Open Access Research BMJ Open: first published as 10.1136/bmjopen-2017-016959 on 8 October 2017. Downloaded from Prevalence of Trichomonas vaginalis infection and protozoan load in South African women: a cross-sectional study Dewi J de Waaij,1,2 Jan Henk Dubbink,1,2 Sander Ouburg,1 Remco P H Peters,3,4 Servaas A Morré1,2 To cite: de Waaij DJ, ABSTRACT Strengths and limitations of this study Dubbink JH, Ouburg S, et al. Objectives Trichomonas vaginalis is thought to be the Prevalence of Trichomonas most common non-viral sexually transmitted infection ► This is the first study about the epidemiology and vaginalis infection and protozoan worldwide. We investigated the prevalence, risk factors load in South African women: a microbiological characteristics of Trichomonas and protozoan load of T. vaginalis infection in South African cross-sectional study. BMJ Open vaginalis in rural South Africa. women. 2017;7:e016959. doi:10.1136/ ► The study has a high sample size. Methods A cross-sectional study of 604 women was bmjopen-2017-016959 ► The number of rectal T. vaginalis infections is conducted at 25 primary healthcare facilities in rural South relatively low. ► Prepublication history for Africa (Mopani district). T. vaginalis DNA was detected in ► A selection bias is possible based on the geographical this paper is available online. vaginal and rectal swabs. In univariate and multivariate To view these files, please visit inclusion. analyses, the infection was investigated in the journal online (http:// dx. doi. T. vaginalis org/ 10. 1136/ bmjopen- 2017- relation to demographic characteristics, medical history 016959). and behavioural factors. The T. vaginalis load was determined as the logarithm of DNA copies per microlitre is symptomatic, women may report a change Received 31 March 2017 sample solution. in vaginal discharge, intermenstrual bleeding Revised 6 September 2017 Results Collected vaginal and rectal swabs were tested or vaginal blood loss during or after sexual Accepted 11 September 2017 for T. vaginalis DNA. Prevalence of vaginal T. vaginalis was contact. Other symptoms that can occur 20% (95% CI 17.0% to 23.4%) and rectal 1.2% (95% CI are vaginal itching, dysuria and abdominal 0.6% to 2.4%). Most women (66%) with a vaginal infection pain. Also, upper reproductive tract disease were asymptomatic. Factors associated with T. vaginalis syndromes can occur, including pelvic inflam- infection were a relationship status of single (OR 2.4; http://bmjopen.bmj.com/ matory disease (PID). Women with a T. vagi- 95% CI 1.5 to 4.0; p<0.001) and HIV positive infection (OR nalis infection have 4.7-fold increase in the 1.6; 95% CI 1.0 to 2.6; p=0.041). Women with vaginal T. 3–5 6 vaginalis infection were more likely to have concurrent risk of PID and tubal pathology. Women Chlamydia trachomatis rectal infection than those without infected with T. vaginalis also have a 1.3-fold 5 1Department of Medical vaginal infection (12%vs3%; p<0.001; OR 4.1). A higher increase in the risk of preterm labour. Microbiology and Infection median T. vaginalis load was observed among women with In women, T. vaginalis infection can Control, Laboratory of observed vaginal discharge compared with those without persist for months and, if left untreated, it Immunogenetics, VU University vaginal discharge (p=0.025). can increase the risk for HIV acquisition Medical Centre, Amsterdam, The on September 23, 2021 by guest. Protected copyright. Conclusions Vaginal trichomoniasis is highly prevalent if exposed.6 7 T. vaginalis could lead to an Netherlands in rural South Africa, especially among single women 2Department of Genetics and increase in the vaginal HIV load and thus and those with HIV infection, and often presents without Cell Biology, Faculty of Health, potentially increase the risk of HIV trans- symptoms. Medicine & Life Sciences, mission to a sexual partner.8 In addition, an Institute for Public Health association between T. vaginalis and Chla- Genomics, Research School 9 GROW (School for Oncology mydia trachomatis infection has been found. & Developmental Biology), INTRODUCTION Having a concurrent chlamydial infection was University of Maastricht, Trichomonas vaginalis is a protozoan para- a predictor of both prevalent and incident T. Maastricht, The Netherlands site that causes trichomoniasis and is mostly vaginalis. 3 Anova Health Institute, sexually transmitted. It is the most common Despite the estimated large burden of T. Johannesburg and Tzaneen, non-viral sexually transmitted infection (STI) vaginalis infection in the African region, South Africa 1 4Department of Medical worldwide. The infection is asymptomatic data on clinical presentation, demographic Microbiology, University of in 85% of the women and in 77% of the and behavioural factors associated with Maastricht, Maastricht, The men.1 WHO describes that T. vaginalis in infection, and microbiological factors are Netherlands the African region occurs 10 times as often relatively limited. However, in other conti- 2 Correspondence to in women than in men. Vaginal T. vaginalis nents, multiple studies have been done in 10–13 Dr Sander Ouburg; infections in the entire African region are the context of T. vaginalis infections. T. s. ouburg@ vumc. nl estimated to be 42.8 million.2 If the infection vaginalis prevalence has been reported for de Waaij DJ, et al. BMJ Open 2017;7:e016959. doi:10.1136/bmjopen-2017-016959 1 Open Access BMJ Open: first published as 10.1136/bmjopen-2017-016959 on 8 October 2017. Downloaded from a few African countries and ranges from 6.5% to 40%.14 of vaginal discharge syndrome, currently being pregnant, One of the explanations for the often high T. vaginalis the use of hormonal contraceptives, administering intra- prevalence is the lack of STI screening programmes and vaginal cleansing, visiting bars, alcohol use, concurrent limited control measurements. The estimated burden of partners, condom use during last sex act, partner >10 disease is significant but available data of symptomatology, years older, experiencing coercion and/or force, and coinfections and pathogen load are limited. C. trachomatis sex for money or for other benefits.17 Coercion reflects and Neisseria gonorrhoeae are other important causative sex inequality in relationships which is associated with micro-organisms for STI, showing an association with the increased risk.19 pathogen load as compared with the clinical presentation Data were double entered in EpiData (Epi Info V.3.5.3) and the development of long-term sequelae.15 16 by two researchers independently, were compared for In this study, we present the epidemiology of T. vaginalis inconsistencies, cleaned and verified. The data were anal- infection including the T. vaginalis load, as measured by ysed using SPSS Statistics V.20.0 (IBM). Dichotomous a validated nucleic acid amplification technique. We data were compared using χ2 test or Fisher’s exact test studied a high HIV prevalence setting in rural South and continuous data using the Mann-Whitney U test. ORs Africa. with 95% CI were calculated. Vaginal T. vaginalis infection was defined symptomatic when either abnormal vaginal discharge during physical examination, intermenstrual METHODS bleeding or vaginal blood loss during or after sexual inter- Population and design course was reported. Univariate analysis was performed This study was part of a cross-sectional study of 604 women to examine possible associations of demographic vari- in rural Mopani District, South Africa, as described previ- ables and clinical and behavioural factors in relation to ously.17 In brief, women aged 18–49 years who reported T. vaginalis. Multivariate analysis was conducted using a sexual activity during the prior 6 months were recruited at backward selection procedure and likelihood ratio tests. 25 primary healthcare facilities, regardless of the reason As a proxy for the T. vaginalis load, we used the median for visiting the facility that day. After informed consent, Cp value, which was calculated for each risk factor and demographic characteristics, medical history and details compared between infected versus non-infected women. of sexual behaviour were collected using a questionnaire. Healthcare workers collected vaginal and rectal swabs for molecular testing of T. vaginalis. RESULTS The Human Ethics Research Committee of the Univer- Demographics of study population sity of the Witwatersrand, South Africa, approved the A quarter of all women (n=154; 25.5%) had an age 18–24, study (Ref M110726). over half of the women were single (334; 55.3%), 26 http://bmjopen.bmj.com/ women reported anal intercourse (4.3%), a high propor- Laboratory testing tion was unemployed and almost 30% reported to be Collected vaginal and rectal swabs were analysed at infected with HIV (177; 29.3%). For more details, see the Laboratory of Medical Microbiology and Infection table 1. Control, VU University Medical Center, Amsterdam, The Netherlands. High Pure PCR Template Preparation Kit Prevalence of T. vaginalis infection (Roche Diagnostics, Basel, Switzerland) was used for the Of the 604 vaginal samples, one sample was excluded extraction of bacterial DNA. Detection of T. vaginalis based on lack of biological material and 28 gave an inde- on September 23, 2021 by guest. Protected copyright. DNA was done with the Prestoplus assay (Microbiome, terminate result (two times Cp values of ≥38), resulting in Amsterdam, The Netherlands) and the LightCycler II 575 samples for analysis. For vaginal infection, 113/575 480 (Roche Diagnostics) was used for the quantitative (20%) samples tested positive. PCR for C. trachomatis, N. gonorrhoeae and T. vaginalis.18 Of the 604 rectal samples, nine samples gave an inde- The Prestoplus assay for C. trachomatis, N. gonorrhoeae and terminate result (two times Cp values of ≥38) and were T. vaginalis is currently for inhouse use only and it will excluded resulting in 595 samples. For rectal infection, be transformed into a Presto-TV assay (Goffin Molecular 7/595 (1.2%) samples tested positive. Technologies, The Netherlands) in 2017 in a 200 reac- Four patients were infected with T.