Comparison of Severe Maternal Morbidities Associated With Delivery During Periods of Circulation of Specific SARS-CoV-2 Variants

Key Points Question Does the association between SARS-CoV-2 infection and severe maternal morbidity (SMM), including nonrespiratory complications, vary by viral strain? Findings In this retrospective cohort study of 3129 patients with SARS-CoV-2 infection and 12 504 patients without infection giving birth in a large US health system between March 2020 and January 2022, the risk of SMM associated with SARS-CoV-2 infection was significantly higher during the phase of the pandemic when the Delta variant was predominant (July 2021-November 2021). This association was also noted specifically for both respiratory and nonrespiratory SMM. Meaning These findings highlight the importance of the prevention of SARS-CoV-2 infection in pregnant individuals and the consideration of infection as a risk factor for adverse peripartum maternal outcomes.


Introduction
Between January 2020 and February 2022, more than 170 000 pregnant people in the US were infected by SARS-CoV-2 and 29 000 pregnant people were hospitalized with COVID-19. 1 Pregnancy can be associated with a hypercoagulable state, altered immune function, and increased susceptibility to hypoxemia, which could potentially worsen maternal outcomes with SARS-CoV-2 infection. 2,3 The findings of studies from early in the COVID-19 pandemic on the impact of SARS-CoV-2 infection on pregnancy outcomes have not been definitive, perhaps in part owing to limited sample sizes. Some studies found higher rates of maternal complications, such as caesarian and preterm deliveries, 4,5 while others did not identify an increase in maternal risk. 6,7 As the pandemic evolved, with greater spread and the emergence of SARS-CoV-2 variants, more studies reported an association between SARS-CoV-2 infection and higher rates of poor maternal outcomes, including eclampsia, preeclampsia, HELLP syndrome (hemolysis, elevated liver enzymes and low platelets), sepsis, intensive care unit admission, preterm birth and low birth weight, and, more recently, maternal death or serious morbidity related to hypertensive disorders of pregnancy, postpartum hemorrhage, or infection other than SARS-CoV-2. [8][9][10][11] The pandemic has been characterized by several waves defined by viral strains that are responsible for the preponderance of infections. In July 2021, the Delta (B.1.617.2) variant of SARS-CoV-2 became the predominant variant in the US. 12 One study 13 from a single hospital found higher COVID-19 caseloads and severity of illness among pregnant people during the Delta surge compared with previous surges. However, this study did not assess adverse maternal outcomes around the time of delivery. Another early study 14 at 1 institution presented findings suggesting serious morbidity and adverse perinatal outcomes associated with the Delta variant in pregnancy. In December 2021, Omicron became the predominant strain in the US, and although this strain continued to be responsible for hospitalizations and deaths, disease severity was generally less than the Delta strain. 15 Maternal outcomes for people infected with the Omicron variant are not well characterized.
The primary objective of our study is to evaluate the association between SARS-CoV-2 infection and severe maternal morbidities (SMM), as defined by the US Centers for Disease Control and Prevention (CDC), in pregnant patients during pandemic periods when the wild-type strain or Alpha (B.1.1.7), Delta, or Omicron (B.1.1.529) variants of SARS-CoV-2 predominated. The CDC defines SMM as unexpected outcomes of labor and delivery, consisting of 21 health indicators, that result in significant short-or long-term consequences to an individual's health. 16 To address this question, we studied pregnant patients delivering in a geographically diverse multistate US health system between March 2020 and January 2022.

Methods
This cohort study was deemed exempt from review and informed consent by the Ascension Seton

Data Sources
SMM outcomes, SARS-CoV-2 infection status, and other patient information were derived from a combination of health system data, including acute care claims data for inpatient encounters and standardized data elements from electronic health records. Race and ethnicity were self-reported for most patients, but the source could not be confirmed for all reviewed records. Race was categorized as Black, White, and other, which included all races other than White or Black, as well as those for whom race was unknown or for whom documentation of race was declined. If race was unknown but ethnicity was Hispanic or Latino, the person was classified as Hispanic or Latino. If ethnicity was unknown, non-Hispanic or Latino ethnicity was assumed. Race and ethnicity were assessed because of potential differential associations with maternal outcomes independent of SARS-CoV-2 infection.
Missing demographic data were not altered and are reported as unknown. For each individual SMM, a null value was considered to be a nonevent and not counted in the totals.

Outcomes
The primary outcome was any SMM event occurring during the hospitalization for delivery. These events are defined by the CDC as including unexpected outcomes of labor and delivery that result in significant short-or long-term consequences to an individual's health. 16

Statistical Analysis
Categorical variables were summarized as frequencies and percentages, and were compared using

All SMM
Patients with SARS-CoV-2 infection had significantly higher rates of any SMM event than those without infection in all time periods, except Omicron (

Secondary Outcomes
In an analysis restricted only to respiratory SMM, patients with SARS-CoV-2 infection had significantly higher observed rates of respiratory SMM for all variants (Table 2). Consistent with the primary    (Figure, B). In an analysis restricted only to nontransfusion SMM, consistent with analysis of the primary outcome, the risk of adverse events was higher among patients with SARS-  (Figure, C).

Sensitivity Analysis
In secondary analyses, the control group was restricted only to patients with a documented negative

JAMA Network Open | Obstetrics and Gynecology
Severe  was universal in the later periods, and because vaccination is generally associated with less severe SARS-CoV-2 infection, the differences in outcomes in the later periods relative to the original period associated with infection would, if anything, be biased toward the null. Second, not all patients early in the wild-type strain period were tested for SARS-CoV-2 infection. In periods when universal testing was used, we were unable to differentiate between incidentally discovered SARS-CoV2 infection vs symptomatic COVID-19. In principal analysis, patients who were not tested were assumed to be negative. This raises the question of potential surveillance bias early in the wild-type strain period, when patients with symptoms were more likely to be tested than those without symptoms.
However, the results remained consistent even when we removed patients who were not tested for SARS-CoV-2 from the analysis. Third, we were unable to identify patients who had experienced a SARS-CoV-2 infection during their pregnancy and had recovered to the point when testing would not identify an active infection. Thus, it is possible that some of the adverse outcomes in patients without a positive test result could have been attributable to an earlier infection. Again, this issue would bias the results of the study toward the null. Fourth, we did not have sequencing data to identify the strain in individuals with SARS-CoV-2 infection. We used time period as a proxy for the dominant strain of SARS-CoV-2. However, epidemiology data suggest that the selected strains (wild-type, Alpha, Delta, and Omicron) were dominant and accounted for more than 70% of infections during each of the 4 time periods used. Fifth, because the study is limited to hospitals within a single health system, the results may not reflect outcomes in the broader US population. However, the population of sites and individuals is geographically and demographically diverse.

Conclusions
In this cohort study in a large US health system, infection with SARS-CoV-2 at the time of delivery throughout the pandemic was associated with higher rates of SMM; however, this association was particularly strong with the Delta strain. This variable association with viral strain was present both for respiratory and nonrespiratory complications, as well as complications not related to blood transfusion. These findings highlight the importance of SARS-CoV-2 prevention in pregnant patients and consideration of infection as an indicator of high risk in the peripartum period.