Social Distancing During the COVID-19 Pandemic and Neonatal Mortality in the US

Abstract

ImportanceNeonatal mortality is a major public health concern that was potentially impacted by the COVID-19 pandemic. To prepare for future health crises, it is important to investigate whether COVID-19 pandemic–related interventions were associated with changes in neonatal mortality.ObjectiveTo investigate whether social distancing during the pandemic was associated with a higher neonatal mortality rate.Design, Setting, and ParticipantsThis cohort study examined maternal-linked birth and infant death records from the National Center for Health Statistics, a population-level US database, from 2016 through 2020. The mortality rates were correlated using machine learning–based autoregressive integrated moving average (ARIMA) models with the social distancing index (SDI). The reference period was January 2016 through February 2020, and the pandemic period was March through December 2020. Statistical analysis was performed from March 2023 to May 2024.ExposuresSDI, computed from 6 mobility metrics.Main Outcomes and MeasuresThe primary outcome was neonatal mortality rate, defined as death at age less than 28 days.ResultsThe study included 18 011 173 births, of which 15 136 596 were from the reference period (7 753 555 [51.22%] male; 11 643 094 [76.92%] with maternal age of 20 to 34 years) and 2 874 577 were from the pandemic period (1 472 539 [51.23%] male; 2 190 158 [76.19%] with maternal age of 20 to 34 years). Through ARIMA-adjusted analyses, accounting for the declining mortality trend in the reference period, the mortality rates during the pandemic period did not significantly differ from the expected rates. SDI did not exhibit significant correlations with neonatal mortality (unadjusted: correlation coefficient [CC], 0.14 [95% CI, −0.53 to 0.70]; ARIMA adjusted: CC, 0.29 [95% CI, −0.41 to 0.77]), early neonatal mortality (unadjusted: CC, 0.33 [95% CI, −0.37 to 0.79]; ARIMA adjusted: CC, 0.45 [95% CI, −0.24 to 0.84]), and infant mortality (unadjusted: CC, −0.09 [95% CI, −0.68 to 0.57]; ARIMA adjusted: CC, 0.35 [95% CI, −0.35 to 0.80]). However, lag analyses found that SDI was associated with higher neonatal and early neonatal mortality rates with a 2-month lag period, but not with infant mortality rate. SDI was also associated with increases in 22-to-27 weeks’ and 28-to-32 weeks’ preterm delivery with a 1-month lag period.Conclusions and RelevanceIn this population-level study of National Center for Health Statistics databases, neonatal, early neonatal, and infant mortality rates did not increase during the initial COVID-19 pandemic period. However, associations were observed between the pandemic period social distancing measures and higher rates of neonatal and early neonatal mortality, as well as preterm birth rate with a lag period, suggesting the importance of monitoring infant health outcomes following pandemic-related population behavior changes.