Pregnancy and neonatal outcomes of COVID-19: The PAN-COVID study

Objective To assess perinatal outcomes for pregnancies affected by suspected or confirmed SARS-CoV-2 infection. Methods Prospective, web-based registry. Pregnant women were invited to participate if they had suspected or confirmed SARS-CoV-2 infection between 1st January 2020 and 31st March 2021 to assess the impact of infection on maternal and perinatal outcomes including miscarriage, stillbirth, fetal growth restriction, pre-term birth and transmission to the infant. Results Between April 2020 and March 2021, the study recruited 8239 participants who had suspected or confirmed SARs-CoV-2 infection episodes in pregnancy between January 2020 and March 2021. Maternal death affected 14/8197 (0.2%) participants, 176/8187 (2.2%) of participants required ventilatory support. Pre-eclampsia affected 389/8189 (4.8%) participants, eclampsia was reported in 40/ 8024 (0.5%) of all participants. Stillbirth affected 35/8187 (0.4 %) participants. In participants delivering within 2 weeks of delivery 21/2686 (0.8 %) were affected by stillbirth compared with 8/4596 (0.2 %) delivering ≥ 2 weeks after infection (95 % CI 0.3–1.0). SGA affected 744/7696 (9.3 %) of livebirths, FGR affected 360/8175 (4.4 %) of all pregnancies. Pre-term birth occurred in 922/8066 (11.5%), the majority of these were indicated pre-term births, 220/7987 (2.8%) participants experienced spontaneous pre-term births. Early neonatal deaths affected 11/8050 livebirths. Of all neonates, 80/7993 (1.0%) tested positive for SARS-CoV-2. Conclusions Infection was associated with indicated pre-term birth, most commonly for fetal compromise. The overall proportions of women affected by SGA and FGR were not higher than expected, however there was the proportion affected by stillbirth in participants delivering within 2 weeks of infection was significantly higher than those delivering ≥ 2 weeks after infection. We suggest that clinicians’ threshold for delivery should be low if there are concerns with fetal movements or fetal heart rate monitoring in the time around infection. The proportion affected by pre-eclampsia amongst participants was not higher than would be expected, although we report a higher than expected proportion affected by eclampsia. There appears to be no effect on birthweight or congenital malformations in women affected by SARS-CoV-2 infection in pregnancy and neonatal infection is uncommon. This study reflects a population with a range of infection severity for SARS-COV-2 in pregnancy, generalisable to whole obstetric populations.


Introduction
Much information on the effect of SARS-CoV-2 infection in pregnancy has been derived from reported outcomes in women with symptomatic infection and/or those hospitalised. However, in view of the lack of testing in the initial months of the pandemic and the high proportion of asymptomatic and mild SARS-CoV-2 infections, the impact of less severe infections in pregnancy has not been captured. It is likely that only a small proportion of women with SARS-CoV-2 infection in pregnancy would be admitted to hospital at the time of a positive PCR test.
In view of this, we designed the PAN-COVID study as a global registry for rapid set-up to recruit women with suspected or confirmed SARS-CoV-2 at any point during their pregnancy. This would include those hospitalised and not hospitalised to gain information on the impact of infection on key perinatal outcomes.
We developed research questions from queries raised by pregnant women in our maternity services, including the impact of infection on miscarriage, stillbirth, fetal growth restriction, pre-term birth and transmission of SARS-CoV-2 to the baby as well as maternal health. We aimed to capture a focussed dataset, feasible for staff working within the constraints of the pandemic, in a range of healthcare settings.
Multiple global registries, e.g. PRIORITY, COVI-PREG, ICMR were established at similar times. As the pandemic progressed, nonpharmaceutical interventions and improvements in treatment for SARS-CoV-2 changed the way pregnant women and their babies were affected by the virus. Variants of the virus were shown to cause more severe disease in women in pregnancy than the wildtype virus [1]. The impact of ethnicity and pre-existing medical conditions on pregnant women with SARS-CoV-2 and the increased risk of pre-term birth became apparent [2][3][4]. We co-reported our interim findings with the AAP SOPNM registry [3] in February 2021 and called for the prioritisation of vaccination for women in or planning pregnancy.

Study design
A detailed description of the PAN-COVID methods has been published [5]; in brief, this was a prospective web-based registry, allowing individual clinical centres to enter data via an online central database. Pregnant women were invited to participate if they had suspected or confirmed SARS-CoV-2 infection between 1st January 2020 and 31st March 2021.
We collected data in UK and centres in other countries (described as non-UK for the purpose of this manuscript) and analysed overall and UK/non-UK data.

Study participants
Eligible participants were aged between 18 and 50 years of age and had suspected or confirmed SARS-CoV-2 infection in pregnancy between 1/1/20 and 31/3/21.

Study conduct
Ethical approval was granted by the North West -Haydock REC, reference 20/NW/0212. Clinical Research Network (CRN) North West London was our coordinating CRN and supported our study management team to undertake, with the assistance of established key networks such as the NIHR Reproductive Health and Childbirth Champions, rapid setup of study centres by our data team at the Centre for Trials Research (CTR), Cardiff University.
Eligible women were approached by study investigators for their consent to participate in the study. Data was collected from participants' and their medical records by study investigators.

Statistical analysis
Pre-specified sample size estimation was not carried out in the PAN-COVID study given that the aim of this observational study was to collate outcomes for all consecutive eligible cases in participating centres during an 18-month period from the start of data collection. Once PAN-COVID closed, all logic check, incomplete entries, and missing data queries were collated and sent out to sites. Any data queries that remained unresolved after several attempted site contacts were denoted missing data.
Gestational age at birth was calculated from the expected due date (EDD) and the date of delivery recorded; the date of the last menstrual period was used when EDD was unavailable. Birth-weight z-scores were calculated according to Fenton et al. [6], were gestational-age and gender adjusted and limited to be within +/− 4.
Sample sizes are given for all outcomes in all tables and vary according to numbers of responses obtained. No hypothesis testing has been included since the study was not designed to determine differences, we have however provided 95 % Confidence Intervals (CI) for selected differences that may be of clinical relevance. It must be noted however that sample sizes for many of the rarer outcomes in the non-UK data are very small and all proportions and CIs must be interpreted in this light.
Participants with confirmed and suspected infections were analysed together as in our interim data paper [3], their pregnancy outcomes were similar and because in the early stages of the pandemic, SARs-CoV-2 testing was only available to those admitted to hospital in the UK. Data cleaning and descriptive analyses were conducted using SPSS V27.0 and Stata V17.0.

Results
Between April 2020 and March 2021, the study recruited 8239 women who had suspected or confirmed SARs-CoV-2 infection episodes between January 2020 and March 2021 (Fig. 1).
Data were collected from a range of healthcare settings, 7395/8239 (89.8 %) participants were recruited by UK centres, 844/8239 (10.2 %) of women from centres in Italy, China, Greece, Indonesia, India, Argentina, China, Czech Republic, Albania, Austria, Egypt and Chile (non-UK centres, Table 1). Multiple, competing registries were established during the pandemic and our original aim of providing a global registry was not fully achieved.
Participants from the UK had a higher mean age, had a higher proportion with European/North American ethnicity and had a higher proportion of women who were current or ex-smokers compared with  * BMI measurement taken either at booking (N = 7,899) or pre-pregnancy (N = 162) (with N = 178 missing time of BMI measurement).
Pre-term delivery, stillbirth and early neonatal death were higher in non-UK participants (Table 4). Delivery by caesarean section was more common in non-UK participants. Spontaneous pre-term birth affected 200/7216 (2.7 %) of UK participants and 20/771 (2.6 %) of non-UK participants. Amongst participants with indicated pre-term delivery, delivery for maternal hypoxia was high in non-UK participants 25/95 (26.3 %) compared with 51/500 (10.2 %) in UK participants.
The proportion of participants affected by stillbirth was higher in those delivering within two weeks of first date of suspected or confirmed infection with SARS-CoV-2 compared with those delivering more than two weeks later, 0.8 % vs 0.2 %, (95 % CI 0.3-1.0) ( Table 6). Participants who delivered more than two weeks after initial infection had a higher proportion affected by fetal growth restriction than those delivering within two weeks of infection, 3

Discussion
Overall, the proportion of participating women affected by stillbirth was not higher than the rate reported in UK ONS data, 3.8/1000 maternities in 2019 and 3.9/1000 maternities in 2020 [7]. However, the 2 week period around the infection with SARS-CoV-2 was associated with a significantly higher proportion of pregnancies affected by stillbirth, approximately double the rate reported in UK ONS data, 3.8/1000 maternities in 2019 and 3.9/1000 maternities in 2020 [7]. Population surveillance data from UK women admitted to hospital with SARS-CoV-2 infection reports stillbirth affecting 2.1 % of symptomatic and 2.4 % of asymptomatic women with SARS-CoV-2 [1].
There is no reported evidence of abnormal growth parameters or Doppler studies in women affected by SARS-CoV 2 infection in pregnancy [8,9]. The assumption that risk of stillbirth in women with SARS-CoV-2 infection could be mitigated by increased fetal ultrasound surveillance remains untested. We would suggest that antenatal care of all women in pregnancy with SARS-CoV-2 should include a low threshold for delivery in the period within 2 weeks of infection if there are any concerns regarding reduced fetal movements or CTG abnormalities. Strategies to mitigate the risk of stillbirth require further evaluation.
Ongoing pregnancy with delivery >2 weeks after SARS-CoV-2 infection was associated with a higher proportion affected by FGR compared to those with delivery within 2 weeks of infection. This may simply be a function of incident FGR diagnosis in ongoing pregnancy. SARS-CoV-2 infection has been shown to cause a villitis with inflammation at the maternal/fetal interface [10] which may impair placental function and lead to FGR.
Consistent with our previous report [3] and others reports [9], and contrary to other reports based predominantly on pregnant women in hospitals [11] we did not find that fetal growth or SGA is affected by SARS CoV-2 infection. In our study, there was a higher than expected proportion of SGA in participants from non-UK centres, which may reflect a focus on recruitment of women hospitalised for pregnancy complications.
Pre-term birth affected 10.9 % of PANCOVID participants, 50 % higher than the background rate of 7.3 % in the UK [12]. Of those participants with pre-term birth, the majority had indicated pre-term delivery, for reasons including fetal distress and maternal hypoxia. SARS-CoV-2 infection was not associated with spontaneous pre-term birth in our participants, the proportion affected byt this was below the background rate for the UK [12]. The high proportion of participants having indicated pre-term delivery may have affected the rate of spontaneous pre-term birth.
Multiple other studies have shown an association of SARS-COV-2 with indicated pre-term birth. INTERCOVID reported pre-term birth in 22.5 % of participants; however it should be noted that in their control group (pregnant women without SARS-CoV-2 infection) 13.6 % of women experienced pre-term birth, suggesting that the study was carried out in centres with high background rates of pre-term birth and may not be generalizable [11]. The UKOSS/ISARIC data from 01/03/2020 to 28/02/2021 reported that 681/5479 (12.4 %) pregnant women admitted to hospital with confirmed SARS-CoV-2 experienced pre-term deliveries (22-36 +6 weeks gestation) [1], perhaps reflecting increased disease severity in that exclusively hospitalised cohort.
Congenital malformations in our participants affected a proportion comparable with the background rate of 213.3 per 10,000 total births in Table 5 Birthweight and fetal growth restriction in participants.  Table 6 Stillbirth and fetal growth restriction in participants delivering within 2 weeks or 2 weeks or more after SARS CoV 2 infection. the UK [13]. Transmission from mother to fetus and neonate was not reliably discernible using our data, as there was a lack of systematic testing of neonates born to mothers affected by SARs-CoV-2 infection in pregnancy, particularly in UK centres. The 1 % of neonates testing positive for SARS-CoV-2 in PANCOVID is comparable with UKOSS data [1] but lower than in the INTERCOVID study [11]. Testing amongst our participants in the UK was not routine and targeted to infants showing respiratory symptoms and stratified by risk status of the mother [14]. The higher proportion of neonates with positive tests in non-UK vs UK centres (3 % vs 1 %) suggests that universal testing may detect cases more reliably than targeted testing. Comparable rates of pneumonia in UK and non-UK participants suggests that this may not impact on immediate management or outcomes for neonates, however the long-term effects of vertical transmission and/or early neonatal infection with SARS-COV-2 are currently unknown.
The risk of vertical or early neonatal transmission appears low, early neonatal death was not increased above background rates for the UK of 0.2 % [15]. There is a need for international consensus on the optimal samples to determine infant infection and testing strategies for infants of women with SARS-CoV-2. Women with SARS-CoV-2 in pregnancy can be generally reassured that infants are unlikely to contract SARS-CoV-2 whilst in-utero, there is no signal apparent for increases in congenital malformations above the background risk of 1 % and infants' respiratory morbidity is similar to that of infants born to women without SARS-CoV-2 infection in pregnancy.
The numbers of women who died were small in our study, particularly from UK centres and this study does not have power to assess whether this proportion was higher than the background rate for maternal mortality in pre-pandemic UK population surveillance data of 8.8 per 100,000 maternities [16]. UKOSS data up to February 2021 in women admitted to hospital with SARS-CoV-2 infection reported case fatality rate (CFR) of 0.6 % (95 % CI 0.3-0.6 %) [1]. INTERCOVID reported a 1.6 % CFR and increased risk of maternal death with SARS-CoV-2 infection compared to those without, RR 22.6 (95 % CI 2.88-172.11) [17].
The proportion of participants affected by pre-eclampsia (4.6 %) was no higher than reported rates of 2-8 % in the latest statement from the International Society for the Study of Hypertension in Pregnancy (ISSHP) [18], is comparable with the non-infected, control group in the INTERCOVID study [11,17] and was higher than the 1.9 % reported by UKOSS [1]. Our data do not support the association between SARS-COV-2 infection and an increase in incidence of pre-eclampsia. However, eclampsia was more common than expected for the UK, 2.7 per 10,000 births [19] suggesting that infection could be associated with more severe manifestations of this vascular syndrome.

Strengths and limitations
This study comprises one of the largest prospective, individual patient datasets of perinatal outcomes among women with suspected or confirmed SARS-CoV-2 infection to date.
Early pregnancy units laegely closed during the first wave of the pandemic limiting data collection, we are not able to assess the miscarriage risk associated with SARS-CoV-2 infection.
The proportion of women affected by maternal deaths in our cohort was low, likely due to our study design requiring consent from the participant or their relative. It is likely that our study underestimates the risk of infection and maternal death.

Conclusions
This study reflects a population with a range of infection severity for SARS-COV-2 in pregnancy, making it generalisable to whole obstetric populations. Infection is associated with indicated pre-term birth, primarily for fetal compromise. Whilst the overall proportions of women affected by SGA and FGR were not higher than expected, there was a significant difference in the proportion affected by Stillbirth and FGR in the participants delivering < 2 weeks/≥2 weeks respectively. We suggest that clinicians' threshold for delivery should be low if there are concerns with fetal movements or fetal heart rate monitoring in this period.
There appears to be no effect on birthweight or increase in congenital malformations in women affected by SARS-CoV-2 infection in pregnancy and neonatal infection is uncommon. The effect of infection on miscarriage was not determined. The rate of pre-eclampsia amongst participants was not higher than would be expected.
We believe a co-ordinated, global study of pandemic viruses' impact on women in pregnancy should be planned now to allow rapid, global response to future pandemics and the avoidance of multiple, parallel studies with differing inclusion criteria and outcome sets. A core outcome set should be developed for this purpose. Study design for future global registries to rapidly assess the impact of pandemic viruses in pregnancy should include methodology for dealing with data from different healthcare settings.

Contribution
What are the novel findings of this work?
Infection is associated with indicated pre-term birth, primarily for fetal compromise. Whilst the overall proportions of women affected by SGA and FGR were not higher than expected, there was a significant difference in the proportion affected by Stillbirth and FGR in the participants delivering < 2 weeks/≥2 weeks after infection respectively.
There appears to be no effect on birthweight or increase in congenital malformations in women affected by SARS-CoV-2 infection in pregnancy and neonatal infection is uncommon. The effect of infection on miscarriage was not determined. The rate of pre-eclampsia amongst participants was not higher than would be expected.

What are the clinical implications of this work?
This study reflects a population with a range of infection severity for SARS-COV-2 in pregnancy, making it generalisable to whole obstetric populations. We suggest that clinicians' threshold for delivery should be low if there are concerns with fetal movements or fetal heart rate monitoring in the two weeks after infection.

Data sharing
PAN-COVID: De-identified participant data will be made available to the scientific community with as few restrictions as feasible, whilst retaining exclusive use until the publication of major outputs. Data will be available via the corresponding author.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.