Clinical characteristics & outcome of SARS-CoV-2 infected neonates presenting to paediatric emergency

Worldwide, more than 278 million patients have COVID-19, with 5.4 million deaths1. As compared to adults, children are less affected and have better prognosis23. Neonates are at risk for acquiring vertical and horizontal SARS-CoV-2 infection. The evidence so far suggests that vertical transmission is uncommon, and a more significant proportion of neonates acquire SARS-CoV-2 infection in the postnatal period through horizontal transmission from either infected mothers or immediate caregivers45. Most neonates remain asymptomatic, and the overall prognosis is good467891011. Most of the studies on neonatal COVID-19 are limited to inborn neonates and focused on vertical transmission. This study was conducted to assess the positivity rate, characteristics, and outcome of COVID-19 among outborn neonates presenting to the paediatric emergency in a tertiary care centre in north India.

Material & Methods

This retrospective study was done at an outborn neonatal unit of the department of Paediatrics, Postgraduate Institute of Medical Education and Research, Chandigarh, India. All outborn neonates up to 44 wk postmenstrual age are admitted to the newborn unit of paediatric emergency. All neonates consecutively referred to the institute from August 1 to December 31, 2020 were enrolled. The details of all the neonates at admission (day of life, history of fever, cough, respiratory distress, day of life at the onset of respiratory distress, admission diagnosis, respiratory support, previous hospitalization, SARS-CoV-2 real-time RT-PCR test, and outcome) were noted. Ethical clearance, including waiver of consent, was obtained from the institute’s ethics committee.

The ICMR national guidelines (https://www.icmr.gov.in) for assessing eligibility for testing, sample processing, and discharge12 were adopted. Oro/nasopharyngeal specimens were obtained using standard technique, and SARS-CoV-2 RT-PCR assay (Cepheid Xpert Xpress SARS-CoV-2 assay, Cepheid India Pvt. Ltd; Rewari, India) was used for COVID-19 testing. COVID-19-positive neonates were managed in a dedicated COVID-19 facility.

During the first three weeks of August 2020, a selective testing policy was adopted in which SARS-CoV-2 RT-PCR testing was limited to those who fulfilled any of the following criteria: (i) mother had confirmed COVID-19 within 14 days before delivery, (ii) the infant had a history of close contact with confirmed COVID-19 patient in the past 14 days, (iii) neonates presenting with severe acute respiratory illness (needing supplemental oxygen or respiratory support to maintain oxygen saturation >90%) with onset at more than 24 h of age with or without cough and fever (>38°C)12. Later (from August 20, 2020, onwards), with a rising number of community cases, universal screening policy was adopted in which all symptomatic patients (irrespective of illness) reporting to the emergency underwent COVID-19 testing (as recommended by ICMR)13. Neonates who tested negative were cared for in the outborn unit, whereas confirmed COVID-19 cases were transferred to the COVID facility of the hospital. The clinical management of these neonates admitted at both places was the same (irrespective of admission site) and was done as per the standard protocols. The samples of accompanying immediate caregivers of COVID-19 confirmed neonates were also sent for contact tracing, if feasible. Rooming-in, breastfeeding, and Kangaroo Mother Care (KMC) were allowed whenever deemed possible. Repeat COVID-19 testing was limited to the neonates needing further care at the institute.

The change from selective testing to universal COVID-19 testing of all neonates presenting to the hospital was not based on neonatal data but was based on extrapolating from the adults. Therefore, the positivity by selective testing in infants admitted after August 20, 2020 was compared with the universal testing to assess how many additional neonates were identified by universal testing. Also, the overall diagnostic accuracy of the selective testing policy was evaluated.

Statistical analysis: The categorical variables were presented as percentages and continuous variables as mean (standard deviation) or median (1^st-3^rd quartile). The categorical variables were compared using a chi-square test or Fisher’s exact test as appropriate and numerical variables using Mann–Whitney U test. Odds ratios (ORs) were calculated with 95 per cent confidence intervals (CIs) for various outcomes among COVID-19-positive neonates compared to COVID-19-negative neonates. SPSS version 21 (IBM SPSS Statistics, IBM Corp. Chicago, IL, USA) software was used for statistical analysis.

Results & Discussion

A total of 1225 neonates were admitted during the study period. The median (1^st-3^rd quartile) age at admission was two (1, 7) days. Of the 1225 neonates, 1127 (92%) were referred from the surrounding States, and 875 (71.4%) had respiratory distress at admission. Most of the neonates (486 of 1225) presented on day one of life. Of the 1225 neonates, 192 (15.7%) either died or left against medical advice. SARS-CoV-2 RT-PCR test was done in 969 of 1225 (79.1%) neonates. The remaining 256 neonates could not be tested for various reasons like not meeting the selective screening criteria in the initial study period, died within four to six hours of admission, or parents not willing for testing/care at our hospital and left against medical advice. Of the 969 neonates tested for SARS-CoV-2 infection, only 17 (1.8%, 95% CI: 1.03-2.79) were positive. Five (29.4%) COVID-19-confirmed neonates died, and none of them left against medical advice.

For this study, only those neonates were analysed whose SARS-CoV-2 testing was done (n=969). The clinical characteristics of COVID-19-positive neonates are presented in Table I. Most of them (11/17) presented in the first two weeks of life, and almost all (14/17) presented with respiratory distress. The median (1^st-3^rd quartile) hospital stay was five (2-7) days. Five (29.4%) COVID-19-positive neonates died during the hospital stay. Two of them had lethal congenital anomalies (Table I), while the rest three died due to severe sepsis (Klebsiella pneumonia - 1, Candida spp. - 1, culture-negative sepsis - 1). Only one neonate had clinical and radiological manifestations (bilateral ground-glass opacities in peripheral lung fields) consistent with COVID-19 pneumonia and received methylprednisolone for five days. In the remaining 16 neonates, the clinical features were either not attributed to the COVID-19, or the disease was not severe enough to warrant specific therapy.

The various risk factors/parameters among the COVID-19 positive and negative neonates were compared (Table II). COVID-19 confirmed neonates had respiratory distress onset at a later age (2 vs. 1 day) and were admitted at an older postnatal age (11 vs. 3 days). Neonates having fever at admission had significantly higher odds of having COVID-19 (OR: 12.5; 95% CI: 3.3-47.9, P=0.004). However, there was no difference in cough, respiratory distress, prior hospitalization, or need for mechanical ventilation. Although neonates with confirmed COVID-19 were at higher risk for all-cause mortality (OR: 3.1; 95% CI: 1.1-8.9, P=0.03), the difference was not significant after adjusting for lethal congenital anomalies (OR: 2.4; 95% CI: 0.7-8.7, P=0.17).

If we had followed the selective testing policy, only 9/17 COVID-19 cases (53%) would have been diagnosed. Therefore, the universal testing policy identified additional eight (47%) cases. The diagnostic accuracy of the selective testing policy was assessed considering SARS-CoV-2 RT-PCR as a gold standard. The sensitivity, specificity, accuracy, positive and negative likelihood ratios were 52.9 (95% CI: 28.5-76.1), 83.3 (80.7-85.6), 82.8 per cent (80.3-85.1), 3.17 (1.98-5.07), and 0.56 (0.34-0.93), respectively. We also attempted to assess the effect of change in the cut-off for the onset of respiratory distress in the selective testing policy. As the cut-offs were increased for the onset of respiratory distress, the sensitivity dropped significantly with marginal improvement in specificity. The sensitivity was 41.2 per cent (18.4-67.1) with 48 h. cut-off and 35.3 per cent (14.2-61.7) with 72 h cut-off, while the specificity was 87.4 per cent (85.1-89.4) at 48 h and 89.4 per cent (87.5-91.3) at 72 h.

As a part of contact tracing, at least one of the parents in 13 neonates (13 mothers and four fathers) could be tested. In five cases mother was COVID-19-positive, whereas in one neonate mother was negative, but the father tested positive. In two cases, both parents were negative, and they were the only caregivers.

In this study, the SARS-CoV-2 positivity rate among the outborn neonates presenting to the emergency and tested was lower (n=17, 1.8%) than older children and adults1415. Though the neonates with COVID-19 were at a higher risk for all-cause mortality, there was no significant difference in the adjusted mortality rate. Though neonates with a positive screen at admission had higher odds of COVID-19, the sensitivity was poor (43% of confirmed COVID cases had a negative screen). These observations support a universal testing screening policy.

Children typically account for 10-12 per cent of laboratory-confirmed SARS-CoV-2 cases, and the hospitalization rates range from 2.5 to 4.1 per cent1617. Neonatal COVID-19 accounts for less than one per cent of all and 5-7 per cent of paediatric COVID-19 cases181920. Previous cohort studies showed that the prevalence of COVID-19 in neonates born to COVID-19 mothers was low, and they were generally asymptomatic72122. Therefore, even if the mother is COVID-19 positive, exclusive breastfeeding and KMC with adequate respiratory and hand hygiene are recommended2324.

A retrospective study from India, where all outborn neonates presenting to the emergency were screened for SARS-CoV-2 infection, reported a higher positivity rate of 4.25 per cent (18 of 423)11. In this study, respiratory distress was the most common clinical presentation, and mortality was 16.6 per cent11. In another case series from India, the authors screened 18 symptomatic outborn neonates presenting to the emergency, of whom only one had SARS-CoV-2 infection9. In a similar study from Bangladesh, 26 of 83 (31.3%) outborn neonates referred to a tertiary center tested positive for SARS-CoV-2 infection10. In their series, most presented in the second week of life, with an unrelated diagnosis. In our study, there was no difference in adjusted mortality rate between COVID and non-COVID cases, inferring that SARS-CoV-2 positivity alone does not increase the risk for mortality25.The practice of universal testing of all outborn neonates requiring admission was not based on neonatal data but rather an extrapolation from older age groups. This study provides some evidence for this practice though the numbers are not large enough to make conclusive recommendations. However, considering the consequences of missing SARS-CoV-2 infection by selective screening and putting others at risk, continuing universal screening in outborn neonates requiring hospitalization may be suggested. Furthermore, this study indicates the possibility of horizontal transmission from immediate caregivers as well as healthcare facilities. In six COVID-19-confirmed cases, the primary caregivers were SARS-CoV-2-negative, and the infants stayed in a healthcare facility since birth. Therefore, there is a need for strict compliance with the precautionary measures in healthcare facilities to prevent horizontal transmission626. Many caregivers were asymptomatic and found to have COVID-19 when tested as a part of contact tracing. This observation reinforces the importance of contact tracing.

Being a retrospective single-centre study, it had many limitations. The primary illness, hospital course, and cause of mortality in COVID-19-negative neonates were not recorded, precluding further analysis. Contact tracing could not be done for all COVID-19 confirmed cases. Though most confirmed COVID-19 neonates had respiratory distress at admission, chest tomography was done in two cases only. The other neonates with respiratory distress might have radiological findings of COVID-19 that went undetected on a plain chest radiograph2728. This study enrolled outborn neonates referred to the emergency because of a clinical illness; therefore, results cannot be generalized to inborn neonates. As it was a retrospective study, formal sample size calculation was not done. Considering the low incidence of COVID-19 in neonates, large prospective multicentric cohort studies are desirable.

In conclusion, the SARS-CoV-2 positivity rate among outborn neonates presenting to the emergency was found to be low. As the clinical findings at admission cannot distinguish COVID-19 from non-COVID illness, universal screening of all outborn neonates requiring hospitalization seems reasonable.