Introduction

Despite ongoing HIV prevention and care efforts [1], Ukraine’s epidemic remains concentrated, with sex workers (SWs) among the most affected key populations [2]. Although the early epidemic was driven primarily by injection drug use, heterosexual transmission has since become dominant [3], accounting for 64.9% of new HIV diagnoses in 2022 [4]. This shift is partly driven by overlapping vulnerabilities between people who inject drugs (PWID) and SWs [1,5], including SWs with injection histories and clients who use drugs [6]. As a result, SWs continue to represent a priority population for HIV prevention and testing programs.

In 2021, an estimated 78,385 SWs were active in Ukraine [7]. Although overall HIV prevalence among SWs declined from 5.2% in 2017 [8] to 3.1% in 2021, prevalence remained high (24.1%) among SWs with a history of injection drug use [9]. This decline likely reflects a combination of mortality, transitions out of sex work, and the expansion of community-based prevention programs led by non-governmental organizations (NGOs) [10]. Nevertheless, HIV prevention coverage among SWs remains below both national and global targets. Program reach increased from approximately 38,000–55,000 SWs between 2017 and 2021 [4], yet the National HIV Strategy aims for 70% coverage [11], while UNAIDS sets a target of 95% by 2025 [12]. Although HIV treatment cascade indicators among SWs improved – from 47% diagnosed, 59% on antiretroviral therapy, and 86% virally suppressed in 2017 [8] to 83-94-80 in 2021 [9] – they still fall short of the 95-95-95 goals. These gaps highlight the need to strengthen HIV testing, which remains a critical entry point to both prevention and treatment services [13], particularly in the context of ongoing war-related population displacement [14] and increased mobility of SWs [15,16], which have disrupted established outreach routes and complicated sustained service delivery across settings.

Delivering HIV prevention and testing services at sex work hotspots through outreach teams or mobile health clinics (vans) is an evidence-based strategy used across diverse settings [17–22]. In Ukraine, the first mobile van was launched in 2004, and rapid HIV testing was introduced in 2006 [23]. These approaches help overcome barriers commonly associated with facility-based services, such as long travel distances, limited awareness of service locations, long waiting times, and inconvenient operating hours [17,24,25]. However, their effectiveness depends on having accurate, up-to-date information about where SWs are actually located.

Hotspot mapping is therefore a foundational component of outreach- and mobile van-based HIV programming [26–29]. It is also essential for time-location sampling frames in integrated bio-behavioral surveys (IBBS), population size estimation (PSE) [30,31], and the design of geographically and temporally appropriate interventions [32–34]. Location-specific and venue-sensitive analyses are also needed to guide effective outreach planning [35], particularly for less visible, informal, and digitally mediated settings [36,37], which is important in the context of uneven and low HIV prevention coverage among SWs [38]. In Ukraine, mapping is typically conducted during IBBS preparation but is rarely validated (often <20%) or used beyond sampling frame development [8,39–44]. Moreover, hotspot data are seldom systematically analyzed or applied to guide service delivery [45]. While individual-level HIV prevention coverage among SWs has been routinely reported through programmatic and surveillance data [46], evidence on prevention service coverage at the level of sex work hotspots in Ukraine remains largely absent, particularly in relation to venue characteristics and service delivery contexts.

To address this gap, we conducted a comprehensive hotspot mapping exercise across eight Ukrainian cities as part of the preparation for the 2021 IBBS round. While the primary purpose of the mapping was to inform recruitment and guide PSE, the resulting dataset also provided an opportunity for systematic analysis to support practical decision-making in HIV prevention – an approach that remains uncommon in Ukraine and is rarely described in the published literature as a basis for routine HIV prevention programming. Beyond identifying and validating active sex work venues, the mapping captured detailed information on hotspot typologies, operating schedules, access modalities, perceived safety, subgroup presence, and outreach activity. This enabled not only a descriptive characterization of where and how sex work is organized, but also an examination of how these operational features relate to HIV prevention service delivery at the hotspot level. By generating granular, context-sensitive data, we aimed to support geographic prioritization, strengthen outreach planning, and improve the equity and efficiency of HIV services for SWs – particularly in the context of limited resources and a dynamic epidemic landscape. Accordingly, this analysis was guided by two research questions: (1) What operational typologies of sex work hotspots can be identified in urban Ukraine based on patterns of sex work organization? (2) Which hotspot characteristics are associated with HIV prevention service coverage?

Materials and methods

Population and settings

Hotspot mapping was conducted between March and July 2021 in eight Ukrainian cities (Cherkasy, Dnipro, Kharkiv, Kropyvnytskyi, Kyiv, Lviv, Mariupol, and Odesa), selected based on HIV prevalence among SWs in 2008–2017, in consultation with the National IBBS Working Group led by the Public Health Center of the Ministry of Health. Table 1 presents city-specific HIV and HCV prevalence and 2019 PSE among SWs.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. HIV prevalence among SWs, HCV prevalence, and population size estimates.

https://doi.org/10.1371/journal.pone.0343803.t001

SWs were defined as adults who exchanged sex for money, goods, or services. Individuals under 18 engaged in commercial sex were classified as sexually exploited minors (SEMs), in line with UNAIDS [47] and WHO [13] guidelines.

Results

During L1, 1,212 secondary KIs identified 3,517 potential hotspots in total; 45.6% were SWs and 18.2% were taxi/truck drivers (Table 2). At L2, 2,581 primary KIs were interviewed on-site, 80.6% of whom were SWs. Validation rates across KI types ranged from 62.4% to 86.8%. Hotspot reports reflected KIs’ professional context: e.g., hotel staff primarily listed hotel sites, while AIDS center staff most often named street-based venues. Apartments were the most frequently identified type.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Typology of key informants and number of identified hotspots across eight Ukrainian cities, 2021.

https://doi.org/10.1371/journal.pone.0343803.t002

Among the 3,517 identified hotspots, 2,581 were unique. Of them, 2,118 (82.1%) were confirmed active (Table 3). The only printed advertisement hotspot failed validation. Apartments comprised 43.2% of hotspots, followed by virtual platforms (11.7%) and street-based sites (11.1%). City-level distributions are provided in S1 Table. Most venues (94.4%) operated year-round, though winter street activity dropped to 71.2%. Escort/on-call venues had the highest 24/7 availability (94.8%), while cafés/bars had the lowest (7.4%). Overall, 63.0% had balanced schedules, 9.2% were daytime, and 27.8% predominantly nighttime venues. A heatmap, provided in S2 Table, showed extended operating hours for office (‘brothel’), virtual, and hotel/motel hotspots.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Characteristics of sex work hotspot types in eight Ukrainian cities, 2021.

https://doi.org/10.1371/journal.pone.0343803.t003

Open access was reported at 88.2% of hotspots; however, entry and contact were often mediated informally by SWs or trusted intermediaries, even in venues classified as “open”. Unsafe conditions (6.8%) – based on observed and/or reported violence or substance use – were concentrated at route/highway hotspots (34.4%) and cafés/bars (21.5%). SWs who inject drugs were reported at 8.3% of hotspots – mostly outdoors and in entertainment venues – and male SWs were present at 4.6%, primarily in pimp-mediated settings.

Hotspots were classified as either exclusive-type (52.6%) or cross-type (47.4%), with cross-venue work most common in pimp‐mediated and escort/on-call settings. EFA produced five factors (Kaiser-Meyer-Olkin measure of sampling adequacy = 0.540; Bartlett’s test of sphericity χ² = 1985.2, df = 66, p < .001), which represent latent groupings based on cross-type working patterns (Table 4). The five factors identified were: (1) Online & Independent Work, loading on virtual and escort hotspots; (2) Street-Apartment Axis, contrasting private apartments and street-based hotspots; (3) Hotel & Leisure Settings, combining hotels, saunas, and massage parlors; (4) Club-Mediated Work, including strip/art clubs, nightlife venues, and pimps; (5) Public-Facing Settings, defined by cafes and bars. Offices and highway hotspots did not align with any factor.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Varimax-rotated factor loadings for sex work hotspot types, 2021.

https://doi.org/10.1371/journal.pone.0343803.t004

HIV service coverage was limited: 13.7% of hotspots were registered in SYREX, 34.1% received services, and 18.9% were reached by mobile vans. Office and street-based sites had the highest engagement; coverage was lowest in apartments, virtual platforms, and pimp-controlled venues.

Regression analysis, presented in Table 5, showed higher odds of HIV prevention service coverage in hotspots with 24/7 operation or non-balanced hours, open access, and the presence of SWs who inject drugs. For example, hotspots perceived as very safe had more than twice the odds of service coverage compared with neutral sites (aOR = 2.3; 95% CI: 1.3–3.9), and hotspots where SWs who inject drugs were present showed substantially higher coverage (aOR = 2.7; 95% CI: 1.6–4.4). Neither the presence of male SWs nor SEMs was significantly associated with coverage.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Associations of hotspot characteristics with HIV prevention service coverage.

https://doi.org/10.1371/journal.pone.0343803.t005

The estimated population of SW reported at hotspots in eight cities was 8,670 (range: 3,837–13,503), with 7,866 (90.7%) observed during 30-minute headcounts (Table 6). Counts varied by hotspot type and city (S1 Table). Apartments hosted the largest numbers (2,897 SWs), yet only 23.2% were estimated to be at registered hotspots, 48.3% at hotspots with reported service provision, and 28.0% at hotspots reached by mobile vans. Virtual venues (2,068 SWs) had the lowest overall coverage.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 6. Estimated number of SWs at hotspots in eight Ukrainian cities, 2021.

https://doi.org/10.1371/journal.pone.0343803.t006

Discussion

This study demonstrated that a multi-informant two-level mapping approach can effectively identify sex work hotspots and generate operationally relevant evidence for HIV prevention programming. Validation of 2,118 active hotspots revealed structural, temporal, and operational heterogeneity, with only about one-third receiving any prevention services. These findings highlight persistent access gaps and reinforce the importance of context-specific resource allocation, in line with UNAIDS’ location-population approach [35] and spatial epidemiological frameworks [28].

Nearly half of identified hotspots involved SWs operating exclusively within a single venue type, underscoring a high degree of operational segmentation in Ukraine’s sex work landscape. Similar patterns have been documented in earlier Ukrainian studies [39], in contrast to settings such as India or Mauritius, where SWs frequently alternate between streets, apartments, and other locations [57,58]. At the same time, cross-type organization was also observed, and factor analysis helped summarize these overlaps into a limited number of shared operational settings, rather than treating each venue category as fully distinct. From a programmatic perspective, these overlaps indicate that HIV prevention strategies organized around single venue categories may fail to capture entire operational networks of sex work. Instead, the findings point to the value of intervention models aligned with broader access environments that share common mechanisms of control, visibility, and risks [59]. For example, the “Hotel & Leisure Settings” factor – linking hotels/motels with massage parlors and saunas – captures a cluster of semi-formal indoor venues characterized by mediated entry and management control, suggesting the value of coordinated venue-level engagement. Such approaches may include negotiated access with venue owners and managers, on-site or proximate delivery of prevention services, and environmental risk reduction strategies that normalize condom use and support safer working conditions [59–62].

Beyond venue clustering, the findings also point to a substantial presence of a highly privatized, decentralized, and low-visibility sex work environment. Historically visible settings – such as street-based sites and brothels – received the majority of programmatic attention, reflecting the long-standing focus of harm reduction efforts on SWs who inject drugs [63,64]. However, several dominant but less visible hotspot settings consistently exhibited markedly lower likelihood of HIV prevention service coverage. In particular, private apartments and Online & Independent Work settings showed the lowest coverage, indicating a systemic and predictable service gap rather than incidental under-reach. This aligns with prior evidence that autonomy, confidentiality, and dispersed organization in such settings limit the feasibility of traditional outreach approaches [63,65], even in contexts of elevated exposure to sexual violence [62]. Similar challenges have been reported across diverse settings, including sub-Saharan Africa, where programmatic mapping exercises have highlighted the limited reach of venue-based models among SWs operating in home-based or digitally mediated contexts [34]. While the global shift toward digitally mediated sex work is well documented across multiple settings and low- and middle-income countries [66–68], empirical evidence on the scale of Online & Independent Work settings and their exclusion from HIV services remains limited [37,68]. To our knowledge, this study is the first in Ukraine to systematically identify and classify digital venues within a broader hotspot mapping framework. Given the consistently low coverage observed in Online & Independent Work settings, these findings suggest that digital strategies should be considered a primary – rather than supplementary – modality of HIV prevention for this segment. However, evidence-based approaches such as virtual outreach through online platforms, HIV self-testing distribution through digital tools, and digital navigation to testing and treatment services [69–71] remain largely absent from Ukraine’s programming [72], with only limited small-scale interventions [11,73,74]. The expansion of digital sex work settings – accelerated by the 2022 full-scale Russian invasion, venue restrictions, and mobility constraints under curfew [16] – is unlikely to reverse, underscoring the urgency of systematic digital outreach to prevent widening service gaps. Taken together, these findings point to a structural policy gap: national HIV programs need to formally recognize digital outreach as a core prevention modality, supported by dedicated funding, performance indicators, and clear guidance on confidentiality, digital safety, and linkage to care.

HIV prevention service coverage also varied systematically across other hotspot characteristics, reflecting broader structural and contextual barriers. Indoor or formalized hotspots – including Hotel & Leisure Settings and Club-Mediated Work environments – were less frequently covered, reflecting the combined effects of managerial control, mediated access, and institutional gatekeeping, a pattern reported in both global [62] and Ukrainian contexts [64]. These access constraints limited outreach to closed venues and contributed to persistent service gaps. In contrast, hotspots where SWs who inject drugs were present were more likely to receive prevention services, indicating that long-standing programmatic prioritization of this subgroup continues to shape service reach across settings. While such individuals were more common at public and highway venues, their presence within manager-controlled environments highlights overlooked prevention needs among less visible segments of this high-risk population. This challenges the assumption – derived from studies in high-income settings such as the UK – that SWs who inject drugs are concentrated exclusively in street-based settings [61], and points to a visibility bias in outreach planning.

Perceived safety emerged as another key determinant of service access. Hotspots described as ‘very safe’ were twice as likely to be reached by prevention teams, underscoring the role of trust, regulation, and perceived risk in enabling outreach [75]. Conversely, cafés and bars (Public-Facing Settings) showed substantially lower coverage despite their public visibility, suggesting that visibility alone does not translate into service access. These venues were frequently described as unsafe environments, where alcohol use is common and condomless sex commands higher prices [62], further constraining outreach feasibility. Importantly, prevention coverage did not correlate with hotspot size: a substantial proportion of Club-Mediated Work settings hosted over 10 SWs, yet these higher-volume locations were less likely to receive services. Together, these findings suggest that structural and gatekeeping barriers may outweigh epidemiological considerations such as venue size or visibility in shaping prevention coverage.

Working hours further influenced prevention service access. Hotspots open 24/7 or primarily during daytime hours were more likely to be reached than those with ‘balanced’ schedules, likely reflecting NGO resource constraints and limited capacity for night-time outreach. Security risks after dark – well documented in other settings including the Democratic Republic of Congo [76] – are particularly acute in Ukraine, where martial law, curfews, and transport limitations restrict safe mobility. Additionally, 24.1–34.2% of cafes, massage parlors, hotels, and nightclubs were inactive at the time of mapping, likely reflecting long-term disruptions to sex work infrastructure because of the COVID-19 pandemic. Comparable COVID-19-related disruptions have been reported across Central and Eastern Europe and Central Asia, where lockdowns and venue closures reshaped sex work organization, reduced safety, and limited outreach feasibility across both indoor and outdoor settings [77]. Taken together, these findings suggest that temporal misalignment between outreach operations and actual hotspot activity constitutes a structural barrier to service coverage. Addressing this gap requires outreach models that are responsive to real operating hours, including adequately resourced evening outreach or flexible-hour models, implemented with appropriate safety measures and coordination with municipal services. In this context, network-based strategies – already implemented in Ukraine [78,79] – have demonstrated promising results in reaching SWs who inject drugs and those operating in concealed or manager-controlled venues [80]. These community-led models adapt to real working conditions rather than rely on fixed schedules or venue-based access. By leveraging peer trust and existing social networks, such approaches extend coverage beyond conventional outreach and help mitigate structural access barriers. A combined model integrating peer recruitment with targeted, time-responsive outreach therefore represents a pragmatic pathway to closing service gaps across diverse sex work settings.

The identified structural challenges also reinforce the critical importance of trust-based partnerships with SW communities. The multi-informant mapping approach, grounded in active participation of SWs as interviewers, KIs, and gatekeepers, achieved a high validation rate (86.8%), with hotspots reported by SWs showing higher confirmation compared to those identified by other informant groups, demonstrating the reliability of community-generated data and the added value of participatory methods. Community engagement was especially effective in identifying hidden or stigmatized venues – a pattern previously documented in community-based screening initiatives in the United States [19] – which were less frequently captured by institutional informants, such as AIDS center staff. Without SW involvement, mapping efforts risk misrepresenting the operational landscape and missing populations most in need of services. Scaling community-led outreach is therefore not only methodologically sound but programmatically strategic. In this study, community participation directly improved data completeness and accuracy, particularly for less visible and privatized settings. Modeling evidence from Ukraine suggests that increasing the share of services delivered by community-based organizations to 65% could avert approximately 2,220 HIV infections among SWs – a 12% reduction [81], aligning with global 2025 targets that emphasize community leadership in HIV prevention and mapping [82]. Integrated models combining hotspot mapping, mobile testing, and community mobilization have shown effectiveness across multiple settings, including Peru [17], sub-Saharan Africa and Thailand [20], and other outreach contexts [76], and could greatly improve reach in Ukraine’s criminalized and under-resourced environments.

Finally, the study revealed substantial underreporting of HIV prevention service coverage within Ukraine’s national monitoring SYREX registry. Only a small fraction (13.7%) of validated hotspots were recorded in SYREX, while KIs reported service delivery at a considerably larger share of locations (34.1%). This discrepancy reflects not only incomplete reporting but also systemic limitations in the data collection system – most notably, the absence of mandatory geolocation. A major contributor is off-site outreach: as in other countries (e.g., in Uganda [25]), Ukrainian teams often meet SWs from indoor or virtual venues at neutral locations (e.g., streets, NGO offices) to protect privacy and minimize legal risk. However, these interactions are rarely linked back to the original hotspot, leaving many serviced venues effectively invisible in official records. Introducing an ‘alternative outreach location’ field with geospatial tagging would improve monitoring accuracy and data completeness. Additional underreporting arises when services are arranged by managers or intermediaries, particularly within structured environments such as the Club-Mediated Work settings [24,39]. In such cases, services may occur outside NGO oversight and remain unregistered. Additionally, time lags between fieldwork (2021) and registry data (2020) may also contribute to this. However, the fact that even highly visible public hotspots – such as highways – were frequently missing in the registry suggests systemic reporting failures. These findings point to the need to revise routine monitoring systems – particularly by introducing mandatory geolocation fields, standardizing reporting for off-site outreach, and strengthening accountability for data completeness – so that resource allocation reflects on-the-ground service delivery.

Conclusion

This study presents the most comprehensive mapping of sex work hotspots in Ukraine to date, revealing numerous previously undocumented venues, diverse operational typologies, and substantial gaps in HIV prevention service coverage. In particular, the low coverage in private apartments, virtual platforms, and other low-visibility settings highlights structural blind spots in current outreach models and underscores the need for digital and network-based prevention strategies. The high validation rate achieved through community engagement demonstrates the value of community-led mapping for identifying hidden and rapidly evolving sex work environments. Evidence from this study suggests that integrating routine hotspot mapping into HIV program monitoring – alongside strengthened community-led outreach and digital service delivery – could substantially improve the alignment between service provision and on-the-ground-realities. As Ukraine’s HIV response continues under rapidly shifting war-related and epidemiological conditions, hotspot mapping should become a core component of adaptive, data-driven HIV programming. Linking mapped hotspot data with service delivery records, including for off-site and digital outreach, is essential to improving resource allocation, reducing service gaps, and accelerating progress toward national and global HIV goals.

Supporting information

Acknowledgments

The authors sincerely thank the SW community members and hotspot staff for their openness, trust, and invaluable contribution to this study. We are also grateful to the key informants and gatekeepers whose cooperation was essential to the successful mapping process. Special thanks go to the regional mapping teams led by Serhii Stratulat, Yulia Tsarevska, Olena Zhuk, Lidiia Zvarych, Olena Duz, Svitlana Sirenko, Natalia Chantseva, and Petro Syvokin, as well as to the NGO specialists involved, for their dedication and proactive support throughout the research. We also deeply appreciate the committed team from Alliance Consultancy, whose expertise and efforts were instrumental in conducting the mapping and data collection.