Level and predictors of receiving essential mother-infant PMTCT services in Tanzania: maternal ART adherence, infant ARV prophylaxis and early diagnosis

Background: Full PMTCT service adherence, including maternal ART, HIV-exposed infant (HEI) nevirapine (NVP) dosing, and early infant diagnosis (EID), is not routinely measured in Tanzania''s PMTCT program. This secondary cohort analysis from 27 PMTCT clinics describes the level and predictors of full PMTCT adherence.
Methods: The Supporting Attendance for Facility Delivery and Infant Health study recruited HIV-positive pregnant women, administering interviews at enrollment and 12-weeks postpartum, with medical-record-abstraction of outcomes. Mother-infant PMTCT adherence--defined as maternal ARV adherence (zero self-reported missed ARV doses in past seven days and not missing two consecutive days of ARV dosing in past six months) HEI NVP birth/6-week dosing, and EID--is described among 584 women-infant pairs. Generalized-estimating-equation regression methods were used to predict PMTCT adherence, adjusting for age, education and site cluster. Hypothesized predictors included known/new HIV-positive, phone access/contact with health providers, ART/MTCT knowledge, facility delivery and clinic volume.
Results: Among HEI overall, 75% received the full NVP regimen; and 70% received both NVP and EID. Maternal ART adherence was 83%, and predicted receipt of HEI services. Eighty-one percent of ART-adherent mothers'' HEI received NVP and EID compared to only 51% of non-adherent mothers (p< .0001). Overall, 67% of mother-infant pairs were PMTCT adherent.
Factors reducing PMTCT adherence were pre-ART status compared to women on ART before pregnancy (known-positive, pre-ART: adjusted odds ratio [AOR] 0.28,p< .0001; newly diagnosed HIV-positive, pre-ART: AOR 0.25,p< .0001); and loss of mobile phone access (AOR 0.48,p=.006) compared to maintaining access from baseline to follow-up. Factors positively associated with PMTCT adherence included knowing that ARVs reduce MTCT risk (AOR 3.18,p< .0001), delivering in a facility (AOR 2.90,p< .0001), having phone contact with a health provider (AOR 2.33,p< .0001), and attending a low-volume ANC facility compared to a high-volume ANC (AOR 5.28,p< .0001; < 90 new patients/quarter vs. >=200).
Conclusions: Increased client-provider interaction may improve PMTCT adherence through promoting ART/PMTCT knowledge and access to facility delivery. Special attention should be focused on newly diagnosed and newly ART-initiated. Facilitating provider-client communication using mobile phones may enhance effectiveness of PMTCT service models. Further investigation is needed to understand why lower volume ANC clinics are associated with better adherence.

G. Antelman1, G. Mbita2, T. Machalo2, A. Maijo2, P. Njau3, J. van 't Pad Bosch2, R. van de Ven2, G. Woelk4
1Elizabeth Glaser Pediatric AIDS Foundation, Strategic Information and Evaluation, Dar es Salaam, Tanzania, United Republic of, 2Elizabeth Glaser Pediatric AIDS Foundation, Dar es Salaam, Tanzania, United Republic of, 3Ministry of Health, Community Development, Gender, Elderly and Children, PMTCT Unit, Dar es Salaam, Tanzania, United Republic of, 4Elizabeth Glaser Pediatric AIDS Foundation, Washington, United States