Background: Effective contraception reduces unintended pregnancies and is a central strategy to reduce vertical HIV-1 transmission for HIV-1-infected women.
Methods: Among 2269 HIV-1-seropositive and 1085-seronegative women from seven African countries who were members of HIV-1-serodiscordant heterosexual partnerships and who were participating in an HIV-1 prevention clinical trial, we assessed pregnancy incidence according to contraceptive method using multivariate Andersen–Gill analysis.
Results: Compared with women using no contraceptive method, pregnancy incidence was significantly reduced among HIV-1-seropositive and HIV-1-seronegative women using injectable contraception [adjusted hazard ratio (aHR) 0.24, P = 0.001 and aHR 0.25, P < 0.001, respectively). Oral contraceptives significantly reduced pregnancy risk only among HIV-1-seropositive women (aHR 0.51, P = 0.004) but not seronegative women (aHR 0.64, P = 0.3), and, for both seropositive and seronegative women, oral contraceptive pill users were more likely to become pregnant than injectable contraceptive users (aHR 2.22, P = 0.01 for HIV-1-seropositive women and aHR 2.65, P = 0.09 for HIV-1-seronegative women). Condoms, when reported as being used as the primary contraceptive method, marginally reduced pregnancy incidence (aHR 0.85, P = 0.1 for seropositive women and aHR 0.67, P = 0.02 for seronegative women). There were no pregnancies among women using intrauterine devices, implantable methods or who had undergone surgical sterilization, although these methods were used relatively infrequently.
Conclusion: Family planning programs and HIV-1 prevention trials need innovative ways to motivate uptake and sustained use of longer acting, less user-dependent contraception for women who do not desire pregnancy.
aDepartment of Obstetrics and Gynaecology, Kenyatta National Hospital
bInstitute of Tropical Medicine and Infectious Diseases, Jomo Kenyatta University, Nairobi, Kenya
cDepartment of Epidemiology
dDepartment of Global Health
eDepartment of Medicine, University of Washington, Seattle, USA
fDepartment of Obstetrics and Gynaecology, University of Nairobi, Nairobi, Kenya
gDepartment of Obstetrics, Gynecology and Reproductive Sciences, University of California San Francisco, San Francisco, USA
hKenya Medical Research Institute, Nairobi, Kenya
iWits Reproductive Health and HIV Institute, University of the Witwatersrand, Johannesburg, South Africa
jDepartment of Reproductive Health, Moi University, Eldoret, Kenya.
Correspondence to Jared M. Baeten, University of Washington, 325 Ninth Avenue, Box 359927, Seattle, WA 98104, USA. Tel: +1 206 520 3808; e-mail: email@example.com
Received 15 June, 2011
Revised 4 November, 2011
Accepted 23 November, 2011
Pregnancy rates are high in sub-Saharan Africa  and almost one-quarter of pregnancies are estimated to be unintended . Reducing unintended pregnancies among HIV-1-seropositive women is a core strategy to reduce vertical HIV-1 transmission . Moreover, pregnancy may increase HIV-1 risk, further emphasizing the benefits of preventing unintended pregnancies [4,5].
For clinical trials of new HIV-1 prevention strategies, such as antiretroviral preexposure prophylaxis, contraceptive use minimizes protocol-required discontinuation of study products during pregnancy and fetal exposure to investigational agents . Indeed, some recent HIV-1 prevention trials have required participants to use an effective contraceptive method in order to qualify for enrollment .
In previous studies among women from Africa, factors related to incident pregnancy have included younger age, fewer children, cohabitation with a male partner and having frequent sex unprotected by condoms [8,9]. Few studies have explored whether the choice of contraceptive method impacts pregnancy incidence among HIV-1-seropositive and HIV-1-seronegative African women. Among women participating in an HIV-1 prevention trial, all of whom were in an HIV-1-serodiscordant partnership (i.e. one partner was HIV-1 infected and the other was HIV-1 uninfected), we sought to determine if pregnancy incidence differed by contraceptive method.
Population and procedures
Between December 2004 and October 2008, 3408 heterosexual HIV-1-serodiscordant couples were enrolled in the Partners in Prevention HSV/HIV Transmission Study, a randomized, double-blind, placebo-controlled clinical trial of daily acyclovir herpes simplex virus type 2 (HSV-2) suppressive therapy provided to HIV-1-seropositive members of couples to reduce HIV-1 transmission risk. The study was conducted at 14 sites in seven countries in eastern and southern Africa (Botswana, Kenya, Rwanda, South Africa, Tanzania, Uganda and Zambia); acyclovir did not decrease HIV-1 transmission risk . Couples were recruited through community outreach activities and referrals from HIV-1 testing and care organizations . This report is a secondary analysis using data from this prospective study.
Eligible couples were at least 18 years of age, reported more than three episodes of vaginal intercourse during the 3 months prior to screening, and intended to remain as a couple. HIV-1-infected partners were seropositive for HSV-2, had a CD4 count of at least 250 cells/μl and were not eligible at enrolment for antiretroviral therapy (ART), according to national guidelines. Couples were followed for up to 24 months, with HIV-1 seropositive partners seen monthly and HIV-1 seronegative partners seen quarterly and were encouraged to come together during the quarterly visit for couples’ risk reduction counseling. HIV-1-infected partners who became eligible for ART initiation during follow-up were actively referred to local HIV-1 care clinics.
Women were not required to use contraception, although HIV-1 seropositive women (who received study medication) could not be pregnant at study enrollment and were encouraged to delay pregnancy until after the study. Current contraceptive method, by self-report, was recorded at each study visit. Contraception was provided on-site or by referral . For hormonal contraceptive methods, specific brand names were not recorded, although combined low-dose oral contraceptives and injectable depot medroxyprogesterone acetate were the most common contraceptive types used in national family planning programs at the time of the study. Urine pregnancy tests were performed quarterly for HIV-1-seropositive women; for HIV-1-seronegative women, pregnancy tests were done based on report of missed menses. HIV-1 seropositive women who became pregnant discontinued study medication for the duration of pregnancy and were referred for prevention of mother-to-child HIV-1 transmission services, and HIV-1-seronegative women who became pregnant were referred for antenatal services. All participants provided written informed consent. The institutional review boards at the University of Washington and all study sites approved the study protocol.
We used multivariate Andersen–Gill modeling for recurrent events (with Efron methods for ties ) to assess the risk of pregnancy among women using different contraceptive methods versus women not using any contraception, analyzed as a time-dependent variable. Adjusted models included socio-demographic and clinical factors. Observations following pregnancy discovery at the study site were censored; observations following pregnancy termination remained in the dataset, as women returned to being ‘at risk’ for another pregnancy. For women experiencing multiple pregnancies, the same censoring rules were used for observations during the second or third pregnancy. For HIV-1-seronegative women, observations were censored following HIV-1 seroconversion. We tested whether the effectiveness of each contraceptive method differed between HIV-1-seropositive and HIV-1-seronegative women using a likelihood ratio test. Analyses were conducted using SAS 9.2 (SAS Institute, Cary, North Carolina, USA).
After excluding 27 women who were determined ineligible for the trial based on confirmatory HIV-1 and HSV-2 testing completed after study completion and 27 women without follow-up data, 3354 women remained for analysis: 2269 (67.7%) were HIV-1-seropositive and 1085 (32.3%) were seronegative. The median age was 30 years and women had one to two children on average. Women reported a median of 4 [interquartile range (IQR) 2–8] sex acts per month (Table 1).
Pregnancy rates during follow-up
Median follow-up for HIV-1-seropositive and HIV-1-seronegative women was 1.64 (IQR 1.21–1.97) and 1.68 (IQR 1.25–2.00) years, respectively. There were 500 pregnancies in 3197.2 woman-years of follow-up among HIV-1-seropositive women and 226 pregnancies in 1542.6 woman-years of follow-up among HIV-1-seronegative women (incidence rates of 15.6 and 14.8 per 100 woman-years, respectively).
Contraceptive use and pregnancy during follow-up
Of the participants, 83.3 and 80.5% of HIV-1-seropositive and HIV-1-seronegative women reported using condoms as contraceptive method at least once during the study, 12.6 and 9.0% ever used oral contraceptives and 33.0 and 20.3% ever used injectable contraception. However, 67.0% of HIV-1-seropositive and 63.5% of HIV-1-seronegative women reported no contraceptive method during at least 1 study visit.
Among women not using contraception, pregnancy incidence was 21.0 and 26.8 per 100 woman-years for HIV-1-seropositive and HIV-1-seronegative women (Table 2). Pregnancy rates were lower among women using condoms only (18.4 and 11.4 per 100 woman-years for seropositive and seronegative women, respectively), oral contraception (15.6 and 14.4 per 100 woman-years), and injectable contraception (6.8 and 6.0 per 100 woman-years). There were no pregnancies observed among women using intrauterine devices (IUDs), implantable methods or who had undergone surgical sterilization, but very few women used these methods.
In multivariate analysis, both HIV-1-seropositive and HIV-1-seronegative women were substantially less likely to become pregnant if they used injectable contraception compared to women who used no contraception [adjusted hazard ratio (aHR) 0.25, P < 0.001 and aHR 0.24, P = 0.001, respectively). Condoms and oral contraceptives also reduced the incidence of pregnancy, but the effect of oral contraceptives was not statistically different from the effect of no contraception among HIV-1-seronegative women (aHR 0.64, P = 0.3) and the effect of condoms was not statistically different from the effect of no contraception among HIV-1-seropositive women (aHR 0.85, P = 0.1). The extent to which condoms prevented pregnancy among HIV-1-seropositive (aHR 0.85, P = 0.1) and HIV-1-seronegative women (aHR 0.67, P = 0.02) was statistically different (P value for interaction=0.002), but the effectiveness of injectables and oral contraceptives was not different (P value for interaction=0.15 and 0.1, respectively). When compared with those using injectable contraception, women were more likely to become pregnant if they used oral contraceptives (aHR 2.06, P = 0.01 for HIV-1-seropositive women and aHR 2.71, P = 0.09 for HIV-1-seronegative women) or condoms only (aHR 3.39, P < 0.001 for HIV-1-seropositive women and aHR 2.82, P = 0.01 for HIV-1-seronegative women).
In the multivariate models, other factors independently associated with incident pregnancy included being younger (aHR 5.08, P < 0.001 for HIV-1-seropositive women and aHR 4.48, P < 0.001 for HIV-1-seronegative women for those ≤24 years of age and aHR 3.16, P < 0.001 and aHR 3.16, P < 0.001 for those 25–34 years when compared with women >35 years old), having at least two children (aHR 1.96, P < 0.001 and aHR 2.49, P = 0.001), having a greater number of sex acts (aHR 1.81–2.32 for categories >0/month as defined in footnote to Table 2; not statistically significant for HIV-1-seronegative women), reporting unprotected sex with the study partner (independent of condom use for contraception, aHR 2.30, P < 0.001 and aHR 5.89, P < 0.001) and having sex with a partner other than the study partner (aHR 2.24, P < 0.001 and aHR 3.26, P < 0.001). Having a curable sexually transmitted infection at the time of study enrollment and ART use by the HIV-1-seropositive member of the couple during study follow-up were not associated with pregnancy.
In this large, prospective, multinational cohort of HIV-1-seropositive and HIV-1-seronegative African women in known heterosexual HIV-1 serodiscordant partnerships, women using injectable contraceptives had significantly lower pregnancy rates than oral contraceptive users. These results are consistent with previous studies reporting that injectable methods are more effective than oral methods in reducing the likelihood of pregnancy [8,9,14]. We also observed high rates of pregnancy for both HIV-1-seropositive and HIV-1-seronegative women (incidence 15.6 and 14.8 per 100 woman-years, respectively) – similar to rates seen in other HIV-1 prevention trials in sub-Saharan Africa [8,9,15]. These high rates of pregnancy have the potential to compromise the power of biomedical HIV-1 prevention trials involving women, as those who become pregnant often are required to discontinue study products.
Notably, we found that no women using IUDs and implantable contraceptives experienced pregnancy, although the number using these methods was small. Previous studies have reported that IUDs are well tolerated and effective for preventing pregnancy among women with and at risk for HIV-1, yet their use is still very low in sub-Saharan Africa [16–18]. Our data provide additional evidence of the potential for well tolerated and effective long-acting user-independent contraceptive methods to reduce unintended pregnancies among HIV-1-seronegative and HIV-1-seropositive African women.
We have previously reported higher rates of condom use and lower rates of incident HIV-1 infection in this cohort (2.7 cases per 100 person-years overall)  compared with similar cohorts in this geographical region [19,20]. In our cohort, condoms alone were not as effective for pregnancy prevention as injectable methods and their effectiveness differed based on HIV-1 status. Qualitative work we have done among HIV-1-serodiscordant couples has described challenges with consistent and correct condom use, with seronegative women reported more consistent condom use to prevent HIV-1 acquisition than seronegative men . As family planning programs consider novel ways to increase access to and uptake of long-acting user-independent methods, it is also important to continue counseling women and men about the importance of condom use for protection against HIV-1 and sexually transmitted infections in conjunction with hormonal contraceptive use for pregnancy prevention [22,23]. In clinical trial settings, efforts to link participants with contraceptive counseling and resources in the community will foster sustained use of contraceptive methods following study closure.
In conclusion, in this study of more than 3300 African women in stable heterosexual partnerships, injectable methods were the most effective contraceptive method for pregnancy prevention. Family planning programs and HIV-1 prevention trials need to develop innovative ways to motivate use of longer-acting, less user-dependent contraception for women with or at risk of HIV-1 who do not desire pregnancy and ensure that couples understand the importance of using dual methods for HIV-1 and pregnancy prevention.
The authors thank the couples who participated in this study, the teams at the study sites and at the University of Washington for work on data and sample collection and management.
K.N., R.H. and J.M.B. conceived the study and wrote the first draft of the manuscript. R.H. and J.M.B. performed the statistical analysis. N.R.M., C.C., C.R.C., J.O., H.R., J.K. and E.W. contributed critical revisions to the analysis and interpretation. All authors contributed to the writing of the final draft.
Partners in Prevention HSV/HIV Transmission Study Team:
University of Washington Coordinating Center and Central Laboratories, Seattle, USA: Connie Celum (principal investigator), Anna Wald (protocol co-chair), Jairam Lingappa (medical director), Jared M. Baeten, Mary Campbell, Lawrence Corey, Robert W. Coombs, James P. Hughes, Amalia Magaret, M. Juliana McElrath, Rhoda Morrow, James I. Mullins.
Study sites and site principal investigators: Cape Town, South Africa (University of Cape Town): David Coetzee; Eldoret, Kenya (Moi University, Indiana University): Kenneth Fife, Edwin Were; Gaborone, Botswana (Botswana Harvard Partnership): Max Essex, Joseph Makhema; Kampala, Uganda (Infectious Disease Institute, Makerere University): Elly Katabira, Allan Ronald; Kigali, Rwanda (Rwanda Zambia HIV Research Group, and Emory University): Susan Allen, Kayitesi Kayitenkore, Etienne Karita; Kisumu, Kenya (Kenya Medical Research Institute, University of California San Francisco): Elizabeth Bukusi, Craig Cohen; Kitwe, Zambia (Rwanda Zambia HIV Research Group, and Emory University): Susan Allen, William Kanweka; Lusaka, Zambia (Rwanda Zambia HIV Research Group, and Emory University): Susan Allen, Bellington Vwalika; Moshi, Tanzania (Kilimanjaro Christian Medical College, Harvard University): Saidi Kapiga, Rachel Manongi; Nairobi, Kenya (University of Nairobi, University of Washington): Carey Farquhar, Grace John-Stewart, James Kiarie; Ndola, Zambia (Rwanda Zambia HIV Research Group, and Emory University): Susan Allen, Mubiana Inambao; Orange Farm, South Africa (Reproductive Health Research Unit, University of the Witwatersrand): Sinead Delany-Moretlwe, Helen Rees; Soweto, South Africa (Perinatal HIV Research Unit, University of the Witwatersrand): Guy de Bruyn, Glenda Gray, James McIntyre; Thika, Kenya (University of Nairobi, University of Washington): Nelly Rwamba Mugo.
Data management was provided by DF/Net Research Inc. (Seattle, USA) and site laboratory oversight was provided by Contract Lab Services (University of the Witwatersrand, Johannesburg, South Africa).
The study is supported by grants from National Institutes of Health (R03 HD068143) and the Bill and Melinda Gates Foundation (grant 26469).
Conflicts of interest
C.C. has received research grant support from GlaxoSmithKline (GSK), which did not include salary support, and has served on an advisory board for GSK and Merck.
1. Sneeringer SE. Fertility transition in sub-Saharan Africa: a comparative analysis of cohort trends in 30 countries
. DHS Comparative Reports No. 23. 2009. Calverton, Maryland: ICF Macro.
2. Hubacher D, Mavranezouli I, McGinn E. Unintended pregnancy in sub-Saharan Africa: magnitude of the problem and potential role of contraceptive implants to alleviate it
3. Wilcher R, Petruney T, Reynolds HW, Cates W. From effectiveness to impact: contraception as an HIV prevention intervention
. Sex Transm Infect
2008; 84 (Suppl 2):ii54–ii60.
4. Mugo NR, Heffron R, Donnell D, Wald A, Were E, Rees H, et al. Increased risk of HIV-1 transmission in pregnancy: a prospective study among African HIV-1-serodiscordant couples
5. Gray RH, 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
6. Institute of Medicine (IOM). Methodological challenges in biomedical HIV prevention trials
. Washington, District of Columbia: The National Academies Press; 2008.
8. Halpern V, Lie C, Feldblum P, Damme L. Predictors of pregnancy in microbicides trials
9. Reid SE, Dai JY, Wang J, Sichalwe BN, Akpomiemie G, Cowan FM, et al. Pregnancy, contraceptive use and HIV acquisition in HPTN039: relevance for HIV prevention trials among African women
. J Acquir Immune Defic Syndr
10. Celum C, Wald A, Lingappa JR, Magaret AS, Wang RS, Mugo N, et al. Acyclovir and transmission of HIV-1 from persons infected with HIV-1 and HSV-2
. N Engl J Med
11. Lingappa JR, Kahle E, Mugo N, Mujugira A, Magaret A, Baeten J, et al. Characteristics of HIV-1 discordant couples enrolled in a trial of HSV-2 suppression to reduce HIV-1 transmission: the Partners Study
. PLoS ONE
2009; 4 Baeten J, Bukusi:e5272.
12. Ngure K, Heffron R, Mugo N, Irungu E, Celum C, Baeten JM, et al. Successful increase in contraceptive uptake among Kenyan HIV-1-serodiscordant couples enrolled in an HIV-1 prevention trial
2009; 23 (Suppl 1):S89–S95.
13. Efron B. Efficiency of Cox's likelihood function for censored data
. J Am Statist Assoc
15. Guthrie B, Gatuguta A, Bosire R, Kiarie JN, Mackelprang RD, Gatuguta A, et al. Predicting pregnancy among HIV-1 discordant couples
. AIDS Behav
16. Clifton D, Kaneda T, Ashford L. Family planning worldwide: 2008 data sheet
. Washington, District of Columbia: Population Reference Bureau; 2008.
17. Stringer EM, Kaseba C, Levy J, Sinkala M, Goldenberg RL, Chi BH, et al. A randomized trial of the intrauterine contraceptive device vs hormonal contraception in women who are infected with the human immunodeficiency virus
. Am J Obstet Gynecol
18. Gutin SA, Mlobeli R, Moss M, Buga G, Morroni C. Survey of knowledge, attitudes and practices surrounding the intrauterine device in South Africa
19. Quinn TC, Wawer MJ, Sewankambo N, Serwadda D, Li C, Wabwire-Mangen F, et al. Viral load and heterosexual transmission of human immunodeficiency virus type 1
. N Engl J Med
20. Fideli US, Allen SA, Musonda R, Trask S, Hahn BH, Weiss H, et al. Virologic and immunologic determinants of heterosexual transmission of human immunodeficiency virus type 1 in Africa
. AIDS Res Hum Retroviruses
21. Ngure K, Mugo N, Celum C, Baeten JM, Morris M, Olungah O, et al.A qualitative study of barriers to consistent condom use among HIV-1 serodiscordant couples in Kenya
. AIDS Care
2011 [Epub ahead of print].
22. Heffron R, Donnell D, Rees H, Celum C, Were E, Mugo N, et al.Hormonal contraceptive use and risk of HIV-1 transmission: a prospective cohort analysis
. Lancet Infect Dis
2011. doi: 10.1016/S1473-3099(11)70247-X. [Epub ahead of print].
23. Holmes KK, Levine R, Weaver M. Effectiveness of condomsin preventing sexually transmitted infections
. Bull World Health Organ