Enhanced Peer-Group strategies to support prevention of Mother-to-Child HIV transmission leads to increased retention in care in Uganda: A Randomized controlled trial

Enrollment and randomization procedures

We enrolled pregnant women living with HIV initiated on PMTCT Option B+ who were: ≥ 18 years old, confirmed pregnant by urine dipstick, confirmed HIV-positive by rapid HIV testing as per National PMTCT Guidelines, being ART-naïve and beginning Option B+ ART within 30 days of enrolment into the study, residing within 20 km of the study clinic, not planning to move out of the clinic catchment area within 2 years, agreeing to be home visited, and providing written informed consent.

Pregnant women testing HIV-positive through the routine National PMTCT Program and starting Option B+ were referred to study counselors by the clinic PMTCT counselors for prescreening and scheduled to return to the study clinics for consent procedures within one month of starting once daily Option B+ ART. HIV rapid testing followed the national and WHO algorithms (23) using three sequential antibody rapid tests: Abbott Determine Rapid HIV 1 Test, Abbott Laboratories, USA; Stat Pak Rapid HIV-1 Test, Chembio Diagnostic Systems, USA; and Unigold HIV Rapid Test, Trinity Biotech, Ireland.

Eligible and consenting women were then randomized 1:1 irrespective of the study sites to SOC control arm or FLC intervention arm. The randomization list was computer-generated by the data manager based at MUJHU using random-sized block groups that included consecutive intervention numbers with corresponding random intervention assignments. Each pregnant woman was assigned unique participant identification number and authors had no access to information that could identify individual participants during or after data collection

After randomization, participants were assessed for socioeconomic status, ARV drug adherence and stigma. A questionnaire addressing participants’ experiences with stigmatization and discrimination at family and community levels was administered at baseline, 12 months post-enrolment and at end of study. An additional needs assessment questionnaire was administered to participants enrolled in the FLC arm at enrolment, 1-year post enrolment and every 6 months thereafter until the end of study. This was to assess individuals’ achievements and to document challenges related to group activities and participation in income-generating activities (IGAs) and its benefits, skills acquired and training needs.

FLC Intervention and SOC controls

The FLC intervention (hereafter ‘FLC’) included community-based enhanced peer group support, IGAs and community-based drug distribution. Participants in the FLC arm formed dynamic peer support groups based on women’s home addresses documented by Geographical Positioning System (GPS). Groups consisted of 8-10 participants who attended monthly meetings in their communities. Sustainable livelihood study assistants and study counselors participated in and documented FLC group activities. Each FLC group was supported by one peer mother who had similar social background and/or life experiences as FLC participants.

During monthly FLC group meetings, study counselors distributed ART and Cotrimoxazole and offered psychosocial support on follow-up clinic visits and drug adherence through individual and group counseling while emphasizing peer support among participants.

FLC groups also offered skills building in a variety of IGAs such as goat rearing, bakery, crafts making, liquid soap making, book binding and charcoal briquette making. These were chosen by the participants and implemented as a group or as individuals with the guidance of a study IGA Advisor. All FLC groups were linked to local government bodies for sustainability, registration and access to Community Development Driven grants funded by the District Local Government to support the implementation of livelihood projects for community-based groups.

SOC arm participants received counseling from clinic PMTCT counselors and collected their drug refills on an individual basis from the PMTCT clinics where they were enrolled as per MOH guidelines (16). SOC arm participants participated voluntarily in family support groups whenever available at their respective health facility as recommended by the MOH (16).

Follow-up procedures

At each study site, all study participants were followed at a dedicated clinic for all scheduled and unscheduled visits and were given transport reimbursement for scheduled study visits. Scheduled visits included a first visit within the initial four weeks post-enrolment, then monthly thereafter until delivery, at delivery, at 6 and 14 weeks postpartum, and quarterly thereafter from 6 months through 24 months postpartum (end of follow-up). At each of these visits, FLC participants were counseled by study counselors on drug adherence and family planning while SOC participants were counseled by PMTCT counselors. Study counselors administered questionnaires on ART adherence, stigma, and socio-economic status to all participants at each scheduled clinic visit.

Participants were terminated from the study if they missed four or more consecutive scheduled study visits, withdrew consent, relocated outside study area, or died.

Infant procedures

All infants born to study participants were enrolled in the study and received daily Nevirapine syrup from birth up to 6 weeks of age regardless of the infant feeding method. Infants were tested routinely at 6 weeks of age for HIV by DNA PCR and at 18 months of age by rapid HIV antibody test as part of the National Early Infant HIV Diagnosis Program (23).

Outcomes and measures

The primary outcomes of interest included adherence to PMTCT clinic appointments at 6 weeks, 12 and 24 months postpartum and adherence to Option B+ ART at 6 weeks, and 6 and 24 months postpartum.

Complete adherence to visits was defined as having attended the 4 scheduled quarterly visits at the end of each of the first and second years of postpartum follow-up as defined by the MOH (16). Scheduled visits were determined to have been adhered to on the basis of a completed visit within a 3-week window on either side of the scheduled visit date.

Retention in care was defined as the proportion of participants who were not terminated from the study prior to the end of follow-up (24 months postpartum). Causes for termination included participant relocation out of the study area, withdrawal from the study, loss to follow-up, miscarriage, referral outside the study facility, and death. The study defined loss to follow-up as not attending 3 consecutive scheduled study visits. Retention in care was calculated as 1 minus the proportion of women terminated for any such reason before the end of follow-up.

As a second primary outcome, we compared maternal ART adherence using self-report at 6 weeks and 6 and 24 months postpartum among all study participants. We validated self-reports by comparing VL measurements for all participants at 6 weeks, 6 and 24 months postpartum.

All participants were asked to self-assess the number of complete ART doses they missed in the past three days. Self-reported ART adherence was then calculated as 1 minus the proportion of self-reported doses missed in the past 3 days.

Laboratory measures of ARV adherence were used to validate self-reports using viral suppression below 400 copies/mL. At the time of writing the study protocol, this limit was considered as viral suppression by Uganda MOH (16). VL measurements were performed at the Infectious Diseases Institute (IDI) Core Laboratory, Kampala, Uganda, a central laboratory certified by the American College of Pathologists. Nucleic acids were extracted and tested for HIV-1 RNA using the COBAS® AmpliPrep/ COBAS® TaqMan® HIV-1 Test procedure following the manufacturer’s instructions (Roche Molecular Systems, Inc., Branchburg, NJ, 08876 USA) (25).

Virological failure was defined as median viral load persistently above 1000 copies/ml for two consecutive viral load measurements within a three-month interval after at least 6 months of initiating ART. The final drug adherence variable at each time point was dichotomized as ‘not adhering’ if viral load was ≥400 copies/ml irrespective of the level of self-reported adherence, or as ‘adhering’ if self-reported adherence in the last 3 days was ≥95% and viral load was <400 copies /ml.

Lastly, the HIV status and the HIV-free survival of children born to study participants were assessed at their 6 week and 18-month postpartum visits as per the National PMTCT program recommendation and practice.

Statistical Analyses

We estimated a target sample size of at least 270 women living with HIV needed to be enrolled in each of the study arms to provide 90% power at 5% significance level to detect a difference of 15% or more in adherence to visits and drugs or in virological suppression between the intervention and control arms. Follow-up time was calculated in person-years, with each participant contributing time in years from enrolment to censoring at the date when a participant was lost to follow-up, died or reached the end of study.

We compared participants’ baseline characteristics in the FLC and SOC arms by testing for differences in proportions using Fisher’s exact p-values. Maternal adherence to at least 4 scheduled PMTCT clinic appointments at the end of each year of postpartum follow-up, adherence to ART and retention in care, as well as the HIV status of children were also summarized using proportions and chi-square p-values. We used the Wilcoxon–Breslow–Gehan test to assess the difference in HIV-free survival of participants’ children between the FLC and SOC arms. Log rank and Chi-Square p-values were used to test the equality of Kaplan-Meier survival probabilities and hazard rates for failure to retain in care for any reason by study arm. Cox Proportional Hazard regression model adjusted for unbalanced baseline variables was used to estimate the hazard rate ratio of failing to retain participants in care with 95% confidence intervals between the FLC and the SOC arms.

Ethical and regulatory considerations

The study was approved by Uganda’s National Council of Science and Technology (UNCST), the Joint Clinical Research Centre (JCRC) Institutional Review Board (IRB), as well as the IRBs of Johns Hopkins University (JHU) and the University of California San Francisco (UCSF). Written informed consent was obtained from all study participants by study counselors in the presence of an impartial witness before enrolment in the study.

Baseline demographic characteristics

Table 1 provides the number of participants, clinic visits made, and person-years of follow-up by baseline socio-demographic characteristics of the study population stratified by study arm. These results show that despite randomization, there were significant differences between the two study arms in participants’ educational level, gravidity, HIV status disclosure to partner and monthly income.

Adherence to PMTCT clinic appointments

There was no statistically significant difference between the intervention and control arms in the number of women who started or completed their postpartum visits at 6 weeks postpartum or in the first or second year of follow-up (Table 2).

Retention in care

Overall, retention in care was high across both study arms with 83.0% of all participants remaining in care at the end of follow-up (Table 3). Significantly, more women were retained in care at the end of follow-up in the FLC arm (86.7%) compared to the SOC arm (79.3%, p= 0.0221). Also, more women were terminated before the end of follow-up due to relocation in the SOC arm (n=27, 10.0%) compared to the FLC arm (n=12, 4.4%).

There was a significant difference between the Kaplan-Meier survival functions for failure to remain in care until the end of study (24 months postpartum visit) between the control and intervention arms (Wilcoxon-Breslow test for equality of survivor functions Chi² (1) = 11.94, p-value = 0.0005) (Figure 2). Overall, the estimated proportion of participants retained in care was high, with 87.0% of all participants remaining in care until the end of the study.

• Download figure
• Open in new tab

Figure 2. Kaplan-Meier survival curves for retention in care up to 24 months postpartum

Adherence to ART

Table 4 shows that participants self-reported optimal adherence to taking >95% of their ART medication in the last 3 days at their 6 weeks, 6 and 24 months postpartum visits. There was no statistically significant difference in optimal adherence to ART between the arms. The median viral loads of participants at each of these time points were <100 copies/ml with no statistically significant differences between the two arms.

Predictors of retention in care

Cox-proportional hazard regression models were used to assess the effect measure of intervention on being retained in care. All baseline variables with p<0.1 at the univariate level were included in the multivariate model and using a backward elimination method, only those variables with p<0.05 were kept in the final model.

Table 5 shows that the rate of dropout from care was 2.5 times greater among participants randomized to the SOC arm versus the FLC arm (aHR=2.498, 95% CI: 1.417 – 4.406, p=0.002). Participants enrolled from rural health facilities and those aged 15-24 years were also significantly less likely to be retained in care compared to their urban (p=0.021) or their older counterparts (p=0.035) respectively. Educational level, gravidity, disclosure status and income were not significant predictors of retention in care and were not included for analysis in the final model.

HIV-free survival of children

Of 492 women with a visit at 18 months postpartum, 8 of their 431 (1.9%) tested infants tested positive for HIV. Most (7/8) of these infants had tested positive to HIV by the 6th week postpartum visit. The proportion of HIV-positive children born to mothers enrolled in the FLC arm (1.4%) was not significantly different from that observed in the SOC arm (2.4%) (p=0.472).

By the 18-month postpartum visit, the HIV-free survival of children born to mothers in the FLC arm was 93.8% [95% CI: 89.9 – 96.2] compared to 83.0% [95% CI: 75.9 – 87.7] among children born to mothers in the SOC arm. The difference between the two arms was significant (p=0.0031).

Limitations

Our study’s main limitation is that our RCT was individually and not cluster randomized, which may have resulted in cross-contamination between the FLC and the SOC arms in the same clinics over the long-term follow-up period. Contamination could have also occurred when study staff attended to both FLC and SOC participants at the same clinic. Additionally, all study participants were given transport reimbursement for their study visits which may have positively influenced adherence to study visits. Lastly, the feasibility of our intervention would depend on its cost and cost-effectiveness, which we did not assess.

Conclusions

Our study found that the FLC intervention significantly increased retention in care and HIV-free survival of children of pregnant mothers living with HIV through 24 months postpartum. These findings are encouraging and suggest that programmatic interventions such as the FLC may contribute to the goal of eliminating MTCT of HIV in Uganda. We recommend scaling up peer group support, community drug distribution and income-generation intervention strategies in the context of PMTCT.