High prevalence of curable sexually transmitted infections among pregnant women in a rural county hospital in Kilifi, Kenya

Background Women attending antenatal care (ANC) in resource-limited countries are frequently screened for syphilis and HIV, but rarely for other sexually transmitted infections (STIs). We assessed the prevalence of curable STIs, defined as infection with either Chlamydia trachomatis or Neisseria gonorrhoeae or Trichomonas vaginalis, from July to September 2015. Methods In a cross-sectional study, women attending ANC at the Kilifi County Hospital, Kenya, had a urine sample tested for C. trachomatis/N. gonorrhoeae by GeneXpert® and a vaginal swab for T. vaginalis by culture. Bacterial vaginosis (BV) was defined as a Nugent score of 7–10 of the Gram stain of a vaginal smear in combination with self-reported vaginal discharge. Genital ulcers were observed during collection of vaginal swabs. All women responded to questions on socio-demographics and sexual health and clinical symptoms of STIs. Predictors for curable STIs were assessed in multivariable logistic regression. Results A total of 42/202 (20.8%, 95% confidence interval (CI):15.4–27.0) women had a curable STI. The prevalence was 14.9% for C. trachomatis (95% CI:10.2–20.5), 1.0% for N. gonorrhoeae (95% CI: 0.1–3.5), 7.4% for T. vaginalis (95% CI:4.2–12.0), 19.3% for BV (95% CI: 14.1–25.4) and 2.5% for genital ulcers (95% CI: 0.8–5.7). Predictors for infection with curable STIs included women with a genital ulcer (adjusted odds ratio (AOR) = 35.0, 95% CI: 2.7–461.6) compared to women without a genital ulcer, women who used water for cleaning after visiting the toilet compared to those who used toilet paper or other solid means (AOR = 4.1, 95% CI:1.5–11.3), women who reported having sexual debut ≤ 17 years compared to women having sexual debut ≥18 years (AOR = 2.7, 95% CI:1.1–6.6), and BV-positive women (AOR = 2.7, 95% CI:1.1–6.6) compared to BV-negative women. Conclusion One in five women attending ANC had a curable STI. These infections were associated with genital ulcers, hygiene practices, early sexual debut and bacterial vaginosis.


Introduction
Sexually transmitted infections (STIs) are a major public health concern globally and especially among women in resource-limited countries [1]. Common curable STIs among women in sub-Saharan Africa (sSA) include Chlamydia trachomatis, Neisseria gonorrhoeae and Trichomonas vaginalis [2]. Additionally, bacterial vaginosis (BV), which has been associated with an increased risk of incident infection with C. trachomatis and N. gonorrhoeae [3,4], is highly prevalent [5]. These curable STIs and BV have been associated with a number of adverse pregnancy outcomes such as spontaneous abortion, ectopic pregnancy, preterm delivery, low birth weight, stillbirth, postpartum sepsis, and congenital infection [6]. Infants born of mothers with genital C. trachomatis are at an increased risk of suffering from neonatal pneumonia and ophthalmia neonatorum [7]. At present the efficacy of neonatal ocular prophylaxis with erythromycin for prevention of C. trachomatis ophthalmia is not clear [8]. These associated sequelae of STIs hence necessitate prevention and management of curable STIs and BV, especially during pregnancy.
Pregnant women attending antenatal care (ANC) in resource-limited countries are frequently screened for syphilis [9] and HIV [10], but rarely for other curable STIs or BV. Moreover, curable STIs and BV are associated with increased acquisition and transmission of HIV [11][12][13]. Among pregnant HIV positive women C. trachomatis is associated with increased risk of HIV mother-to-child transmission [14]. Many resource-limited countries have adopted the World Health Organization's syndromic approach for the management of curable STI [15]. This syndromic approach is deficient in both sensitivity and specificity because STIs may be asymptomatic or have non-specific clinical signs and symptoms [16,17].
There is a paucity of epidemiological and risk factor data on curable STIs among women residing in rural areas, including among pregnant women in coastal Kenya. Most of the recent information about the prevalence of curable STIs among women in Kenya has been from studies done among high-risk individuals [18], women residing in cities [19,20] or from regions with high incidence of HIV [13]. Consequently, we assessed the prevalence and predictors of curable STIs in pregnant women attending ANC at a rural county hospital in Coastal Kenya.

Methods
This was a cross-sectional study conducted from July to September 2015, at the ANC clinic of the Kilifi County Hospital (KCH), Kenya, a clinic attended to by approximately 4,000 women annually [21]. We aimed to enroll a convenience sample of 350 consecutive pregnant women to identify 20-25 women with T. vaginalis for detailed molecular studies of T. vaginalis and assessment of the vaginal microbiome. Here, we report on a sub-study of 202 women who were tested for curable STIs, i.e. C. trachomatis, N. gonorrhoeae, T. vaginalis, and for BV.

Procedures
Recruitment: A nurse at the ANC clinic provided general information about the study to women who had completed their ANC visit. Women interested to know more about the study were invited to participate and up to nine consecutive women were enrolled per day.
Inclusion criteria: Women were eligible if they met the following criteria: age 18-45 years, gestation ! 14 weeks, resident of the Kilifi Health and Demographic Surveillance area [22] and willing to undergo free STI and BV screening procedures.
Study procedures: Consenting women underwent a short private interview with the nurse at the ANC clinic using a structured questionnaire capturing symptoms of vaginal infections. This was followed by a collection of vaginal swabs by the ANC nurse. During the collection of vaginal swabs, the nurse observed for abnormal vaginal discharge and genital ulcers. Three vaginal swabs were collected, by inserting one swab at a time approximately 2 inches into the vaginal opening and gently turning around twice ensuring rubbing the swab against the vaginal wall.
The first swab was used for BV diagnosis and vaginal smears were prepared by rolling the swab onto a microscope glass slide. Slides were air-dried at the clinic before being transported to the laboratory for heat fixation followed by Gram staining and microscopy using the scoring system described by Nugent [23]. Women who self-reported discharge and had a Nugent score of 7-10 were considered BV-positive, and were treated as recommended [8]. Women with a Nugent score of 7-10 but without discharge were considered BV-negative.
The second swab was used to diagnose T. vaginalis using a commercially available T. vaginalis InPouch TM system (BioMed Diagnostics, White City, Oregon, USA). The vaginal swab was inoculated in the upper-chamber of the InPouch at the clinic. The inoculated InPouch was transferred to the laboratory for direct microscopy of the upper chamber, after which the upper chamber was merged with the lower chamber and incubated at 37˚C. Daily microscopy was performed and samples with motile trichomonads within 5 days of culture were considered positive for T. vaginalis.
The third swab was placed in a sterile labeled 2 ml Eppendorf tube, the swab shaft was broken by bending the shaft against the neck of the Eppendorf tube. The bottom portion of the swab with the specimen was transported to the laboratory where it was immediately stored at -80˚C for molecular studies of the vaginal microbiome. No transport or freezing medium was added.
Women provided fresh first catch urine for C. trachomatis and N. gonorrhoeae testing, that was performed by the GeneXpert 1 CT/NG Assay (Cepheid, Sunnyvale, California), according to the manufacturer's instructions.
Results for C. trachomatis, N. gonorrhoeae, for T. vaginalis direct microscopy and for BV were reported on the same day of the visit, and STI and BV treatment was offered according to Kenyan national guidelines [24], except for treatment of N. gonorrhoeae where ceftriaxone was used in place of quinolones [25]. Women who were negative on T. vaginalis direct microscopy but positive on culture were contacted to return to the ANC clinic for treatment immediately after the culture turned positive. Treatment was also offered for the partners of the women with curable STIs.
Syphilis and HIV test results were obtained from each participant's ANC records. Syphilis results were based on testing for treponemal antibodies only, done at the hospital laboratory using a commercial test kit Advanced Quality One Step Anti-TP (InTec products INC, Xiamen, China). Rapid HIV testing was done by the Alere Determine HIV-1/2 kit (Alere Medical Co Ltd, Chiba, Japan), the First response HIV 1-2-0 kit (Premier Medical Corporation Limited, Daman, India) was used to confirm positive results, and the Uni-Gold kit (Trinity Biotech, Wicklow, Ireland) was used to resolve any discordant test results.
After clinical procedures were completed, all study participants were interviewed. The interviews were conducted in Kiswahili by a trained female fieldworker. The questionnaire covered sociodemographic, hygienic and sexual behavior questions. These included age, education level, marital status, number of children, occupation, consumption of alcohol, consumption of tobacco and other substances, age at first sex, number of sexual partners, history of STIs, and hygiene and sanitary practices.

Ethical considerations
The study was approved by the KEMRI Scientific and Ethics Review Unit (#3022). All participants provided written informed consent for study participation. Permission to use hospital records on syphilis and HIV status was obtained from Kilifi County Health department.

Data analysis
Data were entered in the REDCap TM electronic data capture tool, version 6.5.0 (Vanderbilt University, Nashville, Tennessee). Prevalence of STIs was determined as the number of participants with either C. trachomatis, N. gonorrhoeae and/or T. vaginalis out of 202 women that could be tested, expressed as a percentage. Exact 95% binomial confidence intervals (CIs) were calculated for prevalence estimates. Bivariable logistic regression was used to determine associations of sociodemographic, hygienic, behavioral and clinical characteristics with curable STIs. Variables significant at p 0.1 in bivariable analysis were entered into a multivariable model to identify independent associations. Crude odds ratios (COR) and adjusted odds ratios (AOR) were calculated. These statistical analyses were done using STATA version 13.1 (Stata-Corp, College Station, Texas).

Results
From June to September 2015, a total of 373 women were invited to participate in the study. Twenty-three (6.2%) women declined to participate citing no interest in research, lacking time or having no permission from spouse/partner as the main reasons. Of the 350 women enrolled, 148 (42.3%) women did not have C. trachomatis and N. gonorrhoeae tests performed due the unavailability of the tests at the beginning of the study. Therefore, a total of 202 women were tested for C. trachomatis, N. gonorrhoeae, T. vaginalis and BV and were included in this sub-analysis (Fig 1).

Characteristics of the study population
The median age and gestation age including the interquartile range (IQR) for the participants was 26 (22-31) years and 24 (20-28) weeks respectively. The majority of the participants were married (93.6%) and Christian (72.3%). A quarter of the participants had secondary school education and above. The median age and IQR of sexual debut was 18 (16-20) years, although 20.8% of the participants were not sure or did not respond to this question. Forty-two percent of the participants had given birth before and the number of children ranged from 1-7. Sixtytwo percent of the participants had previously experienced signs/symptoms associated with reproductive tract infection and one third had ever received syndromic treatment for genital signs or symptoms of infection. The socio-demographic characteristics of the participants are summarized in Table 1.  (Fig 1). Five (11.9%) women had more than one STI, while 13 (31%) out of the 42 women with a curable STI also had BV. Only two women (1.0%) had asymptomatic BV and none of them had a curable STI (data not shown).

Prevalence of curable STIs, BV, HIV and syphilis
Thirteen women were HIV-positive, a prevalence of 6.4% (95% CI: 3.4-10.8), including four (2.0%; 95% CI: 0.5-5.0) women who were newly diagnosed. Of note, HIV prevalence in 82 women aged 18-24 vs. 120 women aged ! 25 years was 0% vs. 10.8% (p = 0.002). All nine HIV positive women who knew their status were on anti-retroviral therapy. Among the 13 women with HIV, three women had curable STIs, i.e. a prevalence of 23.1% (95% CI: 5.0-53.8). All women with HIV and a curable STI also had BV. The prevalence of curable STIs, or BV in the 13 women with HIV was not statistically significantly different from that of the 189 women without HIV (data not shown).
Two women had treponemal antibodies, a prevalence of 1.0% (95% CI: 0.1-3.6). Prevalence of HIV, treponemal antibodies and T. vaginalis for the 148 women who were not tested for C. trachomatis and N. gonorrhoeae was not statistically significantly different than the prevalence of HIV, syphilis, and T. vaginalis in the 202 tested for C. trachomatis and N. gonorrhoeae in the sub-study.

Clinical signs and symptoms
The most commonly self-reported symptoms amongst all participants were vaginal discharge (73.8%) and lower abdominal pain (46.5%). However, during collection of specimen by the study nurse, only 17.3% of the participants had an abnormal discharge (defined as excess

Risk factors/predictors
In bivariable analysis, curable STIs were more common among participants who were 24 years, had less than 3 children, had a sexual debut 17 years, used water to clean themselves after visiting the toilet, had a genital ulcer or were BV positive ( Table 2).

Discussion
We found that one in five out of 202 pregnant women attending the ANC at the Kilifi County Hospital, Kenya, during the period July to September 2015 had at least one curable STI. The   high burden of curable STIs was mostly due to cases of C. trachomatis (14.9% of women) and T. vaginalis (7.4% of women). We also found approximately 19.4% of the women to have BV, defined as the presence of vaginal discharge in combination with a Nugent score !7. Four out of the 5 women with genital ulcers also had a curable STI. Our study documented that having genital ulcers, cleaning with water as opposed to use of toilet paper after visiting the toilet, early sexual debut, and BV were strongly associated with the presence of curable STIs. The high prevalence (14.9%) of C. trachomatis found in this study is comparable to a prevalence of 13% (95% CI: 8.6-16.8) among women seeking family planning services in Nairobi, Kenya [20] and 9.9% (95% CI: 7.2-13.2) among women in a multi-country (Kenya, Rwanda and South Africa) cross-sectional study [26]. Two other recent studies reported lower prevalence rates of C. trachomatis: 5.6% (95% CI: 4.4-7.0) among 1300 HIV-seronegative ANCattending women at two district hospitals in rural western Kenya [13], and 6.0% (95% CI: 3.6-9.3) in another study among 300 women attending outpatient clinics in Nairobi [27]. Since the high prevalence of C. trachomatis among pregnant women may likely reflect ongoing transmission of C. trachomatis among the general rural population in Kilifi, further studies confirming the high prevalence and ongoing C. trachomatis transmission among this population would be helpful. Chlamydia infection has been found to be an independent predictor of HIV-1 acquisition in a recent study from western Kenya, with also syphilis, yeast infection/colonisation and BV showing strong associations with incident HIV-1 infection [13].
We documented a prevalence of T. vaginalis (7.4%) comparable to 6.0% among HIV-seronegative ANC-attending women in rural western Kenya [13], and 7.7% among women attending a child-health clinic in Mombasa, Kenya [28]. In vitro evidence suggests that T. vaginalis may alter the vaginal microbiome towards BV [29]. T. vaginalis has also been associated with increased viral shedding and increased acquisition of HIV [30,31], potentially contributing to a significant number of additional HIV infections globally [31].
The prevalence of HIV-1 (6.4%) in this study was substantial, with all known and newly diagnosed cases occurring amongst the 120 women !25 years. In Coastal Kenya, HIV-1 prevalence estimates in women aged 15-24 years decreased from 5.7% in 2007 to 2.0% in 2012 (p = 0.026) [32]. Our results suggest a further decline in HIV-1 prevalence, among women 24 years. A relatively low prevalence of N. gonorrhoeae (1%) and treponemal antibodies (1%) were found in our study. The prevalence of treponemal antibodies was similar to the study among HIV-seronegative ANC-attending women in rural western Kenya, [13] while the N. gonorrhoeae prevalence in the latter study was 3.0%.
BV diagnosis in our study was based on an a priori self-report of vaginal discharge in combination with a Nugent score of !7 (while only 2 (1%) women had asymptomatic BV). Women with BV had an almost 3-fold higher risk of infection with a curable STIs than women without BV, comparable to other studies [4,33]. It might be advisable to better document the presence of BV. However, general screening for BV among all women attending ANC, using the Nugent score [23] or Amsel's criteria [34], would require significant commitment of resources and staff. Using simple point of care testing, Madhivanan et al. [35] demonstrated that a vaginal pH >4.5 and a positive whiff test had a sensitivity of 83% but a specificity of only 47% for diagnosing BV. Although, not specific this approach may be adopted for BV screening in resource poor settings. Alternatively, presumptive treatment with intravaginal metronidazole (750 mg) plus miconazole (200 mg) for five consecutive nights each month for 12 months was shown to be effective to improve vaginal health and to decrease susceptibility to bacterial STIs, in a recent trial among non-pregnant women [36], but the frequency of treatment would make it challenging to upscale this intervention.
Our findings reinforce the need to integrate STI services at ANC. Screening for T. vaginalis can be achieved at a relatively affordable cost (US $ 5), making screening for T. vaginalis potentially feasible. Recent acquisition by the Kenyan government of over 120 GeneXpert IV machines for county and sub-county hospitals for tuberculosis testing [37], has potentially expanded the diagnostic capacity for curable STIs using the same testing platform at larger hospitals. However, the cost of the C. trachomatis and N. gonorrhoeae test kits are currently high. Without assay costs becoming considerably less expensive, it seems unlikely that ANCattending women will be routinely tested for C. trachomatis and N. gonorrhoeae. Cost-effective studies have played an important role of informing policy on the gains of syphilis screening in pregnancy while minimizing on the expenditure [38]. Such cost-effectiveness studies of ANC screening and treatment of C. trachomatis, N. gonorrhoeae, and T. vaginalis should be conducted in resource poor settings.
Treatment of STIs in both infected persons and their partner is pivotal for STI management, ensuring reduction in cases of reinfection or non-resolving infections and prevention of adverse outcomes [8]. In our study a presumptive patient delivered partner treatment approach was adopted. Despite the fact that we did not collect information on the number of partners who were treated, this approach has been shown to be acceptable and feasible among pregnant women in Kenya [39].
Early sexual debut was strongly associated with curable STIs in our study population, similar to findings in other studies [40,41]. Kilifi county is one of the poorest counties in Kenya, characterized by low literacy levels mostly among women, young age at marriage of women, and high fertility rates [42]. Structural changes focusing on the economic empowerment of communities, including improved access to education, and legislation protecting women from early marriage will be necessary to reduce the high burden of curable STIs among pregnant women.
A salient predictor of curable STIs in our study population was the use of water to clean after visiting the toilet as compared to those women who reported to use toilet paper or other solid materials. Unfortunately, our hygiene question did not distinguish urinating from defecating, or inquired whether the cleaning mode included both anus and vagina. It is possible that such cleaning transported bacteria from the perianal area to the vagina, as pathogens causing curable STIs may be found in paragenital areas [43]. Moreover, there is a strong correspondence in bacterial species and loads between the vagina and rectum of pregnant women [44]. McClelland et.al [31] showed that vaginal washing increased risk of HIV-1 acquisition among African women. Although, we did not assess whether water used by our study participants was shared, sharing of water may potentially lead to contamination of the vagina [45]. Pereira-Neves and Benchimol [46] showed in vitro that T. vaginalis remained viable and infective in swimming pool water samples for several hours. Further research is required to investigate the potential role that sanitary and hygiene practices may play in women with high prevalence of curable STIs and BV.
Observed genital ulcers had the strongest association with curable STIs, and were found in 2.5% of study participants. Presumptively treating curable STIs on the basis of observed genital ulcers in pregnant women would have treated~10% of the curable STIs in our study. Studies examining microbial aetiology of genital ulcer disease, have shown that genital ulcers can be attributed to several sexually transmitted pathogens with most cases being attributed to Herpes Simplex virus type 2 [47,48]. Genital ulcers were also identified as the strongest predictor of acute HIV acquisition in attendees of STD clinics who initially had HIV-seronegative or discordant rapid test results [49]. While in our study no women with genital ulcers had HIV, observing genital ulcers in women may help initiate risk reduction during ANC counselling and targeted acute HIV screening [50].
Our study had some limitations. First, our study only included a relatively small sample of pregnant women limiting the precision of our prevalence estimates. Second, we only included adult women and were not able to enroll pregnant women who were minors. Minors are potentially more vulnerable and may have presented with different STIs. Third, we included only women residing within the Kilifi Health and Demographic Surveillance Area [22], and therefore will have excluded women from outside the surveillance area who may have different health characteristics. Fourth, our hygiene and sanitary questions were few, and limited [51].
Finally, this was a cross-sectional study, only allowing for an assessment of associations and we did not collect data on pregnancy outcomes.
Despite these limitations, we showed that there is high prevalence of curable STIs among pregnant women attending an antenatal clinic at Kilifi County Hospital, Kenya. Curable STIs were associated with genital ulcers, hygienic practices, early sexual debut and BV. We recommend that antenatal care programs consider strengthening their diagnostic screening for curable STIs and BV.