Factors associated with HIV medication adherence in HIV-positive women enrolled in Option B+ in Zambia: a cross-sectional survey

Background: Every year about 1·4 million women living with HIV become pregnant, and these pregnancies lead to about 220 000 new HIV infections in infants and children. Furthermore, it is estimated that, worldwide, the cumulative total deaths caused by HIV will reach 75 million by 2030. In an effort to eliminate mother-to-child transmission of HIV, WHO has recommended Option B+, which gives HIV-positive pregnant women life-long antiretroviral therapy (ART), regardless of their CD4 cell count and viral load. The purpose of this study is to investigate factors associated with HIV medication adherence in HIV-positive women enrolled in Option B+ in Zambia. Methods: We recruited pregnant and breastfeeding HIV-positive women enrolled in Option B+ and residing in Lusaka (urban) and Sinazongwe (rural) districts of Zambia. Data were collected using a 122-item structured questionnaire prepared in English and then translated into two local languages (Nyanja and Tonga), which are commonly spoken in the Lusaka and Sinazongwe Districts. We used descriptive statistics to characterise ART adherence in the entire sample and by district. Four generalised modified-Poisson regression models were built using a step-wise approach to assess the association between the factors (age, education, marital status, occupation, income, counselling on side-effects, HIV transmission knowledge, and timing of first prenatal visit during pregnancy) and ART adherence. Findings: We included data from 150 participants, 81 (54%) from Lusaka district and 69 (46%) from Sinazongwe district. Mean age was 29 years (SD 6·2). Data were collected between July and September, 2017. A total of 119 respondents (79%) had counselling on the side-effects of ART: 72 (89%) in Lusaka district and 47 (68%) in Sinazongwe district. 36 women (24%) sought a prenatal visit in the first trimester of pregnancy, 93 (62%) in second trimester, and 16 (11%) in the third trimester; five (3%) did not know in which trimester they had first visited the antenatal clinic for care. In all, 37·8% of respondents missed taking their ART at least once in the week before data collection. There was an association between adherence and district of residence (p<0·0001). Also, counselling on the side-effects of ART was associated with adherence to treatment in the overall sample (p=0·01) and the rural district (p