key: cord-0816481-i2x59sf8
authors: GRECHUKHINA, Olga; GREENBERG, Victoria; LUNDSBERG, Lisbet S.; DESHMUKH, Uma; CATE, Jennifer; LIPKIND, Heather S.; CAMPBELL, Katherine H.; PETTKER, Christian M.; KOHARI, Katherine S.; REDDY, Uma M.
title: Coronavirus Disease 2019 (COVID-19) pregnancy outcomes in a racially and ethnically diverse population
date: 2020-10-07
journal: Am J Obstet Gynecol MFM
DOI: 10.1016/j.ajogmf.2020.100246
sha: 91d65537dac47f8989ac015654e2321c20f64eb6
doc_id: 816481
cord_uid: i2x59sf8

Background Older age and medical comorbidities are identified risk factors for developing severe COVID-19. However, there are limited data on risk stratification, clinical and laboratory course, and optimal management of COVID-19 in pregnancy. Objective Our study aims to describe the clinical course of COVID-19, effect of comorbidities on disease severity, laboratory trends, and pregnancy outcomes of symptomatic and asymptomatic SARS-CoV-2 positive pregnant women. Study Design This is a case series of pregnant and postpartum women who tested positive for SARS-CoV-2 between 3/1/2020 and 5/11/2020 within 3 hospitals of the Yale-New Haven Health delivery network. Charts were reviewed for basic sociodemographic and pre-pregnancy characteristics, COVID-19 course, laboratory values, and pregnancy outcomes. Results Out of 1,567 tested pregnant and postpartum women between 3/1/2020 and 5/11/2020, 9% (n=141) had a positive SARS-CoV-2 result. Hispanic women were overrepresented in the SARS-CoV-2 positive group (n=61; 43.8%). Additionally, Hispanic ethnicity was associated with higher rate of moderate and severe disease compared to non-Hispanic (18% (11/61) vs 3.8% (3/78), respectively, OR 5.5 95% CI 1.46-20.7, p=0.01). Forty-four women (31.2%) were asymptomatic, 37 (26.2%) of whom were diagnosed on universal screening upon admission for delivery. Fifty-nine percent (n=83) were diagnosed antepartum, 36% (n=51) upon presentation for childbirth and 5% (n=7) postpartum. Severe disease was diagnosed in 6 cases (4.3%) and there was one maternal death. Obese women were more likely to develop moderate and severe disease than non-obese women (16.4% (9/55) vs 3.8% (3/79), OR 4.96, 95%CI 1.28-19.25, p=0.02). Hypertensive disorders of pregnancy were diagnosed in 22.3% (17/77) of women who delivered after 20 weeks. Higher levels of C-reactive protein during antepartum COVID-19-related admission were more common in women with worse clinical course; this association, however, did not reach statistical significance. Conclusion COVID-19 in pregnancy may result in severe disease and death. Hispanic women were more likely to test positive for SARS-CoV-2 than other ethnic groups. Obesity and Hispanic ethnicity represent risk factors for moderate and severe disease.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), a single-stranded RNA 55 virus, causes coronavirus disease 2019 and is responsible for a global health 56 emergency. This pandemic has led to over 29 million people infected and over 925,000 deaths 57 worldwide (as of September 14, 2020). 1 This health crisis has spared no demographic, causing 58 concern about its impact on vulnerable populations, such as pregnant women. 2,3 59

Since the start of the pandemic, clinicians and researchers have steadily expanded the 60 understanding of COVID-19 in pregnancy. However, the total number of cases reported in the 61 literature remains limited. This study aims to describe the clinical course of pregnant women and 62 their neonates in a large, diverse hospital system in a significantly affected region adjacent to 63

New York City, one of the United States' initial infectious epicenters. Medical comorbidities and 64 sociodemographic factors were examined for association with COVID-19 severity and clinical 65 course. Lastly, we report laboratory trends for SARS-CoV-2 positive pregnant women admitted 66 to the hospital. 67

This is a case series of all pregnant and postpartum women with positive SARS-CoV-2 70 RT-PCR tests between 3/1/2020 and 5/11/2020 from three Yale New Haven Health hospitals 71 (Yale New Haven, Bridgeport, and Greenwich hospitals). Subjects were identified using an 72 electronic health record (EHR) search for an open pregnancy episode and a SARS-CoV-2 RT-73 PCR laboratory result within the timeframe. Ambulatory and inpatient testing was included. 74

Each chart was individually reviewed for current pregnant status (positive pregnancy test with or 75 without ultrasound confirmation) or pregnancy resolution within 6 weeks of SARS-CoV-2 test 76 J o u r n a l P r e -p r o o f for inclusion into the study cohort. Subjects with a positive test were included for analysis. Each 77 case was individually reviewed to collect the following: baseline sociodemographic factors; past 78 medical, surgical and obstetric history; antenatal course; and COVID-19 course including 79 symptoms, laboratory and imaging studies, management, maternal, and neonatal outcomes. The 80 study was approved by Yale University institutional review board with waiver of consent 81 (HIC2000027797). 82

Testing and diagnosis of COVID-19 83 SARS-CoV-2 testing used RT-PCR analysis of nasopharyngeal swab specimens. Testing 84 criteria generally consisted of either 1) patients with symptoms of COVID-19 as deemed by their 85 healthcare provider or the institutional COVID-19 Call Center, or 2) universal testing of all 86 pregnant women who were admitted after April 1, 2020 for delivery or antepartum management. 87

Testing criteria of symptomatic patients evolved during the study period and were set by 88 institutional committees guided by Centers for Disease Control and Prevention (CDC) 89 recommendations. Neonatal testing was indicated for all newborns born to mothers who tested 90 positive for SARS-CoV-2 within 2 weeks of the delivery and was performed by RT-PCR of 91 nasopharyngeal samples between 24 and 48 hours of birth. 4 92 Disease severity was classified per World Health Organization (WHO) into 93 asymptomatic (no current or previous symptoms), mild (symptomatic patients without evidence 94 of viral pneumonia or hypoxia), moderate (clinical signs of pneumonia without signs of severe 95 pneumonia and no need for supplemental oxygen), severe (signs of severe pneumonia i.e. 96 respiratory rate of 30/min or more, blood oxygen saturation of less than 95% [the threshold for 97 oxygen supplementation in pregnancy], severe respiratory distress), and critical (acute 98 respiratory distress syndrome, sepsis or septic shock). 5 Outpatient triage of the pregnant COVID-99 J o u r n a l P r e -p r o o f 19 population was performed per institutional guidelines (Supplemental Figure 1 ). For analysis, 100 severe and critical disease were combined, resulting in a total of 4 groups. Final disease severity 101 was assigned retrospectively according to the above definitions which were set up a priori by a 102 panel of Maternal-Fetal Medicine subspecialists based on the entire course of the disease. 103 Race and ethnicity information were self-reported at the time of hospital registration and 104 abstracted directly from the EHR. Hypertensive disorders of pregnancy (HDP), including 105 gestational hypertension, preeclampsia without and with severe features, eclampsia and HELLP, 106

were identified during individual chart review. All diagnoses were confirmed to meet the 107 American College of Obstetricians and Gynecologists (ACOG) criteria of HDP. 6 Laboratory 108 testing guidelines for admitted patients varied between the hospitals and evolved over time. D-109 dimer and C-reactive protein (CRP) were chosen for analysis as the most consistently tested and 110 trended lab studies. Since the occurrence of birth affects the levels of these lab values, we 111 divided our cohort into two groups for the purpose of analysis: women admitted for delivery 112 (symptomatic and asymptomatic) and women admitted in the antepartum period and discharged 113 undelivered. 114

Patient characteristics including sociodemographics, pregnancy outcomes, comorbidities, and 116 disease severity are reported descriptively and presented as percentages of the total cohort. 117

Continuous variables were not normally distributed and thus reported as median and interquartile 118 range (IQR). Bivariate analysis to evaluate the association between patient characteristics, 119 comorbidities, and disease severity was performed using Fisher exact test. Due to the low 120 number of subjects in the moderate and severe groups, to further examine the association 121 between ethnicity (Hispanic and non-Hispanic) and obesity (pre-pregnancy BMI ≥30 and <30 122 kg/m 2 ) with severity of the disease, the cohort was organized into two groups: 123 asymptomatic/mild disease and moderate/severe disease. Unadjusted odds ratios (OR) with 95% 124 confidence intervals (CI) were calculated for these dichotomous measures. Adjusted OR were 125 unable to be calculated due to small sample size. Tests of association between specific symptoms 126 and disease severity were restricted to those with symptoms (n=97). In this group we evaluated 127 the association between COVID-19 severity as a 3-level categorical measure (mild, moderate, 128 severe) and dichotomous measures of symptoms using the Fisher exact test. Non-parametric 129

Mann-Whitney U test was used to compare non-normally distributed continuous variables (CRP 130 values). P value <0.05 was considered significant. were diagnosed with COVID-19 postpartum after discharge from their childbirth admission. 160

Thirty-one percent of women (44/141) were asymptomatic. Fifty-eight percent of women 161 (82/141) had mild disease; 6.4% (9/141) had moderate disease. Five women had severe or 162 critical disease. One woman died in the Emergency Room. This woman, with a pre-pregnancy 163 BMI of 35 kg/m 2 , was diagnosed with COVID-19 in ambulatory care in the first trimester of 164 pregnancy. She developed respiratory distress at home 13 days after initial symptom onset and 165 arrived at the Emergency Department profoundly hypoxemic, suffering cardiac arrest and 166 ultimately died despite prolonged attempts at cardiopulmonary resuscitation. No autopsy was 167 J o u r n a l P r e -p r o o f performed. Including this case, the rate of severe/critical disease in our population was 4.3% 168 (6/141). Timing of the diagnoses and disease severity are reflected in Figure 1 . 169

Maternal medical comorbidities and their relation to COVID-19 course are presented in 170 Table 1 . Severity of disease was associated with obesity, both as a dichotomous measure and by 171 obesity class (p=0.01 and p<0.01, respectively) but not with any other co-morbidity. Obese 172 women had higher rates of moderate/severe disease than non-obese women (16.4% (9/55) Among symptomatic women, the most common symptoms in our cohort were cough 181 (70.1%), muscle aches (51.6%) and sore throat (47.4%) (Figure 2 ). The most common symptoms 182 in women with severe disease were muscle aches, fever, shortness of breath, nausea, chest pain 183 and abdominal pain. 184 D-dimer and CRP trends, grouped by the type of admission, are presented in Figure 3 . 185

Notably, D-dimer values varied greatly within the group who tested positive for SARS-CoV-2 186 during childbirth admission. However, most had a substantial increase in D-dimer value shortly 187 after birth with a subsequent decline within 48 hours. D-dimer took longer to normalize in one 188 patient (5 days after delivery) whose respiratory status deteriorated in labor necessitating 189 cesarean birth followed by ICU admission for COVID-19-related respiratory failure. There were 190 J o u r n a l P r e -p r o o f no cases of venous thromboembolism diagnosed during the study period. CRP also peaked after 191 delivery. Women admitted antepartum for COVID-19 management who developed severe 192 disease appeared to have higher initial CRP values than those with milder disease. Comparison 193 between these two groups, however, did not reach statistical significance (p=0.057). 194

Pregnancy outcomes were available for 56.7% of women (80/141) (Figure 1 and Table  195 2). Notably, one woman underwent termination via dilation and evacuation at 22 weeks of population is low (4.3%), our cohort includes one maternal death. In our cohort, Hispanic women 216 were disproportionately affected by COVID-19 and appeared to have an increased risk of 217 moderate/severe disease. This finding is unlikely to be related to a disproportionate testing in 218

Hispanic population as all three hospital sites implemented universal SARS-CoV-2 testing upon 219 admission for childbirth. Pre-pregnancy obesity was associated with a higher disease severity 220 category. HDP affected approximately 1 out of every 5 women with COVID-19 after 20 weeks 221 of gestation with the majority diagnosed with preeclampsia with severe features or HELLP 222

and CRP levels in all COVID-19 positive women regardless of symptomatic status. D-dimer 224 returned to predelivery values within 24-48 hours in most women. D-dimer did not appear to be 225 a useful marker to distinguish COVID-19 disease severity category. All newborns born to 226

COVID-19 positive women tested negative for SARS-CoV-2 RNA via nasopharyngeal swab 227 after 24 hours of life; however, there was one case with positive placental SARS-CoV-2 testing. 7 228

Early reports of SARS-CoV-2 infection during pregnancy are encouraging as they failed 230 to demonstrate higher susceptibility or morbidity in pregnant women compared to the general 231 population. 8-11 More recent reports have described severe and critical disease in pregnancy as 232 well as maternal deaths from COVID-19, indicating potential for severe maternal morbidity and 233 mortality. 12-14 The case of maternal death in our series highlights the potential for the disease 234 course to be protracted with seemingly unpredictable and abrupt deterioration in health after 10- the United States. [16] [17] [18] [19] Our data raises concern about the role of social determinants of health and 240 systemic inequities specific to SARS-CoV-2 transmission and healthcare access. Our findings 241 are further supported by a recent study by Moore at al, which demonstrated a disproportionate 242 number of COVID-19 cases among underrepresented racial/ethnic groups (with Hispanic 243 population being the largest affected group) in COVID-19 pandemic hotspots. 20 We demonstrate 244 that pre-pregnancy obesity is associated with more severe COVID-19, which is consistent with 245 prior studies in non-pregnant adults and a small study of pregnant women. 13,21 246

In non-pregnant adults, higher D-dimer levels are associated with increased risk of 247 critical COVID-19 course and death. 22,23 Anticoagulation, guided by D-dimer levels, has been 248 shown to decrease mortality in this population. 24 In both complicated and uncomplicated 249 pregnancies, however, D-dimer levels are known to increase above baseline, though reference 250 ranges are inconsistent. 25 Our study presents novel data on D-dimer trends in SARS-CoV-2 251 positive symptomatic and asymptomatic women in relation to delivery. CRP has emerged as 252 another independent predictor of adverse outcomes in non-pregnant COVID-19 patients 26 . Our 253 data suggests that D-dimer may not be helpful in determining disease severity in a pregnant and 254 peripartum COVID-19 population. Its use for anticoagulation guidance needs to be further 255 evaluated. Similar to D-dimer, there are no well-established reference ranges for CRP in 256 pregnancy and there are limited data for the use of this parameter in pregnant COVID-19 257 positive women. 27 Our data suggest that admission CRP values in antepartum women may 258 emerge as a more helpful in predicting disease severity. 259

Overrepresentation of Hispanic women in our SARS-CoV-2 positive cohort and concern 261 for increased severity of COVID-19 disease in this group indicates an urgent need to further 262 characterize and address the causes of these disparities. Additional larger-scale studies are 263 needed to address the mounting evidence that racial and ethnic disparities are central to the 264 myriad factors (e.g. health care access, housing, and ability to socially distance) that lead to the 265 unequal distribution of SARS-CoV-2 infection and COVID-19 severity and mortality seen 266 throughout the United States. 16 

The CDC guidelines include only severe obesity (BMI >40 mg/m 2 ) as a risk factor for 268 severe illness in the non-pregnant population while our study links pre-pregnancy BMI of ≥30 269 kg/m 2 with worse clinical course during pregnancy. 28 The current ACOG-SMFM COVID-19 270 guidelines do not list obesity as a comorbidity placing pregnant women at risk for more severe 271 disease. 29 Given our findings, consideration should be made to include all classes of obesity as a 272 risk factor in pregnancy for progression to moderate and severe disease. 273

Lastly, larger studies are required to review possible association between SARS-CoV-2 274 infection and HDP. 275

Our study was performed in a diverse health care system consisting of academic and 277 community hospitals with a racially and ethnically diverse population; however, the study is 278 limited to a single geographic location and may not be generalizable to other regions of the 279 country with different patient populations and prevalence of SARS-CoV-2. Additionally, this 280 population is heterogenous with both symptomatic and asymptomatic women being tested for 281 SARS-CoV-2. We acknowledge that many women with symptoms were likely never tested and 282 new commercial tests performed outside of hospital labs emerged during the course of this study, 283 the results of which may have not been incorporated in the EHR and identified for review. 284

Furthermore, testing guidelines as well as management strategies evolved during the study 285 period, thus, contributing to the variation in clinical decision making. WHO COVID-19 severity 286 assignment criteria were used for this study as this classification system was the only one 287 explicitly applicable for pregnancy at the time. Furthermore, we adjusted the oxygen saturation 288 criterion for severe disease from < 90% on room air in non-pregnancy to < 95% in pregnancy. 30 289

The racial and ethnic composition of the tested population may not be an accurate representation 290 of the overall pregnant population. We were unable to compare the rates of HDP in COVID-19 291 positive and negative patients as we had limited access to the data on the latter group. CRP value 292 comparisons between COVID-19 severity groups were limited by small sample size. We were 293 unable to perform a multivariate analysis to assess for confounding due to small sample size. 294

Lastly, unlike other literature, we failed to demonstrate associations between pre-existing 295 hypertension and diabetes with worse COVID-19 course; this may be due to a relatively small 296 sample size. 31 Larger registry studies are needed to examine the risk factors associated with 297 COVID-19 progression in pregnancy. 298

This study demonstrates that the majority of pregnant women with COVID-19 remain 300 either asymptomatic or have mild disease; however, severe illness and death can occur. Pre-301 pregnancy obesity was associated with an increased risk of severe illness. Further, the Hispanic 302 population in this cohort appeared to be at increased risk for severe illness. Large-scale studies 303 are required to develop better risk stratification strategies for COVID-19 in pregnancy. Fisher exact test of association between the 3-level severity and dichotomous symptoms.

Symptoms with P-value <0.05 were marked with *. 

Pérez-Stable EJ. COVID-19 and Racial/Ethnic Disparities

Hospitalization Rates and Characteristics of Patients 351 Hospitalized with Laboratory-Confirmed Coronavirus Disease

Prevalence and Severity of Coronavirus Disease 357 2019 (COVID-19) Illness in Symptomatic Pregnant and Postpartum Women Stratified by 358 Hispanic Ethnicity

Disparities in Incidence of COVID-19 Among 360 Underrepresented Racial/Ethnic Groups in Counties Identified as Hotspots During

Presenting Characteristics, Comorbidities, 364 and Outcomes Among 5700 Patients Hospitalized With COVID-19 in the New York City 365 Area

Clinical course and risk factors for mortality of adult inpatients 367 with COVID-19 in Wuhan, China: a retrospective cohort study

Variations of plasma D-dimer level at 376 various points of normal pregnancy and its trends in complicated pregnancies: A 377 retrospective observational cohort study

Outpatient Assessment and Management for Pregnant Women With 389 Suspected or Confirmed Novel Coronavirus (COVID-19)

Symptoms and Critical Illness Among were discharged undelivered. D&E, dilation and evacuation

PNA, pneumonia; ICU, intensive 467 care unit

Supplemental Figure 1. Yale-New Haven Hospital System Outpatient Management 470 Guidelines of SARS-CoV-2 Positive Pregnant Women. f HIGHLIGHTS: -The majority of pregnant women with SARS-CoV-2 infection are asymptomatic or have mild disease -However, severe COVID-19 disease and maternal death occur