infection in Ukrain: Information Bulletin #54. Public Health Center of the Ministry of Health of Ukraine (PHC), Kyiv, 2023. Accessed: 27 March 2025. [Online]. Available: https://phc.org.ua/sites/default/files/users/user92/HIV_in_UA_54_2023_EN.pdf
- 13.UNAIDS, Global AIDS Monitoring 2019: Ukraine Summary, UNAIDS, Geneva, 2019. Accessed: 23 October 2024. [Online]. Available: https://www.unaids.org/sites/default/files/country/documents/UKR_2020_countryreport.pdf
- 14.Haberer JE, Mujugira A, Mayer KH. The future of HIV pre-exposure prophylaxis adherence: reducing barriers and increasing opportunities. Lancet HIV. 2023;10(6):e404–11. 10.1016/s2352-3018(23)00079-6. [DOI] [PubMed] [Google Scholar]
- 15.Vitruk O, et al. Uptake and persistent use of HIV preexposure prophylaxis among key populations: results from Ukraine’s scaled national preexposure prophylaxis program. J Acquir Immune Defic Syndr. 2024;96(3):241–9. 10.1097/qai.0000000000003432. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Cairns G, Rapid scale-up of PrEP in Ukraine this year, despite the war. Aidsmap. https://www.aidsmap.com/news/oct-2022/rapid-scale-prep-ukraine-year-despite-war (accessed 24 October, 2024).
- 17.Vereshchuk I, On the approval of the Amendments to the List of territories where hostilities are (were) being conducted or temporarily occupied by the Russian Federation. Ministry of Reintegration of the Temporarily Occupied Territories of Ukraine. https://minre.gov.ua/2024/08/09/pro-zatverdzhennya-zmin-do-pereliku-terytorij-na-yakyh-vedutsya-velysya-bojovi-diyi-abo-tymchasovo-okupovanyh-rosijskoyu-federacziyeyu-21/ accessed.
- 18.R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing. http://www.R-project.org/ (accessed 11 September, 2023)
- 19.UNAIDS. AIDSInfo: Global data on HIV epidemiology and response. Factsheets. UNAIDS. https://aidsinfo.unaids.org/ (accessed 24 October, 2024)
- 20.Organization for Economic Cooperation and Development (OECD), What are the integration challenges of Ukrainian refugee women?, In OECD Policy Responses on the Impacts of the War in Ukraine, Paris, 30 May 2023 2023, vol. 2024. [Online]. Available: https://www.oecd.org/en/publications/what-are-the-integration-challenges-of-ukrainian-refugee-women_bb17dc64-en.html
- 21.UN High Commissioner for Refugees (UNHCR). Operational Data Portal: Ukraine Refugee Situation. United Nations High Commissioner for Refugees. https://data.unhcr.org/en/situations/ukraine (accessed 24 October, 2024)
- 22.Zhang J, et al. Discontinuation, suboptimal adherence, and reinitiation of oral HIV pre-exposure prophylaxis: a global systematic review and meta-analysis. Lancet HIV. 2022;9(4):e254–68. 10.1016/S2352-3018(22)00030-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Coy KC, Hazen RJ, Kirkham HS, Delpino A, Siegler AJ. Persistence on HIV preexposure prophylaxis medication over a 2-year period among a national sample of 7148 PrEP users, United States, 2015 to 2017. J Int AIDS Soc. 2019;22(2): e25252. 10.1002/jia2.25252. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Rotsaert A, et al. Pre-exposure prophylaxis (PrEP) use trajectories and incidence of HIV and other sexually transmitted infections among PrEP users in Belgium: a cohort analysis of insurance claims data from 2017 to 2019. BMC Public Health. 2024;24(1):2296. 10.1186/s12889-024-19691-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Le Roux C, et al. Factors associated with PrEP persistence and loss of follow-up: a 5-year historic cohort. Arch Sex Behav. 2024;53(7):2445–52. 10.1007/s10508-024-02862-0. [DOI] [PubMed] [Google Scholar]
- 26.Harsono D, et al. A scoping review of factors associated with HIV acquisition in the context of humanitarian crises. AIDS Behav. 2024. 10.1007/s10461-024-04504-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Bennett BW, Marshall BD, Gjelsvik A, McGarvey ST, Lurie MN. HIV Incidence prior to, during, and after violent conflict in 36 sub-saharan African nations, 1990–2012: an ecological study. PLoS ONE. 2015;10(11): e0142343. 10.1371/journal.pone.0142343. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Vasylyeva TI, Horyniak D, Bojorquez I, Pham MD. Left behind on the path to 90–90-90: understanding and responding to HIV among displaced people. J Int AIDS Soc. 2022;25(11): e26031. 10.1002/jia2.26031. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Lazarus L, et al. Exploring the impact of military conflict on sex work in Ukraine: women’s experiences of economic burden. Glob Public Health. 2023;18(1):2092187. 10.1080/17441692.2022.2092187. [DOI] [PubMed] [Google Scholar]
- 30.Kyselyova G, et al. Young people in HIV care in Ukraine: a national survey on characteristics and service provision. F1000Res. 2019;8:323. 10.12688/f1000research.18573.2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Schaefer R, Peralta H, Radebe M, Baggaley R. Young people need more hiv prevention options, delivered in an acceptable way. J Adolesc Health. 2023;73(6s):S8-s10. 10.1016/j.jadohealth.2023.08.046. [DOI] [PubMed] [Google Scholar]
- 32.Owczarzak J, et al. “Will you need this health at all? Will you be alive?”: using the bioecological model of mass trauma to understand HIV care experiences during the war in Ukraine. J Int AIDS Soc. 2024;27:e26307. 10.1002/jia2.26307. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Lopatina Y, et al. Safeguarding HIV prevention and care services amidst military conflict: experiences from Ukraine. BMJ Glob Health. 2023. 10.1136/bmjgh-2023-014299. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Lazarus L, et al. “...because the social work never ends”: a qualitative study exploring how NGOs responded to emerging needs while upholding responsibility to HIV prevention and treatment during the war in Ukraine. J Int AIDS Soc. 2024;27:e26309. 10.1002/jia2.26309. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Morozova O, et al. Patterns of daily oral HIV PrEP adherence among people who inject drugs in Ukraine: an analysis of biomarkers. J Int AIDS Soc. 2024;27:e26319. 10.1002/jia2.26319. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Dumchev K, et al. Low daily oral PrEP adherence and low validity of self-report in a randomized trial among PWID in Ukraine. Int J Drug Policy. 2024;123: 104284. 10.1016/j.drugpo.2023.104284. [DOI] [PubMed] [Google Scholar]
- 37.Hillis A, Germain J, Hope V, McVeigh J, Van Hout MC. Pre-exposure prophylaxis (PrEP) for HIV prevention among men who have sex with men (MSM): a scoping review on PrEP service delivery and programming. AIDS Behav. 2020;24(11):3056–70. 10.1007/s10461-020-02855-9. [DOI] [PMC free article] [PubMed] [Google Scholar]