SETTING AND SAMPLING. The hospitals in this study were purposively chosen in areas with high and low OB coverage access based on primary care service area distinctions (PCSA). To understand the health care landscape, PCSA distinctions are defined by the collection of counties in which >30% of county residents receive their primary care which range from adequate, at-risk, deficit to no OB services (Xxxxxxxx, 2015). The research team purposively selected hospitals with the emergency departments. Emergency departments were chosen if they had frequent utilization by patients due to its convenience and safety net features. This sampling frame allowed for heterogeneity of hospital distribution across the public health districts of Georgia, in areas of varying access to OB services. Hospitals located in the metropolitan Atlanta area were excluded because metropolitan Atlanta’s is not experiencing OB shortages. Additionally, the varied provider roles demonstrated individual-level insight into the differences and similarities in early pregnancy bleeding management and organizational factors related to the provision of early pregnancy bleeding care. We excluded hospitals without an ED. In sum, 28 hospitals were invited to participate by phone calls, email, and letters to ED directors. We then purposively selected approximately two to four emergency department clinical staff representing diverse roles from eight to ten different hospitals providing heterogeneity in geographic location, hospital type, ED trauma designation and whether the hospital had a labor and delivery unit . Ten hospitals with 32 providers ultimately participated with institutional support. We obtained approval from the Emory Institutional Review Board (IRB) by a review of all study materials and was determined that although it is human subjects research, it is exempt from further IRB review and approval. We then obtained oral consent from all the interviewees. A Certificate of Confidentiality was obtained to recruit providers to alleviate unease discussing reproductive health issues as ED providers and not OB. Individual responses and hospitals were de-identified.
SETTING AND SAMPLING. The location of this investigation was in Soweto, South Africa, the southwest township of Johannesburg. It is estimated that over 1.5 million individuals inhabit the area, most of who are black Africans (GeoNames). Although there is a range of housing situations, the area is primarily characterized by poor infrastructure and high unemployment. Soweto is most internationally known for it’s historical political movements during the apartheid (Johannesburg). The Birth to 20 (BT20) cohort and is the largest and longest running study of child and adolescent development on the African continent. During a seven week period in 1990, 3,273 children were born in public hospitals in the Johannesburg-Soweto area and recruited for a longitudinal study. Over the last 20 years, these children and their families have been monitored by research staff. Attrition over two decades has been comparatively low (30%), mostly occurring during children’s infancy and early childhood. Approximately 2,300 children and their families remain involved with the study. The sample is roughly representative of the demographic parameters of South Africa with equal numbers of male and female participants (Xxxxxxx et al., 2007). The present study was conducted from May to July 2012 in Soweto, South Africa. Participants were purposively sampled from a subset of fifty BT20 cohort members who had been randomly selected and recently participated in the cohort’s periodic Young Adolescent Health Survey (YAHS). Participants were chosen from this subset based on their gender (14 female/9 male), and whether or not they had utilized health services within the last six months (13 users/10 non-users) (Table 1). All participants were either 21 or 22 years of age and were ethnically Zulu.
