Menu
Menu
Menu
Menu
Introduction: It is common in urban African settings for postpartum women to temporarily return to family in distant settings. We sought to explore mobility among peripartum HIV-positive women to understand the timing and motivation of travel, particularly vis-à-vis delivery, and how it may affect healthcare access. Methods: Using the same mobility measurements within three different studies, we examined long-distance travel of mother and infant before and after delivery in three diverse clinics within greater Johannesburg, South Africa (n = 150). Participants were interviewed prior to delivery at two sites (n = 125) and after delivery at one (n = 25). Quantitative and qualitative results are reported. Results: Among 150 women, median age was 29 years (IQR: 26 to 34) and 36.3% were employed. Overall, 76.7% of the participants were born in South Africa: 32.7% in Gauteng Province (Johannesburg area) and 44.0% in other South African provinces, but birthplace varied greatly by site. Almost half (44.0%) planned to travel around delivery; nearly all after delivery. Median duration of stay was 30 days (IQR: 24 to 90) overall, but varied from 60 days at two sites to just 7 days at another. Participants discussed travel to eight of South Africa’s nine provinces and four countries. Travel most frequently was to visit family, typically to receive help with the new baby. Nearly all the employed participants planned to return to work in Johannesburg after delivery, sometimes leaving the infant in the care of family outside of Johannesburg. All expressed their intent to continue HIV care for themselves and their infant, but few planned to seek care at the destination site, and care for the infant was emphasized over care for the mother. Conclusions: We identified frequent travel in the peripartum period with substantial differences in travel patterns by site. Participants more frequently discussed seeking care for the infant than for themselves. HIV-exposed children often were left in the care of family members in distant areas. Our results show the frequent mobility of women and infants in the peripartum period. This underscores the challenge of ensuring a continuity of HIV care in a fragmented healthcare system that is not adapted for a mobile population.
Data were collected by nesting the same mobility‐related questions in data collection tools of three separate studies with diverse study objectives. Table S1 provides additional information on each study; all studies were related to understanding and improving engagement in HIV care among peripartum women. We enrolled 150 adult (age ≥ 18 years) peripartum participants at three study sites in Johannesburg, South Africa; activities and participant eligibility varied slightly at each site according to the objectives of each parent study. Site one is a public health clinic operated by the City of Johannesburg that serves the Ivory Park region of eastern Johannesburg. Care is provided free of charge 21. Site one participants were recruited during routine antenatal care if they were pregnant, HIV positive, and able to speak and understand English. From May 2015 to March 2016, we enrolled 100 participants and conducted a one‐time questionnaire at enrolment. Site two is a large primary healthcare clinic operated by a non‐governmental organization based in Fourways in northern Johannesburg; clinic visits cost R110 (USD~8.50), but fees may be waived if clients are unable to pay. Site two participants were recruited during routine antenatal care and eligible for enrolment if pregnant and HIV positive; we enrolled 25 participants from October 2016 to April 2017. Site three is an academic research clinic located within Chris Hani Baragwanath Hospital in Soweto, on the southwestern edge of greater Johannesburg. Participants were already enrolled in a longitudinal maternal health cohort. Study visits are free and participants receive R150 (USD~11.60) for every completed visit. Women at site three were eligible for the present study if they were postpartum (gave birth 6 to 18 months prior to enrolment), HIV positive, and exhibiting a metabolic disorder (e.g. gestational diabetes). Interviews were conducted from August to December 2016. At sites two and three, we conducted a one‐time, in‐depth interview at enrolment based on a semi‐structured questionnaire guide. All interviews were conducted by a female trained research coordinator in the local language preference of the participant. The residential areas served by site one and two’s patient populations were developed in the 1990s 22, 23; both are densely‐populated areas with vast formal and informal settlements. By comparison, site three serves the population of a township that was formally established in 1963 24. The three diverse urban sites and mix of postpartum HIV‐infected women enabled us to examine migration patterns for diverse clients of a government clinic, a clinic run by an NGO, and a clinic run by an academic medical centre that aided women with metabolic disorders. Study questions are presented in Table S2. At all sites, participants were asked similar initial yes/no questions regarding travel outside of the Johannesburg area before and after delivery. For participants at sites one and two, we asked about intended travel, while we asked about actual travel for the postpartum women enrolled at site three. If travel was noted, we recorded details of the duration, reason, and plans for travel. The questionnaire used at site one collected categorical, short answer, and some open‐ended responses. At sites two and three, in‐depth interviews explored experiences of travel and pregnancy. At site one, questionnaire data were captured on paper forms, then entered into a REDCap (Research Electronic Data Capture) electronic database 25. At sites two and three, interviews were recorded and transcribed. When reporting the timing of travel, travel that was indicated to begin during pregnancy and end after delivery was marked as travel both before and after delivery. For example, if a pregnant participant reported planning to travel prior to delivery and staying through three months post‐delivery, both “before delivery” and “after delivery” travel would be noted for the same participant. All participants provided written informed consent prior to interviewing, and study activities were approved by the institutional review board of Vanderbilt University Medical Center, Boston University (site one), and the Human Research Ethics Committee of the University of the Witwatersrand. SAS® 9.4 (SAS Institute, Cary, NC, USA) was used for statistical analysis of quantitative data. Cohort characteristics are described using counts and proportions for categorical variables, and medians and interquartile ranges (IQR) for continuous data. Mobility within and outside of South Africa was mapped using ArcMap® 10.3.1 (Esri, Inc., Redlands, CA, USA). For the analysis of open‐ended data, coding, analysis, and reporting was completed by following the COREQ guidelines 26. Responses to questions related to mobility during the peripartum period (see list Table S2) were consolidated in REDCap and exported for hand‐coded analysis. Quotes were sorted by category, frequency distributions were examined, then quotes were read in detail to identify higher‐order themes and relationships. The analysis was rooted in the theoretical framework proposed by Phillips and Myer 27, which is an adaptation of the Social‐Ecological Model 28, and asserts that multi‐level factors interact to determine engagement in HIV care among pregnant and postpartum women. We report the themes identified – staying with family, childcare and children separated from the mother, and plans for continuing care – and highlight key, illustrative quotes.
