Neonatal Mortality Risk in Planned Home Birth, 2014-2024.

Abstract

Introduction: Planned home birth increased sharply in the United States after 2019. This study examined the evolution of neonatal mortality risk in planned home births compared with hospital births from 2014 to 2024.

Methods: This is a retrospective observational analysis using US natality detail files from 2014 to 2024. The primary outcome was neonatal death. This study used a pooled logistic regression to estimate the overall odds ratio (OR) for home birth versus hospital birth, interacting home birth with period indicators (2014-2016, 2017-2019, 2020-2021, 2022-2024). The sensitivity analyses also considered how much the individual "low-risk" sample restrictions contributed.

Results: Planned home birth was associated with a higher OR for neonatal death relative to hospital birth (OR, 1.71; 95% CI, 1.23-2.37), but the adjusted OR (aOR) was not statistically significant (aOR, 1.30; 95% CI, 0.92-1.83) from 2014 to 2024. Joint Wald tests indicated that the prepandemic aORs (2014-2016 and 2017-2019) were not statistically different from the 2020-2021 reference period (P = .234), whereas the 2022-2024 aOR was weakly significant (P = .055). In the sensitivity analyses, re-including pregnancies with higher-risk conditions increased the aOR. Adding back noncredentialed midwives to the sample had the largest effect on the aOR (aOR, 1.70; P < .001; q < .001). Adding back breech presentations (aOR, 1.46; P = .014; q = .060), congenital abnormalities (aOR, 1.37; P = .030; q = .087), postterm births (aOR, 1.37; P = .056; q = .125), and missing controls (aOR, 1.36; P = .068; q = .129) increased the aOR as well.

Discussion: For low-risk births attended by certified nurse-midwives/certified midwives or physicians, the neonatal mortality differential between planned home and hospital settings was modest from 2014 to 2024. However, relaxing clinical risk criteria or including less-regulated provider types increased the estimated risk of neonatal mortality.