Introduction
Tremendous progress has been made in identifying HIV-infected women during pregnancy and initiating them on antiretrovirals for prevention of mother-to-child HIV transmission (PMTCT) [1]. These efforts have resulted in a dramatic reduction in infant HIV infections, from 400 000 new infections among children in 2009 to 240 000 in 2013 [2,3]. However, PMTCT interventions predominantly target women with chronic HIV infection. Women in the ‘window period’ prior to HIV seroconversion and those who acquire HIV infection after their first antenatal HIV test may go unrecognized and fail to receive PMTCT interventions.
Several, but not all, studies have noted an increased HIV incidence during pregnancy and/or the postpartum period [4–8]. Epidemiologic approaches to determine the influence of pregnancy and postpartum period on risk of HIV acquisition are challenged by difficulties in defining an appropriate comparison group. A recent meta-analysis using data from 19 cohorts representing 22 803 total person-years of follow-up reported a pooled HIV incidence rate of 3.8/100 person-years [95% confidence interval (95% CI) 3.0–4.6] during pregnancy or postpartum [4]. Increased susceptibility to HIV during pregnancy/postpartum could result from hormonal changes that alter genital mucosal surfaces or distribution of target cells at these surfaces [9–11]. Behavioural factors may also play a role if male partners seek other partners during pregnancy or postpartum and bring HIV back to the relationship [12,13].
Previous studies of HIV acquisition during pregnancy and postpartum have not utilized detailed longitudinal sampling with serial sensitive assays to detect HIV, nor comprehensively evaluated genital infections and sexually transmitted infections (STIs) as risk factors in pregnant or postpartum cohorts. In addition, women presenting to antenatal care (ANC) clinics have distinct social, demographic, sexual networks and partnership characteristics that may influence cofactors of HIV acquisition. Although women are recognized to be at risk for HIV acquisition during and after pregnancy, it remains important to estimate magnitude and define cofactors of risk, the first step to developing sensible preventive strategies for this large population of women. We determined rates and cofactors of incident HIV among HIV-uninfected women to characterize mechanisms for risk during and after pregnancy [14].
Materials and methods
Study design and population
Between May 2011 and June 2013, we recruited HIV-seronegative pregnant women seeking ANC at the Ahero sub-District and Bondo District Hospitals in rural western Kenya into a prospective cohort study. Women were eligible for inclusion if they were pregnant, at least 14 years of age, had a documented HIV-negative rapid HIV test at enrolment or 3 months or less prior to enrollment during routine care, planned to reside in the study area until 9 months postpartum, would provide locator information and have a home visit, and were not participating in other research. Women who were 28–36 weeks gestation, calculated by last menstrual period, were initially eligible; eligibility criteria were expanded to include women 14–28 weeks gestation in June 2011, more than 36 weeks gestation in November 2011 and less than 14 weeks gestation in March 2012. Gestational age eligibility criteria were broadened to increase enrolment and generalize to women throughout the pregnancy/postpartum continuum. Enrolment at less than 14 weeks gestation required a protocol modification to incorporate urine pregnancy test confirmation. Study retention efforts included calling women by phone three times for missed study visits, followed by calling contacts provided by participants and conducting home visits (if necessary). All study procedures were approved by the Kenyatta National Hospital/University of Nairobi Ethical Research Committee and the University of Washington Institutional Review Board prior to study initiation.
Clinical procedures
Eligible women interested in study participation provided written informed consent. Women were scheduled for follow-up during pregnancy (20, 24, 32, 36 weeks gestation) and postpartum (2, 6, 10, 14 weeks; 6 and 9 months). At study visits, women had blood collected for HIV-RNA nucleic acid amplification tests (NAATs), self-collected vaginal swabs, physical examinations and completed questionnaires to assess maternal health, sexual behaviour and vaginal washing and drying practices. At enrolment, women had a vaginal swab collected for chlamydia and gonorrhoea. STIs were treated following microbiologic diagnosis using antimicrobial regimens in Kenyan guidelines [15]. After 19 December 2011, HIV NAAT assays were performed at enrolment; 28 and 36 weeks gestation; and 6, 14, 24, and 36 weeks postpartum. All women with incident HIV were offered lifelong antiretroviral therapy (ART), Option B+, which began being implemented in Kenya in 2013 for all pregnant women.
Laboratory procedures
HIV NAATs were performed at the HIV Research Laboratory at Kenya Medical Research Institute (KEMRI)/Centers for Disease Control and Prevention in Kisumu, Kenya, using the Roche COBAS Ampliprep and COBAS TaqMan HI2CAP platforms. This method of HIV detection is highly sensitive for detection of acute HIV in the ‘window period’ [16,17]. Pools of 10 samples were tested via NAATs; individual samples were retested if pooled results were positive, followed by HIV viral loads to confirm positive results. Self-collected vaginal swabs were tested at the University of Nairobi/Washington Research Laboratory at Coast Hospital in Mombasa, Kenya, for chlamydia and gonorrhoea using PCR technology and Gen-Probe APTIMA Combo2 (chlamydia/gonorrhoea) kits and for trichomoniasis and bacterial vaginosis. Detection of Trichomonas vaginalis was by wet microscopy in clinic. Women with a Nugent score more than 7 were diagnosed with bacterial vaginosis. Syphilis serology (rapid plasma reagin tests) was conducted during routine ANC, or by study staff if results were not available.
HIV incidence definitions
All HIV infections detected during the study were considered incident infections, with three distinct types of incident infection. Pregnant women with a documented negative rapid HIV test 3 months or less prior to enrolment and positive NAAT and rapid HIV tests were classified as having seroconversion at enrolment, whereas pregnant women with HIV rapid negative and NAAT positive results were classified as having acute infection at enrolment. Incidence rates for seroconversions and acute infections at enrolment were calculated according to the Serologic Testing Algorithm for Recent HIV Seroconversion [18]. These incidence rates are calculated as: number with incident infection (Ninc), divided by the window period in days (w), multiplied by 365, and dividing this figure by the number at risk [(365 × Ninc/w) / (number at risk × 1 year)]. HIV-uninfected women and women with a seroconversion or acute infection at enrolment were included. Incidence rates per 100 person-years were based on assay window periods of 48 and 28 days for seroconversions and acute infections at enrolment, respectively [19]. Women with HIV-negative rapid tests and NAAT at enrolment, but NAAT positive at a follow-up visit were classified as having acute infection during follow-up. Incidence rates for acute infection in follow-up were calculated on the basis of number of infections per 100 person-years of follow-up. Overall incidence rates were weighted averages of incidence rates for seroconversion at enrolment, acute infection at enrolment and acute infection during follow-up.
Statistical analysis
We anticipated enrolling 1730 women, accumulating 60 incident HIV infections assuming an incidence of 3% per year. This sample size would give at least 80% power to detect a 2.1 to 2.3-fold difference in HIV risk for an exposure with a prevalence of 25–50%, respectively, assuming a two-sided test (α = 0.05). Due to slower enrolment, the sample was limited to 1304.
Timing of HIV was recoded as the midpoint between the last negative and first positive NAAT and incident HIV infections detected at enrolment were recoded as being infected at 0.5 days for inclusion in Cox proportional hazards models. Univariate Cox proportional hazards regression models were used to identify cofactors associated with maternal HIV acquisition in a complete case analysis. Multivariate imputation by chained equations, which account for within-cluster correlation for time-varying covariates, was used to create 10 datasets to multiply impute missing values for the multivariate model covariates. Missing data were assumed to be missing at random. This approach is flexible in that it does not require the conditional distributions to be normal, and has been shown to perform similarly to a multivariate normal imputation when variously scaled variables are involved [20]. Multivariate Cox proportional hazards regression modelling was performed on each of the 10 imputed datasets, and final coefficient estimates were produced by averaging the 10 imputed regression coefficients for each covariate. Missing values were imputed for the following variables: chlamydia, syphilis, yeast infection, bacterial vaginosis, partner age difference and lifetime number of sex partners. Variables identified as a-priori cofactors (age, STIs and bacterial vaginosis) were included in an initial multivariate model along with any cofactor in univariate analyses with P value less than 0.15; variables not associated with HIV acquisition at the P less than 0.15 level in the multivariate model were dropped from the final model as a group. Coital frequency was calculated as the number of sexual acts women reported in the month before the study visit (and prior to HIV detection) multiplied by the number of months between visits; total coital frequency was calculated as the sum of monthly coital frequency over the duration of the study for each woman who acquired HIV. Per coital act risk of HIV acquisition was modelled for all women using generalized estimating equations with a Poisson distribution, log link and exchangeable covariance matrix. Statistical analyses were performed using Stata version 13.1 (College Station, Texas, USA).
Results
Study population
At the Ahero and Bondo maternal child health (MCH) clinics, 4245 women sought ANC and 3408 (80%) were HIV-seronegative (Fig. 1). Among 2351 HIV-seronegative women eligible for participation, 1304 (55%) were enrolled. Median age was 22 years and median duration of education was 8 years (Table 1). Median gravidity was 2 and age of sexual debut was 16 years. The majority (78%) of women were married, 14% in polygamous relationships, and the median relationship duration was 4 years. Unprotected sex among sexually active women in the month before enrolment was common (92%) and eight (1%) of 782 women reported having multiple partners. Male partners were older than women (median 5 years older). While 879 of 1203 (73%) women reported that their partner was tested for HIV, only 1% of women knew their partner was HIV-infected and 28% did not know their partner's HIV status. Study retention to 9 months postpartum was high (98%).
;)
Fig. 1: Study recruitment, eligibility and enrolment. aIncluded 429 women rebooked to come back at a later date, 176 women who consulted their male partner and 39 women who wanted to think about it. bIncludes women who completed the last study visit or acquired HIV at or prior to study exit; 20 women were lost to follow-up and 5 refused to continue study participation.
Table 1: Baseline characteristics (N = 1304).
HIV incidence
After 1235 person-years of follow-up, 25 women were identified with incident HIV infections: seven seroconversions and three acute infections at enrolment, and 15 acute infections during follow-up (two during pregnancy, 13 postpartum) (Fig. 2). Mean duration of follow-up was 0.95 person-years. The overall incidence rate was 2.31 per 100 person-years (95% CI 0.71–4.10); the incidence rate of acute infection during follow-up was 1.21 (95% CI 0.73–2.01). During pregnancy, the incidence rate was 3.44 per 100 person-years (95% CI 0.60–6.63) compared with 1.40 (95% CI 0.81–2.42) during the postpartum period (P = 0.6). Symptoms of primary infection, including sore throat, swollen glands, headache, malaise, vomiting, diarrhoea, fever or macular papular rash, were reported by seven (28%) women. None of the women with an incident infection reported having an HIV-infected partner at baseline; 12 (48%) reported that their partner was HIV-negative and 13 (52%) had a partner of unknown status. Risk of HIV acquisition was 0.51 per 1000 coital acts during pregnancy, similar to the postpartum period (0.42) (P = 0.8).
Fig. 2: Timing of detection of HIV infection and incidence rates during pregnancy and postpartum.
Sexually transmitted infections and genital tract infections
The most common STIs at baseline were T. vaginalis and chlamydia (both 6%), followed by gonorrhoea (3%) and syphilis (1%). Coinfection with two STIs was identified in 22 women: 12 with T. vaginalis and chlamydia, three with T. vaginalis and gonorrhoea, five with chlamydia and gonorrhoea, one with chlamydia and syphilis, and one with T. vaginalis and syphilis. Nearly one-quarter (23%) of women were diagnosed with bacterial vaginosis at baseline, and 25% had yeast infections.
Cofactors for HIV incidence
Risk factors for HIV incidence are summarized in Table 2. and Fig. 3. Both prior history and diagnosis of STIs in pregnancy were significantly associated with HIV acquisition. Women who reported a history of STIs were three times more likely to acquire HIV than women without a history of STIs (hazard ratio 3.48, 95% CI 1.31–9.27; P = 0.01). Chlamydia (hazard ratio 4.49, 95% CI 1.34–15.0; P = 0.02) and syphilis (hazard ratio 9.18, 95% CI 2.15–39.3; P = 0.003) at enrolment, and bacterial vaginosis (hazard ratio 2.91, 95% CI 1.25–6.76; P = 0.01) and yeast infection (hazard ratio 3.46, 95% CI 1.46–8.19; P = 0.005) at enrolment or follow-up were also associated with HIV incidence. Prevalence of other STIs (gonorrhoea, T. vaginalis) was similar between HIV-infected and HIV-uninfected women. HIV risk was higher among women in shorter relationships (hazard ratio 1.19 for each year), with older male partners (hazard ratio 1.07 for each year older) and with higher lifetime number of sexual partners (hazard ratio 1.19 for each additional partner). Among married women, shorter duration of marriage was also a significant risk factor. Marital status, partner HIV status, partner circumcision status, maternal age, educational level, age at sexual debut, history of commercial sex, and vaginal washing or drying practices did not differ between HIV-infected and uninfected women.
Table 2: Cofactors for maternal HIV acquisition.
Fig. 3: Cumulative risk of HIV acquisition during pregnancy and postpartum, by risk factor (a) age, (b) marriage duration, (c) chlamydia, (d) syphilis, (e) bacterial vaginosis, (f) history of sexually transmitted infection (STI). Parentheses indicate number of HIV infections between each risk period.
We imputed values for independent variables with missing data. The percentage missing values were as follows: syphilis (21.1%), bacterial vaginosis (18.1%), partner age difference (17.8%), yeast infection (17.4%), T. vaginalis (17.4%), chlamydia (0.3%) and lifetime number of sex partners (0.2%). In the multivariate model, having an older partner (hazard ratio 1.06, 95% CI 1.00–1.13; P = 0.04), chlamydia (hazard ratio 4.49, 95% CI 1.34–14.2; P = 0.01) and yeast infection (hazard ratio 2.78, 95% CI 1.17–6.63; P = 0.02) remained significant cofactors for HIV.
Discussion
We found high HIV incidence rates during pregnancy and postpartum, a period when women can transmit the virus to their infants. Our HIV incidence estimate is consistent with other studies in Africa and a recent meta-analysis [5–8], and is comparable to those seen in discordant couples and sex workers, suggesting that pregnancy and postpartum is a period of appreciable susceptibility to HIV [21,22]. Risk per coital act was also high, nearly 0.5 per 1000, considering that partner HIV serostatus was not directly measured and this estimate includes women whose partners were HIV-uninfected. Per coital risk among pregnant and lactating women in serodiscordant couples is nearly 1 per 1000 [5], and estimated at 0.38 per 1000 among women in low-income countries [23]. Importantly, our study identified, for the first time, key risk factors for acquiring HIV in this population. Syphilis was associated with a nine-fold increased risk and chlamydia with a more than four-fold increased risk of HIV in women. In addition, bacterial vaginosis and yeast infection were associated with nearly three-fold increased risk of acquiring HIV. Women with a history of STIs had a higher risk as did women in age-discordant partnerships and shorter duration of marriage.
Women who acquire HIV in pregnancy and postpartum acquire HIV in one of three scenarios: first, increased mucosal susceptibility in the context of a stable partnership with an HIV-infected partner, second, due to a recent partner HIV infection through external partnerships, and third by having a new HIV-infected partner. Our finding that syphilis and chlamydia were risk factors for HIV suggests that some infections were from an external partner. Our finding of increased HIV risk with bacterial vaginosis and yeast infection suggests that for some women, HIV acquisition occurred in an undiagnosed chronic serodiscordant relationship, in which women had increased mucosal susceptibility due to vaginal infections. Age-discordance and short duration of partnership may be proxies for unknown partner HIV status or increased likelihood of external partnerships.
Women with STIs and genital infections were more likely to have incident HIV, with large effect sizes that were highly significant. There is substantial evidence that STIs increase the likelihood of acquiring HIV; however, this has not been studied in context of pregnancy and postpartum [24,25]. Importantly, routine screening for STIs other than syphilis is not performed in ANC in many regions [26]. Laboratory diagnosis for STIs/genital infections is largely unavailable and health providers rely on syndromic management, which has poor diagnostic performance [27,28]. Yeast infections are more prevalent during pregnancy and our finding of a 3.4-fold increased risk for HIV acquisition with yeast infection may reveal one potential mechanism for susceptibility to HIV in pregnancy. Our findings reinforce the need to integrate STI services in MCH clinics, particularly in high HIV prevalence settings, and develop low-cost diagnostics to detect STIs and genital infections, which may contribute to decreased risk of HIV and prevent other adverse MCH outcomes.
Partnership characteristics were associated with HIV acquisition in our study. Women with age-discordant partnerships (young women-older men) had higher HIV risk, as has been seen in other populations [29–31]. Shorter duration of marriage was also associated with HIV risk. Less than 10% of women reported using condoms, which may indicate that women and their partners perceive low risk of HIV acquisition. Male partner HIV testing could identify men who need ART to reduce infectiousness [32]; however, uptake remains low and few men accompany women to ANC for testing [33,34]. In our study, one-third of the women did not know their partners’ HIV status despite repeated encouragement from study clinicians to have their partners tested. Home-based approaches may increase male HIV testing and treatment [35]. It was notable that none of the women in our study who became HIV-infected reported having a seropositive partner. Conversely, none of the women with known HIV-infected partners became infected, likely because of taking precautions to minimize transmission. Thus, in the antenatal/postpartum setting, women in known HIV serodiscordant relationships may be at relatively low risk compared with the large number of women with unknown partner status. Prioritizing partner HIV testing and treatment will continue to be important to protect women from HIV during pregnancy and postpartum.
Primary HIV prevention in women is the first of four prongs recommended by the WHO for effective PMTCT [36]. However, there are challenges in implementing strategies to keep HIV-uninfected women from acquiring HIV in pregnancy or postpartum. Preexposure prophylaxis (PrEP) could be considered for HIV-uninfected pregnant and breastfeeding women, especially those unable to negotiate mutual monogamy or condom use. Although PrEP is effective in preventing HIV acquisition, data from pregnant and postpartum women are limited [21]. In the Partners PrEP study, the most common reason for not dispensing study medication was pregnancy [21]. Given PrEP efficacy among adherent women, but balancing potential concerns regarding the foetus/infant, identification of ‘high-risk’ pregnant/postpartum women may be useful to limit PrEP provision to those who are most likely to acquire HIV, for example women with STIs and shorter relationships. Studies to explore risk scores that identify subgroups to consider for PrEP in pregnancy/postpartum maybe useful.
Our findings of high HIV incidence during pregnancy reinforce the need for repeat HIV testing in PMTCT programmes. Although WHO guidelines recommend repeat HIV testing, few programmes offer repeat HIV testing [37]. Repeat testing is recommended in regions with HIV incidence of more than 1 per 1000 person-years and is highly acceptable and cost-effective [6,38–40]. In a recent U.S. study, 97% of women offered retesting in pregnancy accepted [39]. In Kenya, nearly 95% of more than 2000 women accepted HIV retesting at 6 weeks postpartum [6]. However, repeat testing is infrequently performed due to erratic availability of test kits, concerns of increased workload, late ANC initiation and low emphasis by programmes [6,39]. Addressing these challenges would enhance early detection of incident maternal HIV infections and enable prompt ART initiation to decrease vertical HIV transmission. Clinical detection of acute HIV is poor; symptoms are often nonspecific or absent [41]. In our study, only 28% of women with incident HIV were symptomatic. HIV RNA testing has higher sensitivity and specificity to detect acute HIV than serologic tests [42,43]. Although widespread use of maternal NAAT assays by PMTCT programmes is restricted by cost, this could be lowered by pooling samples ($2 per specimen) [44–46]. However, RNA testing requires sophisticated laboratories and does not have a rapid turn-around time. As early infant diagnosis programmes introduce new point-of-care viral testing for infants in the future, these could be expanded for detection of maternal HIV.
Our study had several strengths. The prospective cohort design enabled detection of incident HIV throughout pregnancy and postpartum, with frequent sampling to determine timing of acquisition. Retention was high (98%) and we used a highly sensitive test, which detects infections soon after infection. We prospectively evaluated demographic, clinical and behavioural characteristics associated with HIV acquisition and included time-varying covariates in our model, using robust imputation methods to account for missing data. However, our study had some limitations. HIV incidence may be underestimated, due to HIV prevention counselling or selection bias among women who elect to participate in a longitudinal study [47]. The number of incident infections detected was also small and may have limited power to detect associations in our adjusted model, particularly those cofactors of lower prevalence. We did not perform longitudinal testing for chlamydia, gonorrhoea and syphilis; thus, associations between these STIs and HIV acquisition are based on baseline STI prevalence. Trichomonas vaginalis diagnosis was based on wet mount microscopy, which has poor sensitivity (60%) compared with PCR assays.
In summary, we found high HIV incidence in pregnant and postpartum women without evidence for increased incidence in pregnancy versus postpartum and with strong associations with genital infections and partner characteristics. Our findings are important to inform new strategies for HIV prevention in this vulnerable population.
Acknowledgements
We would like to acknowledge the significant contributions from study participants and the Mama Salama Study team members. We would also like to acknowledge support from the University of Washington's Global Center for Integrated Health of Women, Adolescents and Children (Global WACh).
The contributions of the authors to the manuscript were as follows: study design (J.K., A.L.D., D.M., B.A.R., C.Z., J.O., R.S.M., G.J.S.), data collection (J.K., A.L.D., D.M., C.Z., L.O., G.J.S.), data analysis (A.L.D., L.O., J.O., B.A.R.), data interpretation (J.K., A.L.D., B.A.R., C.Z., J.O., R.S.M., G.J.S.) and manuscript writing (J.K., A.L.D., B.A.R., C.Z., J.O., R.S.M., G.J.S).
This study was funded through National Institutes of Health grant (P01 HSD 064915) and received assistance from the University of Washington Center for AIDS Research (P30 AI27757). The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript
Conflicts of interest
The authors declare that no competing interests exist.
References
1. UNAIDS.
Global report: UNAIDS report on the global AIDS epidemic 2013. Geneva: UNAIDS; 2013.
2. World Health Organization. Consolidated guidelines on the use of antriretroviral drugs for treating and preventing HIV infection. Geneva: World Health Organization; 2013.
http://www.who.int/hiv/pub/guidelines/arv2013/en/ [Accessed 18 May 2015].
3. UNAIDS.
The gap report. Geneva: UNAIDS; 2014.
4. Drake AL, Wagner A, Richardson B, John-Stewart G.
Incident HIV during pregnancy and postpartum and risk of mother-to-child HIV transmission: a systematic review and meta-analysis.
PLoS Med 2014; 11:e1001608.
5. Gray R, Li X, Kigozi G, Serwadda D, Brahmbhatt H, Wabwire-Mangen F, et al.
Increased risk of incident HIV during pregnancy in Rakai, Uganda: a prospective study.
Lancet 2005; 366:1182–1188.
6. Kinuthia J, Kiarie JN, Farquhar CF, Richardson B, Nduati R, Mbori-Ngacha D, et al.
Cofactors for HIV-1 incidence during pregnancy and postpartum period.
Curr HIV Res 2010; 8:510–514.
7. Moodley D, Esterhuizen TM, Pather T, Chetty V, Ngaleka L.
High HIV incidence during pregnancy: compelling reason for repeat HIV testing.
AIDS 2009; 23:1255–1259.
8. De Schacht C, Hoffman HJ, Mabunda N, Lucas C, Alons CL, Madonela A, et al.
High rates of HIV seroconversion in pregnant women and low reported levels of HIV testing among male partners in southern Mozambique: results from a mixed methods study.
PLoS One 2014; 9:e115014.
9. Wira CR, Fahey JV, Rodriguez-Garcia M, Shen Z, Patel MV.
Regulation of mucosal immunity in the female reproductive tract: the role of sex hormones in immune protection against sexually transmitted pathogens.
Am J Reprod Immunol 2014; 72:236–258.
10. Sheffield JS, Wendel GD, McIntire DD, Norgard MV.
The effect of progesterone levels and pregnancy on HIV-1 coreceptor expression.
Reprod Sci 2009; 16:20–31.
11. Rodriguez-Garcia M, Patel MV, Wira CR.
Innate and adaptive anti-HIV immune responses in the female reproductive tract.
J Reprod Immunol 2013; 97:74–84.
12. Onah HE, Iloabachie GC, Obi SN, Ezugwu FO, Eze JN.
Nigerian male sexual activity during pregnancy.
Int J Gynaecol Obstet 2002; 76:219–223.
13. Lawoyin TO, Larsen U.
Male sexual behaviour during wife's pregnancy and postpartum abstinence period in Oyo state, Nigeria.
J Biosoc Sci 2002; 34:51–63.
14. National AIDS and STI Control Programme, Ministry of Health, Kenya.
Kenya AIDS indicator survey 2012: preliminary report. Nairobi, Kenya: Ministry of Health; 2013.
15.
National Guidelines for reproductive tract infection services. 1st ed. Republic of Kenya, Ministry of Health; 2006.
16. Fiebig EW, Wright DJ, Rawal BD, Garrett PE, Schumacher RT, Peddada L, et al.
Dynamics of HIV viremia and antibody seroconversion in plasma donors: implications for diagnosis and staging of primary HIV infection.
AIDS 2003; 17:1871–1879.
17. Stekler J, Maenza J, Stevens CE, Swenson PD, Coombs RW, Wood RW, et al.
Screening for acute HIV infection: lessons learned.
Clin Infect Dis 2007; 44:459–461.
18. McDougal JS, Pilcher CD, Parekh BS, Gershy-Damet G, Branson BM, Marsh K, et al.
Surveillance for HIV-1 incidence using tests for recent infection in resource-constrained countries.
AIDS 2005; (19 Suppl 2):S25–S30.
19. Pilcher CD, Fiscus SA, Nguyen TQ, Foust E, Wolf L, Williams D, et al.
Detection of acute infections during HIV testing in North Carolina.
N Engl J Med 2005; 352:1873–1883.
20. Lee KJ, Carlin JB.
Multiple imputation for missing data: fully conditional specification versus multivariate normal imputation.
Am J Epidemiol 2010; 171:624–632.
21. Baeten JM, Donnell D, Ndase P, Mugo NR, Campbell JD, Wangisi J, et al.
Antiretroviral prophylaxis for HIV prevention in heterosexual men and women.
N Engl J Med 2012; 367:399–410.
22. Kaul R, Kimani J, Nagelkerke NJ, Fonck K, Nquqi EN, Keli F, et al.
Monthly antibiotic chemoprophylaxis and incidence of sexually transmitted infections and HIV-1 infection in Kenyan sex workers: a randomized controlled trial.
JAMA 2004; 291:2555–2562.
23. Boily MC, Baggaley RF, Wang L, Masse B, White RG, Hayes RJ, et al.
Heterosexual risk of HIV-1 infection per sexual act: systematic review and meta-analysis of observational studies.
Lancet Infect Dis 2009; 9:118–129.
24. Fleming DT, Wasserheit JN.
From epidemiological synergy to public health policy and practice: the contribution of other sexually transmitted diseases to sexual transmission of HIV infection.
Sex Transm Infect 1999; 75:3–17.
25. Laga M, Manoka A, Kivuvu M, Malele B, Tuliza M, Nzila N, et al.
Nonulcerative sexually transmitted diseases as risk factors for HIV-1 transmission in women: results from a cohort study.
AIDS 1993; 7:95–102.
26. Kenya National Bureau of Statistics (KNBS), ICF Macro.
Kenya demographic and health survey 2008–2009. Calverton, MD: KNBS, ICF Macro; 2010.
27. Vuylsteke B.
Current status of syndromic management of sexually transmitted infections in developing countries.
Sex Transm Infect 2004; 80:333–334.
28. Mayaud P, Grosskurth H, Changalucha J, Todd J, West B, Gabone R, et al.
Risk assessment and other screening options for gonorrhoea and chlamydial infections in women attending rural Tanzanian antenatal clinics.
Bull World Health Organ 1995; 73:621–630.
29. Kelly RJ, Gray RH, Sewankambo NK, Serwadda D, Wabwire-Mangen F, Lutalo T, et al.
Age differences in sexual partners and risk of HIV-1 infection in rural Uganda.
J Acquir Immune Defic Syndr 2003; 32:446–451.
30. Glynn JR, Caraël M, Auvert B, Kahindo M, Chege J, Musonda R, et al.
Why do young women have a much higher prevalence of HIV than young men? A study in Kisumu, Kenya and Ndola, Zambia.
AIDS 2001; 15:S51–S60.
31. Gregson S, Nyamukapa CA, Garnett GP, Mason PR, Zhuwau T, Caraël M, et al.
Sexual mixing patterns and sex-differentials in teenage exposure to HIV infection in rural Zimbabwe.
Lancet 2002; 359:1896–1903.
32. Cohen MS, Chen YQ, McCauley M, Gamble T, Hosseinipour MC, Kumarasamy N, et al.
Prevention of HIV-1 infection with early antiretroviral therapy.
N Engl J Med 2011; 365:493–505.
33. Msuya SE, Mbizvo EM, Hussain A, Uriyo J, Sam NE, Stray-Pedersen B.
Low male partner participation in antenatal HIV counselling and testing in northern Tanzania: implications for preventive programs.
AIDS Care 2008; 20:700–709.
34. Farquhar C, Kiarie JN, Richardson BA, Kabura MN, John FN, Nduati RW, et al.
Antenatal couple counseling increases uptake of interventions to prevent HIV-1 transmission.
J Acquir Immune Defic Syndr 2004; 37:1620–1626.
35. Osoti AO, John-Stewart G, Kiarie J, Richardson B, Kinuthia J, Krakowiak D, et al.
Home visits during pregnancy enhance male partner HIV counselling and testing in Kenya: a randomized clinical trial.
AIDS 2014; 28:95–103.
36. UNAIDS.
Global plan towards the elimination of new HIV infections among children by 2015 and keeping their mothers alive. Geneva: UNAIDS; 2011.
37. National AIDS and STI Control Programme.
Guidelines for prevention of mother to child 6 transmission (PMTCT) of HIV/AIDS in Kenya. Nairobi: Ministry of Health; 2012.
38. Sansom SL, Jamieson DJ, Farnham PG, Bulterys M, Fowler MG.
Human immunodeficiency virus retesting during pregnancy: costs and effectiveness in preventing perinatal transmission.
Obstet Gynecol 2003; 102:782–790.
39. Williams B, Costello M, McHugh E, Le Prevost M, Phil-Ebosie A, Tilsed C, et al..
Repeat antenatal HIV testing in the third trimester: a study of feasibility and maternal uptake rates.
HIV Med 2014; 15:362–366.
40. Criniti SM, Aaron E, Levine AB.
Using the rapid HIV test to rescreen women in the third trimester of pregnancy.
J Midwifery Women's Health 2009; 54:492–496.
41. Chu C1 PAS.
Diagnosis and initial management of acute HIV infection.
Am Fam Phys 2010; 81:1239–1244.
42. Hecht FM, Busch MP, Rawal B, Webb M, Rosenberg E, Swanson M, et al.
Use of laboratory tests and clinical symptoms for identification of primary HIV infection.
AIDS 2002; 16:1119–1129.
43. Busch MP, Satten GA.
Time course of viremia and antibody seroconversion following human immunodeficiency virus exposure.
Am J Med 1997; 102:117–124.
44. Morandi P-A, Schockmel GA, Yerly S, Burgisser P, Erb P, Matter L, et al.
Detection of human immunodeficiency virus type 1 (HIV-1) RNA in pools of sera negative for antibodies to HIV-1 and HIV-2.
J Clin Microbiol 1998; 36:1534–1538.
45. Quinn TC, Brookmeyer R, Kline R, Shepherd M, Paranjape R, Mehendale S, et al.
Feasibility of pooling sera for HIV-1 viral RNA to diagnose acute primary HIV-1 infection and estimate HIV incidence.
AIDS 2000; 14:2751–2757.
46. Pilcher CD, McPherson J, Leone PA, Smurzynski M, Owen-O’Dowd J, Peace-Brewer AL, et al.
Real-time, universal screening for acute HIV infection in a routine HIV counseling and testing population.
JAMA 2002; 288:216–221.
47. Baeten JM, Richardson BA, Martin HL Jr, Nyange PM, Lavreys L, Ngugi EN, et al.
Trends in HIV-1 incidence in a cohort of prostitutes in Kenya: implications for HIV-1 vaccine efficacy trials.
J Acquir Immune Defic Syndr 2000; 24:458–464.
