Sexually transmitted infections (STIs) are commonly acquired during adolescence and, if left untreated, can lead to poor sexual and reproductive outcomes, such as pelvic inflammatory disease, infertility and enhanced human immunodeficiency virus (HIV) vulnerability.1 Indigenous peoples, particularly those who live in rural areas, are especially vulnerable to STIs due to unequal access to health care, especially to STI treatment and care.2
In Panama, indigenous peoples make up 12.2% of the total population. The Comarca Ngäbe-Bugle (CNB) is Panama's most populated indigenous region with over 200,000 individuals, with the country's highest levels (93.4%) of multidimensional poverty.3 Prevalence of HIV among 15- to 24-year-old women and men in Panama is estimated at 0.3% and 0.5%, respectively.4 However, in CNB, Ministry of Health (MOH) data show that, between 2013 and 2017, 2.1% (868 of 41,982) of rapid HIV tests performed were positive, with 82.0% of HIV infections in men, and 16.7% in 14- to 19-year-old boys and girls.5 As in other low- and middle-income countries, most STIs are managed syndromically in Panama, with little STI screening or etiological surveillance, except for prenatal MOH data which estimated maternal syphilis as 1.94% in 2017.6 Furthermore, CNB-MOH data indicate low prevalence of hepatitis B virus (HBV), whereas no data exist for Chlamydia trachomatis (CT), Neisseria gonorrhoeae (NG), or high-risk human papillomavirus (HR-HPV).6 Vaccination against HPV genotypes 16 and 18 was rolled out in 2008 through a school-based program targeting 10-year-old girls.7 In 2015, the tetravalent vaccine was substituted in the same age group; in 2016, 10-year-old boys were included.8 Despite universal rollout of HPV vaccination, monitoring of uptake is not available. Although some national STI data are stratified by region, they are not by ethnicity. Specifically, there are no MOH data on STIs among indigenous youth of CNB.
The World Health Organization (WHO) Strategy on STIs 2016 to 2021 defines the global strategy for the prevention and control of STIs.9 The first strategic direction is to collect laboratory diagnosed epidemiological data on STIs, to plan, fund, and undertake effective targeted interventions. To inform this strategic direction, we conducted the first community-based prevalence study of STIs focused on school-going indigenous adolescents in CNB.
Between July and November 2018, we conducted a cross-sectional study among male and female adolescent students aged 14 to 19 years, enrolled in high schools (7th to 12th grades), using 2-stage cluster sample design with random sampling of clusters, and equal probability of selection (Fig. 1).
The sampling frame was the 20 largest CNB high schools, equating to 41.7% of all CNB high school students. Schools were arranged in 5 groups of 4, in decreasing order by enrolment numbers. Two schools were randomly selected from each group. Then, all 7th to 12th grade classrooms in selected schools were ordered by decreasing classroom size, arranged into groups of 10, from which 2 classrooms were selected until about one thousand one hundred 14- to 19-year-old individuals were selected; we assumed two thirds would agree to participate.10 The primary outcome of interest in the study was STI prevalence among adolescents 14 to 19 years who had engaged in sexual intercourse at least once. A final sample size of seven hundred 14- to 19-year-old students was estimated, based on an assumed design effect of 2, 5% precision for STI prevalence estimates, 20% prevalence of the most common STI (CT), assuming two thirds of participants had engaged in sex, based on previous school-based sexual health studies in Panama.10
Guardians of minor students (aged <18 years) from selected classrooms were invited by letter to attend an informational meeting and were requested to bring their child's vaccine card. During the meeting, study objectives and procedures were explained. Guardians were then asked to provide signed informed consent. If available, details of HBV and HPV vaccinations were logged. After the meeting, eligible students were provided with information individually, asked to provide signed informed assent (if <18 years) or consent (≥18 years). Participants were assigned a unique identification code and given an appointment card with the date (within 3–4 weeks) to retrieve their laboratory results and treatment at a designated MOH clinic. At the schools, participants completed a self-administered questionnaire and provided blood and urine samples. All female participants went to the health center to provide additional self-collected genital swabs.
Data were obtained for sociodemographics and sexual behavior including vaginal/anal/oral sex, forced sex, and same-sex sexual activity (further defined in Supplementary Table 1, http://links.lww.com/OLQ/A409) with a self-completed questionnaire using a tablet computer (Kobo Toolbox; Harvard Humanitarian Initiative, Cambridge, MA). All participants, regardless of reported sexual behavior, were asked to give blood and urine samples, and self-collected swab samples (females only). Participants were expected to complete the questionnaire and biological sampling on the same day as consent/assent; however, if necessary, some activities were completed the following day.
Blood samples (8 mL) were tested with rapid HIV and syphilis tests in the field laboratory; confirmatory tests for these infections and HBV and herpes simplex virus (HSV-2) serologies were performed at the Provincial MOH blood bank laboratory in nearby Santiago. Participants were asked to provide a 40-mL first-void urine sample in a sterile cup for CT/NG testing. To limit costs, we only included CT/NG testing in participants who reported sexual experience, or tested positive for any HIV/STI serology (syphilis, HBV, or HSV-2). All female participants were instructed to provide 2 self-collected vaginal swabs,11 a high-vaginal Dacron swab (5 cm insertion) for HPV testing, and a dry-cotton lower-vaginal swab (3 cm insertion) for BV testing. Dacron swabs were placed immediately in a CerviCollect transportation tube (Abbott Molecular, Des Plaines, IL), and transported to the Gorgas Memorial Laboratory in Panama City at weekly intervals. Lower-vaginal swabs (Copan Diagnostics Inc., Murrieta, CA) were rolled onto a glass slide by a health technician, air-dried, fixed with methanol, transported weekly to the MOH Microbiology Laboratory in Santiago. Participants were offered a minimum 3-hour group education session, which included the topics: self-esteem, sexual decision making, dating violence, transmission, and prevention of HIV/STIs.
Within 1 day of submitting their sample, participants testing HIV-positive by rapid test were counseled and referred to the closest HIV clinic for assessment and antiretroviral therapy initiation. All treponemal antibody-positive cases were offered treatment upon receiving confirmation within 5 days of submitting their sample. All participants were reminded by school counselors to collect other test results during their scheduled appointment. All participants with a positive test for a curable STI were provided standardized STI treatment and patient-initiated notification, according to Panama guidelines.12 Participants with positive CT or NG who did not retrieve their results were traced and treated as per MOH guidelines.12
Blood samples were collected at each school. A temporary study laboratory was housed within the closest MOH clinic for HIV and syphilis testing. Human immunodeficiency virus testing was undertaken according to national guidelines13 using 2 rapid tests from different manufacturers: Alere Determine HIV-1/2 Ab (Inverness Medical Japan Co, Chiba, Japan), confirmatory testing using SD Bioline HIV-1/2 3.0 (Standard Diagnostics, Inc., Gyeonggi-do, South Korea). For syphilis testing, a point-of-care treponemal assay (Alere Determine TP, Inverness Medical Japan Co, Chiba, Japan) was used, with confirmation using a nontreponemal assay (Venereal Disease Research Laboratory Weiner Lab, Rosario, Argentina) with assay titration. Treponemal seroreactivity was defined as positivity to the treponemal assay, whereas active syphilis was defined as the dual positivity to treponemal and nontreponemal assays independent of titration, although titers were recorded with reactivity at ≥1:8 concentration considered “high-titer”. Participants with positive treponemal and nontreponemal (any titer) assays were treated. External control for HIV was performed with the College of American Pathologists (CAP) schemes, where concordance with other CAP users was 98.8%. External control of syphilis testing was undertaken using 24 panels from the Curie Laboratory in Panama, with 100% concordance.
At the Santiago MOH Bloodbank Laboratory, serological samples were tested to detect HSV-2 antibodies (Focus Diagnostic, Cypress, CA), HBV surface antigen [HBsAg] and total HBV core antibodies [anti-HBc] (Bio-Rad Laboratories, Hercules, CA). The laboratory participated in quarterly MOH Quality Assurance testing for bloodborne infections (HBsAg, anti-HBc), where performance has been certified as “Excellent” in 9 assessments before the study; HSV-2 testing concordance with other CAP laboratory users was 100%.
At the Santiago MOH Microbiology Laboratory, vaginal swabs were Gram-stained and analyzed using the Nugent's score: 0 to 3 was classified as healthy; 4 to 6 as intermediate microbiota; and 7 to 10 as BV.13 Study microbiologists (M.H. and G.H.) were trained at University of Alabama, read the slides independently, and undertook double reading in ~20% of slides.
At the Gorgas Genomics laboratory in Panama City, urine samples were tested for CT and NG using RealTime CT/NG polymerase chain reaction (Abbott Molecular). High-vaginal swabs were tested with a qualitative high-risk HPV genotypes detection assay (Abbott RealTime High-Risk HPV; Abbott Molecular) that distinguishes HPV16 and/or HPV18 from 12 other high-risk/probable HR-types: 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, 68. The laboratory enrolled in the CAP CT/NG and HPV testing schemes; concordance with other CAP users was 98.2% for HPV, 100% for both CT and NG.
Questionnaires were uploaded into the Kobo Toolbox cloud, imported and analyzed in STATA V15.0 (StataCorp, College Station, TX). Participants who completed the questionnaire and gave at least 1 biological sample were included in the analyses. We used the χ2 test to evaluate the difference in STI prevalence between men and women, age, and key sexual behavior variables, and Fisher exact test where appropriate. Kruskal-Wallis Test was used to evaluate associative trends in age, biological sex, and sexual behavior variables. Missing data were excluded.
Random-effects logistic regression was used to calculate odds ratios (OR) and 95% confidence interval (CI) for each pathogen adjusting for school-level clustering.14 Variables associated with each pathogen at P less than 0.2 in univariable analyses were included in initial multivariable models adjusting for sex and age (a priori). In the model, distal variables were included first, then proximal variables.15 The final model included variables independently associated with outcomes at P less than 0.1.
The research was approved by the Comité Nacional de Bioética de la Investigación de Panamá (EC-CNBI-2016-05-25, November 2017), and the London School of Hygiene & Tropical Medicine, UK (Ref:14558; January 2018).
Overall, 2006 students were assessed for eligibility, of whom 1101 (55%) were eligible (aged 14-19 years). A total of 380 (100%) of 380 eligible minor participants had guardian consent, signed their assent, and were present during the week the study team visited the community; 1 minor subsequently withdrew assent. Separately, a total of 321 (90.7%) of 354 eligible 18 to 19 years adolescents consented to participate and were available during the visit week (Fig. 1). Overall, we included 316 (45.1%) female and 384 (54.9%) male participants (Table 1), median age was 17 years (interquartile range [IQR], 14–19) for female participants, 18 years (IQR, 16–18) for males. The majority of participants (91.3%) were of Ngäbe ethnicity (Table 1).
Few participants had vaccination records and brought them to the study (22.8% [72 of 316] females, 19.0% [73 of 384] males). Of these, 61.1% (44 of 72) girls and 69.9% (51 of 73) men had ≥1 HBV vaccine dose; and 33.3% and 32.9% had full 3-dose vaccination. For HPV vaccination, 38.9%, 15.3% and 11.1% of female participants had full 3, 2, 1 dose, respectively. Male participants were too old to have been included in the country-wide HPV vaccination, therefore none had HPV vaccination records.
Most participants reported previous sexual experience: 74.1% [234 of 316] of female participants, 78.7% [302 of 384] of males. Those who reported sexual experience were older (92.4% of 19 year-olds versus 52.9% of 14 year-olds) and studying in the coastal region (Ño Kribo) compared to those from the mountainous southern regions (86.6% versus 63.2%, P < 0.01) (Table 1). The median age of sexual debut was 15 years (IQR, 14–16 years) among females, and 16 years (IQR, 14–17 years) among males (Table 2). A high proportion of both genders reported forced sex, although female participants were more likely to report this than males (36.2% versus 18.2%, P < 0.01) (Table 2). A history of pregnancy was reported by 11.3% of female participants, whereas 6.0% of male participants reported having impregnated a partner. Ten male participants reported no sexual experience, although they were HSV-2 seropositive, and thus assumed to be sexually experienced in further analyses (Table 3).
HIV and STI Prevalence
All HIV cases (females, 0.4%; males, 1.0%) were new diagnoses. Among all participants, treponemal seroreactivity was more prevalent among males than females (7.7% versus 3.5%; P = 0.05), as was high titer (≥1:8) active syphilis (6.6% vs. 1.3%; P = 0.003), with the following titer distribution: 9.1% (2 of 22) had 1:2, 9.1% (2 of 22) had 1:8, 13.6% (3 of 22) had 1:16, 22.7% (5 of 22) each had 1:32, 1:64, and 1:128 titers. There were no differences in seroprevalence of HBV markers by sex: HBsAg (females, 1.3%; males, 1.4%; P = 0.93), anti-HBc (females, 3.2%; males, 1.4%; P = 0.18), and in HSV-2 seroprevalence (females, 16.1%; males, 16.1%; P = 0.99).
Among participants who were sexually experienced, prevalence of CT was higher in girls (17.4% vs. 10.7%; P = 0.03) (Table 3), whereas NG prevalence was similar in both genders (females, 1.8% vs. males, 1.7%; P = 0.95).
Among girls, no HPV genotypes 16 or 18 were detected; however, 27.9% (95% CI, 22.7–33.9) of all females, and 33.2% (95% CI, 26.8–40.2) of those who reported sexual experience tested positive for other HR-HPV genotypes.
Overall, 100% of NG and 64% of CT-positive cases were treated within 4 weeks, and 93% CT-positive within 8 weeks. Within 4 weeks, 51% of HR-HPV returned for their results, 46% of BV-positive cases returned and were treated.
Factors Associated With STIs
There were no statistical significant associations between participant's age, age of sexual debut, partner's age at sexual debut, or reported condom use at last coitus with any STI (Supplementary Fig. 1, http://links.lww.com/OLQ/A410). After adjusting for age, CT was associated with female gender (adjusted OR [AOR], 2.02; 95% CI, 1.20–3.41) (Supplementary Table 1, http://links.lww.com/OLQ/A409). Active syphilis was strongly associated with being male (AOR, 4.51; 95% CI, 1.17–17.40), weakly associated with reported forced sex (AOR, 2.43; 95% CI, 0.89–6.62), and reported transactional sex (AOR, 2.49; 95% CI, 0.82–7.57, Supplementary Table 2, http://links.lww.com/OLQ/A411). Among male participants, active syphilis was independently associated with reported forced sex (7 [13.5%] of 52 vs. no-forced sex 11 [4.8%] of 231; AOR, 3.11; 95% CI, 1.14–8.45) and weak evidence of association among those who reported transactional sex (5 [12.5%] of 40 vs. no-transactional sex 13 [5.3%] of 247; AOR, 2.57; 95% CI, 0.86–7.65), and those who reported same-sex sex (4/26 [15.4%] vs. no report 11/204 [5.4%], AOR, 1.52; 95% CI, 0.84–2.75).
Region of CNB was strongly associated with HSV-2 seropositivity in both genders, and BV prevalence among female participants, with highest prevalence of both infections among participants from the coastal region of Ño Kribo compared with the southern mountainous regions (Supplementary Tables 3, http://links.lww.com/OLQ/A412 and 4, http://links.lww.com/OLQ/A413).
Among females, there was a strong association between BV and reported sexual experience (AOR, 2.86; 95% CI, 1.39–5.91), number of lifetime sexual partners (≥3 compared with 1, AOR, 3.81; 95% CI, 1.29–11.26), and same-sex sex (compared to nonreport, AOR, 4.68; 95% CI, 0.87–25.11) (Supplementary Table 5, http://links.lww.com/OLQ/A414). HR-HPV infection was associated with reported sexual experience (AOR, 4.05; 95% CI, 1.62–10.09) (Supplementary Table 6, http://links.lww.com/OLQ/A415). In the adjusted analysis, participants who tested positive for HR-HPV had 1.81 times the odds of testing positive for BV than those without HR-HPV (95% CI, 1.02–3.22). Participants with NG had 14.73 times the odds of testing positive with CT (95% CI, 3.47–62.4).
We conducted the first study of HIV/STI prevalence and sexual behavior among indigenous adolescents living in a Comarcal area of Panama and found a very high prevalence of chlamydia, syphilis, BV, and HR-HPV in this population, indicating high vulnerability and great need for interventions. Understanding regional and national STI epidemics through the collection of STI prevalence data is essential to develop interventions targeted for high-risk groups to achieve STI control.9
Syphilis prevalence was remarkably high (5.6% in both genders combined) compared to that found in a similar study conducted in 2015 among sexually experienced school-going adolescents (14-19 years) living in urban areas of Panama (0.0% overall).10 High-titer treponemal infection was particularly high among male adolescents. Results were comparable to, or higher than, adult indigenous peoples of Honduras and the Peruvian Amazon.16 In the absence of reported yaws cases in Panama since 197717 and given the high proportion of high-titer active syphilis found exclusively among sexually experienced adolescents, it can be inferred that this is recent exposure to sexually transmitted treponemal infection. Among males, same-gender sex was weakly associated with active syphilis, which accords with the general epidemiology of syphilis among men who have sex with men (MSM) in Latin America.18 We found high prevalence of forced sex among both genders, and among male participants this activity was strongly associated with active syphilis infection, similar to that found among young men in Ecuador.19 The 0.8% HIV prevalence is a worrying finding in this young population and was higher than among most adult indigenous populations in Latin America.16
Chlamydia prevalence was also high in both genders, although lower for females than in the study of adolescents in Panama City (30.9% in females, 6.2% in males).10 Sex differences are comparable to those observed worldwide.1 Despite the impact of CT on sexual and reproductive health, there are limited control options for chlamydia infection which impact sequelae.1,20 Specifically, there are no CT screening programs in Panama, as CT/NG infections are managed syndromically.12
The prevalence of BV was also high in our study population, greater than among adolescents in Ecuador,21 but comparable to BV prevalence among female adolescents in South Africa and Tanzania.22,23 It is likely that some BV-associated bacteria are sexually transmitted,24 yet school-based studies have reported prevalent BV among female participants who reported never having penile-vaginal sex.23 This may be due to underreporting of sexual activity.25 In our study, BV was strongly associated with reported sexual activity, but was also found among those who did not report any type of sexual activity. Similarly, we found several cases of HSV-2 seropositivity among males who did not report sexual activity, although oral-genital transmission may explain some cases. Study staff emphasized confidentiality of results to both parents and participants, although students may still have feared disclosure. Interestingly, we found BV to be associated with HR-HPV infection. A causal link has been hypothesized in a recent meta-analysis, where BV has been found to be associated with HR-HPV persistence and the development of high-grade cervical neoplasia.26
The absence of HPV16/18 genotypes among female adolescents is encouraging and had not been previously documented in Panama. It suggests possible success of the universal school-based HPV vaccination program that was rolled-out in 2008. However, it is difficult to ascribe a direct effect of the vaccination program, as too few participants produced vaccination records, and a large proportion of female participants had not received the recommended full-dose vaccination regimen, although it is plausible that a suboptimal number of doses may still provide sufficient protection.27 The high prevalence of other HR-HPV genotypes indicates this population is still highly vulnerable, as reported elsewhere.28 Additionally, if elimination of cervical and other HPV-related cancers were to be achieved, Panama will need to include a broader range of high-risk genotypes in the country-wide vaccination.29 Prevalence of HBV was low compared to other studies among adult indigenous populations in Latin America.16 Hepatitis B vaccination has been rolled out among infants in Panama since 2002, although in the few vaccination records produced, compliance with vaccination schedule appeared low. Catch-up vaccination for both HPV and HBV may be advantageous for those who have incomplete vaccination records.
Strengths of our study included testing a large panel of STIs among a highly vulnerable under-researched indigenous population, using population-based sampling and laboratories with excellent EQA results. We used electronic data-capture and self-completed questionnaires to mitigate reporting bias inherent in collecting sensitive behavior data. This study had some limitations. First, we may have encountered selection bias due to the school-based sampling; only about half of guardians consented their minor child's enrolment. More at-risk adolescents may have thus been excluded, leading to an underestimated STI prevalence. Moreover, male participants were over-represented in our sample, as more boys attend school. Finally, for logistical reasons, only the 20 largest schools were included in the sampling, which would have excluded adolescents who do not attend school or attend smaller schools. Second, CT/NG testing was not extended to the entire population, but only to those reporting sexual debut or found to be seropositive for an STI. Among female participants, genital self-swabbing was optional and nearly a quarter (23.1%) declined. To overcome these possible reporting and selection biases, we tested samples of participants who were positive for any of the serological markers, regardless of reported sexual debut. Few participants with HPV (n = 6) or BV (n = 10) were additionally identified. HPV transmission may still be possible in sexually inexperienced females through fomites and non–penile-vaginal sex.30 Third, reporting biases may have arisen from the self-reported nature of the questionnaire. We carefully piloted the use of the tablet computers with consistent results, and other studies have demonstrated high rates of disclosure of sensitive behavior using such methods.31s
In line with the first target of the worldwide WHO Strategy for STI Control, this study has collected data which estimate the epidemiological STI burden among a vulnerable indigenous population in rural Panama. Our results underscore the need for swift coordinated action, focusing on the development of control programs for the WHO-prioritized infections, that is, HIV, syphilis, HBV and HPV. Specifically, programs should be developed for CT screening among at-risk youth. In developing these interventions, it will be imperative to take Ngäbe and Buglé culture into account, as culturally congruent messages are more effective in delivering sexuality education curricula.32s
1. Holmes KK, Sparling PF, Stamm WE, et alSexually Transmitted Diseases: McGraw Hill Medical; 2008:575–594.
5. Ministerio de Salud. Base de Datos Unificado CNB VIH. 2018 ed. Not published. Ministerio de Salud Comarca Ngäbe-Bugle:2018.
6. Ministerio de Salud. Reportes Epidemiologicos, Comarca Ngabe-Bugle. unpublished 2017.
10. Gabster A, Mohammed DY, Arteaga GB, et al. Correlates of sexually transmitted infections among adolescents attending public high schools, Panama, 2015. PLoS One 2016; 11:e0163391.
11. Francis SC, Miiro G, Nakuya K, et al. Self-collection of vaginal swabs among adolescent girls in a school-setting in East Africa. Sex Transm Dis 2019; 46:335–341.
12. Ministerio de Salud, Caja de Seguro Social, Instituto Conmemorativo Gorgas de Estudios de Salud. Normativa Nacional para el Abordaje Integral de las Infecciones de Transmisión Sexual en Panamá, Panamá. 2014.
13. Nugent RP, Krohn MA, Hillier SL. Reliability of diagnosing bacterial vaginosis is improved by a standardized method of gram stain interpretation. J Clin Microbiol 1991; 29:297–301.
14. Bottomley C, Kirby MJ, Lindsay SW, et al. Can the buck always be passed to the highest level of clustering? BMC Med Res Methodol 2016; 16:29.
15. Victora CG, Huttly SR, Fuchs SC, et al. The role of conceptual frameworks in epidemiological analysis: A hierarchical approach. Int J Epidemiol 1997; 26:224–227.
16. Russell NK, Nazar K, Pino S, et al. HIV, syphilis, and viral hepatitis among Latin American indigenous peoples and afro-descendants: A systematic review. Rev Panam Salud Publica 2019; 43:e17.
17. Zoni AC, Saboya-Diaz MI, Castellanos LG, et al. Epidemiological situation of yaws in the Americas: A systematic review in the context of a regional elimination goal. PLoS Negl Trop Dis 2019; 13:e0007125.
18. Kojima N, Klausner JD. An update on the global epidemiology of syphilis. Curr Epidemiol Rep 2018; 5:24–38.
19. Hernandez I, Johnson A, Reina-Ortiz M, et al. Syphilis and HIV/syphilis co-infection among men who have sex with men (MSM) in Ecuador. Am J Mens Health 2017; 11:823–833.
20. Hocking JS, Spark S, Guy R, et al. O8 The Australian chlamydia control effectiveness pilot (ACCEPt): First results from a randomised trial of annual chlamydia screening in general practice. Sex Transm Infect 2012; 88(Suppl 1):A3.
21. Vaca M, Guadalupe I, Erazo S, et al. High prevalence of bacterial vaginosis in adolescent girls in a tropical area of Ecuador. BJOG 2010; 117:225–228.
22. Francis SC, Mthiyane TN, Baisley K, et al. Prevalence of sexually transmitted infections among young people in South Africa: A nested survey in a health and demographic surveillance site. PLoS Med 2018; 15:e1002512.
23. Francis SC, Holm Hansen C, Irani J, et al. Results from a cross-sectional sexual and reproductive health study among school girls in Tanzania: High prevalence of bacterial vaginosis. Sex Transm Infect 2018; 95:219–227.
24. Schwebke JR, Muzny CA. Pathogenesis of bacterial Vaginosis: Discussion of current hypotheses. J Infect Dis 2016; 214(suppl_1):S1–S5.
25. Langhaug LF, Sherr L, Cowan FM. How to improve the validity of sexual behaviour reporting: Systematic review of questionnaire delivery modes in developing countries. Trop Med Int Health 2010; 15:362–381.
26. Brusselaers N, Shrestha S, Van De Wijgert J, et al. Vaginal dysbiosis, and the risk of human papillomavirus and cervical cancer: Systematic review and meta-analysis. Am J Obstet Gynecol 2018; 221:9–18.e8.
27. Kreimer AR, Herrero R, Sampson JN, et al. Evidence for single-dose protection by the bivalent HPV vaccine—Review of the Costa Rica HPV vaccine trial and future research studies. Vaccine 2018; 36(32, Part A):4774–4782.
28. Watson-Jones D, Baisley K, Brown J, et al. High prevalence and incidence of human papillomavirus in a cohort of healthy young African female subjects. Sex Transm Infect 2013; 89:358.
30. Houlihan CF, Baisley K, Bravo IG, et al. Human papillomavirus DNA detected in fingertip, oral and bathroom samples from unvaccinated adolescent girls in Tanzania. Sex Transm Infect 2019; 95:374–379.
For further references, please see “Supplemental References,” http://links.lww.com/OLQ/A420.
Supplemental Digital Content
© Copyright 2019 American Sexually Transmitted Diseases Association