TY - JOUR T1 - Pregnancy-related healthcare utilisation in Agincourt, South Africa, 1993–2018: a longitudinal surveillance study of rural mothers JF - BMJ Global Health JO - BMJ Global Health DO - 10.1136/bmjgh-2021-006915 VL - 6 IS - 10 SP - e006915 AU - Daniel E Sack AU - Ryan G Wagner AU - Daniel Ohene-Kwofie AU - Chodziwadziwa W Kabudula AU - Jessica Price AU - Carren Ginsburg AU - Carolyn M Audet Y1 - 2021/10/01 UR - http://gh.bmj.com/content/6/10/e006915.abstract N2 - Introduction Pregnancy-related health services, an important mediator of global health priorities, require robust health infrastructure. We described pregnancy-related healthcare utilisation among rural South African women from 1993 to 2018, a period of social, political and economic transition.Methods We included participants enrolled in the Agincourt Health and Socio-Demographic Surveillance System in Mpumalanga Province, South Africa, a population-based longitudinal cohort, who reported pregnancy between 1993 and 2018. We assessed age, antenatal visits, years of education, pregnancy intention, nationality, residency status, previous pregnancies, prepregnancy and postpregnancy contraceptive use, and student status over the study period and modelled predictors of antenatal care utilisation (ordinal), skilled birth attendant presence (logistic) and delivery at a health facility (logistic).Results Between 1993 and 2018, 51 355 pregnancies occurred. Median antenatal visits, skilled birth attendant presence and healthcare facility deliveries increased over time. Delivery in 2018 vs 2004 was associated with an increased likelihood of ≥1 additional antenatal visits (adjusted OR (aOR) 10.81, 95% CI 9.99 to 11.71), skilled birth attendant presence (aOR 4.58, 95% CI 3.70 to 5.67) and delivery at a health facility (aOR 3.78, 95% CI 3.15 to 4.54). Women of Mozambican origin were less likely to deliver with a skilled birth attendant (aOR 0.42, 95% CI 0.39 to 0.45) or at a health facility (aOR 0.43, 95% CI 0.41 to 0.46) versus South Africans. Temporary migrants reported fewer antenatal visits (aOR 0.35, 95% CI 0.33 to 0.38) but were more likely to deliver with a skilled birth attendant (aOR 1.91, 95% CI 1.66 to 2.2) or at a health facility (aOR 1.4, 95% CI 1.24 to 1.58) versus permanent residents.Conclusion Pregnancy-related healthcare utilisation and skilled birth attendant presence at delivery have increased steadily since 1993 in rural northeastern South Africa, aligning with health policy changes enacted during this time. However, mothers of Mozambican descent are still less likely to use free care, which requires further study and policy interventions.Data are available on reasonable request, after approval of any requisite Agincourt Health and Demographic Surveillance Systems processes. All data cleaning and analysis code are available at https://github.com/dannysack/pregnancy_agincourt. ER -