Women with HIV have higher risk of loss to follow-up and death after delivery

Elsa from Mozambique. Photo by Médecins sans Frontières
This article is more than 8 years old. Click here for more recent articles on this topic

Despite the widespread implementation of Option B+ (lifelong antiretroviral therapy (ART) for pregnant and breastfeeding women regardless of CD4 cell count) postpartum HIV-positive women remain at higher risk of loss to follow-up and death, studies from South Africa and Botswana presented at the 21st International AIDS Conference (AIDS 2016) in Durban last week show.

South Africa: postpartum virologic failure

HIV infection remains a significant problem among South African women of reproductive age and an important contributor to maternal death.

Many conceive while already on ART. Long-term ART success depends on attaining and maintaining virologic suppression. Adherence and remaining in care after delivery can be challenging. Scant information exists on postpartum virologic outcomes.

The implementation of Option B+ in South Africa in over 90% of all sites for prevention of mother-to-child transmission (PMTCT) and maternal, newborn and child health warranted a closer look at postpartum treatment outcomes.


person years

In a study “100 person years of follow-up” could mean that information was collected on 100 people for one year, or on 50 people for two years each, or on ten people over ten years. In practice, each person’s duration of follow-up is likely to be different.

loss to follow up

In a research study, participants who drop out before the end of the study. In routine clinical care, patients who do not attend medical appointments and who cannot be contacted.


Expresses the risk that, during one very short moment in time, a person will experience an event, given that they have not already done so.

treatment failure

Inability of a medical therapy to achieve the desired results. 

retrospective study

A type of longitudinal study in which information is collected on what has previously happened to people - for example, by reviewing their medical notes or by interviewing them about past events. 

Postpartum HIV-positive women, particularly those who become pregnant while on ART (incident pregnancy), are at high risk of treatment failure, Dina Onoya, on behalf of the Health Economics and Epidemiology Research Office (HERO), University of Witwatersrand and the University of Boston, told delegates. She was presenting findings from a retrospective cohort study of over 6900 HIV-positive women aged 15 to 49 years who started ART between 2004 and 2013 at ten public sector ART clinics in Johannesburg, South Africa.

The longer women were on ART before pregnancy, the more likely they were to remain in care. Virologic failure seemed to be linked to poor health at delivery suggesting the need for targeted adherence support for ART-experienced women at risk of adverse outcomes after delivery.

Among those who left care after delivery the apparent low risk of virologic failure was due to not being in the system long enough for the failure to be recorded. Dr Onoya and colleagues suggest that adherence and support among women diagnosed in the third trimester or at delivery be strengthened.

Overall 563 (8.1%) had virologic failure at a rate of 5.0 per 100 person-years (95% CI: 4.6-5.5). Cohorts differed slightly: controls (HIV-positive and not pregnant): 5.0 per 100 person-years; incident pregnancy: 5.7 per 100 person-years; and prevalent pregnancy (already pregnant when starting ART): 4.2 per 100 person-years.

Predictors of treatment failure after giving birth among the incident pregnancy cohort included being anaemic at delivery, adjusted hazard ratio (aHR): 1.5 (95%CI: 1.1-2.1); World Health Organization stage 3 HIV disease at delivery, aHR: 1.6 (95% CI: 1.1-2.5); and experiencing virologic failure in pregnancy, aHR: 2.1 (95% CI:1.5-3.0). Older age (30-39), aHR: 0.6, (95% CI: 0.3-1.0) and a CD4 cell count above 350 cells/mm3 at delivery, aHR: 0.2 (95% CI: 0.1-0.3 were protective against treatment failure.

Among the prevalent pregnancy cohort a higher CD4 cell count (≥ 350 cells/m3) was protective against postpartum virologic failure.

Virologic failure increased with time. A prevalent pregnancy with a CD4 cell count under 350 cells/mm3 at delivery and an incident pregnancy with a CD4 cell count over 350 cells/mm3 resulted in faster time to virologic failure.

Overall close to one-quarter (1645 of 6971) of women were lost to follow-up (LTFU) by 24 months at a rate of 8.6 per 100 person-years (95% CI: 8.2-9.0).

The prevalent pregnancy cohort (27.9%) had the highest LTFU rate of 11.2 per 100 person-years (95% CI: 10.3-12.2) and the incident pregnancy cohort (19.4%) the lowest, at 6.5 per 100 person-years (95% CI: 5.9-7.2).

The risk of LTFU 24 months postpartum was not associated with age, employment status, CD4 level or time on ART before delivery.

Botswana: maternal mortality

Postpartum HIV-positive women were five times more likely, compared to HIV-negative women, to die within 24 months of delivery regardless of uptake of ART and independent of CD4 cell count, according to a study comprising 3000 mothers (of which 1499 were HIV-positive) who had delivered at five public hospitals throughout Botswana between February 2012 and March 2014, Rebecca Zash reported at the same session.

Dr Zash and her colleagues wanted to look at how ART use during pregnancy, and whether or not it continued during the postpartum period, affected death rates in a setting with widespread uptake of ART and PMTCT services.

HIV-positive and HIV-negative mothers enrolled within 48 hours of giving birth were followed up by cell phone at one and three months then every three months until 18 and 24 months postpartum. One of the approved contacts the mother gave at enrolment reported maternal deaths.

At baseline median age was 29 and 24 years among HIV-positive and HIV-negative mothers, respectively. This was a first birth for a larger number of HIV-negative mothers than for HIV-positive mothers, 675 (45%) and 244 (16%), respectively. Over a third (36%) of HIV-positive mothers had had at least four pregnancies compared to 13% of HIV-negative mothers. 

Hospitalisations and modes of delivery were similar, while HIV-positive mothers had slightly lower markers of socioeconomic status.

Using Cox proportional hazard models, risk factors for maternal survival were assessed.

In the 24-month postpartum period 26 mothers died, 22 among HIV-positive mothers (758 per 100,000 person-years) and four among HIV-negative mothers (138 per 100,000 person-years). More deaths occurred in the first twelve months.

The overwhelming majority of deaths (59% or 13) occurred among those having taken ART in pregnancy and throughout the follow-up period.

There were considerably fewer deaths among those who stopped ART after giving birth but started again in the postpartum period: two deaths, aHR: 0.9 (95% CI: 0.3-6.4); among those who got ART during pregnancy but stopped postpartum: four deaths, aHR: 1.7 (95% CI: 0.6-5.1); and those not having taken ART during pregnancy or follow-up: three deaths, aHR: 1.6 (95% CI: 0.2-15.2).

Neither CD4 cell count during pregnancy, nor a longer time on ART before delivery were linked with mortality.

Despite the absence of adherence data, Dr Zash proposed closer clinical monitoring to improve maternal outcomes, and further research to understand the increased risk of death among HIV-positive postpartum women.


Onoya D et al. Timing of pregnancy among HIV-positive women, postpartum retention and risk of virologic failure. 21st International AIDS Conference, Durban, abstract WEAB010, 2016.

View the abstract on the conference website.

Download the presentation slides from the conference website.

Watch the webcast of this presentation on YouTube.

Zash R et al. High proportion of death attributable to HIV among post-partum women despite widespread uptake of ART. 21st International AIDS Conference, Durban, abstract WEAB0104, 2016.

View the abstract on the conference website.

Download the presentation slides from the conference website.

Watch the webcast of this presentation on YouTube.