Clinical characteristics, prognostic factors, and maternal and neonatal outcomes of SARS‐CoV‐2 infection among hospitalized pregnant women: A systematic review

Abstract Background Pregnant women represent a potentially high‐risk population in the COVID‐19 pandemic. Objective To summarize clinical characteristics and outcomes among pregnant women hospitalized with COVID‐19. Search strategy Relevant databases were searched up until May 29, 2020. Selection criteria Case series/reports of hospitalized pregnant women with laboratory‐confirmed COVID‐19. Data collection and analysis PRISMA guidelines were followed. Methodologic quality was assessed via NIH assessment tools. Main results Overall, 63 observational studies of 637 women (84.6% in third trimester) with laboratory‐confirmed SARS‐CoV‐2 infection were included. Most (76.5%) women experienced mild disease. Maternal fatality, stillbirth, and neonatal fatality rates were 1.6%, 1.4%, and 1.0%, respectively. Older age, obesity, diabetes mellitus, and raised serum D‐dimer and interleukin‐6 were predictive of poor outcomes. Overall, 33.7% of live births were preterm, of which half were iatrogenic among women with mild COVID‐19 and no complications. Most women underwent cesarean despite lacking a clear indication. Eight (2.0%) neonates had positive nasopharyngeal swabs after delivery and developed chest infection within 48 hours. Conclusions Advanced gestation, maternal age, obesity, diabetes mellitus, and a combination of elevated D‐dimer and interleukin‐6 levels are predictive of poor pregnancy outcomes in COVID‐19. The rate of iatrogenic preterm birth and cesarean delivery is high; vertical transmission may be possible but has not been proved.


| INTRODUCTION
Coronavirus disease 2019  is caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and was first reported in Wuhan, Hubei Province, China, in December 2019. The infection has spread swiftly across the globe to 188 countries, affecting over thirteen million individuals with a case fatality rate of 5%, and was declared a pandemic by the WHO on March 11, 2020. 1 The initial data on COVID-19 outcomes in pregnancy were derived from small numbers of patients, first from China and then from Italy, with uncertainty regarding the extent to which these findings might be extrapolated to inform obstetric practice elsewhere in the world.
Small international data sets have subsequently emerged, but these are often from different institutions with heterogeneous testing policies, obstetric management, and patient populations. Population-level studies are therefore urgently required to provide robust data on the incidence of COVID-19 in pregnancy and its effects on the pregnant woman, developing fetus, and newborn.
The objective of the present review was to evaluate clinical characteristics and maternal, fetal, and neonatal outcomes among pregnant women admitted to hospital with laboratory-confirmed SARS-CoV-2 infection by conducting a large global comprehensive review of data from various epicenters. By addressing practical issues surrounding the parameters that predict prognosis across the disease spectrum, the review aims to inform clinical practice and guide policy in maternity services for this and future global pandemics affecting pregnant women and those wishing to conceive.

| MATERIALS AND METHODS
The present systematic review was conducted in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. 2 PubMed, Ovid Medline, Web of Science, and China Academic Literature Database were searched for studies on pregnant women with COVID-19 infection from database inception until May 29, 2020. The search strategy combined terms for SARS-CoV-2, COVID-19, pregnancy, maternal mortality, maternal morbidity, complications, miscarriage, preterm birth, neonatal morbidity, intrauterine fetal demise, and neonatal mortality. Studies from mainland China were included. Three studies published in Mandarin were translated into English by two of the researchers (JWJL and PD). [3][4][5] Studies in languages other than English or Mandarin were excluded.
The review included case series and case reports of pregnant women with SARS-CoV-2 infection confirmed by either quantitative real-time polymerase chain reaction (RT-PCR) or dual fluorescence PCR assessment. Unpublished reports, articles in which the date and location of the study were not specified, women with suspected COVID-19 that was not confirmed by laboratory tests, and studies that did not report maternal, fetal, or neonatal outcomes were excluded.
Studies were independently retrieved and reviewed for eligibility by two authors (OT and RK). The following data were extracted: Severity of COVID-19, gestational age at presentation, presenting symptoms, laboratory parameters, admission to intensive care unit (ICU), need for Invasive mechanical ventilation (IMV), gestational age at delivery, mode of delivery, admission to NICU, stillbirth, neonatal death and maternal death.
Data analysis, including prevalence and incidence levels, was performed by using SPSS version 26.0 (IBM, Armonk, NY, USA). Categoric variables were expressed as number (percentage).
The methodologic quality of studies was assessed by using National Institutes of Health quality assessment tools. 6 Each study was classified as low (≥7), moderate (5-6), or high risk of bias (≤4).

| Included studies
In total, 261 articles were identified in the initial search from March 26 to May 29, 2020; of these, 63 observational studies met the inclusion criteria (Fig. 1). There were 28 case series, 31 case reports, and 4 retrospective cohort studies. All 63 observational studies were eligible for qualitative synthesis (Fig. 1).
Overall, the mean risk for publication bias was moderate (6.4 out of 9) across all case series; 4 (14%) and 8 (29%) of case series reported high or moderate risk of publication bias, respectively. Subgroup analysis indicated that a high percentage of case series (n=25; >85%) adequately described the research questions and study population, and ensured that study subjects were comparable. Only 12 (43%) and 15 (54%) of case series fully described statistical methods and patient F I G U R E 1 PRISMA flowchart showing selection of studies for review. The reasons for excluding studies were (1) type of study (only case reports, case series, and retrospective cohort analyses were included); (2) language (non-English and non-Mandarin publications were excluded); (3) access to full text (abstract-only publications were excluded); and (4) confirmation of SARS CoV-2 infection (only laboratory-confirmed cases were included). results, respectively. Consequently, reporting of statistical methods and full analysis of study data posed the greatest risk of publication bias.

| Maternal outcomes
The incidence of adverse maternal outcomes was assessed for all 637 pregnant women ( Table 1). The rate of intensive care unit (ICU) admission was 9.6% (n=61) ( was still on extra-corporeal membrane oxygenation, 4 remained intubated, and 46 had been successfully extubated. Overall, the incidence of maternal mortality was 1.6% (10/637); all of these women were admitted to the ICU with severe or critical COVID-19 requiring IMV. 15 Two of the 10 women who died were 40 years or older 15 ; two women had diabetes mellitus and obesity. 15,26 Hemorrhagic complications were observed in 10 cases (Table 1).
Four women presented at hospital with antepartum hemorrhage described as one of their presenting symptoms. 45,51 In addition, six women experienced postpartum hemorrhage (estimated blood loss >1000 mL), all of whom had undergone cesarean delivery. Three of these women had severe or critical COVID-19: one had developed consumptive coagulopathy requiring treatment with fibrinogen concentrate, whereas the other two had mild COVID-19 disease. 18,20,32 Six women had abnormal coagulation parameters, all of which were associated with critical infection and multi-organ dysfunction; three of three women subsequently died. 18,20,44,59 Coagulation parameters returned to normal during the postpartum period among the three women who survived.
Three thrombotic events were reported among the study women included in the review. All three events occurred despite the administration of thromboprophylaxis with low molecular weight heparin.
One woman was admitted with mild COVID-19 symptoms at 29 gestational weeks 60 ; however, she deteriorated rapidly and developed acute respiratory failure that required continuous positive airway pressure therapy. She did not have pre-existing co-morbidities except obesity or a family history of thrombophilia or venous thromboembolism.
The second women, aged 29 years, was admitted at 31 gestational weeks with mild COVID-19 symptoms and had diabetic ketoacidosis with a background medical history of obesity, insulin-dependent diabetes mellitus, renal tubular acidosis, and bronchial asthma. 26 After rapid respiratory compromise, she was admitted to the ICU for IMV. with non-occlusive thrombi. 45,52,54,55 Notably, all placentae were from women with clinically mild COVID-19, among whom there were seven preterm deliveries, one small for gestational age neonate, one case of placental abruption, and one case of second-trimester miscarriage.

| Fetal and neonatal outcomes
Miscarriage or termination of pregnancy was reported in six studies. 32 with only one positive result. 54 Breast milk PCR testing was reported in five studies 8,24,43,50,51,54  The review found that 9.6% of women were admitted to ICU and the overall maternal mortality was 1.6%. In a recent report from the US Center of Disease Control and Prevention of COVID-19 among women aged 15-44years, 31.5% of pregnant women were hospitalized as compared with 5.8% of non-pregnant women. 68 After adjustment for cofounding factors, pregnant women were also more likely to be admitted to the ICU and receive mechanical ventilation, but there was a similar death rate of 0.2% among the two groups of women. 68 However, data were not available to determine whether hospitalization and ICU admission were due to COVID-19 or pregnancy-related indications. A lower threshold for hospital admission in pregnancy may explain a higher rate of hospitalization; in turn, this may render a comparison of morbidity and mortality with a non-pregnant population impossible because the majority of pregnant women admitted to hospital have mild disease, whereas individuals with COVID-19 are generally admitted to hospital only with severe or critical infection. In the present review, the mortality rate among women with severe/critical COVID-19 and those admitted to ICU was 6.7% and 16.4%, respectively. Women in their third trimester manifested the greatest need for ICU admission and IMV, and had the highest risk of death. Furthermore, there was a high incidence of key risk factors among pregnant women with poor maternal outcomes. The prevalence of obesity, diabetes mellitus, and advanced maternal age (≥40 years) was 40% among all maternal mortalities. Therefore, women with pre-existing co-morbidities are likely to be at highest risk of morbidity and mortality from SARS-Cov-2 infection in the third trimester, which warrants increased vigilance from healthcare providers.
COVID-19 is a novel infection with rapidly evolving scientific data regarding its management. This has a significant impact on the psychosocial wellbeing of women during pregnancy. In a study from Ireland, There remains a paucity of data on the impact of COVID-19 on pregnancy in the first and second trimester. The present review, in line with UK Obstetric Surveillance data, 75 shows that women in early pregnancy account for a minority of hospitalized pregnant women with COVID-19. Whether they are less likely to have severe infection, or whether clinicians have a lower threshold to admit women in the latter part of pregnancy is unknown. In the review, the rate of miscarriage was approximately 16% in first trimester and 4% in the second trimester among women with SARS-CoV-2 infection. The latter seems to be higher than the baseline abortion rate of 1%-2% in the second trimester. 76,77 However, the number of women with COVID-19 during the second trimester of pregnancy reported in the review was small, and bias cannot be excluded because women with COVID-19 who miscarry are potentially more likely to be reported in studies. It is important to ascertain the actual abortion rates among affected women with different disease severities.
This will require the capture of data from women not admitted to hospital, which were not collected in the review. This is an area that requires further investigation to assess placental response, considering the multi-organ impact of COVID-19 and emerging evidence of its resemblance to complement-mediated thrombotic microangiopathies. 78  During the early surge of COVID-19, there was a lack of information on the vertical transmission potential of SARS-CoV-2, resulting in understandable anxiety among women and obstetricians. This is reflected in the high rates of preterm birth (~34% of all births) and cesarean delivery in the review. Although deteriorating maternal condition and fetal distress accounted for ~50% of all preterm births, the remainder seemed to be iatrogenic among women with mild COVID-19 with no maternal or fetal compromise.
It seems unlikely that COVID-19 itself increases the risk of spontaneous preterm labor; the rate in the review was approximately 4% (equally distributed among women with mild and severe/critical disease) as compared with a rate of 5%-8% reported for the general population. 84 The overall high rate of preterm birth is likely to con- The possibility of vertical transmission of SARS-CoV-2 virus has been widely debated and the picture remains unclear. In the present study, only 8 (2%) of 405 neonates who underwent naso-and oropharyngeal PCR testing for SARS-CoV-2 virus had a positive test. Some neonates were also found to be positive for throat swabs immediately after cesarean delivery, despite measures to isolate the neonate from the mother on delivery. Recently, the visualization of SARS-CoV-2 virions in syncytiotrophoblasts and microvilli of the placenta suggests that the placenta, which is known to express the ACE2 receptor (the putative receptor for SARS-CoV-2), may be a target for virion entry. 88 However, those findings are not confirmatory of vertical transmission.
A major strength of the review is the inclusion of data from a large number of pregnant women with laboratory-confirmed SARS-CoV-2 infection from various regions of the world, including Asia, Europe, and the United States. However, the analysis also has limitations because COVID-19 management decisions, as well as variations in healthcare resources, differ from country to country. In addition, datasets were obtained from retrospective observational studies that are prone to recall and/or misclassification bias, and limited the ability to explore risk factors. However, the overall risk for publication bias was moderate and more than 85% of the case series in the review adequately described the research questions and study population, and ensured that study women were comparable.
In conclusion, as our collective understanding grows, we believe that obstetricians will become less interventional, especially in preterm gestations, when managing mild COVID-19 disease, thereby reducing the burden of iatrogenic neonatal morbidity. The data support previously documented findings of reassuring maternal outcomes for mild COVID-19 infection, with poor maternal and fetal outcomes among those with severe or critical disease, predisposed by obesity, diabetes mellitus, advanced maternal age, and advanced pregnancy. This, we believe, should inform the counselling of women diagnosed with COVID-19 in pregnancy.

AUTHOR CONTRIBUTIONS
OT was responsible for conceptualization, project administration, investigation, methodology, validation, data curation, formal analysis, and drafting and revising the manuscript. AH was responsible for conceptualization, methodology, formal analysis, data curation, and drafting and revising the manuscript. PD was responsible for investigation, validation, and drafting and revising the manuscript. WJL was responsible for investigation, validation, and drafting and revising the manuscript. AW was responsible for and drafting and revising the manuscript. RA-K was responsible for conceptualization, methodology, validation, supervision, and revision of the manuscript.

CONFLICTS OF INTEREST
The authors have no conflicts of interest.