Racial/Ethnic Disparities in Very Preterm Birth and Preterm Birth Before and During the COVID-19 Pandemic

Key Points Question Was the first wave of the coronavirus disease 2019 (COVID-19) pandemic associated with exacerbated racial/ethnic disparities in preterm birth in New York City? Findings This cross-sectional study found that racial/ethnic disparities in very preterm birth and preterm birth among 8026 women were similar during the first wave of the COVID-19 pandemic in New York City compared with the same period the year prior. Meaning Monitoring of racial/ethnic disparities in adverse birth outcomes as the COVID-19 pandemic continues is warranted.


Introduction
Black women and infants experience persistent disadvantage in birth outcomes in the US. Black infants are 50% more like to be born preterm and twice as likely to be born very preterm. 1 In New York City (NYC), Latina women are also at increased risk of delivering preterm. 2 The coronavirus disease 2019 (COVID-19) pandemic threatens to exacerbate existing preterm birth (PTB) and very preterm birth (VPTB) disparities, yet data are scarce to inform this pressing concern.
The COVID-19 pandemic is replicating existing structures of inequality and disproportionately harming communities of color. Black and Latina women are more likely to be infected with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) at delivery 3 and more likely to experience pandemic-related psychosocial and economic impacts during pregnancy. 4 Research thus far on obstetric outcomes during the pandemic typically report a modest increased risk of PTB among women with COVID-19 5 and little to no increased risk among women who are asymptomatic but test positive for SARS-CoV-2. 6 To date, research has not adequately examined the association of the COVID-19 pandemic with PTB from a health equity perspective.
To fill this gap, we conducted a difference-in-differences (DID) analysis of electronic medical records of 8026 women from 2 hospitals in a NYC health system, which draws patients from the Bronx, Manhattan, Queens, and Brooklyn. We compared racial/ethnic differences in PTB during the first wave of the pandemic with the year prior.

Study Design
We created a pandemic cohort of 3834 women who delivered between March 28, 2020, the date universal testing of women undergoing labor and delivery began, and July 31, 2020. We did not include births prior to universal testing owing to unknown SARS-CoV-2 status. We created a prepandemic cohort of 4192 women who delivered children from March 28 to July 31, 2019 (eFigure in the Supplement). We followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cross-sectional studies. The institutional review board of the Icahn School of Medicine at Mount Sinai approved the study and waived the requirement for obtaining informed consent because risk to participants was considered minimal and the study could not reasonably been conducted otherwise. No one received compensation or was offered any incentive for participating in this study.
Reverse transcription polymerase chain reaction tests for the presence of SARS-CoV-2 were performed using samples obtained via nasopharyngeal swab. We used electronic medical records to ascertain all variables. Participant race and ethnicity were self-reported on admission and classified according to the US Office of Management and Budget standards. We estimated VPTB (<32 weeks completed gestation) and PTB (<37 weeks completed gestation) using the clinician's best estimate of gestational age.

Statistical Analysis
We used log binomial regression to estimate a DID equation with main effects for Black vs White risk difference, pandemic vs prepandemic cohort risk difference, and an interaction term representing the DID estimator. The DID estimator estimates the additional disparity resulting from the pandemic beyond disparities that had previously existed. We repeated the model for Latina vs White women, restricting the pandemic cohort to positive or negative SARS-CoV-2 test status, and singleton births.
We estimated multivariable models adjusting DID estimates for age, insurance type, prepregnancy body mass index, and parity. The DID approach is typically robust to confounding given the balance of covariates between treatment groups is constant over time. In multivariable analyses, we excluded observations with missing values (<4% for body mass index, <3% for polymerase chain reaction, and <1% all others). All analyses were conducted using SAS, version 9.4 (SAS Institute Inc). A 2-sided P < .05 was considered statistically significant.    (Table 3). Covariate-adjusted estimates were similar (Table 3). Analyses stratified by SARS-CoV-2 status found DID estimators in the SARS-CoV-2negative group were similar to the overall cohort ( Table 4). We did not conduct DID analyses in the SARS-CoV-2-positive group owing to low counts of outcomes ( Table 5). The DID estimators for VPTB and PTB were similar for singleton births.

Discussion
We found no evidence that the first wave of the COVID-19 pandemic increased racial/ethnic disparities in preterm birth in NYC. Results were similar by SARS-CoV-2 status.
Our findings should be considered in the context of a current hypothesis that the lockdown has lessened the risk of PTB for women. [7][8][9] In contrast to this hypothesis, in Philadelphia, PTB did not change, 10 whereas in California, VPTB increased slightly among Latina mothers. 11 Unlike other recent reports, 10,11 we explicitly tested racial/ethnic disparities with a robust DID design and were able to stratify our results by active SARS-CoV-2 infection.
Researchers have proposed potential reasons for a decrease in PTB during the COVID-19 pandemic, such as a decrease in known risk factors for PTB, including occupational environment, pollution, or stress. 9 However, any benefit from COVID-19-related restrictions may be less prevalent among Black and Latina women in NYC, who may be more likely to be essential workers 12 and to experience higher rates of COVID-19 pandemic-related stress, anxiety, and food insecurity. 4,13,14 Black and Latina women are also more likely than White women to experience loss and trauma due to  Decreased access to prenatal care, increased incidence of pregnancy complications, or decreased control of chronic conditions may also play a role. Despite these potential mechanisms, we did not find an increase in racial/ethnic differences in PTB. Regardless, given the known inequitable repercussions of COVID-19 in Black and Latinx populations, continued monitoring of racial/ethnic disparities in preterm birth is warranted.

Limitations and Strengths
The limitations of our study include the lack of information on maternal comorbidities and SARS-CoV-2 infections prior to delivery. If healthy White women disproportionately left NYC to deliver their infants during the pandemic, this selection bias would cause a greater proportion of preterm births in the White pandemic cohort, and underestimate an increase in PTB disparities. Evaluation of covariates by race/ethnicity and cohort suggested this bias was minimal. Another limitation is the lack of precision to calculate DID estimators among women who tested positive for SARS-CoV-2. Our