Abstract

Introduction

Pregnant women are more susceptible to viral respiratory infections owing to immunologic and physiological adaptations of pregnancy.1 An early Chinese report of coronavirus disease 2019 (COVID-19), the disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), found that the risk of severe disease in pregnant patients was similar to the general population.2 This was also observed in initial studies in the United States, including a report from New York3 and Washington.4

Recent data from the National Notifiable Diseases Surveillance System, reported by the Centers for Disease Control and Prevention (CDC) on June 25, 2020, compared outcomes of 8207 pregnant and 83,205 nonpregnant women with COVID-19 (January 22 to June 7, 2020).5 The authors included all laboratory-confirmed infections with SARS-CoV-2 among women at the age of 15 to 44 years from all of the United States and Washington, DC. The main findings were that pregnant women with COVID-19 were more likely to be hospitalized, require intubation and mechanical ventilation, and be admitted to an intensive care unit (ICU) than nonpregnant women. The authors concluded that pregnant women should be counseled about the potential for severe COVID-19 disease, despite a low absolute risk of ICU admissions and the need for mechanical ventilatory support. This report, which generated substantial press,6 ^, 7 could not distinguish hospitalizations related to COVID-19 from obstetrical indications and could not ascertain whether complications from COVID-19 or pregnancy resulted in the escalation of medical care and increased morbidity.1

To date, medical care and guidance from professional societies during the pandemic have been largely based on case reports and case series and epidemiologic studies that compared outcomes of pregnant women with COVID-19 with nonpregnant women. Most of these studies lacked an appropriate control group, a common limitation that has complicated our understanding of COVID-19’s impact on pregnancy. Therefore, to quantify the maternal and neonatal risks associated with COVID-19 in pregnancy and to describe the epidemiology and risk factors for morbidity associated with COVID-19, we undertook a matched case-control study. Associations were evaluated for all COVID-19 patients and for diseases classified as mild vs severe or critical diseases.2 We also characterized the epidemiology and identified risk factors for morbidity associated with COVID-19 in pregnancy.

Materials and Methods

We performed a matched case-control study at the Robert Wood Johnson University Hospital, a regional perinatal center in New Brunswick, NJ. The Institutional Review Board of the Rutgers Robert Wood Johnson Medical School, NJ, granted the ethics approval under a waiver of informed consent (PRO2020000854). The study followed the Strengthening the Reporting of Observational Studies in Epidemiology reporting guideline for case-control studies.

Consecutive pregnant patients with COVID-19 who were admitted to the hospital were enrolled from March 11 to June 11, 2020; this period corresponds to the first 3 months of the SARS-CoV-2 pandemic.8 COVID-19 testing to detect SARS-CoV-2 infection was performed by nasopharyngeal swab and quantitative polymerase-chain-reaction test. Before April 10, 2020, patients admitted to the labor and delivery unit were tested if they had symptoms of SARS-CoV-2 infection, had a recent travel to high-risk countries with a prevalent disease, or had a direct contact with someone who traveled to high-risk countries or who had COVID-19 (eg, considered to have a high-risk exposure). On April 10, 2020, we implemented universal COVID-19 testing for all pregnant patients at the time of hospital admission, regardless of symptoms or exposure history.

Consecutive patients with COVID-19 were prospectively identified in a clinical database. Patients were considered to be cases if they had a positive COVID-19 test result and delivered between 16.0 and 41.6 weeks’ gestation. Cases were categorized as mild, severe, or critical disease according to previously published criteria.9 Mild disease was defined as nonpneumonia and mild pneumonia; severe disease was defined as dyspnea, respiratory rate of ≥30/minute, blood oxygen saturation of ≤93%, partial pressure of arterial oxygen to fraction of inspired oxygen ratio of <300, and lung infiltrates of >50% on chest X-ray; and critical disease was defined as respiratory failure, septic shock, and multiple organ failure. Patients were eligible to be cases if the initial COVID-19 testing was negative, but the subsequent testing during the delivery hospitalization became positive. Patients were excluded if they were persons under investigation (PUIs) without confirmatory testing or had a negative COVID-19 test result or if they were hospitalized but discharged before delivery.

Each COVID-19 case was matched to 2 controls by delivery date. Before April 10, controls were selected as the first 2 patients who delivered between 16.0 and 41.6 weeks’ gestation on the same date as the cases if they were asymptomatic or had a negative COVID-19 test result. After April 10, controls were selected if they had a negative COVID-19 test result and delivered on the same date as the cases. On days with 2 or more cases, we identified the next 2 eligible controls as potential matches.

All patient data were abstracted from the electronic medical record. The primary outcomes were composites of adverse maternal outcomes and adverse neonatal outcomes. The maternal composite included preeclampsia (defined according to the American College of Obstetricians and Gynecologists’ Task Force on Hypertension in Pregnancy10), venous thromboembolism, antepartum admission (defined as hospital admission for obstetrical or nonobstetrical indications for inpatient management for >48 hours), ICU admission, need for mechanical ventilation, supplemental oxygen, or maternal death. The neonatal composite included respiratory distress syndrome (RDS; defined as the need for supplemental oxygen and the presence of typical radiographic findings in the absence of other causes for respiratory distress), intraventricular hemorrhage (IVH; defined as grade 1–4 hemorrhages), necrotizing enterocolitis (defined as radiographic or operative findings consistent with perforation), 5-minute Apgar score of <5, persistent category 2 fetal heart rate tracing despite intrauterine resuscitation, or neonatal death. The primary maternal and neonatal outcomes were examined as a composite owing to a small number of patients with the specific complication, and importantly, all of the individual outcomes were regarded as “competing risks.”

Secondary maternal outcomes included components of the composite outcome and preterm delivery (defined as delivery before 37 weeks’ gestation, <34 weeks’ gestation, and <28 weeks’ gestation), mode of delivery (cesarean or vaginal delivery), intrauterine fetal demise, length of hospital stay, and chorioamnionitis. Other neonatal outcomes included birthweight, neonatal ICU (NICU) admission, 1- and 5-minute Apgar scores, and length of hospital stay. Per institutional policy, all babies born to COVID-19–positive mothers during the study period were considered PUI; these babies were admitted to the NICU until a negative COVID-19 test result at 36 hours of life. Persistent category 2 fetal heart rate tracing despite intrauterine resuscitation, chorioamnionitis, and other clinical outcomes were defined by the providers managing each patient. If the provider believed that the patient met the criteria for 1 of these diagnoses, it was noted in the medical record.

Gestational age was based on the best obstetrical estimate.11 Maternal demographics were abstracted from the electronic medical record, including age, gravidity, parity, body mass index at delivery, race, and medical comorbidities. Renal disease was defined as baseline proteinuria of >300 mg/24 hours or creatinine of >1.1 mg/dL. An immunocompromised state was defined as HIV infection or chronic steroid use. Anemia was defined as admission hemoglobin of <10.5 mg/dL.

Data related to COVID-19 symptoms, laboratory evaluations, and clinical management were also abstracted. Maternal symptoms included fever (defined as temperature of >100.4°F), cough, shortness of breath, chest pain, diarrhea, myalgias, and sore throat. Laboratory evaluation included the results of routine complete blood counts and comprehensive metabolic panels (the latter when ordered for clinical purposes). Clinical management included supplemental oxygen, hydroxychloroquine, remdesivir, antibiotic agents, bronchodilators, mechanical ventilation, steroids, and ICU admission.

Results

During the 3-month study period, there were 61 pregnant patients diagnosed as having COVID-19 who delivered at our institution and met the inclusion criteria (cases). Each case was matched to 2 controls by delivery date; 11 cases (18%) were enrolled in the first month, 28 (45.9%) were enrolled in the second, and 22 (36.1%) were enrolled in the third. Among the cases, disease severity was mild (n=54 [88.5%]), severe (n=6 [9.8%]), and critical (n=1 [1.6%]). Demographic characteristics for cases and controls are presented in Table 1 . Overall, the groups were well matched, but there were more white women with COVID-19 (58.3% vs 42.7%). Notably, 142 (77.6%) patients were healthy and without medical comorbidities. However, compared with controls, patients with severe or critical disease had higher rates of medical comorbidities (42.9% vs 24.6%), such as a diabetes mellitus (28.6% vs 16.4%), chronic hypertension (28.6% vs 4.9%), renal disease (14.4% vs 0%), and anemia (14.3% vs 3.3%). In addition, cases with severe or critical disease were more likely to be Hispanic (57.1% vs 26.2%) and black (14.1% vs 6.6%).

Overall, 61.1% of patients with mild COVID-19 were asymptomatic. Of the 11 patients who were enrolled during the first month of the study period, 1 patient (9.1%) was asymptomatic; testing of the asymptomatic patient was because of a high-risk exposure. During the latter 2 months of the study period, 32 patients (64%) were asymptomatic. The most common symptoms for mild disease were cough, fever, and myalgias (Table 2 ). In contrast, all patients with severe or critical disease reported symptoms, with cough, shortness of breath, and fever being the most common symptoms. All patients with severe or critical disease required supplemental oxygen, and some also received other interventions such as hydroxychloroquine (n=4 [57.1%]) in the early part of the study period and corticosteroids (n=4 [57.1%]) in the latter part. In contrast, only 1 patient with mild disease received treatment; the patient was treated with antibiotic agents for a bacterial pneumonia.

Laboratory results are presented in Table 3 . Patients with mild disease had similar mean white blood cell and platelet counts and median lymphocyte counts and transaminases. In contrast, cases with severe or critical COVID-19 had higher risks of white blood cell count of <9.5 cells/L, platelets of <150,000/mm³, lymphocytes of <10⁹ cells/L, and elevated alanine aminotransferase of >45 units/L or aspartate transaminase of >35 units/L than controls.

Obstetrical and neonatal outcomes are presented in Table 4 . Cases with mild disease had similar obstetrical outcomes compared with controls. However, cases with severe or critical disease had more adverse obstetrical outcomes, including earlier gestational age of delivery (34.0 vs 38.7 weeks; mean difference, 4.8 weeks; 95% CI, 2.6–6.9). Cases were more likely to deliver preterm at <37, <34, and <28 weeks’ gestation than controls. Patients with severe or critical COVID-19 also had higher risks of antepartum admissions, cesarean delivery, chorioamnionitis, preeclampsia, and persistent category 2 fetal heart rate tracing despite intrauterine resuscitation and required longer hospital stays than controls.

Notably, 4 patients with severe or critical disease required antepartum admissions compared with 1 control patient. All 4 cases were admitted for COVID-19 management and required delivery during their admissions. One was admitted at 38 weeks’ gestation with COVID-19 symptoms and severe disease. After a period of maternal stabilization, she required cesarean delivery at 39 weeks’ gestation for a worsening respiratory status. The other 3 patients had clinician-initiated preterm deliveries at 26 weeks’ gestation (critical disease with emergent cesarean delivery in the ICU for refractory hypotension and persistent category 2 fetal heart rate tracing despite intrauterine resuscitation), at 29.3 weeks’ gestation (repeat cesarean delivery for severe COVID-19 in the context of superimposed preeclampsia with severe features), and at 34.6 weeks’ gestation (induction of labor for severe COVID-19 with worsening respiratory status). The control patient was admitted with preeclampsia with severe features and hemolysis, elevated liver enzymes, and a low platelet count syndrome at 28 weeks’ gestation. She underwent inpatient expectant monitoring for 48 hours and then had a successful induction of labor.

Driven primarily by the gestational age at delivery, the offspring of pregnant patients with severe or critical COVID-19 had lower birthweights and higher rates of NICU admission, RDS, and IVH than controls (Table 4). Neonatal outcomes were similar for pregnant patients with mild COVID-19 vs controls.

Associations of COVID-19 and composites of adverse maternal and neonatal outcomes are presented in Table 5 . Comparing all COVID-19 cases with controls, the unadjusted ORs of adverse maternal and neonatal outcome were 2.7 (95% CI, 1.0–10.0) and 1.4 (95% CI, 0.6–3.6), respectively. After adjusting for advanced maternal age, obesity, race, and comorbid medical problems, the adjusted odds of adverse maternal and neonatal outcomes were 3.4 (95% CI, 1.2–13.4) and 1.7 (95% CI, 0.8–4.8), respectively. In analyses stratified by disease severity, the odds of adverse maternal and neonatal outcome were similar for mild COVID-19 cases vs controls (Table 5). These results suggest that the morbidity associated with COVID-19 in pregnancy is largely driven by the severe or critical phenotype.

Comment

Footnotes

References