Efavirenz-based regimens among womenof reproductive age receiving ART in Johannesburg

Polly Clayden, HIV i-Base

South African guidelines recommend efavirenz-based regimens as the preferred first line.

Out of concern about efavirenz use in pregnancy the 2004 guidelines only recommend its use in women of reproductive age when they are using injectable contraception plus condoms; the 2010 revision amended this to “reliable contraception”. Nevirapine is recommended for the remainder, who are “unable to guarantee reliable contraception”.

Three posters authored by researchers from the Reproductive Health Unit, University of the Witwatersrand and Johns Hopkins Bloomberg School of Public Health, showed findings from investigations into how well these guidelines were being followed, the fertility intentions of women receiving efavirenz based and other antiretroviral regimens, and whether providers discuss these issues with women of reproductive age receiving ART. [1, 2, 3]

Colleen Hanrahan and colleagues looked at the application of guidelines. This prospective cohort study enrolled non-pregnant women on ART aged 18-35 years in September 2009-January 2010 in four Johannesburg clinics. The investigators conducted baseline interviews to determine demographics, contraceptive use and the fertility intentions of the women. A record review was used to confirm ART regimens.

They classified women correctly assigned to first line regimens at baseline according to the 2004 guidelines, which were in use at the time of the evaluation, and two interpretations of the 2010 guidance, “reliable contraception”. They used logistic regression to determine predictors of “inappropriate assignment”.

The investigators reported, out of a cohort of 805 women on first line ART, 44.6% (95% CI, 41.2–48.0%) were receiving efavirenz-based regimens at baseline. Overall 26.5% (95%CI: 23.4–29.5%) of women were receiving hormonal contraception.

Of those receiving efavirenz, 90% (95% CI, 86–93%) were incorrectly assigned according to 2004 guidance, but only 11% (95% CI 9–15%) were wrongly assigned to nevirapine. These proportions reduced to 77% for efavirenz and 27% for nevirapine and 24% for efavirenz and 73% for nevirapine interpreting the 2010 guidance as using hormonal contraception and hormonal contraception or consistent condom use respectively.

In a multivariate analysis including: age, time on ART, CD4 count, number of living children, relationship and employment status and enrollment site, none were significant predictors of incorrect assignment to efavirenz. For nevirapine, each additional child gave a two-fold increase in the odds of incorrect assignment, AOR 1.95 (95% CI 1.34–2.84), p<0.001.

In a related study Sheree Schwatz and colleagues presented data from a cross-sectional analysis of the same baseline interviews to compare differences in current and future fertility intentions. This analysis included a total of 851 women; the 805 on first line regimens described above plus a small proportion (5.4%, n=46) receiving second line regimens with a boosted PI.

Multivariate analysis revealed women on efavirenz-based regimens were older and had more living children, both p<0.001. Of these 39% were either currently trying to conceive or planned to do so in the next year. Women receiving nevirapine were more likely to be currently trying to conceive than those receiving efavirenz, p=0.025, but were no more likely to plan to in the next year, p=0.17.

In the third study describing communication between providers and HIV-positive women receiving ART about fertility and reproduction, less than half (40.7%) of the 851 women enrolled reported that providers had talked to them about future pregnancy options.

Older women and those with higher income were more likely to have fertility discussions with providers in multivariate analysis whereas parity, CD4 count, time on ART, regimen, marital status or fertility intentions were not associated with the likelihood of these discussions.

The investigators also found that PMTCT knowledge was significantly higher if their providers has discussed this with them p<0.001. Only about a third (35.4%) understood that efavirenz is contraindicated if trying to conceive and this was not associated with efavirenz use, p=0.774. A small proportion (6.4%) said a provider had told then not to have more children and 36% were unsure whether their provider approved of them having children. Discussion about contraception varied by type: 93.5% of women reported that their providers had discussed male condoms, 71.6% female condoms, 45.2% injectable contraceptives, 41.6% oral contraceptives and 18.8% sterilisation.


These data make the important point that no matter what labeling or guidelines recommend concerning use of efavirenz in pregnancy there is a strong likelihood that it will happen.


  1. Hanrahan et al. South African antiretroviral treatment guidelines for women: how well are they being followed and for whom? Poster abstract THPEO120.
  2. Schwartz S et al. Fertility intentions by antiretroviral regimen among women in inner-city Johannesburg. Poster abstract THPE0784.
  3. Schwartz S et al. Communication between patients and providers about HIV and fertility in inner city Johannesburg, South Africa. Poster abstract THPE0650.

Links to other websites are current at date of posting but not maintained.