Background: This study examined the effect of COVID-19 on health facility delivery in Ethiopia, using longitudinal data with a pre-post exposure to the pandemic to assess change in facility delivery patterns nationally and by urban and rural strata, covering six regions of Ethiopia representing 91% of the population.

Methods: We conducted design-based multivariable multinomial logistic regression comparing home, lower-level facility, and hospital delivery with date of birth as a spline term, with a knot indicating births before and on/after April 8, 2020. Women were eligible to participate if they were currently pregnant or less than six weeks postpartum and were recontacted at six weeks, six months, and one year postpartum; 2,889 women were eligible and 2,855 enrolled, with data in this paper coming from the six-week interview (follow-up rate of 88.9%, 2,537 women).

Results: In urban areas, women who delivered during the COVID-19 pandemic had a 77% reduced relative risk of delivering in a hospital relative to women who delivered prior to the pandemic (aRRR: 0.23, 95% CI: 0.07-0.71). There were no significant differences between the pre- and COVID-19 periods within rural strata, where the majority of women deliver at home (55.6%). Overall, the effect was non-significant at a national level. Among women who delivered during the COVID-19 pandemic, 20.0% of urban women said COVID-19 affected where they delivered relative to 8.7% of rural women.

Conclusions: Delivery patterns in urban areas changed during the early months of the COVID-19 pandemic, but there was no evidence of large-scale declines of hospital delivery at the national level. Concerns about COVID-19 transmission in health facilities and ensuring lower-level facilities are equipped to address obstetric emergencies are critical to address.