Syphilis among Female Sex Workers: Results of Point-of-Care Screening during a Cross-Sectional Behavioral Survey in Burkina Faso, West Africa

Background Syphilis among female sex workers (FSW) remains a public health concern due to its potential impact on their health and the possibility of transmission to their clients, partners, and children. Recent data on the prevalence of syphilis in the population in West Africa are scarce. The objective of this study was to measure the seroprevalence of syphilis serological markers among female sex workers in Burkina Faso. Methods We conducted a cross-sectional survey among FSW between February 2013 and May 2014. Participants were recruited using respondent-driven sampling (RDS) methods in five cities of Burkina Faso (Ouagadougou, Bobo-Dioulasso, Koudougou, Ouahigouya, and Tenkodogo). FSW were enrolled and screened for syphilis using a syphilis serological rapid diagnostic test. Data from all cities were analyzed with Stata version 14.0. Results A total of 1045 FSW were screened for syphilis. Participants' mean age was 27.2 ± 0.2 years. The syphilis serological markers were detected in 5.6% (95% CI: 4.4–7.2) of the participants whereas active syphilis was seen in 1.4% (95% CI: 0.9–2.4). RDS weighted prevalence of syphilis serological markers and active syphilis by city were, respectively, estimated to be 0.0% to 11.0% (95% CI: 8.1–14.7) and 0.0% to 2.2% (95% CI: 1.1–4.4). No syphilis markers were found among Ouahigouya FSW. Low education level and high number of clients were factors associated with syphilis markers among the FSW. Conclusion The prevalence of syphilis markers was high during this study among FSW. This highlights the need to reinforce the comprehensive preventive measures and treatment of syphilis in this population.


Background
Syphilis is a sexually transmitted disease (STD) due to Treponema pallidum, a spirochete bacteria belonging to the Spirochaetaceae family [1,2]. e infection is often asymptomatic but can cause genital ulceration associated with enhanced risk of transmission or acquisition of HIV infection [3][4][5][6][7]. Syphilis remains a worldwide public health problem. Despite the availability of effective preventive measures and relatively low treatment cost, the World Health Organization (WHO) estimates that more than 10 million persons are newly infected each year [8].
A number of studies around the world have reported high prevalence of syphilis among risk groups including female sex workers (FSW) [9][10][11][12]. In Africa, the prevalence of syphilis among FSW was 2.2% in Togo [13] and between 1.5 and 89% in Sudan [14]. Studies conducted in the 1990s have reported a prevalence of 42.1% in South Africa [15] and 15% in Burkina Faso [16]. Despite this high prevalence in Africa, FSW have limited access to STIs/STDs like syphilis screening and care [17,18]. Stigma and criminalization of sex work in some countries have complicated their access to care and have prevented the planning of efficient health systems [19][20][21][22]. FSW are identified as a population at high risk for HIV acquisition [23,24]. Biologic risk factors for HIV include physiological factors [25] such as the presence of STDs, including syphilis [26,27]. Recent data on syphilis prevalence in FSW are scarce in West Africa.
e main objective of this study was to determine the serological prevalence and correlates of syphilis with rapid diagnostic testing among FSW in Burkina Faso.

Study Design.
We conducted a cross-sectional survey between February 2013 and May 2014 among FSW in five cities in Burkina Faso. Participants were recruited using respondent-driven sampling (RDS) to be part of HIV seroprevalence and behavioral studies. RDS is an approach to recruit hidden and hard-to-reach populations [28].

Setting.
e study was conducted in five cities of Burkina Faso: Ouagadougou (the capital city in the centre), Bobo-Dioulasso (in the Hauts-Bassins region), Koudougou (in the middle west), Ouahigouya (in the north), and Tenkodogo (in the middle east). ese areas were selected based on the urbanization level, the prevalence of HIV, and the geographical location.

Study Participants.
Inclusion criteria were aged ≥ 18 years, assigned female sex at birth, earned at least 50% of her income in the last 12 months from sex work, lived in the selected city during the last three months, and provided informed consent. Recruitment methods were fully described elsewhere [29]. e enrolment of FSW in each city started with an eligible seed who was enrolled and then trained to recruit other participants.

Sample Size.
e sample size was based on expected HIV prevalence and calculated to recruit 345 FSW in the two major cities of Ouagadougou and Bobo-Dioulasso and 126 FSW in the other medium cities of Koudougou, Ouahigouya, and Tenkodogo. To calculate the sample size, we hypothesized that populations who use condoms would have 75% lower HIV prevalence than that of those not using condoms, assuming condom efficacy as 80% (73% to be cautious). A design effect of 1.5 was used associated with the RDS methods; we set 0.05 for type I errors and 80% power.

Data
Collection. Data were collected from February to August 2013 in Ouagadougou and Bobo-Dioulasso and from February to May 2014 in Koudougou, Ouahigouya, and Tenkodogo. Upon informed consent and completion of a behavioral questionnaire, participants received counseling for HIV and syphilis screening. Participant information was anonymized and identified with a unique code number to protect confidentiality. Data collected in the questionnaire included demographic and socioeconomic information, the number of sexual partners, sexual behavior (including condom use), HIV-related knowledge, and STD screening and symptoms. In each city, one site was used for data collection. e team was trained on the study procedures and on health research ethics prior to the onset of the survey. e staff included one site manager, one coupon manager, two data collectors for questionnaire administration, a nurse for HIV and syphilis pre-and posttest counseling, and a lab technician for blood collection and sample processing for HIV and syphilis screening. Participants with positive syphilis test results were treated in situ while those with positive HIV test results were referred to dedicated staff for appropriate care. Figure 1 shows the flowchart of the syphilis screening among study participants. Blood samples were centrifuged, and serum obtained was used in situ for syphilis screening. Alere Determine ™ Syphilis TP ("Alere Ltd.," USA) rapid diagnostic test was used to screen the syphilis serological markers as per manufacturer recommendations. Positive samples were consecutively tested with a Rapid Plasma Reagin (RPR-CHARBON ® ) test kit (BIOLABO SAS, French), a nontreponemic test [30]. Alere Determine ™ Syphilis TP is a qualitative in vitro immunechromatographic test detecting antibodies against the Treponema pallidum bacteria in patient's serum, plasma, and whole blood [31]. Positivity to the Alere Determine ™ Syphilis TP test reveals the presence of antibodies against the Treponema pallidum bacteria indicating the presence of either previous, acute, or treated syphilis. e nontreponemic rapid plasma reagin (RPR) test was used to determine the syphilis stage (active or not).

Data Entry and Statistical
Analysis. Data were double entered into EpiData 3.1 (the EpiData Association, Odense, Denmark) software and analysed with Stata 14 (StataCorp, College Station, TX). RDS-weighted analysis was only used to estimate active syphilis and syphilis serological markers seroprevalence for each city. To have sufficient number of positive cases of syphilis serological markers for statistical analysis, data from all study cities were pooled. e pooled data were not analyzed using RDS weighting because there was no overlap in recruitment between study cities. Baseline survey data and biological test results of participants were summarised with descriptive statistics. Bivariate and multivariate logistic regression analyses were conducted to identify factors associated with syphilis serological markers. Associations were expressed as odds ratios (ORs) and adjusted odds ratios (aORs) with 95% confidence intervals (CIs). Results significant at the p < 0.2 level in the bivariate analyses were included in the multivariate analysis. Results were considered statistically significant at the p < 0.05 level.

Ethical Issues.
e National Ethics Committee for Health Research in Burkina Faso reviewed and approved the study protocol prior to participant recruitment. Collected data were anonymized with a unique identifier code so that no participant could be identified by their name. All FSW detected with positive antibodies were treated in situ with one intramuscular dose of benzathine benzylpenicillin G 2.4 MUI and kept under observation over 30 minutes to capture any immediate adverse event. Positive cases presenting allergy received doxycycline 100 mg tablet over 2 weeks. FSW were sensitized on safe sex (protection measures against STDs), were encouraged to undergo a serological test for syphilis every 3 to 6 months, and were provided with condoms and lubricants free of charge.

Results
A total of 1045 FSW were screened for syphilis in the five study cities: 349 FSW in Ouagadougou, 350 in Bobo-Dioulasso, 115 in Koudougou, 102 in Ouahigouya, and 129 in Tenkodogo. Approximately 44.7% (463/1036) of those interviewed self-reported a history of STD-related symptoms during the last twelve months prior to the survey. Only 1.3% (14/1042) underwent syphilis serological testing during the last twelve months.

Prevalence of Syphilis Serological Markers.
e prevalence of syphilis serological markers was 5.6% (95% CI: 4.4-7.2) among FSW overall, and active syphilis was found among 1.4% (95% CI: 0.9-2.4) of the participants. RDS weighted prevalence of syphilis serological markers and active syphilis by city were, respectively, estimated between 0.0% and 11.0% (95% CI: 8.1-14.7) and 0.0% and 2.2% (95% CI: 1.1-4.4) ( Table 2). No syphilis markers were found among FSW in Ouahigouya. As shown in Table 3, the prevalence of syphilis markers increased with age (p � 0.025) and decreased with education level (p < 0.001). (ii) No FSW was known or reported to be allergic to penicillin, and no allergy was noted during the 30 minutes postinjection observation.
(iii) Participants were provided with education about safe sex and consistent condom use.  International Journal of Microbiology Active syphilis was more prevalent among FSW with no education (2.8%) compared to those with a primary school education (0.8%) or secondary school education (0.6%) (p � 0.037). e country of origin (Burkina Faso and other countries) was not significantly associated with the presence of serological markers among FSW.

Factors Associated with Syphilis Serological Markers.
e factors associated with the syphilis serological markers among FSW in bivariate and multivariate analyses are presented in Table 4.

Bivariate Analysis.
In the bivariate analysis, the risk of syphilis serological markers was 80% lower in participants with at least a secondary school education compared to FSW with no education (OR � 0.20; p < 0.001). Participants with at least 22 clients per week were significantly more likely to have syphilis serological markers. FSW who reported 22 to 28 clients or at least 29 clients per week were, respectively, 2.97 (OR � 2.97; p � 0.021) and 3.34 (OR � 3.34; p < 0.001) times more at risk of serological marker carriage than those who reported less than 15 clients per week. FSW with a history of pregnancy were 4.31 times more likely to carry the markers compared to those who did not (OR � 4.31; p � 0.014). Participants who sold sex for at least 6 years were more likely to carry syphilis markers than those who sold sex for less than one year (OR � 3.85; p � 0.004). FSW in Bobo-Dioulasso had higher risk of syphilis serological marker carriage (OR � 2.87; p � 0.001).

Multivariate Analysis.
In multivariate analysis, after adjustment for age, education level, marital status, number of clients per week, history of pregnancy, use of condoms, and the study city, the number of clients and education level were independently associated with syphilis serological marker carriage. No education was associated with an 77% increase in the risk of syphilis serological marker carriage as compared to a secondary school level of education (aOR � 0.23; p � 0.002). FSW having more than 29 clients per week were 2.62 times more likely to be syphilis serological marker positive compared to those with less than 15 clients per week (aOR � 2.62; p � 0.008).

Discussion
is study found that the prevalence of syphilis serological markers and active syphilis were, respectively, 5.6% and 1.4% among FSW in Burkina Faso. e prevalence varied between study cities. To our knowledge, there are no recent data on syphilis among FSW in Burkina Faso. Overall, it appears that the prevalence of active syphilis (acute) was lower than results found in previous studies in Sub-Saharan Africa. A systematic review of congenital syphilis reported a decrease of 3.01% in syphilis prevalence between 1990 and 1999 and 1.48% between 2000 and 2009 [32]. Other studies have reported a similar reduction from 8.9% in 1993 to 1.5% in 1999 in Benin [33] and from 12.5% in 1991 to 0.8% from 1998 to 2000 in Burkina Faso [34]. is decrease of syphilis prevalence could be explained by several factors including STD � sexually transmitted disease; * p < 0.05, * * p < 0.01, and * * * p < 0.001. 6 International Journal of Microbiology International Journal of Microbiology the impact of sustained campaigns on the prevention of HIV and STDs over the last two decades in most countries [35]. is can also be explained by the introduction of syndromic management of STDs [36][37][38], and the widespread use of antibiotics in Africa may have contributed to the decrease. In addition to penicillin, the Treponema pallidum (bacterial cause of infectious syphilis) is also treated with other antibiotics such as cephalosporins (for example, cephaloridin), tetracyclins, macrolids (erythromycin and spiramycin), and chloramphenicol [2]. ese antibiotics probably used for other indications have significantly contributed to reducing the incidence and evolution of infectious syphilis, including symptomatic and severe chronic forms [1].
Paradoxically, the prevalence of syphilis in our sample was increased, compared to the 2000 data [34]. is could be related to increase in sexual risk-taking, especially unprotected sex following high-active antiretroviral therapy availability and its effectiveness in HIV prevention that could lead to increase acquisition of other sexualtransmitted diseases such as syphilis [39,40].
In addition, the prevalence of active syphilis in our study remains high in comparison with the estimate of 0.24% found in pregnant women in 2000 [41] and 1.2% in 2011 [42]. FSW are a leading group at high risk of syphilis infection due to several factors such as the multiple sexual partners and barriers to condom use [43]. Similar to prior research, in our study, FSW reporting more than 29 clients per week were three times more likely to be infected compared to those having less than 14 clients per week. Lower education was also a risk factor for syphilis infection in our study and more generally, for STDs in prior research [44,45]. e control of STDs and particularly syphilis infection is a public health challenge [8]. Indeed, syphilis infection is very often asymptomatic and therefore untreated; thus, the potential for transmission is enhanced, and the risk of complication (including cervical cancer, higher genital tract infections, sterility, chronic pain, ectopic pregnancy, congenital abnormalities, spontaneous abortion, and associated neonatal and maternal mortality) is increased [46]. Syphilis-related genital ulcerations among FSW are important risk factors for HIV transmission [3], highlighting the relevance of early screening and treatment. In this study, all FSW who screened positive marker test were treated with a single intramuscular dose of benzathine benzylpenicillin G slow release 2.4 MUI [47] as per WHO recommendations for syphilis screening and treatment in high-risk groups, including FSW [30].

Limitations.
In our study, the number of cases of active syphilis was low, which limited the ability to further identify factors associated with the acute infection through multivariate analysis. We instead used syphilis serological markers for analysis purposes. Furthermore, self-reporting sexual behavior in a context of social stigmatization of and discrimination toward FSW may have led to social desirability bias; however, this did not have impact on the prevalence of syphilis, as this was evaluated through syphilis serological markers. Our study provides updated data on syphilis serological markers in Burkina Faso.

Conclusion
is study provided updated epidemiological data on syphilis based on rapid diagnostic testing in situ among sex workers in Burkina Faso. We found that syphilis prevalence among FSW was relatively high and requires targeted preventive measures. Screening with rapid diagnostic tests followed by a single-dose administration of penicillin could be an effective approach to control syphilis among FSW and thus contribute to the prevention of HIV transmission. Safer sex and HIV prevention activities for FSW offer an opportunity to integrate screening and early treatment of syphilis.
Data Availability e datasets used during the current study are available from the corresponding author on reasonable request.

Conflicts of Interest
e authors declare that there are no conflicts of interest regarding the publication of this article.

Authors' Contributions
S. Baral, S. Kouanda, H. G. Ouedraogo, O. Ky-Zerbo, and A. Grosso conceptualized the study and contributed to study design, implementation, and manuscript writing; I.B. Meda and I. Zongo assisted in manuscript writing; B. C. Samadoulougou and G. Tarnagda helped in study implementation, data cleaning, and manuscript review; K. Cissé, A. Sondo, N. Sawadogo, Y. Traore, and Nicolas Barro conducted the manuscript review.