Antiretroviral therapy (ART) is critical to reducing morbidity and mortality in people living with HIV (PLHIV) [1,2], and preventing mother-to-child HIV transmission (MTCT) during pregnancy, delivery and breastfeeding [3,4]. With increasing ART access in sub-Saharan Africa (SSA) since the early 2000s, the proportion of pregnant and postpartum women receiving ART has markedly increased, with subsequent declines in perinatal infections . HIV treatment guidelines in Kenya and Uganda have recommended lifelong ART for all pregnant and postpartum women (’Option B+’) since 2012 [6,7], predating current guidelines that recommend ART for all PLHIV. Despite this progress, MTCT rates remain above global targets in SSA, with over half of HIV infections among children in East Africa in 2018 occurring during breastfeeding and approximately one-third of infections attributable to women stopping or being unable to access ART during pregnancy .
Pregnant and postpartum women living with HIV (WLHIV) in SSA face multiple barriers to achieving viral suppression. Inadequate access to antenatal care (ANC) due to distance, cost or lack of education regarding benefits of ANC results in missed opportunities for HIV testing and early diagnosis . Many women who access ANC only attend one ANC visit during pregnancy, resulting in missed opportunities to identify women who seroconvert after an initial antenatal visit . Once diagnosed with HIV, pregnant/postpartum women face challenges with ART access and adherence distinct from other PLHIV. For some, these barriers include fear and stigma when coping with a new HIV diagnosis alongside a new awareness of pregnancy – particularly if unintended – creating the challenge of dual disclosure to partners and family members . Once on ART, adherence may be limited by symptoms during pregnancy, side effects associated with antiretrovirals and fear of stigma, particularly in situations wherein women have not disclosed their HIV status or are facing interpersonal violence or lack of support from partners [8,11]. Postpartum, transitioning from antenatal to HIV clinics can be stigmatizing and has been associated with loss to follow-up . Competing priorities, including caring for a new child, and reduced concern regarding MTCT following an initial negative infant test, may contribute to increased ART nonadherence during breastfeeding compared with pregnancy . These distinct barriers occur in addition to those faced by PLHIV in SSA, such as long distances to and waiting times at clinics, low perceived HIV risk and costs in accessing services [14,15].
Several recent universal HIV ‘test and treat’ (UTT) trials in SSA have tested the effects of population-wide HIV testing with universal ART on HIV care cascade outcomes, morbidity, mortality and incidence [16–19]. The Sustainable East Africa Research in Community Health (SEARCH) UTT trial compared an intervention of annual population-wide HIV testing with universal ART with a control of baseline population-wide HIV testing with ART by country standard, including ART eligibility for pregnant/postpartum women, in communities in Kenya and Uganda . SEARCH's universal testing intervention was designed to increase knowledge of HIV status for all and re-engage PLHIV who knew their status but were out of care . SEARCH's universal treatment intervention was designed to increase viral suppression among all PLHIV by reducing barriers to ART . As such, the SEARCH intervention, in addition to increasing viral suppression among high-CD4+ cell count individuals, offered the potential to improve the prevalence of viral suppression among pregnant/postpartum women, through frequent, community-based testing and re-engagement, and by reducing barriers to ART for all women of reproductive age. The trial's primary results have been published . In this secondary analysis, we sought to determine whether the SEARCH intervention increased population-level viral suppression among pregnant/postpartum women, above and beyond baseline population-wide testing and the ‘Option B+’ strategy of universal ART eligibility for pregnant /postpartum women offered in control communities throughout the trial.
Materials and methods
The SEARCH UTT trial (NCT:01864603) was a community cluster randomized trial that compared annual population-wide, HIV testing at multidisease health fairs with universal ART eligibility via patient-centred care (intervention) to baseline population-wide HIV testing at multidisease health fairs with national-guideline based ART eligibility (control), in 32 rural communities in Kenya and Uganda over three years. The trial was conducted from 2013 to 2017, and the trial methods and primary outcome results have been published . In brief, study communities underwent pair-matched randomization at baseline (2013–2014), with 16 communities randomized to intervention and 16 to control. We conducted door-to-door census enumeration of community residents at baseline, followed by 2-week health fairs that offered universal HIV testing integrated with multidisease services. Adult (≥15 years) residents who did not attend fairs were offered HIV testing at home, or locations of their choice .
In SEARCH intervention communities, population-wide testing occurred annually, and all PLHIV were given appointments to link to care for ART within 7 days of testing HIV positive, or within 48 h if pregnant or breastfeeding by self-report. All PLHIV were offered ART regardless of CD4+ cell count, using a patient-centred, streamlined approach that included 3-month visit intervals for stable patients, reduced waiting time and welcoming staff, mobile phone access to providers and flexible hours for accessing care . This approach also included clinician-guided transitions between HIV clinics and ANC for pregnant WLHIV, and the option to continue receiving ART at HIV clinics while attending ANC for women already engaged in HIV care. In control communities, population-wide HIV testing was performed at baseline and after 3 years, and PLHIV were offered ART according to country guidelines, which changed over the course of the study, ultimately expanding to ART for all PLHIV in Kenya and Uganda . However, Kenya and Uganda Ministry of Health guidelines recommended lifelong ART initiation for all HIV-infected pregnant women (Option B+) throughout the trial [6,7]. Therefore, in both intervention and control communities, we offered rapid referral with ART initiation within 48 h of HIV diagnosis for all pregnant/breastfeeding women during population-wide testing. During each round of population-level testing, all women aged 15–45 years were asked about current pregnancy (at time of testing) or any live births over the prior year, and WLHIV had plasma HIV viral load measured. We defined HIV testing coverage at baseline and at year 3 of the trial as the proportion of women of reproductive age who participated in HIV testing services provided by SEARCH (i.e. health fairs and home-based testing) at baseline and year 3, respectively.
In this post-hoc analysis, we first sought to determine whether the SEARCH UTT intervention resulted in increased prevalence of viral suppression compared with control among WLHIV of reproductive age who reported current pregnancy or live birth over the prior year at trial completion (year 3). Although these women were ART eligible in both trial arms, we hypothesized that viral suppression would be higher among pregnant/postpartum women in intervention than control due to annual, population-wide testing with rapid linkage to streamlined care. Second, we compared population-viral suppression at year 3 among all women of reproductive age (15–45 years) with HIV in intervention and control communities and third, evaluated predictors of viral suppression among these women by arm. For this latter analysis, we hypothesized that pregnancy or live birth over prior year would increase the probability of viral suppression in control communities, but not intervention communities wherein all HIV-infected women were ART eligible, regardless of current or prior pregnancy.
Our primary outcome was the population-level proportion of pregnant or postpartum WLHIV who were virally suppressed (HIV RNA <500 copies/ml) at trial completion. In each community separately, we first estimated the following HIV care cascade outcomes among female residents (inclusive of in-migrants), aged 15–45 years who reported a current pregnancy or live birth in the last year: proportion of PLHIV who knew their status, proportion of PLHIV with known status who had initiated ART, proportion of PLHIV with ART use who were virally suppressed and proportion of all PLHIV who were virally suppressed . Targeted maximum likelihood estimation (TMLE) was used to adjust for differences in the characteristics of women with known versus unknown HIV status, and known versus missing viral suppression status, as previously described [17,23]; TMLE incorporates machine learning to avoid model misspecification bias and offers efficiency gains over alternative approaches, such as inverse-weighting . In secondary analyses, we calculated unadjusted cascade estimates among women with known HIV status and measured viral load. We obtained estimates of cascade outcomes and population-level suppression at baseline and year 3 in both arms, and at years 1 and 2 in the intervention arm. Universal testing was not conducted at years 1 and 2 in the control arm.
We compared year 3 estimates by arm with community-level TMLE, accounting for the matched design and with data-adaptive selection of adjustment variables, as previously described [17,25]; prespecified candidates were limited to baseline suppression and proportion of young women (age 15–24 years) to avoid over-fitting in analyses with 16 matched pairs of communities. We repeated these analyses among all women of reproductive age (15–45 years). Finally, we used TMLE to assess arm-specific predictors of viral suppression among all women of reproductive age living with HIV at year 3. Further details are available in the Supplementary Materials, https://links.lww.com/QAD/B771.
We obtained approval from the ethics committee of the University of California, San Francisco Committee on Human Research, Kenya Medical Research Institute Ethical Review Committee, Ugandan National Council on Science and Technology and Makerere University School of Medicine Research and Ethics Committee in Uganda. Verbal consent was obtained at enrolment; written consent was obtained for persons in the intervention arm receiving ART not yet indicated by country guidelines.
Testing, HIV prevalence and pregnancy/live birth at trial baseline
Of 150 395 adult (≥15 years) residents enumerated in 32 communities at baseline, 62 066 (41%) were women of reproductive age (15–45 years): 32 954 women in intervention and 29 112 women in control communities (Table 1). Among enumerated women of reproductive age at baseline, SEARCH achieved 91% (30 074/32 954) and 92% (26 895/29 112) HIV testing coverage in intervention and control communities, respectively. Baseline HIV prevalence among women who tested varied by region: 23.8% in Western Kenya, 7.7% in Western Uganda and 3.7% in Eastern Uganda, with an overall baseline HIV prevalence of 12.4% (7047/56 969) among women who tested. At the time of baseline HIV testing, 93% (57 813/62 066) of all women of reproductive age and 99% (6995/7047) of HIV-positive women responded to the following pregnancy/live birth questions: ‘are you pregnant now?’ and ‘how many live births have you had in the past year?’ Among HIV-positive women aged 15–45 years, 33% (1252/3741) in intervention communities and 33% (1081/3254) in control communities reported a current pregnancy or at least one live birth in the prior year at baseline.
Testing, HIV prevalence and pregnancy/live birth at year 3 of the trial
During year 3 of the trial, there were 77 862 women age 15–45 years enumerated in the 32 study communities, inclusive of in-migrants and young women (12–14 years old at baseline) who turned 15 years during follow-up and excluding women who had aged out, died or out-migrated: 41 598 women in intervention and 36 264 women in control communities. Among these women, SEARCH achieved 80% (33 326/41 598) and 84% (30 282/36 264) HIV testing coverage in intervention and control communities, respectively, at year 3 of the trial. Year 3 HIV prevalence among women age 15–45 years who tested was 10.6% (3540/33 326) and 10.2% (3086/30 282) in intervention and control communities, respectively. At the time of year 3 testing, 81% (63 413/77 862) of all women of reproductive age and 99% (6557/6626) of HIV-positive women responded to the pregnancy/live birth questions. Among HIV-positive women, 14% (481/3505) reported a current pregnancy or a live birth in the prior year at year 3 of the trial in intervention communities, compared with 16% (487/3052) in control communities.
Prevalence of viral suppression among pregnant/postpartum women
Among 15 to 45-year-old WLHIV who reported current pregnancy or live birth in the prior year at baseline, HIV viral suppression was 42% [95% confidence interval (CI): 36–47] in intervention, and 44% (95% CI: 41–47) in control communities, after adjusting for missingness in HIV status and viral load (Fig. 1). During subsequent annual rounds of offering universal HIV testing in intervention communities only, viral suppression estimates among pregnant/postpartum women were 77% (95% CI: 73–80) and 80% (95% CI: 77–84) in follow-up years 1 and 2, respectively.
Among 15 to 45-year-old WLHIV reporting a current pregnancy or live birth in the prior year during year 3 of the trial, 95% (95% CI: 92–98) and 91% (95% CI: 89–93) knew their HIV status in intervention and control communities, respectively. Of those with a prior diagnosis of HIV, 98% (95% CI: 97–99) and 95% (95% CI: 94–97) were on ART in intervention and control communities, respectively. Of those on ART, 88% (95% CI: 85–91) and 87% (95% CI: 84–91) were virally suppressed, respectively. At year 3, overall population-level prevalence of viral suppression among 15 to 45-year-old women reporting a current pregnancy or live birth in the prior year had increased to 81% (95% CI: 78–84) in intervention compared with 76% (95% CI: 72–80) in control communities (Fig. 1). At year 3, population-level prevalence of viral suppression among women reporting current pregnancy or live birth in the prior year was 6% higher in intervention versus control communities (adjusted relative prevalence: 1.06, 95% CI: 1.01–1.12; P = 0.02).
Prevalence of viral suppression among women of reproductive age
In comparison, population-level viral suppression estimates for all 15 to 45-year-old WLHIV (regardless of reported pregnancy or live birth) at year 3 was 77% (95% CI: 74–80) in intervention versus 68% (95% CI: 66–70) in control communities (adjusted relative prevalence: 1.13, 95% CI: 1.08–1.19; P < 0.001).
Comparison of predictors of viral suppression by SEARCH trial arm
When determining predictors of viral suppression at year 3 among 15 to 45-year-old WLHIV, viral suppression at year 3 was not associated with self-reported current pregnancy or live birth in the prior year in intervention communities (aRR: 1.01, ref: no pregnancy/live birth, 95% CI: 0.98–1.05, P = 0.43), whereas in control communities, women with a current pregnancy or live birth in the prior year were more likely to have viral suppression at year 3 than women without current pregnancy or live birth in prior year (aRR: 1.06, 95% CI: 1.01–1.12, P = 0.016). In both intervention and control communities, older age consistently predicted viral suppression at year 3 compared with age 15–19 years (Fig. 2).
The SEARCH UTT intervention resulted in a significantly higher prevalence of HIV viral suppression after 3 years among HIV-infected pregnant and recently postpartum women in intervention communities that offered repeat out-of-facility testing with rapid linkage to streamlined HIV care compared with standard clinical care in control communities, following baseline universal HIV testing in all communities. Although the absolute difference in prevalence of viral suppression in intervention versus control was relatively modest (6%), these gains occurred at a population level, in a group of PLHIV (pregnant/postpartum women) who were ART-eligible throughout the trial, and in communities (in both arms) that achieved prevalence of viral suppression in pregnant/postpartum women beyond the ‘73%’ threshold of the UNAIDS ‘90–90–90’ targets for 2020 . The ‘90–90–90’ targets aim to have 90% of PLHIV know their HIV status, 90% of PLHIV who know their status receive ART and 90% of those on ART achieve viral suppression: resulting in a goal that at least 73% of all PLHIV are virally suppressed. In a post 90–90–90 context, modest absolute gains in viral suppression will likely require testing and treating some of the hardest-to-reach subgroups and optimizing each ‘90’. Our results provide one of the first estimates of population-level prevalence of viral suppression among pregnant/postpartum women, independent of ANC access, in Kenya and Uganda. These findings add to the evidence demonstrating population-level health benefits conferred by UTT interventions, including reductions in HIV-associated mortality, tuberculosis and incidence [16,17,19].
There are several potential explanations for how the SEARCH intervention achieved a higher prevalence of viral suppression among peripartum women than control. First, annual, out-of-facility, population-wide HIV testing provided a platform to diagnose new infections that had occurred over the prior year, identify new in-migrants to the community who had not been diagnosed previously and re-engage PLHIV who had stopped ART. Community-based testing can also reach women who do not access ANC, deliver at home or do not engage in HIV testing while pregnant/postpartum. Thus, the proportion of peripartum women in year 3 of the trial who knew their HIV status was higher in intervention (95%) than control (91%) communities, even though in both, knowledge of HIV status exceeded 90%. Second, referrals to HIV care following annual testing prioritized pregnant and breastfeeding women by making appointments and starting ART within 48 h of testing positive, introducing women to clinic staff at testing sites, and closely following up women who missed appointments. As a result, the proportion of pregnant/postpartum women on ART was higher in intervention than control, even though in both arms this proportion was very high (98 and 95%, respectively). Third, the SEARCH intervention's streamlined model of care sought to reduce barriers to care, by reducing frequency of clinic visits and waiting times, providing a welcoming environment with flexible appointment scheduling and mobile phone access to clinicians, and actively tracking those lost to follow-up. For pregnant/postpartum women, particularly those caring for an infant, reducing these barriers to care and easing transitions between ANC and HIV clinics may explain the higher proportion of women on ART and suppressed in intervention communities. Finally, universal ART eligibility in intervention communities resulted in higher prevalence of suppression among women of reproductive age, providing an opportunity for viral suppression preconception, which also likely contributed to the higher prevalence of viral suppression in pregnant/postpartum women compared to control.
At present, as universal ART eligibility is the global standard and differentiated models of care are being widely adopted across SSA, the need for rapid, universal (i.e. nontargeted) testing initiatives and how often such initiatives may be needed remains under discussion. The higher knowledge of HIV status and higher levels of viral suppression achieved in pregnant/postpartum women may be potential benefits to implementing intermittent, large-scale HIV testing initiatives, in medium to high-prevalence settings such as rural Uganda and Western Kenya. Even as SEARCH control communities exceeded the UNAIDS ‘90–90–90’ target of 73% among pregnant/breastfeeding women – with 76% viral suppression in this subgroup eligible for ART throughout the trial in control – the SEARCH intervention was able to achieve significantly higher levels of suppression (81%). It remains unclear whether efforts to push viral suppression further among the remaining 20–25% of unsuppressed PLHIV in communities that have surpassed ‘90–90–90’ in SSA will require targeted outreach versus intermittent, nontargeted, population testing, or combination approaches. The higher levels of suppression achieved among pregnant/postpartum women in intervention communities suggest that nontargeted, population testing can ‘move the needle’ in knowledge of HIV status and ART access, even among a group that was eligible for ART and targeted for HIV testing at ANCs throughout the trial.
Our results provide a population-level prevalence of viral suppression among pregnant/postpartum women, and all women of reproductive age, including measures of each step of the HIV care cascade. One advantage of population measures, compared with clinic-based measures, is that they are not conditional on accessing antenatal or postpartum HIV care. Published data on the prevalence of viral suppression in pregnant/postpartum women in SSA are sparse, and largely limited to clinic-based cohorts, with a large range of suppression estimates (30–98%) reported . Our findings highlight ongoing challenges in adherence, even after exceeding the first two ‘90s’, with 88% of pregnant/postpartum women on ART achieving viral suppression in intervention communities. In a recent study from South Africa, the vast majority (>90%) of instances of nonsuppression during pregnancy among women on ART were attributable to nonadherence rather than pretreatment drug resistance , demonstrating the need for adherence support interventions in pregnant/postpartum women. In addition, our finding of lower prevalence of viral suppression in younger women is consistent with prior literature [13,28], and emphasizes the need for specific, and integrated, interventions to support young women with both HIV and reproductive healthcare.
Our study has several limitations. First, classification of current pregnancy/live birth over the prior year was based on self-report. This may have resulted in some misclassification, though it seems unlikely that self-report would be inaccurate for pregnancy (particularly in second and third trimester) or the birth of a child, or that misclassification would have differed by trial arm. Second, incomplete viral load measures among PLHIV may have resulted in overestimates of population viral suppression. However, both HIV testing coverage and viral load measures among PLHIV were high (>80%), and we adjusted for differences between persons with and without viral load measures. Third, this analysis did not include measures of neonatal and childhood HIV incidence; analyses are ongoing to determine whether changes in maternal viral suppression resulted in reduced vertical transmission. However, maternal viral load during pregnancy, delivery and breastfeeding is clearly associated with likelihood of vertical transmission and increasing ART access for pregnant women has resulted in significant declines in the number of new child infections in multiple settings, including Kenya and Uganda .
In conclusion, the SEARCH UTT intervention resulted in significant gains in the prevalence of viral suppression among pregnant and postpartum women, compared with an active control that offered one-time universal testing and ART eligibility for all pregnant/postpartum women. The intervention also increased viral suppression among all women of reproductive age, extending the promise of ART to improve maternal health and further reduce new childhood HIV infections. As countries seek to eliminate MTCT and improve viral suppression among WLHIV, our findings provide insights into how to move closer towards achieving these important goals.
We thank the residents of the 32 SEARCH Trial communities for their generous participation in our study. We also thank the Ministries of Health of Uganda and Kenya, our research and administrative teams in San Francisco, Uganda and Kenya, and collaborators and advisory boards.
J.K., L.L.B. and G.C. contributed to the study design, data analysis and interpretation, literature search, tables, figures and writing (first draft and revising) of the manuscript. H.S., J.A., D.K., M.A., E.D.C., T.D.C., C.A.K., T.R., E.A.R., C.R.C., P.M., M.L.P., D.V.H. and M.R.K. contributed to the study design, data interpretation, literature search and writing (revising) of the manuscript.
This work was supported by grants from the Division of AIDS, NIAID of the National Institutes of Health (NIH) (UM1AI068636 and U01AI099959; D.V.H.), Fogarty International Center of the National Institutes of Health (D43 TW010037; to M.R.K.) and in part by the President's Emergency Plan for AIDS Relief and Gilead Sciences.
Conflicts of interest
The authors have no competing interests.
1. Lundgren JD, Babiker AG, Gordin F, Emery S, Grund B, Sharma S, et al. Initiation of antiretroviral therapy in early asymptomatic HIV infection
. N Engl J Med
2. Group TAS, Danel C, Moh R, Gabillard D, Badje A, Le Carrou J, et al. A trial of early antiretrovirals and isoniazid preventive therapy in Africa
. N Engl J Med
3. Kesho Bora Study G, de Vincenzi I. Triple antiretroviral compared with zidovudine and single-dose nevirapine prophylaxis during pregnancy and breastfeeding for prevention of mother-to-child transmission of HIV-1 (Kesho Bora study): a randomised controlled trial
. Lancet Infect Dis
4. Fowler MG, Qin M, Fiscus SA, Currier JS, Flynn PM, Chipato T, et al. Benefits and risks of antiretroviral therapy for perinatal HIV prevention
. N Engl J Med
5. UNAIDS. UNAIDS global AIDS update: communities at the centre
. Geneva: UNAIDS; 2019.
6. Uganda Ministry of Health. Uganda integrated national guidelines on antiretroviral therapy, prevention of mother to child transmission of HIV, and infant & young child feeding
7. Kenya Ministry of Health. Guidelines for prevention of mother to child transmission (PMTCT) of HIV/AIDS in Kenya
. 4th Edition. National AIDS and STI Control Programme (NASCOP). 2012.
8. Hodgson I, Plummer ML, Konopka SN, Colvin CJ, Jonas E, Albertini J, et al. A systematic review of individual and contextual factors affecting ART initiation, adherence, and retention for HIV-infected pregnant and postpartum women
. PLoS One
9. Benova L, Dennis ML, Lange IL, Campbell OMR, Waiswa P, Haemmerli M, et al. Two decades of antenatal and delivery care in Uganda: a cross-sectional study using Demographic and Health Surveys
. BMC Health Serv Res
10. Crankshaw TL, Voce A, King RL, Giddy J, Sheon NM, Butler LM. Double disclosure bind: complexities of communicating an HIV diagnosis in the context of unintended pregnancy in Durban, South Africa
. AIDS Behav
2014; 18 Suppl 1:S53–S59.
11. Buregyeya E, Naigino R, Mukose A, Makumbi F, Esiru G, Arinaitwe J, et al. Facilitators and barriers to uptake and adherence to lifelong antiretroviral therapy among HIV infected pregnant women in Uganda: a qualitative study
. BMC Pregnancy Childbirth
12. Phillips T, Thebus E, Bekker LG, McIntyre J, Abrams EJ, Myer L. Disengagement of HIV-positive pregnant and postpartum women from antiretroviral therapy services: a cohort study
. J Int AIDS Soc
13. Phillips TK, Myer L. Shifting to the long view: engagement of pregnant and postpartum women living with HIV in lifelong antiretroviral therapy services
. Expert Rev Anti Infect Ther
14. Njau B, Covin C, Lisasi E, Damian D, Mushi D, Boulle A, et al. A systematic review of qualitative evidence on factors enabling and deterring uptake of HIV self-testing in Africa
. BMC Public Health
15. Musheke M, Ntalasha H, Gari S, McKenzie O, Bond V, Martin-Hilber A, et al. A systematic review of qualitative findings on factors enabling and deterring uptake of HIV testing in sub-Saharan Africa
. BMC Public Health
16. Hayes RJ, Donnell D, Floyd S, Mandla N, Bwalya J, Sabapathy K, et al. Effect of universal testing and treatment on HIV incidence - HPTN 071 (PopART)
. N Engl J Med
17. Havlir DV, Balzer LB, Charlebois ED, Clark TD, Kwarisiima D, Ayieko J, et al. HIV testing and treatment with the use of a community health approach in rural Africa
. N Engl J Med
18. Iwuji CC, Orne-Gliemann J, Larmarange J, Balestre E, Thiebaut R, Tanser F, et al. Universal test and treat and the HIV epidemic in rural South Africa: a phase 4, open-label, community cluster randomised trial
. Lancet HIV
19. Makhema J, Wirth KE, Pretorius Holme M, Gaolathe T, Mmalane M, Kadima E, et al. Universal testing, expanded treatment, and incidence of HIV infection in Botswana
. N Engl J Med
20. Chamie G, Clark TD, Kabami J, Kadede K, Ssemmondo E, Steinfeld R, et al. A hybrid mobile approach for population-wide HIV testing in rural east Africa: an observational study
. Lancet HIV
21. Kwarisiima D, Kamya MR, Owaraganise A, Mwangwa F, Byonanebye DM, Ayieko J, et al. High rates of viral suppression in adults and children with high CD4+ counts using a streamlined ART delivery model in the SEARCH trial in rural Uganda and Kenya
. J Int AIDS Soc
2017; 20: (Suppl 4): 21673.
22. Joint United Nations Programme on HIV/AIDS (UNAIDS). 90-90-90 An Ambitious Treatment Target to Help End the AIDS Epidemic.; 2014. http://www.unaids.org/en/resources/documents/2014/90-90-90
23. Balzer L, Havlir D, Schwab J, van der Laan M, Petersen M. Statistical analysis plan for search phase I: health outcomes among adults. arXiv:1808.03231 [Preregistered Protocol]. 2018. https://arxiv.org/abs/1808.03231
[Accessed 28 April 2020].
24. van der Laan M, Rose S. Targeted learning: causal inference for observational and experimental data
. New York Dordrecht Heidelberg London: Springer; 2011.
25. Balzer LB, van der Laan MJ, Petersen ML, Collaboration S. Adaptive prespecification in randomized trials with and without pair-matching
. Stat Med
26. Abuogi LL, Humphrey JM, Mpody C, Yotebieng M, Murnane PM, Clouse K, et al. Achieving UNAIDS 90-90-90 targets for pregnant and postpartum women in sub-Saharan Africa: progress, gaps and research needs
. J Virus Erad
2018; 4: (Suppl 2): 33–39.
27. Myer L, Redd AD, Mukonda E, Lynch BA, Phillips TK, Eisenberg A, et al. Antiretroviral adherence, elevated viral load and drug resistant mutations in HIV-infected women initiating treatment in pregnancy: a nested case-control study
. Clin Infect Dis
28. Omonaiye O, Kusljic S, Nicholson P, Manias E. Medication adherence in pregnant women with human immunodeficiency virus receiving antiretroviral therapy in sub-Saharan Africa: a systematic review
. BMC Public Health