Accuracy of gestational age estimation from last menstrual period among women seeking abortion in South Africa, with a view to task sharing: a mixed methods study

BACKGROUND: The requirement for ultrasound to establish gestational age among women seeking abortion can be a barrier to access. Last menstrual period dating without clinical examination should be a reasonable alternative among selected women, and if reliable, can be task-shared with non-clinicians. This study determines the accuracy of gestational age estimation using last menstrual period (LMP) assessed by community health care workers (CHWs), and explores providers' and CHWs' perspectives on task sharing this activity. The study purpose is to expand access to early medical abortion services. METHODS: We conducted a multi-center cross-sectional study at four urban non-governmental reproductive health clinics in South Africa. CHWs interviewed women seeking abortion, recorded their LMP and gestational age from a pregnancy wheel if within 63 days. Thereafter, providers performed a standard examination including ultrasound to determine gestational age. Lastly, investigators calculated gestational age for all LMP dates recorded by CHWs. We compared mean gestational age from LMP dates to mean gestational age by ultrasound using t-tests and calculated proportions for those incorrectly assessed as eligible for medical abortion from LMP. In addition, in-depth interviews were conducted with six providers and seven CHWs. RESULTS: Mean gestational age was 5 days (by pregnancy wheel) and 9 days (by LMP calculation) less than ultrasound gestational age. Twelve percent of women were eligible for medical abortion by LMP calculation but ineligible by ultrasound. Uncertainty of LMP date was associated with incorrect assessment of gestational age eligibility for medical abortion (p = 0.015). For women certain their LMP date was within 56 days, 3% had ultrasound gestational ages >70 days. In general, providers and CHWs were in favour of task sharing screening and referral for abortion, but were doubtful that women reported accurate LMP dates. Different perspectives emerged on how to implement task sharing gestational age eligibility for medical abortion. CONCLUSIONS: If LMP recall is within 56 days, most women will be eligible for early medical abortion and LMP can substitute for ultrasound dating. Task sharing gestational age estimation is feasible in South Africa, but its implementation should meet women's privacy needs and address healthcare workers' concerns on managing any procedural risk