Effect of coronavirus disease 2019 (COVID‐19) on maternal, perinatal and neonatal outcome: systematic review

ABSTRACT Objective To evaluate the effect of coronavirus disease 2019 (COVID‐19) on maternal, perinatal and neonatal outcome by performing a systematic review of available published literature on pregnancies affected by COVID‐19. Methods We performed a systematic review to evaluate the effect of COVID‐19 on pregnancy, perinatal and neonatal outcome. We conducted a comprehensive literature search using PubMed, EMBASE, the Cochrane Library, China National Knowledge Infrastructure Database and Wan Fang Data up to and including 20 April 2020 (studies were identified through PubMed alert after that date). For the search strategy, combinations of the following keywords and medical subject heading (MeSH) terms were used: ‘SARS‐CoV‐2’, ‘COVID‐19’, ‘coronavirus disease 2019’, ‘pregnancy’, ‘gestation’, ‘maternal’, ‘mother’, ‘vertical transmission’, ‘maternal–fetal transmission’, ‘intrauterine transmission’, ‘neonate’, ‘infant’ and ‘delivery’. Eligibility criteria included laboratory‐confirmed and/or clinically diagnosed COVID‐19, patient being pregnant on admission and availability of clinical characteristics, including at least one maternal, perinatal or neonatal outcome. Exclusion criteria were non‐peer‐reviewed or unpublished reports, unspecified date and location of the study, suspicion of duplicate reporting and unreported maternal or perinatal outcomes. No language restrictions were applied. Results We identified a high number of relevant case reports and case series, but only 24 studies, including a total of 324 pregnant women with COVID‐19, met the eligibility criteria and were included in the systematic review. These comprised nine case series (eight consecutive) and 15 case reports. A total of 20 pregnant patients with laboratory‐confirmed COVID‐19 were included in the case reports. In the combined data from the eight consecutive case series, including 211 (71.5%) cases of laboratory‐confirmed and 84 (28.5%) of clinically diagnosed COVID‐19, the maternal age ranged from 20 to 44 years and the gestational age on admission ranged from 5 to 41 weeks. The most common symptoms at presentation were fever, cough, dyspnea/shortness of breath, fatigue and myalgia. The rate of severe pneumonia reported amongst the case series ranged from 0% to 14%, with the majority of the cases requiring admission to the intensive care unit. Almost all cases from the case series had positive computed tomography chest findings. All six and 22 cases that had nucleic‐acid testing in vaginal mucus and breast milk samples, respectively, were negative for severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2). Only four cases of spontaneous miscarriage or termination were reported. In the consecutive case series, 219/295 women had delivered at the time of reporting and 78% of them had Cesarean section. The gestational age at delivery ranged from 28 to 41 weeks. Apgar scores at both 1 and 5 min ranged from 7 to 10. Only eight neonates had birth weight < 2500 g and nearly one‐third of neonates were transferred to the neonatal intensive care unit. There was one case of neonatal asphyxia and death. In 155 neonates that had nucleic‐acid testing in throat swab, all, except three cases, were negative for SARS‐CoV‐2. There were no cases of maternal death in the eight consecutive case series. Seven maternal deaths, four intrauterine fetal deaths (one with twin pregnancy) and two neonatal deaths (twin pregnancy) were reported in a non‐consecutive case series of nine cases with severe COVID‐19. In the case reports, two maternal deaths, one neonatal death and two cases of neonatal SARS‐CoV‐2 infection were reported. Conclusions Despite the increasing number of published studies on COVID‐19 in pregnancy, there are insufficient good‐quality data to draw unbiased conclusions with regard to the severity of the disease or specific complications of COVID‐19 in pregnant women, as well as vertical transmission, perinatal and neonatal complications. In order to answer specific questions in relation to the impact of COVID‐19 on pregnant women and their fetuses, through meaningful good‐quality research, we urge researchers and investigators to present complete outcome data and reference previously published cases in their publications, and to record such reporting when the data of a case are entered into one or several registries. © 2020 The Authors. Ultrasound in Obstetrics & Gynecology published by John Wiley & Sons Ltd on behalf of the International Society of Ultrasound in Obstetrics and Gynecology.


INTRODUCTION
With over 3 million individuals infected worldwide as of 2 May 2020 1 , the coronavirus disease 2019 , caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), is a global public health crisis 2,3 . Most cohort studies have focused on evaluating the effects of COVID-19 on the general population [2][3][4] and there are insufficient data on its impact on vulnerable populations, such as pregnant women.
It is recognized that pregnant women are at an increased risk of acquiring viral respiratory infection and developing severe pneumonia, due to the physiologic changes in their immune and cardiopulmonary systems 5,6 . Lessons learned from the two previous notable coronavirus outbreaks, the severe acute respiratory syndrome coronavirus (SARS-CoV) and the Middle East respiratory syndrome coronavirus (MERS-CoV), suggest that pregnant women are particularly susceptible to adverse outcome, including the need for endotracheal intubation, admission to an intensive care unit (ICU), renal failure and death [7][8][9] . The first study describing the clinical characteristics and investigating the possibility of vertical transmission of SARS-CoV-2 in nine pregnant women, with laboratory-confirmed COVID-19, demonstrated that the severity of COVID-19 in pregnant women was similar to that in non-pregnant adults; and that there was no evidence of vertical transmission, as SARS-CoV-2 was not detected in amniotic-fluid, cord-blood and neonatal throat-swab samples in six cases 10 . To date, the largest series reporting on both pregnancy and neonatal outcomes, including a total of 99 SARS-CoV-2-infected pregnant women, demonstrated that COVID-19 during pregnancy was not associated with an increased risk of adverse outcomes, such as spontaneous preterm birth 11 . None of the 100 neonates born to these women was infected with SARS-CoV-2. Based on very scarce data to date, conflicting evidence from nucleic acid-based testing and antibody testing, in neonates born to mothers with COVID-19, has raised further controversy in relation to the risk of vertical transmission during pregnancy 12,13 .
The objective of this study was to perform a systematic review of available published literature on pregnancy affected by COVID-19, in order to evaluate the effect of COVID-19 on maternal, perinatal and neonatal outcome.
Of note, at the peak of the SARS-CoV-2 outbreak in the Hubei province, China, cases with relevant symptoms, significant epidemiological history and typical computed tomography (CT) chest finding were clinically diagnosed as COVID-19 pneumonia, without the need for laboratory confirmation. Therefore, eligibility criteria included laboratory-confirmed and/or clinically diagnosed COVID-19, patient being pregnant on admission and availability of clinical characteristics, including at least one maternal, perinatal or neonatal outcome. Exclusion criteria were non-peer-reviewed or unpublished reports, unspecified date and location of the study, suspicion of duplicate reporting and unreported maternal or perinatal outcomes. No language restrictions were applied.

Study selection
Relevant titles were selected from the first screening and abstracts of citations were reviewed independently by two reviewers (J.J. and M.M.G.) to identify all potentially relevant articles. We identified a high number of case reports and case series. After the first screening of titles and abstracts, we decided to repeat the screening procedure and exclude case reports from China or case series that included fewer than 10 cases from China, in order to avoid duplication of cases as there have since been several cohort series published. The potentially relevant articles were evaluated by the same reviewers. The reference lists of the relevant original and review articles were searched for additional reports. Full-text articles were retrieved for further consideration for inclusion. Disagreements were resolved by a third author (L.C.P.), who also reviewed independently the final included articles to confirm they met the inclusion criteria. The review was registered in PROSPERO on 23 April 2020, prior to data extraction (registration number: CRD42020181557) 14 .

Data extraction, quality assessment and outcome measures
Two authors (J.J. and M.M.G.) extracted the information (population, outcome, study design and results) from the selected studies. A modified version of the Cochrane public health group data extraction and assessment template 15 , which was previously piloted by the researchers, was used to tabulate the findings of the included articles. The methodological quality of the studies was assessed independently by the same two authors using the Joanna Briggs Institute (JBI) tool for case series and case reports 16 .
Publication bias was considered high since all the included studies were either case series or case reports. The quality of this review was validated using the preferred reporting items for systematic reviews and meta-analyses (PRISMA) tool 17 .
The following information was extracted from the included studies: author names; institution and country; study design; sample size; maternal age; gestational age at admission; symptoms at admission; pregnancy complications; gestational age at delivery; mode of delivery; disease severity; laboratory and radiological findings; maternal and neonatal outcomes; and sample collection (amniotic fluid, cord blood, placenta, maternal vaginal secretion, urine, feces and breast milk, neonatal pharyngeal swab, neonatal blood, neonatal urine, neonatal feces, neonatal gastric juice). Any evidence of maternal-fetal transmission of SARS-CoV-2 was also recorded. Not all studies reported on all assessed variables. The denominators reported in the results were generated from the papers that provided such data. Studies that did not report on a specific outcome were recorded as not reported. We contacted directly the corresponding authors of potentially eligible studies, when further clarification on their data was needed; such as overlapping cases in different studies published by the same group or to complete outcome data. Details of the requested data and responses of the authors, or lack of response, are provided in Table S1.

RESULTS
reported on a total of 295 pregnant patients 11,19,[21][22][23][24][25][26]  We recorded data related to maternal and perinatal characteristics, and the clinical manifestations of COVID-19 at admission, including laboratory testing, treatment received and maternal, perinatal and neonatal outcome. The majority of the cases originated from China 11,21-25 , but cases from Australia (one case report) 38 , Canada and France (one case report) 40 , Korea (one case report) 37 , Iran (two case reports and one case series) 20,35,41 , Italy (one case series and one case report) 26,29 , Peru (one case report) 28 , Spain (two case reports) 27,31 , Sweden (one case report) 30 , Turkey (one case report) 34 and USA (one case series and four case reports) 19,32,33,36,39 were also included. We differentiated cases that were laboratory-confirmed from those that were diagnosed clinically, but data from the two types of diagnosis were combined for presentation. A laboratory-confirmed case of COVID-19 was defined as a positive result on quantitative reverse transcriptase polymerase chain reaction (qRT-PCR) assay of maternal pharyngeal swab specimens. As mentioned, at the peak of the COVID-19 outbreak in Hubei province, China, cases with relevant symptoms, significant epidemiological history and typical chest CT finding were diagnosed clinically as COVID-19, because the viral nucleic acid test was reported 42 to have a false-negative rate of 30%.
Quality assessment of the included studies is reported in Tables S3 and S4. The inclusion of eight case series with consecutive cohorts partly mitigated the high publication bias expected from case reports.

Clinical characteristics
In the combined data from the eight consecutive case series 11,19,[21][22][23][24][25][26] , the maternal age ranged from 20 to 44 years and the gestational age on admission ranged from 5 to 41 weeks ( Table 1). The most common symptoms at presentation were fever, cough, dyspnea/shortness of breath, fatigue and myalgia. Based on the case series by Yan et al. 11 and Breslin et al. 19 , up to one-third of pregnant patients with COVID-19 were asymptomatic  27 Kalafat (2020) 34 Zamaniyan (2020) 41 Only first author is given for each study. Data are presented as n or n/N (%). *Hantoushzadeh (2020) 20 excluded from calculation because not consecutive case series. †Additional information for this study retrieved from Buonsenso (2020) 53 . ‡Seven cases received oxygen support (nasal cannula, continuous positive airway pressure); authors did not specify invasive or non-invasive ventilation. §Cases from Ferrazzi (2020) 26 not included. ICU, intensive care unit; NR, not reported. on admission. The rate of severe pneumonia reported amongst the consecutive case series ranged from 0% to 14%, with the majority of the cases requiring ICU admission, of which only a few cases received invasive mechanical ventilation ( Table 2). The majority of cases (70.7%) received antibiotic therapy, whilst only 82 of 217 (37.8%) and 31 of 176 (17.6%) cases received antiviral therapy and corticosteroids, respectively. Only two cases were reported to have received hydroxychloroquine. There were no cases of maternal death in the consecutive case series. Seven maternal deaths were reported in the Iranian case series 20 of nine non-consecutive cases with severe COVID-19. Almost all women in the case series had positive CT chest findings, including patchy shadowing or ground-glass opacity. On admission, the majority of the cases with laboratory test results had normal or low leukocyte count (146/182; 80.2%), while just under half had lymphocytopenia (85/197; 43.1%) and increased C-reactive protein (CRP) (90/197; 45.7%) ( Table 3). Of note, all six cases that had nucleic-acid testing in vaginal mucus and all 22 cases that had nucleic-acid testing in breast milk samples were negative for SARS-CoV-2.
Of the 20 pregnant women with COVID-19 included in the case reports, two maternal deaths were reported, one each in the case reports of Karami et al. 35 and Zamaniyan et al. 41 from Iran ( Table 2). The first case was a 27-year-old woman at 30 weeks' gestation who complained of fever, cough and myalgia for 3 days. Her admission laboratory tests showed leukopenia and thrombocytopenia, accompanied by elevated CRP and lactate dehydrogenase levels 35 . Soon after admission, her temperature was noted to be 40 • C and her respiratory rate was 55 breaths per min, accompanied by suprasternal and intercostal retraction. Immediate blood tests showed metabolic alkalosis while the patient was under non-invasive ventilation. She was eventually intubated

21
for mechanical ventilation due to worsening acute respiratory distress syndrome (ARDS), based on clinical and radiological findings. On day 2 of admission, the patient had spontaneous onset of labor and delivered vaginally a cyanotic neonate with no signs of life, that did not respond to neonatal cardiopulmonary resuscitation. One day after birth, the mother developed multiorgan failure (ARDS, acute kidney injury and septic shock) and died 35 .
The second case was a 22-year-old woman at 32 weeks' gestation who experienced dyspnea, myalgia, anorexia, nausea, non-productive cough and fever for 4 days 41 . On admission, she was treated with Azithromycin, Ceftriaxone, Kaletra, Tamiflu and hydroxychloroquine. Based on the CT chest finding, lymphopenia, worsening pneumonia symptoms and an unfavorable cervix for induction, a Cesarean delivery was indicated at 33 weeks' gestation. A preterm female infant, weighing 2350 g, was delivered uneventfully, with Apgar scores of 8 and 9 at 1 and 5 min, respectively. The amniotic fluid tested positive for SARS-CoV-2 by qRT-PCR. The immediate post-delivery nasal and throat swabs of the newborn tested negative for SARS-CoV-2, however, repeat testing 24 h later was positive. Such results raised the possibility of vertical transmission. The mother underwent peritoneal dialysis due to ARDS on days 4 and 6 postpartum, and required intubation and mechanical ventilation due to sudden oxygen desaturation to 70% on day 10. She developed emphysema after intubation that resolved spontaneously on day 12; however, her condition deteriorated dramatically and she died on day 15 postpartum 41 . Gidlöf (2020) 30 González Romero (2020) 31 Iqbal (2020) 32 Juusela (2020) 33 Karami (2020) 35 Kelly (2020) 36  Only first author is given for each study. Data are presented as n, n/N, range or n/N (%) unless indicated otherwise. *Hantoushzadeh (2020) 20 excluded from calculation because not consecutive case series. †Additional information for this study retrieved from Buonsenso (2020) 53 . ‡Including two women with intrauterine fetal death (IUFD), who remained undelivered at time of maternal death (one with twin pregnancy). §Including one IUFD. ¶IUFD. CS, Cesarean section; GA, gestational age; GDM, gestational diabetes mellitus; HPD, hypertensive disorder of pregnancy; HT, hypothyroidism; NR, not reported; PE, pre-eclampsia; PP/PA, placenta previa/accreta.

Pregnancy and neonatal outcomes
Based on the consecutive case series, the rates of gestational diabetes, hypertensive disorders of pregnancy and pre-eclampsia did not appear to be higher in pregnant women with COVID-19 compared to pregnant women without (Table 4). There were only a few cases with hypothyroidism and placenta previa/accreta. A quarter of the cases (72/295) had not been delivered at the time of reporting. Only four cases of spontaneous miscarriage or termination were reported. In the 219 cases that were delivered, including two with twin pregnancy, the majority had a Cesarean section. The gestational age at delivery ranged from 28 to 41 weeks. The Apgar scores at both 1 and 5 min ranged from 7 to 10 ( Table 5).
Only eight neonates had birth weight < 2500 g. Nearly one-third of the neonates were transferred to the neonatal intensive care unit (NICU), due mainly to the need for investigation and monitoring as a result of maternal infection. There was one case of neonatal asphyxia and death in the consecutive case series. Of 19 neonates for which laboratory test results were reported, only four and two neonates had increased leukocyte count and CRP, respectively (  26 , were negative for SARS-CoV-2. In the Iranian case series of nine non-consecutive cases with severe COVID-19, there were two cases of intrauterine fetal death (IUFD), that remained undelivered at the time of maternal death (one with twin pregnancy), and two further cases of IUFD amongst the remaining seven cases 20 . An additional two neonatal deaths occurred, as part of a twin pregnancy 20 .
In one case report, the neonatal throat swab tested positive for SARS-CoV-2 (Table 6) 28 . The mother was a 41-year-old woman with pre-existing diabetes mellitus and significant COVID-19 exposure from immediate family members. She presented at 33 weeks' gestation with a 4-day history of malaise, fatigue, low-grade fever and progressive shortness of breath. The nasopharyngeal swab of the patient was positive

Case reports
Alonso Díaz (2020) 27  Juusela (2020) 33 Kalafat (2020) 34 1 Karami (2020) 35 Kelly (2020) 36  Only first author is given for each study. Data are presented as n, range or n/N (%) unless indicated otherwise. *Hantoushzadeh (2020) 20 excluded from calculation because not consecutive case series. †Additional information for this study retrieved from Buonsenso (2020) 53 . ‡Including one twin pregnancy. §Twin pregnancy. ¶Mean value. **Two very preterm newborns had 5-min Apgar score < 7. NICU, neonatal intensive care unit; NR, not reported; RT, respiratory tract. delivery, the neonatal nasopharyngeal swab tested positive for SARS-CoV-2 by qRT-PCR, which was repeated 48 h later and remained positive. However, anti-SARS-CoV-2 immunoglobulin-M (IgM) and -G (IgG) were negative at birth. The possibility of postpartum neonatal infection cannot be completely excluded because of the delay in testing. The newborn required ventilatory support for 12 h, after which he was extubated and placed on continuous positive airway pressure, with favorable outcome.

Risk of bias
All case series, except one 20 , had low risk of bias due to clear inclusion criteria, proper evaluation of the disease status, consecutive and complete enrolment, clarity of the reported variables and appropriate statistical analysis. Additionally, although most of the case reports failed in the reporting of other adverse or unanticipated events besides the main case presentation, all presented the clinical case clearly and provided takeaway lessons.

Main findings
This systematic review has demonstrated that, first, the most common reported symptoms of COVID-19 are fever, cough, dyspnea/shortness of breath, fatigue and myalgia; second, on admission, most cases have patchy shadowing or ground-glass opacity on CT of the chest, while normal or low leukocyte count, lymphocytopenia and raised CRP were the most common laboratory findings observed in pregnant patients with COVID-19; third, the rate of severe pneumonia reported amongst the case series ranged from 0% to 14%; fourth, of the 324 pregnant women included, seven maternal deaths were reported in a case series of nine non-consecutive cases with severe COVID-19 and two amongst the case reports (as of 28 April 2020); fifth, COVID-19 does not appear to increase the risk of adverse pregnancy outcome, such as pre-eclampsia; sixth, only a few cases of spontaneous miscarriage or termination have been reported in pregnancies with COVID-19; seventh, there is a lack of completeness of pregnancy outcome data; eighth, based on women who had delivered at the time of reporting, the gestational age at delivery ranged from 28 to 41 weeks and the majority of cases had Cesarean delivery; and ninth, among the consecutive case series, there were three reported cases of neonates testing positive for SARS-CoV-2 and among the case reports, there was one case each with positive SARS-CoV-2 in amniotic fluid and neonatal throat swab.

Comparison with previous studies
At present, there is much controversy relating to the possibility of vertical mother-to-baby transmission of SARS-CoV-2. In two earlier studies, with a combined total of 10 pregnant women with COVID-19 in the third trimester, amniotic-fluid, cord-blood and neonatal throat-swab samples tested negative for SARS-CoV-2, suggesting there was no evidence of vertical transmission in women who developed COVID-19 pneumonia in late pregnancy 10,43 . In another series, a neonate born to a pregnant woman with COVID-19 tested positive for SARS-CoV-2 in the pharyngeal swab sample 36 h after birth, but it was subsequently confirmed that qRT-PCR testing of the placenta and cord blood was negative for SARS-CoV-2, suggesting that intrauterine vertical transmission might not have occurred 44,45 . Two research letters 12,13 have suggested the possibility of vertical transmission of SARS-CoV-2, based on the presence of IgM antibodies in blood drawn from three neonates born to mothers with COVID-19. However, for all three, the respiratory samples tested negative for SARS-CoV-2. Furthermore, one neonate had repeat testing for SARS-CoV-2 IgG and IgM antibodies, and the observed rapid decline of IgG levels within 14 days, along with a decline in IgM antibodies, strongly suggested that neonatal SARS-CoV-2 IgG antibodies were derived transplacentally from the mother and not actively induced by the presumed neonatal infection.
The findings of the case of maternal death reported by Zamaniyan et al. 41 , in which amniotic fluid and neonatal nasal and throat swabs tested positive for SARS-CoV-2, and the case report by Alzamora et al. 28 , in which the neonatal throat swab tested positive for SARS-CoV-2, have again raised the possibility of vertical transmission. The authors of the former case report could not measure specific antibodies in the neonate, which might have provided supplementary evidence for vertical transmission of SARS-CoV-2. This case is of particular interest because the pregnant patient might have had virulent COVID-19, which was not immediately apparent prior to delivery, as her clinical course deteriorated dramatically only after delivery. The common features of these two cases of neonatal SARS-CoV-2 infection are that the pregnant patients contracted the virus preterm and had severe COVID-19. There appear to be at least two ways through which SARS-CoV-2 can cause intrauterine infection by vertical transmission. Angiotensin-converting enzyme 2 (ACE2), which has been indicated recently as the putative surface receptor of sensitive cells for SARS-CoV-2 46 , has been shown to be expressed in human placentas 47 . This opens up the possibility that SARS-CoV-2 could spread transplacentally through ACE2. Specifically, the viral surface spike glycoprotein (S-protein) of SARS-CoV-2 is cleaved by transmembrane protease serine 2 (TMPRSS2) to facilitate efficiency of entry and viral replication, and there is emerging evidence that SARS-CoV-2 co-opts and recruits additional host proteases for transmissibility 48,49 . It is, however, unknown whether there is placental expression of host proteases (such as TMPRSS2 and others) necessary for cleavage of the S-protein and receptor priming. On the other hand, placental barrier COVID-19 and pregnancy 25 damage caused by severe maternal hypoxemia in women with COVID-19 could also be one potential way through which SARS-CoV-2 can cause intrauterine infection. There is an urgent need to investigate further the possibility of vertical transmission 50 of SARS-CoV-2.
The second case of maternal death, that was reported by Karami et al. 35 , underwent autopsy and histopathologic evaluation of paraffin-embedded lung tissue that showed alveolar spaces with focal hyaline membrane, pneumocyte proliferation and metaplastic changes. There was evidence of viral pneumonia (viral cytopathic effect, including multinucleation and nuclear atypia and a mild increase in alveolar wall thickness). These findings are comparable to observations in non-pregnant cases 51 . The Iranian case series reported the maternal death of seven pregnant women presenting with severe COVID-19, during the latter second or third trimester, over a 30-day interval 20 . All cases received a three-drug regimen, which included oseltamivir 75 mg PO, every 12 h for 5 days, hydroxychloroquine sulfate 400 mg PO daily or chloroquine sulfate 1000 mg tablet PO, as a single dose, and lopinavir/ritonavir 400/100 mg PO, every 12 h for 5 days. There were two key features in these cases. First, the average maternal age in this series was higher than that in others (36.7 ± 7.3 vs 30.3 ± 3.6 years) 20 . Second, according to the authors, none of the pregnant patients had pre-existing comorbidities, such as hypertension, cardiovascular disease and asthma 20 . The authors believed that the COVID-19 pregnant patients did not receive suboptimal care and raised concern regarding the potential for maternal death among pregnant women diagnosed with COVID-19 in the latter trimester. Notably, the authors did not provide details of the cause(s) of death 20 .

Strengths and limitations
This systematic review has several strengths. We undertook a meticulous process in identifying duplicate reporting: (i) case reports from China were excluded; (ii) when a hospital had published their cases more than once, only the paper with the most data or published on the later date was included; (iii) we were able to identify the duplicates from our own publication 11 , which is the biggest cohort series with available pregnancy outcome data from China; and (iv) when necessary, we contacted the corresponding authors of included studies using their native language, i.e. Chinese and Spanish, thus increasing the chance of receiving a response from them. Moreover, we did not apply any language restrictions in order to capture a global picture of the impact of COVID-19 on pregnancy and perinatal outcomes, and most importantly, we had the capability to ensure that data from China could be accurately interpreted and recorded.
This study also has several major limitations. The main limitation arises from the nature of the included studies. Although case reports are an important source of knowledge, they are not suitable for inference statistics and are expected to suffer from high publication bias. On the other hand, the main limitation of case series is the lack of internal controls; thus, they typically yield very low-quality evidence. However, at this point in time when higher-quality studies on COVID-19 in pregnancy are yet to be conducted, both case reports and case series provide valuable and necessary information to guide clinicians in their decision making. We included papers that reported at least one maternal, perinatal or neonatal outcome. Given the severity of the pandemic and the urgent need for knowledge sharing, we felt that it would still be valuable to include papers that had provided limited pregnancy outcome data. In view of the low rate of miscarriage, it is likely that asymptomatic cases with SARS-CoV-2 infection resulting in a miscarriage were not identified and, therefore, are under-reported. In our opinion, it was simply a failure of appropriate diagnosis and there was no reporting/publication bias. We were unable to include the second largest cohort series with available pregnancy outcome data from China because the data were pooled from a national registry 52 . Despite contacting the corresponding author directly, it was not possible to identify the actual source of cases. All 68 pregnant patients with pregnancy outcome data were cases from Wuhan and the authors did not make reference to previously published cases. In our publication 11 , we included 77 cases from Wuhan and we had strong reasons to believe that there was significant overlap between the two series; therefore, we regrettably had to exclude the data from Chen et al. 52 in order to avoid inflating the number of SARS-CoV-2-infected pregnant cases from China. We also learned that pregnant women with COVID-19 could have been transferred to other hospitals, which made it difficult to determine duplicate reporting as cases could have been reported by both hospitals in which the woman received care. We decided to combine laboratory-confirmed and clinically diagnosed COVID-19 cases in our analysis because both groups had similar outcomes 11 . We believed it was important to include the clinically diagnosed cases in this review, as such cases had typical COVID-19 chest CT findings and significant epidemiological exposure. As of 28 April 2020, maternal death had been reported in seven and two cases in the non-consecutive case series 20 and amongst the case reports 35,41 , respectively. Traditionally, any maternal death requires an extensive process of investigation and therefore it is not immediately reported and made public. It was not possible to determine COVID-19-related maternal case fatality rate and such data will come to light only at the end of the pandemic.

Conclusions
Despite the increasing number of published studies on COVID-19 in pregnancy, there are insufficient good-quality data to draw unbiased conclusions with regard to the severity of COVID-19 or specific complications of the disease in pregnant women, as well as vertical transmission and perinatal and neonatal complications. In order to answer specific questions in relation to the impact of COVID-19 on pregnant women and
their fetuses through meaningful good-quality research, we urge researchers and investigators to present complete outcome data and reference previously published cases in their publications, and to record such reporting when the data of a case are entered into a registry or several registries.

SUPPORTING INFORMATION ON THE INTERNET
The following supporting information may be found in the online version of this article: Appendix S1 Systematic literature search This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.
Conclusiones A pesar del creciente número de estudios publicados sobre COVID-19 en embarazos, no hay suficientes datos de buena calidad para sacar conclusiones no sesgadas con respecto a la gravedad de la enfermedad o a las complicaciones específicas de COVID-19 en las mujeres embarazadas, así como a la transmisión vertical y a las complicaciones perinatales y neonatales. A fin de responder a preguntas específicas en relación con los efectos de COVID-19 en mujeres embarazadas y sus fetos, mediante una investigación significativa de buena calidad, se insta a los investigadores y a los encargados de investigaciones a que presenten en sus publicaciones datos completos sobre los resultados y casos de referencia publicados anteriormente, y a que registren esos informes cuando los datos de un caso se introduzcan en uno o varios registros.