key: cord-0865451-tr8eiz97
authors: GIULIANI, Francesca; OROS, Daniel; GUNIER, Robert B.; DEANTONI, Sonia; RAUCH, Stephen; CASALE, Roberto; NIETO, Ricardo; BERTINO, Enrico; REGO, Albertina; MENIS, Camilla; GRAVETT, Michael G.; CANDIANI, Massimo; DERUELLE, Philippe; GARCÍA-MAY, Perla K.; MHATRE, Mohak; Ado USMAN, Mustapha; ABD-ELSALAM, Sherief; ETUK, Saturday; NAPOLITANO, Raffaele; LIU, Becky; PREFUMO, Federico; SAVASI, Valeria; Silva, Marynéa; BAAFI, Eric; Ariff, Shabina; MAIZ, Nerea; Baffah AMINU, Muhammad; CARDONA-PEREZ, Jorge Arturo; CRAIK, Rachel; TAVCHIOSKA, Gabriela; BAKO, Babagana; BENSKI, Caroline; HASSAN-HANGA, Fatima; SAVORANI, Mónica; SENTILHES, Loïc; Carola CAPELLI, Maria; TAKAHASHI, Ken; VECCHIARELLI, Carmen; IKENOUE, Satoru; THIRUVENGADAM, Ramachandran; SOTO CONTI, Constanza P.; CETIN, Irene; NACHINAB, Vincent Bizor; ERNAWATI, Ernawati; DURO, Eduardo A.; KHOLIN, Alexey; TEJI, Jagjit Singh; EASTER, Sarah Rae; SALOMON, Laurent J.; AYEDE, Adejumoke Idowu; CERBO, Rosa Maria; AGYEMAN-DUAH, Josephine; ROGGERO, Paola; ESKENAZI, Brenda; LANGER, Ana; BHUTTA, Zulfiqar A.; KENNEDY, Stephen H.; PAPAGEORGHIOU, Aris T.; VILLAR, Jose
title: Effects of prenatal exposure to maternal COVID-19 and perinatal care on neonatal outcome: results from the INTERCOVID Multinational Cohort Study
date: 2022-04-19
journal: Am J Obstet Gynecol
DOI: 10.1016/j.ajog.2022.04.019
sha: 34f45324beda45a5b8f0efd02b21db290646112b
doc_id: 865451
cord_uid: tr8eiz97

Background The effect of COVID-19 in pregnancy on maternal outcomes and its association with preeclampsia and gestational diabetes has been reported; however, a detailed understanding of the effect of maternal positivity, delivery mode and perinatal practices on fetal and neonatal outcomes is urgently needed. Objective To evaluate the impact of COVID-19 on fetal and neonatal outcomes and the role of the mode of delivery, breastfeeding and early neonatal care practices on the risk of mother-to-child transmission. Study Design In this cohort study that took place from March 2020 to March 2021, involving 43 institutions in 18 countries, 2 unmatched, consecutive, not-exposed women were concomitantly enrolled immediately after each infected woman was identified, at any stage of pregnancy or delivery, and at the same level of care to minimize bias. Women and neonates were followed up until hospital discharge. COVID-19 in pregnancy was determined by laboratory confirmation of COVID-19 and/or radiological pulmonary findings or 2 or more predefined COVID-19 symptoms. The outcome measures were indices of neonatal and perinatal morbidity and mortality, neonatal positivity and its correlation with mode of delivery, breastfeeding and hospital neonatal care practices. Results A total of 586 neonates born to women with COVID-19 diagnosis and 1535 neonates born to women without COVID-19 diagnosis were enrolled. Women with COVID-19 diagnosis had a higher rate of cesarean section (52.8% compared to 38.5% for those without COVID-19 diagnosis, p<0.01) and pregnancy related complications such as hypertensive disorders of pregnancy and fetal distress, all with p-value < 0.001, compared to women without COVID-19 diagnosis. Maternal diagnosis of COVID-19 carried an increased rate of preterm birth (p ≤0.001) and lower neonatal weight (p ≤0.001), length, and head circumference at birth. In mothers with COVID-19 diagnosis, the length of in-utero exposure was significantly correlated to the risk of the neonate testing positive (OR, 4.5; 95% CI 2.2-9.4 for length of in-utero exposure > 14 days). Among neonates born to mothers with COVID-19 diagnosis, birth via cesarean section was a risk factor for them testing positive for COVID-19 (OR 2.4, 95% CI 1.2-4.7), even when severity of maternal conditions was considered and after multivariable logistic analysis. In the subgroup of neonates born to women with COVID-19 diagnosis, the outcomes worsened when the neonate also tested positive, with higher rates of Intensive Care Unit admission, fever, gastrointestinal and respiratory symptoms and death, even after adjusting for prematurity. Breastfeeding by mothers with COVID-19 diagnosis, as well as hospital neonatal care practices including immediate skin-to-skin contact and rooming-in, were not associated with an increased risk of newborn positivity. Conclusions In this multinational cohort study, COVID-19 in pregnancy was associated with increased maternal and neonatal complications. Cesarean section was significantly associated with newborn COVID-19 diagnosis. Vaginal delivery should be considered the safest mode of delivery if obstetrical and health conditions allow it. Mother to child skin-to-skin contact, rooming-in and direct breastfeeding did not represent risk factors for newborn COVID-19 diagnosis, thus well-established best practices can be continued among women with COVID-19 diagnosis.

COVID-19 diagnosis in pregnancy and the postnatal period carries a substantial risk of neonatal morbidity and 103 mortality, as compared with their not-exposed counterparts, with the most severe effects observed in test-104 positive neonates born to women with COVID-19 diagnosis. 105

SHORT TITLE 106

Association of prenatal exposure to maternal COVID-19 and perinatal care with neonatal outcome. 107 108 AJOG AT A GLANCE

Why was this study conducted? 110

This study aimed to describe and quantify any association between COVID-19 during pregnancy and newborn 111 outcomes, and to assess the safety of perinatal care practices, including breastfeeding, in mothers with a 112 What are the key findings? 114 COVID-19 diagnosis in pregnancy and the postnatal period carries a substantial risk of neonatal morbidity and 115 mortality, as compared with their not-exposed counterparts, with the most severe effects observed in test-116 positive neonates born to women with COVID-19 diagnosis. Cesarean section was significantly associated with 117

neonatal positivity. Vaginal delivery should be considered as the preferred mode of delivery even in 118 symptomatic women, when obstetric and general health conditions allow it. Mother to child skin-to-skin 119 contact, rooming-in and direct breastfeeding do not represent risk factors for neonatal test positivity; thus, well-120 established best evidence-based practices can be continued among women with COVID-19 diagnosis. 121

What does this study add to the already known? 122 COVID-19 in pregnancy is associated with adverse newborn outcomes; unless otherwise indicated cesarean 123 section should not be the preferred mode of delivery in positive mothers. Skin-to-skin and breastfeeding should 124 be encouraged. 125 J o u r n a l P r e -p r o o f Abstract 126

BACKGROUND The effect of COVID-19 in pregnancy on maternal outcomes and its association with 127 preeclampsia and gestational diabetes has been reported; however, a detailed understanding of the effect of 128 maternal positivity, delivery mode and perinatal practices on fetal and neonatal outcomes is urgently needed. 129

OBJECTIVE To evaluate the impact of COVID-19 on fetal and neonatal outcomes and the role of the mode of 130 delivery, breastfeeding and early neonatal care practices on the risk of mother-to-child transmission. 131

STUDY DESIGN In this cohort study that took place from March 2020 to March 2021, involving 43 132 institutions in 18 countries, 2 unmatched, consecutive, not-exposed women were concomitantly enrolled 133 immediately after each infected woman was identified, at any stage of pregnancy or delivery, and at the same 134 level of care to minimize bias. Women and neonates were followed up until hospital discharge. COVID-19 in 135 pregnancy was determined by laboratory confirmation of COVID-19 and/or radiological pulmonary findings or 136 2 or more predefined COVID-19 symptoms. present report focuses on the impact of COVID-19 on neonatal outcomes and the role of the mode of delivery, 190 breastfeeding and early neonatal care practices on the risk of mother-to-child transmission 17-19 . 191 192 MATERIALS AND METHODS

From 2 March 2020 to 18 March 2021, we enrolled women from 43 institutions in 18 countries (Argentina, 195 Brazil, Egypt, France, Ghana, India, Indonesia, Italy, Japan, Mexico, Nigeria, North Macedonia, Pakistan, 196 Russia, Spain, Switzerland, UK, and the US). The distribution by country is presented in Supplemental Figure  197 1. Data on ethnicity were not collected. 198 J o u r n a l P r e -p r o o f

We enrolled a total of 742 women, aged ≥18 years old, at any stage of pregnancy or at delivery, with a COVID-199 19 diagnosis based on: a) laboratory confirmation of SARS-CoV-2 infection by RT-PCR (n=687); b) ≥2 200 predefined COVID-19 symptoms or signs, without laboratory confirmation (n=55). When a woman with 201 COVID-19 diagnosis was identified antenatally, two immediately concomitant women without COVID-19 202 diagnosis aged ≥18 years old of similar gestational age (±2 weeks), receiving standard antenatal care, were 203 enrolled that day in order to create an unbiased sample of all pregnant individuals without COVID-19 diagnosis 204 in these institutions. If not possible or if the women without COVID-19 diagnosis were lost to follow-up, we 205 enrolled two women without COVID-19 diagnosis who were admitted at the same level of care and delivered 206

immediately after the woman with COVID-19 diagnosis. The same selection strategy was employed when a 207 woman with COVID-19 diagnosis was identified at hospital admission and delivery was likely during that 208 admission. As a quality check, we sought confirmation, from a bi-weekly random 10% sample, that the two 209 women without COVID-19 were appropriately chosen; we excluded five women with COVID-19 diagnosis and 210 the corresponding women without COVID-19 diagnosis where such confirmation was missing 12 . 211

For the present analysis, we excluded mother/newborn dyads when the neonate was not tested for COVID-19 212 even if clinically indicated, or when the reason was not clearly described. 213

Live and stillborn, singleton and multiple pregnancies were included, even those with congenital anomalies. In 214 keeping with reporting requirements during the pandemic, we excluded mothers/newborns from the final 215 analysis if their data had already been published in any comparative study with women without COVID-19 216 diagnosis, other than INTERCOVID related papers. 217

The Oxford Tropical Research Ethics Committee and all local ethics committees approved the study. Informed 218 consent (oral or written) was obtained from participants according to local requirements, except when a 219 waiver/exemption of such consent was granted by a local committee. We adhered to the Declaration of Helsinki 220 J o u r n a l P r e -p r o o f and Good Clinical Practice guidelines. The study protocol, including the laboratory tests used, has been 221 previously published 12 . 222

Outcomes definition 223

The primary outcome was the association between maternal COVID-19 exposure and neonatal positivity; the 224 secondary outcome was the association of time of exposure, mode of delivery, breastfeeding and neonatal care 225 practices with neonatal outcomes. 226

Maternal and pregnancy history and delivery mode were collected together with indication for caesarean 227 section, newborn outcomes, and feeding practices with standardized forms as used in the INTERGROWTH-21 st 228

Project 20 . In addition, we recorded detailed data on each mother's health and condition at admission, perinatal 229 management, and in-hospital practices (e.g., skin-to-skin contact, isolation from the neonate, and the practice by 230 mother and hospital staff of using masks and hand washing). We also recorded information regarding the timing 231

and results of SARS-CoV-2 testing and COVID-19-related symptoms for mother and neonate. 232

Gestational age estimation was based on ultrasound measurement of fetal crown-rump length (<14 weeks' 233 gestation) 21 or, if early ultrasound dating was not carried out, the "best obstetric" estimate was used based on 234 all clinical and ultrasound data available at the time of delivery. 235

The total time of exposure to SARS-CoV-2 was defined as the number of days between the woman testing 236 positive or the onset of symptoms and delivery. We chose a 10-day cut off to study the risk in different 237 populations (i.e., women still infectious during labour versus women most probably not-infectious during 238 labour) as the horizontal infectiousness of patients with symptoms or a positive test more than 10 days earlier 239 appears very low 22,23 . The maternal symptoms severity score was defined as a continuous variable made up of 240 the sum of pre-set values attributed to each maternal COVID-19-related symptom, according to the severity of 241 the symptom. 242 J o u r n a l P r e -p r o o f

In the data collection form the indications for delivery, that are often used in medical records, were recorded. 243

For the analyses, in mothers who delivered by cesarean section, those indications were grouped into those 244 potentially COVID-19-related vs others. We included in the potentially COVID-19-related indications 245 hypertensive disorders of pregnancy 15 , fetal distress, fetal growth restriction, suspected small for gestational 246 age (SGA) or fetal growth restriction 10 , premature rupture of membranes (PROM), and infections. SGA was 247 defined as being born with weight below the 10 th percentile based on INTERGROWTH-21 th International 248

Standards for newborn weight 24 . 249

Newborn weight, length and head circumference were assessed against the international INTERGROWTH-21 st 250 standards following a standardized protocol. Measurement instruments were regularly calibrated and used by 251 trained staff. Neonatal health outcomes, diagnostics and treatments were collected in detail and then presented 252 as categories: 1) Neurological problems including seizures, hydrocephalus, neurological disorders, any hypoxic-253 ischaemic encephalopathy and grade 3 or 4 per Papile criteria periventricular haemorrhage/Leukomalacia; 2) 254

Gastrointestinal conditions including no enteral feeding for > 24 hours, necrotising enterocolitis, stoppage of 255 enteral feeding for more than three consecutive days, gastro-esophago-pharyngeal reflux, persistent vomiting, 256 and diarrhoea; 3) Infections including sepsis, hypotension requiring inotropic-steroids and pneumonia/acute 257 respiratory infections; 4) Respiratory conditions including pneumonia/bronchiolitis, apnea of prematurity, 258 bronchopulmonary dysplasia (BPD) and corticosteroids for BPD. 259

Detailed data regarding feeding were recorded and included: the type of feeding, i.e., any breastfeeding (defined 260 as exclusive or partial breastfeeding) and no breastfeeding (defined as exclusive formula or only parenteral 261 nutrition); mode of feeding, i.e., direct breastfeeding, bottle feeding, or tube feeding. Also, information 262 regarding hospital newborn care practices, including immediate skin-to-skin contact, rooming-in and hygiene 263

measures were recorded for neonates tested for COVID-19. All data were collected on newborn care forms 264 during hospital stay and at discharge. 265 J o u r n a l P r e -p r o o f Statistical analysis 275 We used chi-square tests for proportions and t-tests for continuous variables to compare maternal baseline 276 characteristics and early outcomes between neonates born to mothers with, and without a COVID-19 diagnosis; 277 similarly, for neonatal characteristics and other outcomes, we compared the three groups of neonates. We used 278 negative binomial models to calculate relative risks for neonatal outcomes among the three groups; neonates 279 born to mothers without COVID-19 diagnosis were the reference group. We adjusted for the following 280 covariates that were selected using directed acyclic graphs 25 : maternal age, tobacco use, parity, history of 281 pregnancy complications and gestational age. To complement the crude, unadjusted analysis, we explored 282 logistic regression models to calculate odds ratios (OR) and 95% confidence intervals (CI) for neonates testing 283 positive for COVID-19 stratified by the number of days between maternal diagnosis and delivery and adjusting 284

for mode of delivery for comparison (Supplementary Table 3 ). 285

Among neonates tested for COVID-19 and born to women with COVID-19 diagnosis, we collected complete 286 information from newborn care forms to determine if factors during delivery and after birth were related to the 287 neonates testing positive. We used chi-square tests to compare the reasons for cesarean section among neonates 288 J o u r n a l P r e -p r o o f that tested positive vs negative for COVID-19 born to women with COVID-19 diagnosis. We used logistic 289 regression models to calculate ORs and 95% CIs for predictors of the neonates testing positive for We stratified by the time between diagnosis and delivery (≤ 24 hours or > 24 hours) and used chi-square tests to 291 evaluate delivery outcomes, neonatal outcomes and newborn care practices. Finally, as a sensitivity analysis we 292 assessed the associations between neonatal COVID-19 status and neonatal outcomes among neonates born to 293 mothers with a positive COVID-19 test only. 294 295 RESULTS

We enrolled a total of 742 women with a COVID-19 diagnosis based on: a) laboratory confirmation of SARS-297

CoV-2 infection by RT-PCR (n=687); b) ≥2 predefined COVID-19 symptoms or signs, without laboratory 298 confirmation (n=55). Mother/newborn dyads in which the neonate was not tested for COVID 299

were excluded (n=180 neonates and 173 mothers). 300

We therefore included in this analysis 569 women with, and 1500 women without COVID 19 diagnosis. Since 301 multiple pregnancies were included, a total of 586 newborns of mothers with COVID-19 diagnosis and 1535 302 newborns of mothers without COVID-19 diagnosis were included, all with broadly similar demographic 303 characteristics as described in previous papers.. Supplemental Figure 2 provides the study enrollment flowchart. 304 neonates who tested negative (p<0.01). The reason for cesarean section did not significantly differ between 319 groups, either individually or when grouped by COVID-19-related indications vs any other indications (see 320

Supplemental Table 1 ). In a multivariable logistic regression analysis (see Supplemental Table 2 ), including 321 time of exposure and immediate mother/newborn skin-to-skin contact, birth via cesarean section was 322 statistically significantly associated with neonates testing positive for COVID-19 (aOR 2.4, 95% CI 1.2-4.7). 323

Moreover, we investigated if caesarean was independently associated with neonatal positivity and found no 324 interaction between direct breast feeding and cesarean section (p-interaction = 0.93). Also the interaction term 325 between skin to skin contact and cesarean section is marginally significant (p-interaction = 0.17). With skin to 326 skin contact and the interaction between skin to skin and cesarean in the model, the odds ratio for neonate 327 testing positive with cesarean section increases to 3.4 (1.4, 8.2), but the confidence intervals are much wider. 328

As presented in Table 1 , fetal distress was lowest in neonates of women without COVID-19 diagnosis, higher 329

among COVID-19 test-negative neonates of women with COVID-19 diagnosis, and highest among COVID-19 330 test positive neonates whose mothers also had a COVID-19 diagnosis. 331 Figure 1 shows the ORs and 95% CIs for the COVID-19 test-positive neonates by the time elapsed between 339 maternal diagnosis and delivery, adjusted for cesarean section. The aORs increase with the time between 340 diagnosis and delivery, particularly after 7 days (aOR 2.0, 95% CI 1-3.7, p=0.04) and 14 days of exposure (aOR 341 4.5, 95% CI 2.2-9.4, p <0.001) (see Supplemental Table 3 ). 342

As shown in Table 2 , we did not observe any significant differences regarding the severity and number of 343 maternal symptoms across the three neonatal groups with mothers with COVID-19 diagnosis. Those test-344 positive neonates born to a woman with COVID-19 diagnosis had, on average, a gestational age at birth more 345 than one week less than those neonates born to a woman without COVID-19 diagnosis (Table 2) . Thus, birth 346 weight, length and head circumference were, on average, lower among test-positive neonates born to women 347 with a COVID-19 diagnosis than among those whose mothers without COVID-19 diagnosis. The rates of fetal 348 distress in labour, NICU admission and early neonatal complications and morbidities among test-positive 349 newborns of women with COVID-19 diagnosis were also higher than in those from mothers without COVID-19 350 diagnosis. NICU admission and early neonatal complications were also higher in test-negative newborns born to 351 women with COVID-19 diagnosis, compared to women without COVID-19 diagnosis (Table 2 ). 352 Table 3 shows outcomes up to hospital discharge of test-negative, test-positive and not-tested neonates of 353 women with COVID-19 diagnosis. A NICU stay longer than 7 days occurred significantly more frequently in 354 test-positive than negative neonates. The proportion of any breastfeeding did not differ significantly between 355 those who tested negative vs positive. However, a higher proportion who breastfed, both during hospital stay 356 and discharge, was observed in not-tested neonates, in whom the rate of respiratory problems and infections was 357 significantly lower than in test-negative neonates of women with COVID-19 diagnosis. In contrast, test-358 J o u r n a l P r e -p r o o f positive neonates showed a significantly higher rate of complications such as fever, infections, respiratory 359 problems or need for respiratory support compared to test-negative neonates. (Table 3)  360  Table 4 shows the increased relative risks for most neonatal outcomes comparing neonates born to a mother 361 with COVID-19 diagnosis, to those born to a mother without COVID-19 diagnosis. As expected, relative risks 362

were higher in the subgroup of neonates who tested positive after correction for maternal risk factors and 363 gestational age. In particular, we found a higher risk of respiratory (OR 3.4, CI 95% 2.2-5.3), neurological (OR 364 4.9, CI 95% 1.7-14.1) and gastrointestinal (OR 5.9, CI 95% 2.1-16.6) signs and the need for a NICU stay longer 365 than 7 days (OR 5.4, CI 95% 3.2-9.1) among test-positive neonates compared to those with a mother without 366

COVID-19 diagnosis. The results were similar, although the confidence intervals were wider, when we 367 restricted this analysis only to mothers who tested positive for COVID-19 (Supplemental Table 5 ). 368 Table 5 provides data regarding care practices for the neonates of mothers with COVID-19 diagnosis. 369

Immediate skin-to-skin contact was lower in test-positive than in test-negative neonates. Rooming-in with the 370 mother; mother and hospital staff practices of wearing masks and washing their hands before touching the 371 neonate, and the proportion of neonates who received breast milk did not differ. We specifically explored the 372 association between feeding human milk regimens and neonatal COVID-19 test positivity and the risk of 373 transmission of SARS-CoV-2 by breastfeeding, as compared to feeding expressed human milk. Any 374 breastfeeding compared to exclusive formula or no oral feeding was not associated with neonatal test positivity. 375 We did not find any differences in the risk of being test-positive between neonates who received direct 376 breastfeeding and those receiving donor milk or extracted mother's breast milk administrated by bottle. 377 378 COMMENT

This large-scale, prospective, multinational study assessed the association between COVID-19 diagnosis in 381 J o u r n a l P r e -p r o o f pregnancy and maternal and neonatal outcomes. We have previously provided evidence of the risk associated 382 with a COVID-19 diagnosis during pregnancy 12 . Here, we concentrate on the role of neonatal and perinatal 383 practices on outcomes, with a particular focus on topics of interest for clinical practice such as the indication for 384 mother-newborn separation after birth in case of the mother testing positive, the effectiveness of preventive 385 measures and the safety of breastfeeding. We also present data regarding the association between in utero 386 exposure, type of delivery and the neonatal risk of testing positive for COVID-19, as well as the association 387 between maternal COVID-19 diagnosis and neonatal morbidity. A COVID-19 diagnosis in pregnancy and the 388 postnatal period carries a substantial risk of neonatal morbidity and mortality. Cesarean section was 389 significantly associated with neonatal COVID-19 test positivity. Mother to child skin-to-skin contact, rooming-390 in and direct breastfeeding do not represent risk factors for neonatal test positivity. 391

Overall, a maternal diagnosis of COVID-19 greatly influenced perinatal and neonatal outcomes, with increased 393 rates of preterm birth and lower weight, length, and head circumference at birth. Respiratory signs and NICU 394 admission were also more common among neonate born to women with COVID-19 diagnosis. Hence, we have 395 demonstrated a direct impact on the newborn, secondary to maternal infection, independent of neonatal test 396 positivity or negativity. Moreover, as expected, COVID-19 test-positive neonates of women with COVID-19 397 diagnosis, compared to neonates that tested negative, had increased rates of prolonged NICU stay, fever, 398

gastrointestinal and respiratory problems, and death, even after adjusting for prematurity, which suggests a 399 direct effect of SARS-CoV-2 infection on neonatal morbidity. 400

In women with COVID-19 diagnosis, there was a significant correlation between the length of in-utero 401 exposure and risk of the neonate testing positive. In women with COVID-19 diagnosis, the gestational age at 402 maternal diagnosis was significantly lower for neonates who then tested positive at birth, compared with those 403 who tested negative (35.3 weeks versus 37 weeks). However, the time between maternal diagnosis and delivery 404 was significantly longer in test-positive compared to test-negative neonates (13.3 days versus 6.4 days) resulting 405 J o u r n a l P r e -p r o o f in a similar mean gestational age at birth. 406

The pathogenic mechanisms that could explain the correlation between the total time of exposure and risk of 407 neonatal positivity are yet to be elucidated 26 . In general, it is considered that vertical transmission with SARS-408

CoV-2 does not occur prenatally. However, the fact that SARS-CoV-2's cellular receptor, the angiotensin 409 converting enzyme-2 receptor (ACE-2), has been detected in the placenta, albeit at a low level, raises the 410 possibility of transplacental transmission 27 in some rare cases. Once SARS-CoV-2 binds to the ACE-2 411 receptor, the Transmembrane protease serine 2 enzyme (TMPRSS2) is activated and allows the virus to pass 412 into the cell; TMPRSS2 is expressed after 24 weeks' gestation 28 . Conflicting data exist related to the extent of 413 co-expression 29,30 . Viraemia is also associated with vascular damage, including hypercoagulability and poor 414 vascular perfusion 31 ; the resulting placental damage could facilitate such vertical transmission 26 . 415

The cesarean section rate was significantly higher in women with, compared women without COVID-19 416 diagnosis, possibly because obstetricians adopted a more interventional approach to the affected women. 417

However, when we focused on women with COVID-19 alone, the cesarean section rate was still significantly 418 higher in the test-positive (71.4%) compared to negative (48.9%) neonates. Analysis of both cesarean section 419 COVID-19-related indications and the severity of maternal conditions did not show any differences between the 420 test-positive and negative neonates, which reinforces the independence of cesarean section in determining 421 neonatal positivity, as confirmed also by multivariable logistic analysis. There is no clear explanation for this 422 observation, although one interesting hypothesis is that neonates born by cesarean section have less immediate 423 contact with the mother with consequent less intake of colostrum, very rich in immunological protective factors 424 32 , which would increase the risk of SARS-CoV-2 infection. At present these exploratory data do not support a 425 recommendation for cesarean section in mothers with COVID-19 diagnosis. 426

Another important finding was that breastfeeding in mothers with COVID-19 diagnosis was not associated with 427 an increased risk for neonatal test positivity. Therefore, given the additional well-known benefits of the 428 J o u r n a l P r e -p r o o f mother's own milk on neonatal health, we strongly recommend that all measures to promote, protect and sustain 429 breastfeeding should be maintained in mothers with COVID-19 diagnosis, as indicated by WHO and CDC 430 guidelines 33, 34 . Interestingly, in this large and multicultural study, rates of breastfeeding during hospital stay 431

and at discharge were similar in test-positive and negative neonates. Considering the initial uncertainty in the 432 setting of a global pandemic, this is a positive message about the commitment to breastfeeding in our 433

populations and allowed us to have a good number of breastfed newborns in this study. 434

Finally, the data we collected on neonatal care practices showed that immediate skin-to-skin contact and 435

rooming-in did not increase the risk of neonatal test positivity in settings where mothers wore masks and 436

washed their hands before touching their neonates and the hospital staff used gloves and masks. This is an 437 important result because some hospitals have adopted policies that discourage immediate skin-to-skin contact or 438 kept the neonate isolated from mothers with COVID-19 diagnosis, especially early in the pandemic 35,36 . Our 439 data show that these are unnecessary practices and can deprive the mother and her neonate of the well-440 recognised beneficial effects of early contact such as closer bonding, early initiation and continuation of 441 breastfeeding, and reduced infections 37 . 442

Strengths and limitations 443

Our study has expected limitations. Regarding selection of the population, by selecting a reference group of 2 444 women recruited immediately after each woman with COVID-19 diagnosis, at the same level of care, we were 445 able to obtain results rapidly and reduce systematic bias despite the lack of widely available COVID-19 tests 446 until late 2020. However, we recognise that a few asymptomatic affected women may have been included in the 447 control group, but this conservative bias would eventually underestimate the effect of the COVID-19 infection; 448

in our opinion, this gives even more strength to the differences identified between the groups. 449

We acknowledge a risk of ascertainment bias in reporting maternal and neonatal morbidity as the newborns of 450 women with COVID-19 diagnosis compared to newborns of women without wrote the manuscript with input from all co-authors. JV and ATP had full access to all the 477 data in the study and take responsibility for the integrity of the data and the accuracy of the 478 data analysis. All co-authors read the manuscript and made suggestions on its content. We 479 thank all the contributing institutions and local researchers involved in the study. Appendix 480 1 contains their details as well as details of the study committees. 481 482

We are very grateful to the contributing institutions and local researchers involved in the study. The Appendix 483 contains their details as well as details of the study committees. 484 J o u r n a l P r e -p r o o f 

COVID-19 Vaccine in Pregnant and Lactating Women: A Review 486 of Existing Evidence and Practice Guidelines

COVID-19 vaccination during pregnancy: coverage and safety

Epidemiology of coronavirus disease 2019 in pregnancy: risk factors 491 and associations with adverse maternal and neonatal outcomes

Clinical characteristics of 494 confirmed COVID-19 in newborns: a systematic review

An update on COVID-19 and pregnancy

Outcome of coronavirus spectrum infections

COVID-19) during pregnancy: a systematic review and meta-analysis

Pregnancy and perinatal outcomes of women with severe acute 502 respiratory syndrome

Severe acute respiratory syndrome (SARS) in neonates and children

SARS-CoV, MERS-CoV and SARS-CoV-2 506 infections in pregnancy and fetal development

Outcome of coronavirus spectrum infections (SARS, MERS, 509 COVID-19) during pregnancy: a systematic review and meta-analysis

Coronavirus disease 2019 (COVID-19) pandemic and 512 pregnancy

Maternal and Neonatal Morbidity and Mortality Among Pregnant 514 Women With and Without COVID-19 Infection: The INTERCOVID Multinational Cohort Study. JAMA 515 Pediatr

COVID-19 infection in pregnant women: Review of maternal and fetal 517 outcomes

Pregnancy and COVID-19

Preeclampsia and COVID-19: results from the 521 INTERCOVID prospective longitudinal study

Diabetes mellitus, maternal adiposity, and insulin dependent 524 gestational diabetes are associated with COVID-19 in pregnancy: the INTERCOVID study

NIH/NICHD) SARS-CoV-2 placental infection workshop

Maternal SARS-CoV-2 infection during pregnancy: possible 535 impact on the infant

International standards for fetal growth based on 538 serial ultrasound measurements: the Fetal Growth Longitudinal Study of the INTERGROWTH-21st Project

Predicting Infectious Severe Acute Respiratory Syndrome Coronavirus 541 2 From Diagnostic Samples

The duration of infectiousness of individuals infected with 543 SARS-CoV-2

International standards for newborn weight, length, and 545 head circumference by gestational age and sex: the Newborn Cross-Sectional Study of the INTERGROWTH-546 21st Project

Use of directed acyclic graphs (DAGs) to identify 548 confounders in applied health research: review and recommendations

SARS-CoV-2 Targets by the pscRNA Profiling of ACE2

COVID-19: clinical presentation and implications. A primer for obstetricians

SARS-CoV-2 ACE-receptor detection in 557 the placenta throughout pregnancy

Does the human placenta express the canonical cell entry 560 mediators for SARS-CoV-2? Elife

A structured review of placental morphology and histopathological 562 lesions associated with SARS-CoV-2 infection

Human milk composition: nutrients and bioactive factors

Breastfeeding and Breast Milk Feeds in the Context of COVID-19

Breastfeeding and COVID-19, Scientific Brief

Safe delivery for pregnancies affected by COVID-19

Review of guidelines and recommendations from 17 countries 573 highlights the challenges that clinicians face caring for neonates born to mothers with COVID-19. Acta 574 Paediatr

Abbreviations: ICU = intensive care unit

SD = standard deviation

SDS = standardized score

01 comparing each category from mother with COVID-19 diagnosis (neonate without signs not tested and neonate positive) separately to negative neonates born to mother with COVID-19 diagnosis †

001 comparing each category from mother with COVID-19 diagnosis (neonate negative, neonate without signs not tested and neonate positive) separately to neonates born to mother without COVID-19 diagnosis

Abbreviations: NICU = neonatal intensive care unit

SD = standard deviation

001 compared to mother with COVID-19 diagnosis, child COVID-19 negative

a Neurological problems includes seizures, hydrocephalus, neurological disorders, hypoxic-ischaemic encephalopathy, periventricular haemorrhage/leukomalacia