Primary HIV prevention in pregnant and lactating Ugandan women: A randomized trial

Abstract

Background: The 'Primary HIV Prevention among Pregnant and Lactating Ugandan Women' (PRIMAL) study aimed to assess the effectiveness of an enhanced HIV counseling intervention for preventing HIV acquisition among HIV-uninfected mothers during pregnancy and throughout the breastfeeding period.

Methods: We conducted an unblinded randomized control trial between 22 February 2013 and 22 April 2016 to assess the effectiveness of an extended repeat HIV testing and enhanced counseling (ERHTEC) intervention aimed at preventing primary HIV infection among HIV-uninfected pregnant and lactating women in Uganda. HIV-uninfected pregnant women aged 15-49 were enrolled 1:1 individually or in couples together with their partner. Enrolled women and couples were randomized 1:1 to an intervention (ERHTEC) or control (extended repeat HIV testing and standard counseling) group and followed up to 24 months postpartum or six weeks past complete cessation of breastfeeding, whichever came first. Both groups were tested for sexually transmitted infections (STIs) and HIV at enrollment, delivery, 3 and 6 months postpartum and every 6 months thereafter until the end of follow-up. The intervention group received enhanced HIV prevention counseling every 3 months throughout follow-up. The control group received standard counseling at the time of HIV retesting. Both intervention and control couples were offered couple HIV testing and counseling at all study visits.

Main outcome measures: Frequency of condom use and incidence of HIV, syphilis, gonorrhea, chlamydia and trichomoniasis over follow-up.

Results: Between February 2013 and April 2014, we enrolled 820 HIV-uninfected pregnant women presenting for antenatal care individually (n = 410) or in couples (n = 410 women and 410 partners) in one urban and one rural public Ugandan hospital. Women's median age was 24 years (IQR 20-28 years). At baseline, participants did not differ in any socio-demographic, reproductive health, HIV testing history, sexual behavior, medical history or STI status characteristics; 96% (386/402) of couples were tested and counseled for HIV together with their partners at enrolment, 2.1% (7/329) of whom were found to be HIV-infected. Six hundred twenty-five (76%) women completed follow-up as per protocol (S1 Protocol). Women were followed for an average of 1.76 years and cumulated 1,439 women-years of follow-up or 81% of the maximum 1,779 women-years of follow-up assuming no dropouts. Men were followed for an average of 1.72 years. The frequency of consistent condom use and the proportion of women who used condoms over the last 3 months or at last vaginal sex increased substantially over follow-up in both arms, but there were no statistically significant differences in increases between the intervention and control arms. During follow-up, on average 42% (range 36%-46%) of couple partners were counseled together. Between 3.8% and 7.6% of women tested positive at any follow-up visit for any STI including syphilis, gonorrhea, C. trachomatis or T. vaginalis. Four women (two in each arm) and no enrolled men became infected with HIV, representing an overall HIV incidence rate of 0.186 per 100 person-years. Three of the women seroconverters had enrolled individually, one as a couple. At or before seroconversion, all four women reported their partners had extramarital relationships and/or had not disclosed their suspected HIV-infected status. There were no statistically significant differences between study arms for STI or HIV incidences.

Conclusions: A sustained enhanced HIV prevention counseling intervention for up to 2 years postpartum among pregnant and breastfeeding women did not have a statistically significant effect on condom use or HIV incidence among these women. However, in both study arms, condom use increased over follow-up while STI and HIV incidence remained very low when compared to similar cohorts in and outside Uganda, suggesting that repeat HIV testing during breastfeeding, whether with enhanced or standard counseling, may have had an unintended HIV preventive effect among pregnant and lactating women in this setting. Further research is needed to verify this hypothesis.

Trial registration: ClinicalTrials.gov NCT01882998.

Fig 1. Trial profile.

¹ None of these women presented with any particular risk profile according to their screening data; ² Unrelated to the intervention ³ LTFU: lost to follow-up.

Fig 2. Baseline characteristics of intervention and control women at enrolment.

1. IQR: Interquartile range 2. Actual values are based on 631/818 (77%) of women who had a gestational age captured at enrolment–midwives had no access to ultrasound imaging, had problems with last normal menstrual period recall as well as fundus height measurement. Estimated gestational age was based on time between enrolment and delivery dates and an estimated average pregnancy duration of 40 weeks. 3. Questions on mother-to-child HIV transmission (MTCT) explored knowledge about transmission during pregnancy, delivery, breastfeeding and post-breastfeeding 4. Questions about prevention of MTCT explored knowledge about HV antiretroviral therapy prophylaxis, delivery in health facilities, baby nevirapine prophylaxis, exclusive breastfeeding, duration of breastfeeding, and formula feeding, 5. Questions about prevention of primary maternal HIV acquisition explored knowledge about abstinence, condom use, risk of transfusions and injections, partner testing, breastfeeding avoidance and breastfeeding duration, health facility delivery, faithfulness, and testing for HIV. 6. Symptoms probed or examined included: Abnormal vaginal discharge, bleeding, itching, swelling, vesicles, pustules, odor, genital or anal ulcers/sores/warts, frequent or painful urination, pain or bleeding during intercourse, and lower abdominal pain.

Fig 3

a. Proportion of sexually active women since last visit, by visit Difference in average change by treatment assignment: p = 0.120 Difference in average follow-up level by treatment assignment: p = 0.952 b. Median frequency of vaginal intercourse in the last 3 months, by visit Difference in average change by treatment assignment: p = 0.615 Difference in average follow-up level by treatment assignment: p = 0.592.

Fig 4. Proportion of women who practiced safer sex through condom use in the last 3 months, by visit.

Difference in average change by treatment assignment: p = 0.665 Difference in average follow-up level by treatment assignment: p = 0.026.

Fig 5

a—Proportion of women who reported consistent or intermittent condom use in the last 3 months, by visit Difference in average change by treatment assignment: p = 0.665 Difference in average follow-up level by treatment assignment: p = 0.056 b—Average frequency of condom use as percent of vaginal sex episodes in the last 3 months Difference in average change by treatment assignment: p = 0.209 Difference in average follow-up level by treatment assignment: p = 0.137 c—Median frequency of vaginal sex without condom in the last 3 months Difference in average change by treatment assignment: p = 0.534 Difference in average follow-up level by treatment assignment: p = 0.148 d—Proportion of women who used condoms at last vaginal sex Difference in average change by treatment assignment: p = 0.353 Difference in average follow-up level by treatment assignment: p = 0.029.

Fig 6

a. Proportions of women testing positive for syphilis antibodies over follow-up Difference in average change by treatment assignment: p = 0.136 Difference in average follow-up level by treatment assignment: p = 0.266 b. Proportions of women testing positive for Neisseria gonorrhea over follow-up Difference in average change by treatment assignment: p = 0.632 Difference in average follow-up level by treatment assignment: p = 0.785 c. Proportions of women testing positive for Chlamydia trachomatis over follow-up Difference in average change by treatment assignment: p = 0.543 Difference in average follow-up level by treatment assignment: p = 0.595 d. Proportions of women testing positive for Trichomonas vaginalis over follow-up Difference in average change by treatment assignment: p = 0.051 Difference in average follow-up level by treatment assignment: p = 0.367 e. Proportions of women testing positive for any of the STIs tested over follow-up Difference in average change by treatment assignment: p = 0.681 Difference in average follow-up level by treatment assignment: p = 0.753.