The effects of revised peer-counselor support on the PMTCT cascade of care: results from a cluster-randomized trial in Kenya (the EMMA study)

Background This study evaluated the effect of revisions to existing peer-counselor services, called Mentor Mothers (MM), at maternal and child health clinics on medication adherence for women living with HIV ( WLWH) in Kenya and on early infant HIV testing. Methods The Enhanced Mentor Mother Program study was a 12-site, two-arm cluster-randomized trial enrolling pregnant WLWH from March 2017 to June 2018 (with data collection through September 2020). Six clinics were randomized to continued MM-supported standard care (SC). Six clinics were randomized to the intervention arm (INT = SC plus revised MM services to include more one- on- one interactions). Primary outcomes for mothers were defined as: (PO1) the proportion of days covered (PDC) with antiretroviral therapy (ART) >= 0.90 during the last 24-weeks of pregnancy; and (PO2) >= 0.90 PDC during the first 24-weeks postpartum. Secondary outcomes were infant HIV testing according to national guidelines (at 6, 24, and 48 weeks). Crude and adjusted risk differences between study arms are reported. Results We enrolled 363 pregnant WLHV. After excluding known transfers and subjects with incomplete data extraction, data were analyzed for 309 WLWH (151 SC, 158 INT). A small share achieved high PDC during the prenatal and postnatal periods (0.33 SC/0.24 INT achieved PO1; 0.30 SC/0.31 INT achieved PO2; crude or adjusted risk differences were not statistically significant). In addition, similar to 75% in both study arms completed viral load testing during year two after enrollment, with > 90% suppressed in both arms. For infants, >= 90% in both arms had at least one HIV test through study follow up (76 weeks) but testing on schedule according to PMTCT guidelines was uncommon. Conclusions While national guidelines in Kenya recommended that all HIV-infected pregnant women take a daily antiretroviral regimen for life following a HIV diagnosis, results presented here indicate that a minor share achieved high medication coverage during the prenatal and postnatal periods analyzed. In addition, adjustments to Mentor-Mother services showed no improvement in study outcomes. The lack of effect for this behavioral intervention is relatively consistent with the existing literature to improve mother-infant outcomes along the PMTCT care cascade.