In March 2020, as COVID-19 was first recognized in the United States, reports emerged about how severe COVID-19 could be among pregnant individuals, and clinicians had to speculate about its potential effects on newborns. Out of concern for patient, clinician, and neonatal safety, system-based hospitals—hospitals owned by an organization operating outpatient practices, including primary care—rapidly implemented COVID-19 policies, restricting companions during childbirth and separating parents with COVID-19 from their babies. , Fear of COVID-19 and commensurate changes in hospital policies may have led some pregnant individuals to choose to deliver in independent hospitals, in birthing centers, or at home. , Our objectives were to compare place of birth between 2020 and years before the COVID-19 pandemic and to investigate whether changes in place of birth differed between system-owned and independent hospitals and urban and rural regions in Oregon.
We connected 2 Oregon data sources (Oregon Health Authority Vital Statistics and Oregon Perinatal Collaborative) with the Johns Hopkins COVID-19 data tracker and the novel Health Systems Provider Database to examine overall trends in place of birth (in the hospital, in a birthing center, or at home) between system-owned and independent hospitals and between urban and rural areas. Birth location was defined as rural or urban using county-type designation from the Centers for Medicaid & Medicare Services. The institutional review boards of the Boston Children’s Hospital and Harvard University approved this study.
We used regression modeling and Student t tests to compare predicted birth numbers with actual birth numbers by facility type, system-owned or independent status, and urban or rural location between 2020 and previous years in 2 ways: predicted 2020 values using (1) 2010–2019 data and (2) an average of 2 preceding years, 2018–2019. These models produced (1) a predicted number of births by place for 2020 vs previous years had births followed existing trends and (2) a percentage change in place of births from 2018 to 2020. We considered a 2-sided P value of.05 to be significant.
Comparing actual births in 2020 with predicted births in 2020, decreases in hospital birth were statistically significantly from March 2020 to December 2020, the period affected by COVID-19 ( P <.001 for all months), and the decrease remained greater than predicted by secular trends. Decreases in system-owned hospital births were significant from March 2020 to December 2020 ( P ≤.001 for all months). Increases in birthing center and at-home births were statistically significant in the spring of 2020 ( P <.01) during the first wave of COVID-19. ( Table 1 )
|Month||Year||Facility type||Actual births in 2020||Predicted births in 2020||Difference (actual vs predicted)||Student t test P value|
|Jan.||2020||Birthing center||38||50||−12||<.001 a|
|April||2020||Birthing center||69||57||+12||<.001 a|
|Oct.||2020||Birthing center||65||52||+13||<.001 a|