key: cord-1016880-nv96x2p7
authors: Di Mascio, Daniele; Khalil, Asma; Saccone, Gabriele; Rizzo, Giuseppe; Buca, Danilo; Liberati, Marco; Vecchiet, Jacopo; Nappi, Luigi; Scambia, Giovanni; Berghella, Vincenzo; D’Antonio, Francesco
title: Outcome of Coronavirus spectrum infections (SARS, MERS, COVID 1 -19) during pregnancy: a systematic review and meta-analysis
date: 2020-03-25
journal: Am J Obstet Gynecol MFM
DOI: 10.1016/j.ajogmf.2020.100107
sha: ee2254bc1c223852d30fe7d8fcf44dee6c8dd854
doc_id: 1016880
cord_uid: nv96x2p7

ABSTRACT Objective The aim of this systematic review was to report pregnancy and perinatal outcomes of Coronavirus (CoV) spectrum infections, and particularly COVID-19 disease due to SARS-COV-2 infection during pregnancy. Data sources Medline, Embase, Cinahl and Clinicaltrials.gov databases were searched electronically utilizing combinations of word variants for “coronavirus” or “severe acute respiratory syndrome” or “SARS” or “Middle East respiratory syndrome” or “MERS” or “COVID-19” and “pregnancy”. The search and selection criteria were restricted to English language. Study eligibility criteria Inclusion criteria were pregnant women with a confirmed Coronavirus related illness, defined as either SARS, MERS or COVID-19. Study appraisal and synthesis methods We used meta-analyses of proportions to combine data and reported pooled proportions. The pregnancy outcomes observed included miscarriage, preterm birth, pre-eclampsia, preterm prelabor rupture of membranes, fetal growth restriction, and mode of delivery. The perinatal outcomes observed were fetal distress, Apgar score < 7 at five minutes, neonatal asphyxia, admission to neonatal intensive care unit, perinatal death, and evidence of vertical transmission. Results 19 studies including 79 women were eligible for this systematic review: 41 pregnancies (51.9%) affected by COVID-19, 12 (15.2%) by MERS, and 26 (32.9%) by SARS. An overt diagnosis of pneumonia was made in 91.8% and the most common symptoms were fever (82.6%), cough (57.1%) and dyspnea (27.0%). For all CoV infections, the rate of miscarriage was 39.1% (95% CI 20.2-59.8); the rate of preterm birth < 37 weeks was 24.3% (95% CI 12.5-38.6); premature prelabor rupture of membranes occurred in 20.7% (95% CI 9.5-34.9), preeclampsia in 16.2% (95% CI 4.2-34.1), and fetal growth restriction in 11.7% (95% CI 3.2-24.4); 84% were delivered by cesarean; the rate of perinatal death was 11.1% (95% CI 84.8-19.6) and 57.2% (95% CI 3.6-99.8) of newborns were admitted to the neonatal intensive care unit. When focusing on COVID-19, the most common adverse pregnancy outcome was preterm birth < 37 weeks, occurring in 41.1% (95% CI 25.6-57.6) of cases, while the rate of perinatal death was 7.0% (95% CI 1.4-16.3). None of the 41 newborns assessed showed clinical signs of vertical transmission. Conclusion In mothers infected with coronavirus infections, including COVID-19, >90% of whom also had pneumonia, PTB is the most common adverse pregnancy outcome. Miscarriage, preeclampsia, cesarean, and perinatal death (7-11%) were also more common than in the general population. There have been no published cases of clinical evidence of vertical transmission. Evidence is accumulating rapidly, so these data may need to be updated soon. The findings from this study can guide and enhance prenatal counseling of women with COVID-19 infection occurring during pregnancy.

25.6-57.6) of cases, while the rate of perinatal death was 7.0% (95% CI 1. 4-16.3 ). None of the 41 84 newborns assessed showed clinical signs of vertical transmission. 85

Conclusion: In mothers infected with coronavirus infections, including COVID-19, >90% of whom 86 also had pneumonia, PTB is the most common adverse pregnancy outcome. Miscarriage, 87 preeclampsia, cesarean, and perinatal death (7-11%) were also more common than in the general 88 population. There have been no published cases of clinical evidence of vertical transmission. 89

Evidence is accumulating rapidly, so these data may need to be updated soon. The findings from 90 this study can guide and enhance prenatal counseling of women with COVID-19 infection 91 occurring during pregnancy. 

This review was performed according to a priori designed protocol recommended for systematic 125 reviews and meta-analysis. 9-11 Medline, Embase, Cinahl and Clinicaltrials.gov databases were 126 searched electronically on 03/13/2020, utilizing combinations of the relevant medical subject 127 heading (MeSH) terms, key words, and word variants for "coronavirus" or "severe acute respiratory 128 syndrome" or "SARS" or "Middle East respiratory syndrome" or "MERS" or "COVID-19" and 129 "pregnancy". The search and selection criteria were restricted to English language. Reference lists 130 of relevant articles and reviews were hand searched for additional reports. PRISMA and MOOSE 131 guidelines were followed. 12-14 132 Inclusion criteria were pregnant women with a confirmed Coronavirus spectrum illness, defined as 133 either SARS, MERS or COVID-19 infection. 134

The pregnancy outcomes observed were: 135

• Preterm birth (PTB) (either before 37 or 34 weeks of gestation) 136

• Pre-eclampsia (PE) 137

• Preterm prelabor rupture of membranes (pPROM) 138

• Fetal growth restriction (FGR) 139

• Miscarriage, as defined by authors 140

• Cesarean mode of delivery 141

The perinatal outcomes observed were: 142

• Fetal distress (as defined by original authors) 143

• Apgar score < 7 at five minutes 144

• Neonatal asphyxia (as defined by original authors) 145

• Evidence of vertical transmission, defined as the presence of clinical signs of mother-to-148 child transmission in the antenatal or perinatal period 149 150 Furthermore, we aimed to perform a sub-group analysis according to the trimester of pregnancy at 151 infection and the type of Coronavirus. 152 Data from studies reporting the incidence of these outcomes in pregnancies with CoV spectrum 153 infections were considered eligible for analysis. For the purpose of the analysis, we included only 154 full-text articles with data of pregnant women who already delivered; we excluded data regarding 155 on-going pregnancies. Furthermore, as these are relatively rare infections occurring during 156 pregnancy with the majority of data coming from studies with small sample sizes, case reports and 157 case series were also included in the analysis. Studies reporting cases of infective pneumonia or 158 other respiratory disorders during pregnancy caused by other viral agents were excluded. We also 159 excluded studies pediatric series on newborns and children from which maternal and pregnancy 160 information could not be extrapolated. 161

Two authors (DDM, GS) reviewed all abstracts independently. Agreement regarding potential 162 relevance or inconsistencies was reached by consensus or resolved by discussion with a third 163 reviewer (FDA). Full text copies of applicable papers were obtained, and the same reviewers 164 independently extracted relevant data regarding study characteristics and pregnancy outcome. If 165 more than one study was published on the same cohort with identical endpoints, the report 166 containing the most comprehensive information on the population was included to avoid 167 overlapping populations. 168 169

We used meta-analyses of proportions to combine data and reported pooled proportions (PP). 171

Funnel plots (displaying the outcome rate from individual studies versus their precision (1 per SE)) 172 were carried out with an exploratory aim. Tests for funnel plot asymmetry were not used when the total number of publications included for each outcome was <10. In this case, the power of the tests 174 is too low to distinguish chance from real asymmetry. 175

Between-study heterogeneity was explored using the I 2 statistic, which represents the percentage of 176 between-study variation that is due to heterogeneity rather than chance. A value of 0% indicates no 177 observed heterogeneity, whereas I 2 values ≥50% indicate a substantial level of heterogeneity. A 178 random effect model was used to compute the pooled data analyses. All proportion meta-analyses 179 were carried out by using StatsDirect version 2.7.9 (StatsDirect, Ltd, Altrincham, Cheshire, United 180 Kingdom). 181

Quality assessment of the included studies was assessed using the methodological quality and 182 synthesis of case series and case reports described by Murad et al. 15 According to this tool, each 183 study is judged on four broad perspectives: the selection of the study groups, the ascertainment and 184 Table 3) . 208

The majority of women affected by CoV infections were usually treated first with broad spectrum 209 antibiotics in 89.3% of cases (49/52) and then with antiviral therapy and steroids in 67.7% (37/51) 210 and 29.8% (12/31) of cases (Table 3; Supplementary Table 4) . 211

The results of the quality assessment of the included studies are presented in Supplementary 212 Table 5 . 213 214

In the overall population of pregnancies infected with CoV, The rate of miscarriage for CoV 216 infections was 39.1% (8/21 -95% CI 20.2-59.8). The rates of PTB < 37 and 34 weeks of gestation 217 were 24.3% (14/56 -95% CI 12.5-38.6) and 21.8% (11/56 -95% CI 12.5-32.9), respectively; 218 pPROM occurred in 20.7% (6/34 -95% CI 9.5-34.9), while the rate of pregnancies experiencing 219 PE and FGR was 16.2% (2/19 -95% CI 4.2-34.1) and 11.7% (2/29 -95% CI 3.2-24.4), 220

respectively. The rate of CD was 83.9% (50/58 -95% CI 73.8-91.9) ( Table 4 ; Table 5 ). The rate of 221 perinatal death was 11.1% (5/60 -95% CI 84.8-19.6) including three stillbirths and two neonatal 222 deaths (further details are provided in Supplementary Table 6 ). Thirty-four point six percent (15/44 223 -95% CI 20.3-49.5) of fetuses suffered from fetal distress and 57.2% (3/12 -95% CI 3.6-99.8) of 224 newborns was admitted to NICU. The rate of Apgar score < 7 at five minutes was 6.1% (1/48 -95% CI 1.3-13.9), but no case of neonatal asphyxia were reported. Finally, none of the newborns 226 showed signs of vertical transmission during the follow-up period (Table 6; Table 7) . 227

Six studies 16-21 reported information on COVID-19 infection during pregnancy. There was no data 229 on miscarriage for COVID-19 infection occurring during the first trimester. The rates of PTB < 37 230 and 34 weeks of gestation were 41.1% (14/32 -95% CI 25.6-57.6) and 15% (4/32 -95% CI 3.9-231 31.7), respectively. pPROM occurred in 18.8% (5/31 -95% CI 0.8-33.5), while the rate of 232 pregnancies experiencing PE was 13.6% (1/12 -95% CI 1.2-36.0), with no reported cases of FGR. 233

The rate of CD was 91% (38/41 -95% CI 81.0-97.6) ( Table 5 ). The rate of perinatal death was 7% 234

(2/41 -95% CI 1.4-16.3) including one stillbirth and one neonatal death; 43% (12/30 -95% CI 235 15.3-73.4) of fetuses had fetal distress and 8.7% (1/10 -95% CI 0.01-31.4) of newborns were 236 admitted to NICU. The rate of Apgar score < 7 at five minutes was 4.5% (1/41 -95% CI 0.4-12.6) 237 and no case of neonatal asphyxia was reported. Finally, none of the newborns showed signs of 238 vertical transmission during the follow-up period (Table 7) . 239

Seven studies 22-28 reported information on MERS infection during pregnancy. There was no data on 242 miscarriage for MERS infection occurring during the first trimester. The rate of PTB was 32.1% 243

(3/11 -95% CI 10.0-59.8), all occurring before 34 weeks of gestation. Preeclampsia occurred in 244 19.1% (1/7 -95% CI 1.1-51.3) respectively, while no case of pPROM or FGR was reported in these 245 studies. The rate of CD was 61.8% (5/8 -95% CI 32.7-86.9) ( Table 5 ). The rate of perinatal death 246 was 33.2% (3/10 -95% CI 11.2-59.9) including two stillbirths and one neonatal death (four hours 247 after birth of an extremely preterm infant). No case of fetal distress, Apgar score < 7 at five 248 minutes, neonatal asphyxia, and admission to NICU was reported. Finally, none of the newborns 249 showed signs of vertical transmission during the follow-up period (Table 7) . 250

Six studies 29-34 reported information on SARS infection during pregnancy. The rate of miscarriage 253 for MERS infection was 39.1% (8/21 -95% CI 20.2-59.8). The rate of PTB < 37 and 34 weeks of 254 gestation was 15% (1/15 -95% CI 0.3-45.6) and 28.9% (4/15 -95% CI 10.7-51.6), respectively. 255 pPROM and FGR occurred in 50% (1/2 -95% CI 0.5-95.3) and 18.5% (2/15 -95% CI 4.4-39.5) 256 respectively, while no cases of preeclampsia were reported. The rate of CD was 72.2% (7/9 -95% 257 CI 44.1-93.1) ( Table 5 ). Fetal distress occurred in 35.9% (3/9 -95% CI 12.0-64.4) of pregnancies, 258 while no case of perinatal death, Apgar score < 7 at five minutes, and neonatal asphyxia was 259

reported. There were no data on rates of admission to the NICU of infants born to infected mothers. 260

Finally, none of the newborns showed signs of vertical transmission during the follow-up period 261 (Table 7) . 262 263 It was not possible to perform a comprehensive pooled data synthesis on the incidence of pregnancy 264 and perinatal outcomes according to the trimester of pregnancy at infection due to the very small 265 number of included studies for each trimester of pregnancy. 266

The findings from this systematic review show that more than 90% of hospitalized pregnant women 270 affected by CoV infections present radiological signs suggestive for pneumonia, detected either at 271 chest x-ray or computerized tomography and the most common symptoms are fever, cough and 272 lymphopenia. Pregnancies affected by CoV infections have high rates of PTB before 37 and 34 273 weeks, and miscarriage when the infection is acquired earlier in pregnancy. Preeclampsia and 274 cesarean delivery are also more common than in the general population. The rate of perinatal 275 mortality is about 10%, while the most common adverse perinatal outcome is fetal distress, with 276 more than half of the newborns admitted in NICU. Importantly, clinical evidence of vertical 277 transmission was found in none of the newborns included. 278

To the best of our knowledge, this is the first systematic review exploring pregnancy and perinatal 281 outcomes of CoV infections occurring during pregnancy. This comprehensive meta-analysis 282 included all series published so far on this topic. 283

The small number of cases in some of the included studies, their retrospective non-randomized 284 design, and the lack of standardized criteria for the antenatal surveillance, management and timing 285 of delivery of pregnancies affected by CoV infections represent the major limitations of this 286 systematic review, thus making it difficult to draw any convincing evidence on this clinical 287 management strategies. Furthermore, there is a possibility that some patients were included in more 288 than one report, although two authors independently reviewed all the included studies, carefully 289 focusing on the different Institutions reporting outcomes. Moreover, when focusing on the 290 outcomes of COVID-19 infection, and particularly perinatal outcomes, reported data are intuitively 291 limited to a very short-term follow-up period and thus infectious that occurred proximate to the 292 underestimate more longitudinal risks such as FGR. Additionally, it was not possible to extrapolate 294 data about the rate of both spontaneous and iatrogenic PTB and indications for CD, that was 295 performed in the majority of cases; furthermore, few outcomes, i.e. "fetal distress", were not clearly 296 defined, thus leading to some discrepancies in the results, like the rate of PTB < 34 weeks (15%) 297 and the rate of newborns admitted to NICU (9%), particularly in COVID-19 infection. Another 298 limitation of the present review was the lack of stratification of the analysis according to the 299 gestational age at CoV infection due to the very small number of included studies for each trimester 300 of pregnancy. We cannot assume that the rate of miscarriage and PTB should be attributed solely to 301 the virus / infection, since there are no comparable control groups of uninfected women from the 302 same time. It may be that the stress of the situation in the community contributed to some of these 303

outcomes. Finally, we also included case reports and case series, thus facing a higher risk 304 publication bias and decreasing the level of the evidence of our findings. The lack of knowledge about COVID-19 infection has raised urgent questions among physicians 316 regarding clinical management and expected outcomes of the affected patients, and therefore, there 317 is currently a compelling need of data to guide clinical decisions.

Regarding pregnancy, the findings from this study found that radiological features suggestive for 319 pneumonia can be found in almost all of the hospitalized pregnant women, usually presenting with 320 fever, cough and lymphopenia similar to the non-pregnant population. Of note, serious conditions 321 requiring admission to ICU and mechanical ventilation are significantly less common when 322 compared with the two previous CoV infections (MERS and SARS). Similarly, we found no case of 323 maternal death related to COVID-19 infection, while MERS and SARS infections caused a 324 mortality rate in pregnant women ranging from 25% to 30%. 325

In this systematic review, women affected by COVID-19 disease had higher rates of miscarriage, 326 preterm birth, preeclampsia, while the babies had higher rates of perinatal mortality (7-11%) and of 327 admission to NICU. 328

Furthermore, as all the included studies reported data on hospitalized women, the reported rate of 329 infection-related adverse outcomes, including either pregnancy and perinatal outcomes, might not 330 reflect the overall population of pregnant when who got infected with SARS-COV-2, and there may 331 be a cohort of patients with no or mild symptoms whose pregnancy outcome is, as of yet, 332 unknown. 36 333 More importantly, it should be emphasized that there are no known neonatal symptoms and 334 therefore no clinical evidence suggestive for vertical transmission, particularly when COVID-19 335 infection occurs later in pregnancy. Unfortunately, the lack of data of first and early second 336 trimester infection does not allow to determine whether in this case the infection may cause more 337 severe perinatal outcomes and how to monitor the pregnancy once the infection has passed. 1 338 Based on the limited information from this study, COVID-19 cannot be considered as an indication 339 for delivery and therefore the timing and mode of delivery should be individualized according to 340 maternal clinical conditions or obstetric factors as usual (and not COVID-19 status alone), and the 341 decision should involve a multidisciplinary team including maternal fetal doctors, neonatologists, 342 anesthesiologists, and infective disease specialists. 343

In summary, with the limited data reported to date, mothers infected with coronavirus infections, 346 including COVID-19, >90% of whom also had pneumonia, are at increased risks of miscarriage, 347 preterm birth, preeclampsia, cesarean delivery, and their babies at higher risk of perinatal death and 348 admission to the NICU, compared to the general population. There have been no published cases of 349 clinical evidence of vertical transmission. Evidence is accumulating rapidly, so these data may need 350 to be updated soon. (20) 

COVID-19) during pregnancy: a case series Overlapping risk with Liu 2020 (included with preliminary, pre-peer review data) Chen

Same Institution of Zhu 2020 that was included in this systematic review Li 2005 Severe acute respiratory syndrome in neonates and children Review about pediatric outcomes Ng

Infant born to mothers with SARS It is likely that all -or the majority -of the cases presented in this series are also included in Wong

Author Year Type of CoV Diagnosis

Sars-CoV-2 Typical sign of viral respiratory infection at CT Wang 2020

Sars-CoV-2 Bilateral ground glass opacities at CT Zhu 2020

Sars-CoV-2 Bilateral ground glass opacities, patchy consolidation, blurred borders at CT Li 2020

Sars-CoV-2 Ground glass opacities, crazy paving, consolidations at CT Jeong

Diffuse opacity in the lower lung area at CXR Alserehi

Bilateral infiltrates and lower lobe opacity at CXR Assiri

Bilateral infiltrates and lower lobe opacity at CXR Malik

Bilateral consolidations at CT Park

Patchy lobe infiltrates at CXR Wong

Features suggestive for progressive air-space disease at CXR

Features suggestive for atypical pneumonia at CXR Robertson

Bilateral lower lobe infiltrates at CXR Schneider

Progressive pulmonary infiltrates at CXR

Sars-CoV Diffuse infiltrates at CXR

For more information, visit www.prisma-statement.org. Mers-CoV 1 stillbirth: At 34 weeks, the mother complained shortness of breath since 3 days and was admitted for elevated blood pressure and 3+ proteinuria consistent with preeclampsia, and pneumonia was diagnosed by means of chest radiography. Fetal heart tones were absent, and intrauterine fetal demise was suspected. A stillborn infant was delivered the same day. 1 neonatal death: At 24 weeks gestation, the mother presented to the hospital on 23 October with cough and myalgia, and chest radiography at admission showed a right lower lobe opacity. Her respiratory status deteriorated during hospitalization, and she was admitted to the ICU on 28 October for ARDS requiring intubation and mechanical ventilation. On 31 October, the patient delivered a 240-gram infant by cesarean delivery. The infant died 4 hours after birth. Payne 2015 Mers-CoV