HIV testing behaviour and HIV prevalence among female sex workers in Ukraine: findings from an Integrated Bio-Behavioural Survey, 2013-2014.

OBJECTIVES: Ukraine has one of the largest HIV epidemics in Europe, with high prevalence among female sex workers (FSWs). We aimed to identify factors associated with HIV testing and receipt of the test result in the last 12 months, HIV prevalence and self-reported positive status among FSWs in Ukraine.
METHODS: We used data from an Integrated Bio-Behavioural Survey among FSWs conducted in 2013-2014. The survey methodology combined three sampling strategies: time and location sampling, respondent-driven sampling and key informant recruitment. We used multivariable regression to identify factors associated with self-reported HIV testing in the last 12 months, HIV prevalence and self-reported positive status among FSWs living with HIV. Explored factors included: age, age at first sex, age at entry into sex work, education, marital status, employment status beside sex work, condom use with last paying or non-paying sexual partner, drug or alcohol consumption and sex work venue.
RESULTS: Recent HIV testing was low overall with only 63.2% of FSWs reported having tested and received their test result in the last 12 months prior to the survey. HIV prevalence was 7.1% overall, but only 45.0% of FSWs living with HIV were aware of their HIV status. Testing in the last 12 months with receipt of test result was less common among FSWs who used drugs ever in life (adjusted OR (AOR) 0.7, 95% CI 0.6 to 0.9), women soliciting clients indoors (AOR 0.8, 95% CI 0.7 to 0.9) and those not using a condom with last paying sexual partner (AOR 0.3, 95% CI 0.2 to 0.5). HIV positivity was associated with history of ever using drugs (AOR 2.3, 95% CI 1.4 to 3.6) and soliciting clients outdoors (AOR 1.5, 95% CI 1.1 to 2.0). Women working indoors were less aware of their positive status (AOR 0.1, 95% CI 0.1 to 0.9).
CONCLUSION: HIV prevalence is high among FSWs in Ukraine, and testing and knowledge of one's status remain insufficient. HIV testing programmes need to expand with strategies to reach specific subgroups of FSWs. © Author(s) (or their employer(s)) 2019. Re-use permitted under CC BY-NC. No commercial re-use. See rights and permissions. Published by BMJ.

Introduction

Ukraine has one of the largest HIV epidemics in Europe with current estimates of HIV prevalence in the general population at 0.9%,1 and higher rates among key affected populations including female sex workers (FSWs) and people who inject drugs. There is an estimated 80 000 FSWs in Ukraine,2 of whom 7.0% are living with HIV.3

Early diagnosis of HIV and sustained and effective immediate HIV treatment for SWs living with HIV could contribute to a more effective HIV response.4 In 2006, the Alliance for Public Health in Ukraine launched the first HIV testing intervention with rapid diagnostic tests across 12 pilot sites as a part of a comprehensive Harm reduction (HR) programme for key affected populations. The intervention was subsequently scaled up countrywide in 2007 and by 2014 was implemented in 25 oblasts (administrative regions).5 6

Despite a widespread testing programme, a large proportion of people are still unaware of their HIV status.7 8 At the beginning of 2017, a total of 127 620 people living with HIV/AIDS (PLHV) were officially diagnosed and enrolled in care at the Ukrainian AIDS clinics, which constituted up to 54.0% of estimated population size of PLHV.1 There are no data on how many of those diagnosed and enrolled in care are FSWs, as this type of data is not collected in Ukraine.

To improve the design and implementation of HIV testing and care in Ukraine for FSWs, our aim was to identify factors associated with self-reported HIV testing and receiving test result in the last 12 months, HIV prevalence and knowledge of positive status among FSWs.

Methods

Study setting and population

We performed secondary analysis of cross-sectional data collected within an Integrated Bio-Behavioural Survey (IBBS) among FSWs in Ukraine (IBBS, 2013–2014). Initial analysis was performed by Alliance for Public Health, Ukraine9; detailed description of the IBBS methodology is presented elsewhere.9 10 The study was developed and conducted by the international charitable foundation ‘Alliance for Public Health’ under support of the Global Fund to Fight AIDS, Tuberculosis and Malaria. Survey sites were selected to include geographically representative samples of FSWs; thus, data were collected in 25 capital cities of each oblast (administrative regions) of Ukraine. The field phase of the IBBS was completed during October to December 2013, followed by data validation, analysis and reporting during January to March 2014. In an effort to recruit diverse subpopulations of FSWs in terms of sociodemographic characteristics (age, marital status, education, income, and so on) and sex work venues (indoors, outdoors, via internet), the IBBS methodology combined three sampling strategies: time and location sampling (TLS), respondent-driven sampling (RDS) and key informant recruitment. For each particular survey site (city) the choice of sampling strategy was based on formative assessment, which was done among FSWs, brothel owners/managers, human rights activists, social workers and healthcare workers, who had sufficient experience and knowledge of the sex work scene in the city. For each study site, formative assessment included reviewing the geographical location of existing sex work venues of soliciting clients, their organisational structure and estimated size of population and socioeconomic background of women working there. Based on this assessment, the method for recruitment was selected.

Study procedures

Study staff (interviewers and healthcare workers) who had previous experience working with SWs were recruited and trained on the study methods and tools for the IBBS. FSWs were eligible to participate if they were women aged ≥14 years who had exchanged sex for money/drugs/services or goods in the last 6 months; held Ukrainian citizenship; and provided written or verbal informed consent to participate. Participants completed an interviewer-administered questionnaire which collected information on sociodemographic characteristics, self-reported health status, use of public health services, sex work experience, alcohol, drug use and sexual behaviours, experiences of violence and HIV testing history. Finger-prick samples were collected by healthcare workers to perform HIV rapid diagnostic test using CITO TEST HIV 1/2/07, Abon Biotech/Hangzhou, PR China (prequalified by WHO11). HIV testing and pretest and post-test counselling were provided as per the National HIV Testing guidelines by the Ministry of Health of Ukraine. Individuals who tested positive were referred for confirmatory laboratory diagnostics and linkage to care at nearby government HIV clinics. Participants were able to decline testing and around 10% of eligible participants refused to participate in the study overall (90.4% agreed). Participants received compensation of US$10 for their time.

Outcome variables

We examined HIV testing in terms of three binary (Yes/No) outcome measures: (1) self-reported HIV testing and receipt of test result in the last 12 months (based on the Joint United Nations Programme on HIV/AIDS recommendations)7; (2) HIV prevalence; and (3) self-reported HIV positive status (knowledge of one’s HIV status). HIV prevalence was determined by positivity based on rapid test at time of bio-behavioural survey. Self-reported HIV positive status was defined as those FSWs who were diagnosed HIV positive by rapid test and who had reported that they were living with HIV during the survey.

Explanatory variables

Although the choice of explanatory variables was limited by available data, we based our decisions on a systematic review our team conducted summarising evidence on predictors and barriers to HIV testing in FSWs.12 Based on these findings, we applied classification at the mesolevel and microlevel of the socioecological framework of acquisition of HIV.13 At a mesolevel we examined sex work venues. Sex work venues’ most common method for soliciting clients was classified as outdoors (‘on the street’, ‘on the highway’ and ‘at the stations’) or indoors (‘in the sauna’, ‘in the hotel’ and ‘at the night club, bar, restaurant, discothèque’) or phone/internet. At a microlevel, variables included: age, age at first sex, age at start of sex work, education, marital status, employment beside sex work, condom use with last paying/non-paying sexual partner, ever in life drug use, injection drug use in the last 12 months and alcohol consumption in the last 30 days. We analysed variables of age at first sex and age at start of sex work as variables with 5-year intervals per each category starting from the lowest presented value.

Statistical analysis

We calculated descriptive statistics, and excluded data as missing if responses were ‘hard to answer’ and ‘refuse to answer’. Missing data were not included in the bivariate or multivariable analyses. Where data are missing, it is reported as missing in the tables. In bivariate analysis we compared categorical variables by χ2 test. We used multivariable logistic regression analysis to examine associations between explanatory variables and testing in the last 12 months, HIV prevalence and self-reported HIV positive status. We employed a backward stepwise technique using Wald χ2 test to select the final model of the best fit. Explanatory variables were removed one at a time if they were not associated with an outcome at 5% level of significance. We used SPSS (SPSS V.21) to conduct analyses. We presented both bivariate and multivariable results where associations were significant in the Results section; online supplementary annex files include all bivariate and multivariable results.

Results

Descriptive characteristics

Sociodemographic characteristics are shown in table 1. Overall, the median age of participants was 28 years (IQR 19–37). The median age at first sex was 16 years (IQR 14–18) and the median age at first sex work was 20 years (IQR 14–26). Almost every second woman had 11 years of schooling. The majority of participants were unmarried and lived alone. Approximately 10% of participants had a permanent job other than SW and about half of participants were officially recognised by the state social support system as unemployed. Almost all women used a condom with last paying partner (96.8%), while only about 1 in 5 reported condom use with non-paying partner (18.8%). The majority had never used drugs (71.4%), and 46.4% reported weekly alcohol use. About one-half of participants (53.0%) solicited clients at multiple types of venues, overall the most common place of solicitation was either outdoor (51.0%) or indoor venues (52.8%). Internet/phone solicitation was less common (33.0%).

Table 1

Descriptive characteristics among FSWs in Ukraine (n=4906)

Variables	n	%
Age (years)
14–18	132	2.7
19–23	1039	21.2
24–28	1497	30.5
29–33	1191	24.3
34+	1047	21.3
Median (IQR)	28 (20;38)
Age at first sexual intercourse (years)
6–13	356	7.3
14–18	4252	86.7
19–25	259	5.3
Median (IQR)	16 (15;19)
Missing	40	0.8
Age at first sex work (years)
12–13	9	0.2
14–18	1242	25.3
19–23	2227	45.4
24–28	1067	21.7
29–33	252	5.1
34+	70	1.4
Median (IQR)	20 (14;26)
Missing	42	0.8
Education level
<9 years of schooling	105	2.1
9 years of schooling	563	11.5
11 years of secondary school	2283	46.5
Vocational school	1621	33.0
University degree	334	6.8
Marital status
Married living with husband	214	4.4
Married living with other SP	137	2.8
Married living alone	405	8.3
Unmarried living with SP	922	18.8
Unmarried living alone	3228	65.8
Employment status (besides sex work)
Schoolgirl/student	351	7.2
Have casual earnings	1414	28.8
Unemployed*	1997	40.7
Housekeeper	536	10.9
Incapacitated†	24	0.5
Have a permanent job	565	11.5
Missing	19	0.4
Condom use with last paying partner
No	150	3.1
Yes	4753	96.9
Missing	3	0.1
Condom use with last non-paying partner
No	967	19.7
Yes	922	18.8
I had no such partner	3018	61.5
Alcohol use in the last 30 days
Every day	1171	23.9
More than once a week	2274	46.4
Less than once a week	1114	22.7
Never	347	7.1
Drug use ever in life
Yes	545	11.1
I used before, now I don't	847	17.3
No, never used drugs	3504	71.4
Missing	10	0.2
Injection drug use in the last 12 months
Yes	292	6.0
No	4614	94.0
Sex work venue: outdoors
No	2402	49.0
Yes	2504	51.0
Sex work venue: indoors
No	2315	47.2
Yes	2591	52.8
Sex work venue: phone/internet
No	3286	67.0
Yes	1620	33.0

*Unemployed by official records as sex work is not considered to be official occupation and cannot give official employment status (recognised by the government of Ukraine).

†Incapacitated—those who are receiving social security because of the limited employment due to disability.

FSWs, female sex workers; SP, sexual partner.

Self-reported testing and receipt of test result in the last 12 months

In total, 63.2% (95% CI 61.8 to 64.5) of FSWs reported an HIV test in the last 12 months with receipt of test result (table 2). In the bivariate model, condomless sex with last paying sexual partner (OR 0.3, 95% CI 0.2 to 0.5), any alcohol use in the last 30 days (OR 0.3, 95% CI 0.2 to 0.5), previous experience of drug use (OR 0.7, 95% CI 0.6 to 0.8), soliciting clients indoors (OR 0.8, 95% CI 0.7 to 0.9) and solicitation via phone/internet (OR 0.7, 95% CI 0.6 to 0.8) were each associated with a lower likelihood of testing in the last 12 months. In the final multivariable model, the following variables were independently associated with a lower likelihood of testing and receipt of test result: condomless sex with last paying sexual partner (adjusted OR (AOR) 0.3, 95% CI 0.2 to 0.5); any alcohol use in the last 30 days (AOR 0.3, 95% CI 0.2 to 0.5); previous experience of drug use (AOR 0.7, 95% CI 0.6 to 0.9); and soliciting clients indoors (AOR 0.8, 95% CI 0.7 to 0.9) and via phone/internet (AOR 0.7, 95% CI 0.6 to 0.9).

Table 2

Bivariate and multivariable analysis of self-reported HIV testing and receipt of test result in the last 12 months among FSWs in Ukraine*

Sociodemographic and behavioural characteristics	Bivariate analysis			Multivariable analysis
	n (%)	OR (95%CI)	Pvalue	n (%)	AOR (95%CI)	Pvalue
Age (years)
14–18	66 (1.6)	1.4 (0.7 to 2.6)	0.24	66 (1.7)	1.4 (0.7 to 2.8)	0.29
19–23	790 (19.3)	1.5 (1.2 to 1.8)	<0.001	783 (19.7)	1.4 (1.0 to 1.8)	0.02
24–28	1308 (32.1)	1.0 (0.8 to 1.3)	0.38	1288 (32.5)	1.0 (0.7 to 1.2)	0.90
29–33	1025 (25.1)	1.1 (0.9 to 1.4)	0.17	998 (25.1)	1.1 (0.8 to 1.4)	0.30
34+	888 (21.8)	1.0 (ref)		864 (21.8)	1.0 (ref)
T	829			907
Condom use with paying SP
No	100 (2.5)	0.3 (0.2 to 0.5)	<0.001	3902 (97.5)	0.3 (0.2 to 0.5)	<0.001
Yes	3974 (97.5)	1.0 (ref)		97 (2.4)	1.0 (ref)
Missing data	832			907
Alcohol use in the last 30 days
Every day	959 (23.5)	0.3 (0.2 to 0.5)	<0.001	950 (23.7)	0.3 (0.2 to 0.5)	<0.001
More than once a week	1865 (45.7)	0.3 (0.2 to 0.4)	<0.001	1823 (46)	0.3 (0.2 to 0.4)	<0.001
Less than once a week	952 (23.4)	0.4 (0.2 to 0.6)	<0.001	939 (23.6)	0.4 (0.2 to 0.6)	<0.001
Never	301 (7.4)	1.0 (ref)		287 (7.2)	1.0 (ref)
Missing data	829			907
Drug use ever
Yes	484 (11.9)	0.9 (0.7 to 1.1)	0.65	472 (11.8)	1.0 (0.7 to 1.3)	0.84
I used before, now I don't	712 (17.5)	0.7 (0.6 to 0.8)	<0.001	698 (17.4)	0.7 (0.6 to 0.9)	0.01
No, I never used drugs	2874 (70.6)	1.0 (ref)		2829 (70.7)	1.0 (ref)
Missing data	836			907
Sex work venue: outdoors
Yes	2210 (54.2)	1.3 (1.2 to 1.5)	<0.001	2154 (53.8)	1.4 (1.1 to 1.6)	<0.001
No	1867 (45.8)	1.0 (ref)		1845 (46.2)	1.0 (ref)
Missing data	829			907
Sex work venue: indoors
Yes	2047 (50.3)	0.8 (0.7 to 0.9)	0.01	2016 (50.5)	0.8 (0.7 to 0.9)	0.01
No	2030 (49.7)	1.0 (ref)		1983 (49.5)	1.0 (ref)
Missing data	829			907
Sex work venue: phone/internet
Yes	1289 (31.6)	0.7 (0.6 to 0.8)	<0.001	1267 (31.6)	0.7 (0.6 to 0.9)	0.01
No	2788 (68.4)	1.0 (ref)		2732 (68.4)	1.0 (ref)
Missing data	829			907

AOR, adjusted OR; CI, confidence interval; FSWs, female sex workers; OD, odds ratio; SP, sexual partner.

HIV prevalence

Overall, 7.1% (95% CI 6.4 to 7.8) of FSWs were HIV positive. In the bivariate model, HIV prevalence was higher among unmarried women who were living with a sexual partner (OR 2.0, 95% CI 1.1 to 3.9); had casual earnings other than sex work (OR 2.8, 95% CI 1.6 to 4.6); were unemployed (OR 2.5, 95% CI 1.5 to 4.2), working as a housekeeper (OR 3.0, 95% CI 1.7 to 5.2) or incapacitated (OR 13.9, 95% CI 5.2 to 37.3); used drugs (OR 5.9, 95% CI 4.4 to 7.9); and solicited clients outdoors (OR 2.6, 95% CI 2.0 to 3.3).

In the final multivariable model (table 3), the following variables were independently associated with greater likelihood of HIV infection: being unmarried and living alone (AOR 2.7, 95% CI 1.2 to 5.9) or being unmarried but living with sexual partner (AOR 2.6, 95% CI 1.2 to 5.7); having a non-permanent job including casual earnings other than sex work (AOR 1.8, 95% CI 1.0 to 3.1), being a housekeeper (AOR 2.3, 95% CI 1.2 to 4.2), being unemployed (AOR 1.7, 95% CI 1.0 to 2.9) or incapacitated (AOR 3.8, 95% CI 1.2 to 11.8); used drugs ever in life (AOR 2.3, 95% CI 1.4 to 3.7) or injected drugs in the last 12 months (AOR 2.9, 95% CI 1.8 to 4.8); and soliciting clients outdoors (AOR 1.5, 95% CI 1.1 to 2.0).

Table 3

Bivariate analysis and multivariable analysis of HIV prevalence among FSWs in Ukraine

Sociodemographic and behavioural characteristics	Bivariate analysis			Multivariable analysis
	n (%)	OR (95%CI)	Pvalue	n (%)	AOR (95%CI)	Pvalue
Age (years)
14–18	133 (2.7)	–		131 (2.7)	–
19–23	1030 (21.4)	0.2 (0.1 to 0.3)	<0.001	1022 (21.6)	0.2 (0.1 to 0.4)	<0.001
24–28	1471 (30.6)	0.4 (0.3 to 0.6)	<0.001	1446 (30.7)	0.4 (0.3 to 0.6)	<0.001
29–33	1148 (23.8)	0.7 (0.6 to 1.0)	0.10	1117 (23.7)	0.8 (0.6 to 1.1)	0.35
34+	1024 (21.3)	1.0 (ref)		999 (21.2)	1.0 (ref)
Missing data	100			191
Education level
<9 years of schooling	97 (2)	6.1 (2.7 to 13.4)	<0.001	94 (1.9)	3.8 (1.5 to 9.2)	0.01
9 years of schooling	554 (11.5)	2.5 (1.3 to 5.0)	0.01	540 (11.4)	2.3 (1.1 to 4.8)	0.02
11 years of schooling	2243 (46.6)	2.2 (1.2 to 4.2)	0.01	2199 (46.6)	2.4 (1.2 to 4.7)	0.01
Vocational school	1576 (32.7)	2.1 (1.1 to 4.1)	0.15	1551 (32.9)	2.0 (1.0 to 4.0)	0.04
University degree	336 (6.9)	1.0 (ref)		331 (7)	1.0 (ref)
Missing data	100			191
Marital status
Unmarried living alone	3144 (65.4)	1.5 (0.7 to 2.8)	0.21	3090 (65.5)	2.7 (1.2 to 5.9)	0.01
Married living with other SP	136 (2.8)	0.6 (0.1 to 2.0)	0.42	134 (2.8)	0.8 (0.2 to 2.8)	0.73
Married living alone	400 (8.3)	1.6 (0.7 to 3.4)	0.18	396 (8.4)	2.0 (0.8 to 4.9)	0.09
Unmarried living with SP	912 (18.9)	2.0 (1.1 to 3.9)	0.04	888 (18.8)	2.6 (1.2 to 5.7)	0.01
Married living with husband	214 (4.4)	1.0 (ref)		207 (4.4)	1.0 (ref)
Missing data	100			191
Employment status (besides sex work)
Schoolgirl/student	350 (7.3)	0.1 (0.0 to 0.7)	0.02	347 (7.4)	0.3 (0.0 to 1.6)	0.19
Have casual earnings	1380 (28.8)	2.8 (1.6 to 4.6)	<0.001	1362 (28.9)	1.8 (1.0 to 3.1)	0.02
Unemployed*	1965 (41)	2.5 (1.5 to 4.2)	<0.001	1926 (40.8)	1.7 (1.0 to 2.9)	0.03
Housekeeper	522 (10.9)	3.0 (1.7 to 5.2)	<0.001	516 (10.9)	2.3 (1.2 to 4.2)	0.01
Incapacitated†	23 (0.4)	13.9 (5.2 to 37.3)	<0.001	22 (0.5)	3.8 (1.2 to 11.8)	0.02
Have a permanent job	547 (11.4)	1.0 (ref)		542 (11.5)	1.0 (ref)
Missing data	119			191
Condom use with non-paying SP
No	949 (19.7)	0.7 (0.5 to 0.9)	0.04	932 (19.7)	0.5 (0.4 to 0.8)	0.01
I had no such partner	2961 (61.6)	0.7 (0.5 to 0.9)	0.03	2910 (61.7)	0.7 (0.5 to 1.0)	0.08
Yes	896 (18.6)	1.0 (ref)		873 (18.5)	1.0 (ref)
Missing data	110			191
Drug use ever
Yes	522 (10.9)	5.9 (4.4 to 7.9)	<0.001	509 (10.8)	2.3 (1.4 to 3.7)	<0.001
I used before, now I don't	811 (16.9)	5.0 (3.8 to 6.5)	<0.001	795 (16.9)	3.9 (2.9 to 5.2)	<0.001
No	3463 (72.2)	1.0 (ref)		3411 (72.3)	1.0 (ref)
Missing data	100			191
Injecting drug use in the last 12 months
Yes	280 (5.8)	6.2 (4.6 to 8.2)	<0.001	270 (5.7)	2.9 (1.8 to 4.8)	<0.001
No	4526 (94.2)	1.0 (ref)		4445 (94.3)	1.0 (ref)
Missing data	100			191
Sex work venue: outdoors
Yes	2526 (52.5)	2.6 (2.0 to 3.3)	<0.001	2458 (52.1)	1.5 (1.1 to 2.0)	0.01
No	2280 (47.4)	1.0 (ref)		2257 (47.9)	1.0 (ref)
Missing data
Sex work venue: indoors
Yes	2457 (51.1)	0.7 (0.6 to 0.9)	0.01	2418 (51.2)	0.9 (0.7 to 1.2)	0.99
No	2349 (48.8)	1.0 (ref)		2297 (48.7)	1.0 (ref)
Missing data	100			191

*Unemployed by official records as sex work is not considered to be official occupation and cannot give official employment status (recognised by the government of Ukraine).

†Incapacitated—those who are receiving social security because of the limited employment due to disability.

AOR, adjusted OR; CI, confidence interval; FSWs, female sex workers; OR, odds ratio; SP, sexual partner.

Self-reported HIV status

Only 45.0% (95% CI 39.7 to 50.0) of FSWs diagnosed with HIV reported they were HIV positive prior to testing in the survey. In the bivariate model, drug use ever in life (OR 8.3, 95% CI 4.0 to 17.1) and injection drug use in the last 12 months (OR 4.7, 95% CI 2.2 to 9.9), as well as soliciting clients outdoors (OR 1.7, 95% CI 1.0 to 3.0) predicted self-reported HIV positive status (table 4).

Table 4

Bivariate analysis and multivariable analysis of HIV self-reported status among FSWs in Ukraine*

Sociodemographic and behavioural characteristics	Bivariate analysis			Multivariable analysis
	n (%)	OR(95%CI)	Pvalue	n (%)	AOR(95%CI)	Pvalue
Age (years)
14–18	–	–	–	–	–
19–23	16 (8.2)	0.1 (0.0 to 0.5)	<0.001	15 (8)	0.2 (0.0 to 1.2)	0.09
24–28	51 (26.4)	0.3 (0.1 to 0.6)	0.01	50 (26.9)	0.4 (0.1 to 1.4)	0.18
29–33	48 (24.8)	0.8 (0.4 to 1.6)	0.62	46 (24.7)	0.7 (0.2 to 2.0)	0.58
34+	78 (40.4)	1.0 (ref)		75 (40.3)	1.0 (ref)
Missing data	157			164
Age at first sex work (years)
12–13	1 (0.5)	–	–	1 (0.5)	–	–
14–18	33 (17.4)	2.8 (0.2 to 28.6)	0.38	32 (17.2)	5.3 (0.2 to 135.0)	0.31
19–23	93 (49.2)	3.6 (0.3 to 36.5)	0.26	91 (48.9)	6.5 (0.2 to 155.3)	0.24
24–28	40 (21.1)	9.5 (0.8 to 101.0)	0.06	40 (21.5)	12.6 (0.5 to 320.6)	0.12
29–33	18 (9.5)	28.2 (1.9 to 417.3)	0.01	18 (9.7)	54.8 (1.7 to 1709.7)	0.02
34+	4 (2.1)	1.0 (ref)		4 (2.1)	1.0 (ref)
Missing data	161			164
Alcohol use in the last 30 days
Every day	64 (33.1)	1.8 (0.6 to 5.3)	0.27	63 (33.8)	5.9 (1.1 to 31.7)	0.03
More than once a week	48 (24.8)	1.9 (0.6 to 6.1)	0.23	69 (37.1)	6.1 (1.0 to 38.2)	0.05
Less than once a week	70 (36.2)	1.6 (0.5 to 4.7)	0.36	46 (24.7)	7.6 (1.2 to 45.1 )	0.02
Never	11 (5.6)	1.0 (ref)		8 (4.3)	1.0 (ref)
Missing data	157			164
Drug use ever
Yes	62 (32.1)	8.3 (4.0 to 17.1 )	<0.001	59 (31.7)	12.6 (2.3 to 68.0 )	0.01
I used before, now I don't	63 (32.6)	2.9 (1.5 to 5.3)	<0.001	60 (32.2)	4.9 (1.8 to 12.9)	<0.001
No, I never used drugs	68 (35.2)	1.0 (ref)		67 (36)	1.0 (ref)
Missing data	157			164
Injecting drug use in the last 12 months
Yes	51 (26.4)	4.7 (2.2 to 9.9)	<0.001	49 (26.3)	1.0 (0.1 to 6.0)	0.93
No	142 (73.5)	1.0 (ref)		137 (73.6)	1.0 (ref)
Missing data	157			164
Sex work venue: outdoors
Yes	140 (72.5)	1.7 (1.0 to 3.0)	0.04	135 (72.6)	1.0 (0.4 to 2.6)	0.86
No	53 (27.4)	1.0 (ref)		51 (27.4)	1.0 (ref)
Missing data	157			164
Sex work venue: indoors
Yes	82 (42.4)	0.3 (0.2 to 0.6)	<0.001	79 (42.5)	0.4 (0.1 to 0.9 )	0.02
No	111 (57.5)	1.0 (ref)		107 (57.5)	1.0 (ref)
Missing data	157			164
Sex work venue: phone/internet
Yes	44 (22.7)	1.3 (0.7 to 2.3)	0.28	43 (23.1)	3.4 (1.4 to 8.4)	0.01
No	149 (77.2)	1.0 (ref)		143 (76.8)	1.0 (ref)
Missing data	157			164

*Denominator—those FSWs who were diagnosed HIV positive with the rapid test.

AOR, adjusted odds ratio; CI, confidence interval; FSWs, female sex workers; OR, odds ratio.

In the final multivariable model, women aged 29–33 years (AOR 54.8, 95% CI 1.7 to 1709.7), who reported consuming alcohol in the last 30 days (AOR 5.9, 95% CI 1.1 to 31.7) or ever using drugs (AOR 12.6, 95% CI 2.3 to 68.0) and SWs soliciting clients via phone/internet (AOR 3.4, 95% CI 1.4 to 8.4) were more likely to report their HIV positive status. SWs who solicited indoors were less likely to report their positive status (OR 0.3, 95% CI 0.2 to 0.6).

Discussion

To our knowledge, this is the first paper on factors associated with HIV testing behaviours, prevalence and self-reported HIV status among FSWs in Ukraine. Our study demonstrates that recent HIV testing remains low with only 63.2% of FSWs reported having tested and received their test results in the last 12 months. Further, only 45.0% of FSWs living with HIV were aware of their status.

Overall, FSWs who reported inconsistent condom use were less likely to test for HIV. As other studies have demonstrated, consistency of condom use might change with longer duration of relationships with sexual partners as FSWs might trust their partners more and perceive themselves to be at lower risk.14 15 HIV testing might be subsequently influenced by lower risk perceptions of women in longer relationships.

It is important to note that testing was lower among FSWs with previous experience of drug use and those soliciting clients indoors. HR programmes in Ukraine have now begun to prioritise the most vulnerable key populations, including people who currently use drugs and those working outdoors as an important population for testing. In other regions, such as Vietnam, Russia, Uzbekistan and Kenya, similar results of lower testing uptake among FSWs who use drugs have been shown,16–18 however, this is not consistent globally.19 20

Overall HIV prevalence in our study was 7.1% among FSWs. This is lower than previous bio-behavioural surveys where prevalence was 10% among FSWs in 2011.21 As previously shown in Ukraine, we found that women who reported injection drug use and those working outdoors were more likely to be HIV positive.4 10 Our results support previous findings that HIV prevalence among FSWs in Ukraine may be driven by overlapping networks of SWs and people who use drugs.1 22

We identified a significant gap between self-reported positive HIV status and diagnosed HIV status with approximately half of FSWs living with HIV aware of their HIV status. This difference could be due to the high level of stigma associated with both HIV and sex work in the region,17 22 23 as stigma might prevent FSWs from testing and/or it may also affect their willingness to disclose a positive test result. Currently, prohibitive administrative (Statute 181/1)24 and criminal (Statutes 130, 302, 303)25 laws may fuel stigmatisation of SWs as well as lead to a violation of their human rights. Several recent public debates around the development of the first non-discriminative sex work policies in Ukraine were called off as they faced deep resistance from government and civil society to recognise sex work as a profession. As other research studies have demonstrated, prohibition of sex work can lead to stigma and discrimination, social exclusion, unsafe working condition, poor occupational health, low self-esteem and restriction on housing, travel and parenting.26–28 Thus, even though testing is available in Ukraine, stigma and criminalisation may constrain its accessibility and uptake, or may impact disclosure of positive HIV status, resulting in forced invisibility of FSWs and their needs. Moreover, women soliciting clients indoors were less likely to self-report positive HIV status, which might be due to fear that brothel owners/managers would know and would reject further employment. Our findings suggest that a range of implementation of outreach strategies with innovative delivery methods, including expanding HR programmes with a range of interventions such as needle/syringe exchange programme, condom delivery programmes, peer-to-peer support, innovative testing approaches and strategies to reach those ‘left behind’ or missing is urgently required.

Limitations

Approximately 10% of eligible participants declined to participate and these individuals may be different from the women who chose to participate in our study. The cross-sectional study design does not allow us to determine casual relationships between outcomes and explanatory variables. As both RDS and TLS sampling strategies were concurrently applied across different IBBS rounds, the conclusions may not be generalisable to the entire population of FSWs in Ukraine. Sensitivity of the questions and stigma surrounding HIV and sex work may have resulted in informational bias. However, we argue that any informational bias may have been limited due to the engagement of trained study personnel who were experienced in working with FSWs. We were unable to verify self-reported HIV testing history and therefore, the proportion of previously tested may not be an accurate reflection because of the recall or self-reported bias. HIV positivity was determined by rapid test and not based on confirmatory serological assays. However, the rapid test used has a sensitivity and specificity over 99%11 and has been used in other surveys and programmes in Ukraine.29 We also acknowledge limitation of the missing data which could reduce the statistical power of the study and can produce biased estimates. An additional limitation to this study is that these data were collected during 2013–2014 years, and thus might not fully reflect the current situation with HIV testing. However, we believe that our findings are still very relevant as HIV testing policies and HIV testing programmes among FSWs in Ukraine have been approved by the National HIV/AIDS Strategy for the period of 2013–2018 years, and thus HIV testing programmes and testing availability and uptake have likely not changed significantly during this time. Finally, our study did not address macrolevel factors (eg, HIV testing policies, healthcare funding, criminalisation of sex work or drug use) as we focused our analysis on bio-behavioural survey data. We acknowledge that these macrolevel factors are very important for contextualising the study findings and this is an important area for future research.

Conclusion

HIV prevalence is high among FSWs in Ukraine; however, HIV testing as well as knowledge of one’s status remains insufficient. Over half of FSWs living with HIV were unaware of their status. HIV prevention programmes in Ukraine need to expand with strategies to engage more difficult to reach subgroups of FSWs to ensure access, engagement and uptake of prevention services including testing and linkage to care for those FSWs living with HIV.

HIV prevalence is high among FSWs in Ukraine, yet HIV testing as well as knowledge of one’s status remain insufficient.

Women soliciting clients indoors were less likely to test for HIV and less likely to be aware of their positive status.

FSWs who have used drugs were less likely to test and had higher HIV prevalence.

FSWs who inject drugs should remain a priority for HIV prevention programs.

HIV prevention programs need to expand testing strategies to ensure high levels of testing coverage to all sex workers regardless of place of solicitation.