Pediatric Hospitalizations and ICU Admissions Due to COVID-19 and Pediatric Inflammatory Multisystem Syndrome Temporally Associated With SARS-CoV-2 in England

Key Points

Question  What are the risks associated with serious SARS-CoV-2 infection among children and adolescents, and have associated risks changed as new variants have emerged?

Findings  In this cohort study of hospitalizations with COVID-19 and pediatric inflammatory multisystem syndrome temporally associated with SARS-CoV-2 among children and adolescents in England, risk of intensive care unit admission and death remained very low across the first 2 years of the pandemic and decreased as new variants emerged. Children and adolescents with complex medical problems and neurodisability had the highest risk regardless of SARS-CoV-2 variant.

Meaning  The findings in this study provide further reassurance regarding the low risk of severe illness after SARS-CoV-2 infection in children and adolescents and highlight that children and adolescents with comorbidities and neurodisability should be prioritized in terms of COVID-19–related health interventions.

Importance  Investigating how the risk of serious illness after SARS-CoV-2 infection in children and adolescents has changed as new variants have emerged is essential to inform public health interventions and clinical guidance.

Objective  To examine risk factors associated with hospitalization for COVID-19 or pediatric inflammatory multisystem syndrome temporally associated with SARS-CoV-2 (PIMS-TS) among children and adolescents during the first 2 years of the COVID-19 pandemic and change in risk factors over time.

Design, Setting, and Participants  This population-level analysis of hospitalizations after SARS-CoV-2 infection in England among children and adolescents aged 0 to 17 years was conducted from February 1, 2020, to January 31, 2022. National data on hospital activity were linked with data on SARS-CoV-2 testing, SARS-CoV-2 vaccination, pediatric intensive care unit (PICU) admissions, and mortality. Children and adolescents hospitalized with COVID-19 or PIMS-TS during this time were included. Maternal, elective, and injury-related hospitalizations were excluded.

Exposures  Previous medical comorbidities, sociodemographic factors, and timing of hospitalization when different SARS-CoV-2 variants (ie, wild type, Alpha, Delta, and Omicron) were dominant in England.

Main Outcomes  PICU admission and death within 28 days of hospitalization with COVID-19 or PIMS-TS.

Results  A total of 10 540 hospitalizations due to COVID-19 and 997 due to PIMS-TS were identified within 1 125 010 emergency hospitalizations for other causes. The number of hospitalizations due to COVID-19 and PIMS-TS per new SARS-CoV-2 infections in England declined during the second year of the COVID-19 pandemic. Among 10 540 hospitalized children and adolescents, 448 (4.3%) required PICU admission due to COVID-19, declining from 162 of 1635 (9.9%) with wild type, 98 of 1616 (6.1%) with Alpha, and 129 of 3789 (3.4%) with Delta to 59 of 3500 (1.7%) with Omicron. Forty-eight children and adolescents died within 28 days of hospitalization due to COVID-19, and no children died of PIMS-TS (PIMS-S data were limited to November 2020 onward). Risk of severe COVID-19 in children and adolescents was associated with medical comorbidities and neurodisability regardless of SARS-CoV-2 variant. Results were similar when children and adolescents with prior SARS-CoV-2 exposure or vaccination were excluded.

Conclusions  In this study of data across the first 2 years of the COVID-19 pandemic, risk of severe disease from SARS-CoV-2 infection in children and adolescents in England remained low. Children and adolescents with multiple medical problems, particularly neurodisability, were at increased risk and should be central to public health measures as further variants emerge.

Evidence from the first year of the COVID-19 pandemic shows that most children and adolescents develop mild illness following SARS-CoV-2 infection and that serious outcomes are rare.^1-4 However, as new SARS-CoV-2 variants have emerged, additional data on risk of severe disease are needed to inform clinical and public health responses.^5-8

Assessing the severity of different SARS-CoV-2 variants is complex and dependent on multiple factors that may vary by age.^9,10 These include immunity from prior infection, patterns of healthcare utilization, changes to shielding advice, and eligibility and effectiveness of vaccination.

An increase in hospitalizations among children and adolescents with the Delta (B.1.617.2) variant in summer 2021, and particularly after the emergence of the Omicron (B.1.1.529) variant in December 2021,^7,8 led to concerns that children and adolescents may have become more vulnerable to serious disease. However, rise in hospitalizations may reflect variation in community infection rates^7,11 and higher incidental infections among hospitalized children and adolescents rather than more severe disease.^12,13 Although monitoring hospitalizations is critical for COVID-19 surveillance in children and adolescents, we must also examine other indicators of disease severity, such hospitalizations requiring pediatric intensive care unit (PICU) admission and deaths.

Previous work assessing severity across different SARS-CoV-2 variants among children and adolescents has been based on notification systems within hospitals or on limited regional data and are open to bias. National population-level analyses are lacking for more recent variants.^8,14,15 We used a unique dataset including all hospitalizations among children and adolescents aged 0 to 17 years in England linked to national data on SARS-CoV-2 testing, SARS-CoV-2 vaccination, PICU admissions, and deaths. We describe numbers of hospitalizations, PICU admissions, and deaths after hospitalization for both COVID-19 and pediatric inflammatory multisystem syndrome temporally associated with SARS-CoV-2 (PIMS-TS)¹⁴ and assess whether previously identified risk factors associated with serious illness changed with the emergence of different variants.

We used Secondary Use Services data provided by National Health Service England, a dataset covering 98% of national inpatient activity. We analyzed all hospitalizations in England among children and adolescents aged 0 to 17 years from February 1, 2020, to January 31, 2022. The study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cohort studies. Ethics approval was provided after review by Yorkshire and the Humber South Yorkshire National Health Service Research Ethics Committee on June 10, 2021 (reference 21/YH/0127). Control of Patient Information (COPI) regulations provided a legal basis for linking these datasets without consent.¹⁶

Hospitalization data were deterministically linked with the following English national datasets: all community and hospital-based polymerase chain reaction and lateral flow tests for SARS-CoV-2 provided by the UK Health Security Agency (UKHSA); data on all UK PICU admissions provided by the Pediatric Intensive Care Audit Network; death registrations provided by the Office for National Statistics; death notifications provided by the National Child Mortality Database; and SARS-CoV-2 vaccination data provided by UKHSA. We also obtained authoritative modeled monthly estimates of new infections with SARS-CoV-2 in individuals younger than 18 years by age group in England from UKHSA.^17,18

We used PICU admission after hospitalization with COVID-19 or PIMS-TS as our main indicator for severe SARS-CoV-2 infection. We also examined deaths within 28 days of hospitalization with COVID-19 or PIMS-TS from death registration data (excluding deaths due to injury) or notification data within the National Child Mortality Database to account for delay in death registration.¹⁹ National Child Mortality Database data were only available to the end of November 2021, and underlying cause of death was not available in these data.

We defined COVID-19 hospitalizations as those where the primary reason for hospitalization was recorded using an International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) COVID-19 code (ie, U071, U072, U073, or U074). This highly specific definition (hereafter, primary COVID-19) may miss unclear diagnoses or children and adolescents with multiple diagnoses. To account for this, we also identified hospitalizations where a COVID-19 code was recorded as either the primary or secondary diagnosis or where the children and adolescents had a positive polymerase chain reaction or lateral flow test for SARS-CoV-2 within 7 days of admission or discharge. This definition (hereafter, primary or secondary COVID-19) is likely to be highly sensitive but will capture hospitalizations for other reasons, including incidental SARS-CoV-2 infection. We defined hospitalizations due to PIMS-TS where the relevant ICD-10 code (ie, U075) was the primary reason for admission, as this condition was unlikely to be recorded as a secondary diagnosis due to its severity. This code was introduced in November 2020 and so we restricted this analysis to November 1, 2020, to January 31, 2022. For all definitions, we excluded elective and maternal hospitalizations and those where the primary reason for admission was an injury (ICD-10 chapters 19 and 20).

We defined time periods where different variants were dominant in England using UKHSA data. These periods were defined as follows: wild type, February 1, 2020, to January 30, 2021; Alpha, January 31, 2021, to May 15, 2021; Delta, May 16, 2021, to December 11, 2022; and Omicron, December 12, 2021, to January 31, 2022.

We used hospitalization data from March 1, 2015, onward to identify medically recorded comorbidities in children and adolescents and to populate sociodemographic variables. We grouped comorbidity diagnoses by body system and identified children and adolescents with life-limiting conditions using established code lists.^20,21 We categorized age as 0 to 4, 5 to 11, and 12 to 17 years. As outcomes for SARS-CoV-2 among children and adolescents have been shown to vary by ethnicity,³ we included this in our sociodemographic variables. Ethnicity was defined using categories provided within the Secondary Uses Service, namely: Asian, Black, White, mixed, and other.

We described hospitalizations with COVID-19 and PIMS-TS by sociodemographic factors and previous comorbidities and examined hospitalizations per 100 000 monthly new infections with SARS-CoV-2 in England. We then examined differences in sociodemographic characteristics among hospitalizations with dominant SARS-CoV-2 variants using standardized residuals from χ² statistics.

We described numbers of hospitalized children and adolescents with COVID-19 and PIMS-TS who were admitted to PICUs or died within 28 days of hospitalization. We then repeated this among children and adolescents with and without prior SARS-CoV-2 exposure or vaccination. We considered children and adolescents to have been infected with SARS-CoV-2 from the date of first positive SARS-CoV-2 test result or hospitalization with either primary or secondary COVID-19 or PIMS-TS, whichever was sooner. We considered children and adolescents to be vaccinated beginning 14 days after receiving any SARS-CoV-2 vaccine.

We then modeled the association between sociodemographic factors and presence of comorbidities among hospitalizations with COVID-19 and PIMS-TS with the odds of PICU admission. We used generalized estimation equations to account for multiple admissions within the same children and adolescents, using a logit link and specifying the covariance structure as exchangeable. We then compared odds ratios (ORs) and predicted probability for PICU admission during the whole study period (February 1, 2020, to January 31, 2022) with those during different variants. We conducted analyses in Stata version 16 (StataCorp).

Between February 1, 2020, and January 31, 2022, there were 1 125 010 nontraumatic emergency hospitalizations in children and adolescents aged 0 to 17 years in England. Among these, there were 10 540 hospitalizations with primary COVID-19 among 10 298 children and adolescents (4891 [46.4%] female and 5649 [53.6%] male); 9827 (93.2%) of these were among children and adolescents without prior exposure to SARS-CoV-2 infection or vaccination (Table 1). There were 33 597 hospitalizations with primary or secondary COVID-19 among 29 262 children and adolescents. Patterns of hospitalizations were similar with the 2 definitions (eFigure 1 in Supplement 1). We focus on primary COVID-19 here with findings for primary or secondary COVID-19 shown in eFigure 5 and eTables 1-3, 5, and 7 in Supplement 1.

Monthly COVID-19 hospitalizations tracked closely with estimated infections (eFigure 6 in Supplement 1). The highest number of monthly hospitalizations due to primary COVID-19 was 2304 in January 2022, with 1631 (71%) of these occurring in children and adolescents younger than 5 years. Monthly hospitalizations with primary COVID-19 per 100 000 infections was highest in May 2020 at 345, with another peak of 219 in February 2021. Monthly hospitalizations per 100 000 new infections were highest and more variable among children younger than 5 years and lowest among those aged 5 to 9 years. In those aged 12 to 17 years, hospitalization rates declined steadily after August 2021 with a small increase in January 2022 (eFigures 2 and 3 in Supplement 1).

Overall, 6546 hospitalizations for primary COVID-19 (62.1%) were among children younger than 5 years. The age distribution varied over time, with those younger than 5 years forming a higher proportion during Omicron than during previous dominant variant periods. Children and adolescents from deprived neighborhoods and those who were Asian, Black, or of mixed ethnicity were overrepresented in hospitalizations throughout. We identified medical comorbidities in 4627 children (43.9%) and adolescents hospitalized with primary COVID-19, ranging from 1908 (29.1%) among children younger than 5 years to 1521 (69.5%) among those aged 12 to 17 years (eTable 2 in Supplement 1).

Overall, 448 primary COVID-19 hospitalizations (4.3%) required PICU admission, and 1055 primary or secondary hospitalizations (3.1%) required PICU admission. Among primary COVID-19 hospitalizations, PICU admission was lowest in those younger than 5 years (141 [2.2%]) compared with those aged 5 to 11 years (142 [7.9%]) and 12 to 17 (165 [7.5%]) years (Table 2). The proportion of hospitalizations requiring PICU admission declined as the pandemic progressed, from 162 of 1635 (9.9%) during wild type, 98 of 1616 (6.1%) during Alpha, 129 of 3789 (3.4%) during Delta, and 59 of 3500 (1.7%) during Omicron periods, with similar patterns seen in all age groups. Patterns of PICU admissions and hospitalizations due to primary COVID-19 were similar among children and adolescents without prior exposure to SARS-CoV-2 (eTable 7 in Supplement 1).

Among children and adolescents with comorbidities hospitalized with primary COVID-19, 396 (8.6%) required PICU admission compared with 52 (0.9%) in children and adolescents without comorbidities (eTable 3 in Supplement 1). This increased from 78 (3.9%) with comorbidities within 1 body system to 318 (12.2%) in those with comorbidities in more than 1 body system and 99 (17.0%) for those with life-limiting neurodisability. The proportion of children and adolescents requiring PICU admission declined as the pandemic progressed among those with and without comorbidities.

After adjusting for the presence of any comorbidity, children and adolescents aged 5 to 11 years and 12 to 17 years had significantly higher odds of PICU admission for primary COVID-19 compared with those younger than 5 years, as did children and adolescents who were Asian, Black, or of unspecified ethnicity compared with White children and adolescents. Associations between sociodemographic factors and odds of PICU admission remained similar across different variants, with older children and adolescents and those who were Asian, Black, or of unknown ethnicity most at risk. Being in the least deprived quintile of deprivation was associated with increased odds of PICU admission during the Alpha period only. Adjusted ORs for PICU admission with primary COVID-19 associated with comorbidity status by dominant variant are shown in the Figure. Compared with children and adolescents with no comorbidities, the odds of PICU admission were significantly higher among children and adolescents with any comorbidity (OR, 7.66; 95% CI, 5.58-10.49), with only 1 body system involved (OR, 3.71; 95% CI, 2.56-5.37), children and adolescents with 2 or more body systems involved (OR, 11.39; 95% CI, 8.18-15.88), and children and adolescents with life-limiting neurodisability (OR, 15.13; 95% CI, 10.06-22.76). Children and adolescents with any form of comorbidity had higher odds of PICU admission than those without comorbidities across all variants. While odds appeared highest during the Omicron period, confidence intervals were wide and overlapping (eTable 4 and 5 in Supplement 1).

Between November 1, 2020, and January 31, 2022, there were 997 hospitalizations with PIMS-TS among 961 children and adolescents, 928 of which (93.9%) were among children and adolescents without prior SARS-CoV-2 infection or vaccination. The number of hospitalizations per 100 000 SARS-CoV-2 infections was highest during Alpha, peaking at 87.6 in February 2021 (eFigure 4 in Supplement 1). From May 2021 to the end of the study period, there were fewer than 20 monthly hospitalizations due to PIMS-TS per 100 000 SARS-CoV-2 infections. A total of 614 hospitalizations with PIMS-TS (61.6%) were among male children and adolescents, 583 (58.5%) were among children and adolescents aged 5 to 11 years, and 291 (29.2%) were among children and adolescents aged 12 to 17 years (Table 3). A total of 474 hospitalizations (47.5%) were among White children and adolescents compared with 795 238 (71%) among all nontraumatic emergency hospitalizations during this period. There was little change in demographic characteristics of PIMS-TS hospitalizations across dominant variants, although when Alpha was the dominant variant, there were higher numbers of children and adolescents younger than 5 years, lower numbers of children and adolescents who were White, and higher numbers of children and adolescents from the most deprived quintile than expected. When Omicron was dominant, there were also higher than expected numbers of White children and adolescents and fewer children and adolescents from the most deprived neighborhoods. A total of 695 PIMS-TS hospitalizations (69.7%) were among children and adolescents with an identified comorbidity; however, a high proportion of these may have been acute complications. After excluding hematological and noncongenital cardiac conditions (common acute manifestations of PIMS-TS), 167 hospitalizations (35.6%) had a previous comorbidity. Further, 546 PIMS-TS hospitalizations (54.8%) occurred among children and adolescents who had no previous admissions to hospital recorded in our data (ie, since March 2015).

Overall, 437 hospitalizations with PIMS-TS (43.8%) required PICU admission, highest among those aged 12 to 17 years (135 [46.4%]) and lowest among those younger than 5 years (42 [34.1%]). Similar to COVID-19, this proportion declined during the pandemic from 20 (58.8%) during wild type to 173 (51.8%) during Alpha, 154 (45.3%) during Delta, and 90 (31.1%) during Omicron (Table 2). Patterns of hospitalizations due to PIMS-TS requiring PICU admission were similar when only children and adolescents without prior infection or vaccination were included (eTable 7 in Supplement 1).

After adjusting for the presence of any comorbidity, female sex was associated with increased odds of PICU admission, as were age 12 to 17 years compared with those younger than 5 years and Asian ethnicity compared with White ethnicity (eTable 6 in Supplement 1). The odds of PICU admission increased in those with any comorbidity compared with those with no comorbidity (OR, 2.07; 95% CI, 1.55-2.78) and those with comorbidities in more than 1 body system (OR, 3.11; 95% CI, 2.23-4.33). Associations of comorbidities with risk of PICU admission were similar across dominant variants (eTable 6 in Supplement 1).

Forty-eight children and adolescents died within 28 days of hospitalization with primary COVID-19, all of whom had documented comorbidities; 44 (91.7%) had comorbidities affecting 2 or more body systems, and 26 (59%) had life-limiting neurodisability. We did not identify any children and adolescents who died within 28 days after admission with PIMS-TS. Further analysis of deaths was not possible due to low numbers.

In this cohort study, we present robust national data on risk factors for hospitalization and severe disease related to SARS-CoV-2 infection during the first 2 years of the COVID-19 pandemic in England. We identified 10 540 hospitalizations with primary COVID-19 across the study period, 448 of which (4.3%) were admitted to PICUs. Using a broader more sensitive but less specific primary or secondary COVID-19 definition, we identified 33 597 hospitalizations, 1055 of which (3.1%) required PICU admission. Monthly hospitalizations closely tracked community infection rates. Hospitalizations per 100 000 new SARS-CoV-2 infections were highest at the start of the pandemic, with a second peak during the Alpha variant period across all age groups. Despite the highest community infection rates being associated with Omicron, rates of hospitalizations and PICU admissions were lowest across all age groups with this variant.

We found that the proportion of children and adolescents with COVID-19 requiring PICU admission declined steeply from the start of the pandemic across all age groups, despite later higher community infection rates. By the Omicron period, the proportion of children and adolescents admitted to PICUs for COVID-19 was 1.7%. These falls may reflect multiple mechanisms, including lower disease severity with more recent variants, changes in clinical practice and thresholds for hospitalization and PICU admission as learning occurred across the pandemic, and higher incidental infections in hospitalized children and adolescents at times of high community infection rates.

We found no changes in major medical risk factors associated with severe COVID-19 disease across variants or compared with our previous report from the first pandemic year.³ Most children and adolescents who required PICU admission had medical comorbidities, as did all those who died within 28 days of hospitalization for COVID-19. We found deaths after hospitalization to be extremely uncommon in children and adolescents, as we have previously reported,²² although we acknowledge deaths occurring outside hospital are not included in our dataset.

We found differences in associations of demographic factors with COVID-19 hospitalizations as the pandemic progressed. Risk factor associations early in the pandemic, such as older age, non-White ethnicity, and lower socioeconomic status, tended to attenuate across later variants. These findings may reflect differences between variants, higher levels of immunity due to prior infection in those at greater risk, or the impact of vaccination.^23-25 Vaccination is only likely to have impacted those aged 12 to 17 years in our study, as this age group was eligible for vaccination in the UK from September 2021 onward, while younger children were considered eligible after January 2022. Of note, almost all hospitalized children and adolescents included in our study had no prior exposure to SARS-CoV-2 infection or vaccination.

The pattern of decline in the proportion of hospitalizations requiring PICU admission across the pandemic was also found for PIMS-TS, and we identified no deaths within 28 days of hospitalization after November 2020. This likely reflects changes in clinical practice as knowledge and experience increased. Further, we found a steep decline in the number of monthly hospitalizations with PIMS-TS per 100 000 SARS-CoV-2 infections after the Alpha period and no evidence of an increase as subsequent variants emerged. This is consistent with regional data from the UK and elsewhere showing the incidence of PIMS-TS to have reduced during the Omicron and Delta periods compared with the Alpha period.^14,26

Our results support previous findings showing low overall risk of serious outcomes after SARS-CoV-2 infection in children and adolescents^3,22 and a decline in disease severity during the pandemic.^7,15 Wang and colleagues¹⁵ reported a lower risk of emergency department attendance, hospitalization, and intensive care support in children younger than 5 years who were infected with Omicron compared with Delta in the US. Similarly, Swann et al⁸ found severity among children and adolescents hospitalized with COVID-19 across the UK to have declined during the Alpha period compared with early months of the pandemic. An analysis of deaths in children and adolescents in England between March 2020 and December 2021 attributed 81 deaths to COVID-19 (1.2% of all deaths over this period), with lower infection fatality rates during Delta than for Alpha or wild type.⁴

We used multiple linked national datasets to describe COVID-19 and PIMS-TS hospitalizations during the first 2 years of the COVID-19 pandemic. The study provides a robust analysis of risks of serious disease not possible with regional or hospital datasets.

Limitations include variable diagnostic coding within Secondary Use Services and challenges in identifying true COVID-19 hospitalizations from incidental or past infection. Further, although SARS-CoV-2 testing of all emergency hospitalizations was introduced from the end of April 2020, limited testing availability earlier in the pandemic may have affected identification of COVID-19 hospitalizations, and children and adolescents with multiple comorbidities may have been more likely to have been tested.²⁷ The ICD-10 code for PIMS-TS cases was only introduced in November 2020 and we did not attempt to examine hospitalizations prior to this using proxy codes. We used chronic condition codes described by Hardelid et al²⁰ recorded from 2015 to identify children and adolescents with comorbidities. These may have also captured acute complications of PIMS-TS and COVID-19, including among children and adolescents who died. However, as almost all children and adolescents who died had comorbidities in multiple organ systems and more than half had life-limiting neurodisability, this is unlikely to have substantially affected our results. We found odds of PICU admission with COVID-19 increased with age, but did not specifically examine risks in infants and neonates, and further population-level analyses to explore this are required.²⁸ We considered COVID-19 admissions to be due to variants that were dominant in England at that time but were unable to confirm this. We used PICU admission as a proxy for disease severity, but did not have data on the level of intensive support required. Thresholds for PICU admission may have changed during the pandemic, and children and adolescents with complex needs may have been admitted to PICU as a precaution.³ Our estimate for COVID-19 deaths after hospitalization is likely an overestimate, as we were unable to attribute cause of death in this analysis.^4,22 During the first pandemic year, we identified 29 children and adolescents who died within 28 days of hospitalization with COVID-19; 8 of these deaths were found to be attributable to COVID-19 after case note review.³ Further, we were unable to exclude deaths due to injury for those only reported through the National Child Mortality Database, which was also restricted until November 2021.

The risk of severe disease from SARS-CoV-2 infection in children and adolescents in England remained low and decreased across the first 2 years of the pandemic. Children and adolescents with multiple medical problems were most at risk regardless of SARS-CoV-2 variant, and should be central to public health measures as further variants emerge.

Accepted for Publication: April 3, 2023.

Published Online: July 31, 2023. doi:10.1001/jamapediatrics.2023.2357

Correction: This article was corrected on September 25, 2023, to add missing degrees for author Peter J. Davis.

Corresponding Author: Joseph L. Ward, PhD, University College London Great Ormond St. Institute of Child Health, 30 Guilford St, London WC1N 1EH, United Kingdom (joseph.ward@ucl.ac.uk).

Author Contributions: Dr Ward had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Ward, Harwood, Kenny, Clark, Davis, Hargreaves, Gent, Luyt, Turner, Fraser, Viner.

Acquisition, analysis, or interpretation of data: Ward, Kenny, Cruz, Davis, Draper, Ladhani, Gent, Williams, Luyt, Whittaker, Bottle, Fraser, Viner.

Drafting of the manuscript: Ward, Turner, Viner.

Critical revision of the manuscript for important intellectual content: Ward, Harwood, Kenny, Cruz, Clark, Davis, Draper, Hargreaves, Ladhani, Gent, Williams, Luyt, Whittaker, Bottle, Fraser, Viner.

Statistical analysis: Ward, Harwood, Davis, Fraser, Viner.

Obtained funding: Kenny, Draper, Hargreaves, Luyt, Bottle, Viner.

Administrative, technical, or material support: Harwood, Cruz, Clark, Luyt.

Supervision: Kenny, Gent, Luyt, Turner, Bottle, Viner.

Conflict of Interest Disclosures: Dr Kenny reported grants from the National Institute for Health Research during the conduct of the study. Dr Luyt reported that the National Child Mortality Database Programme, including this work, is funded by National Health Service England and commissioned by the Healthcare Quality Improvement Partnership as part of the National Clinical Audit and Patient Outcomes Programme. Dr Bottle reported consulting fees from AstraZeneca and Eli Lilly outside the submitted work. Drs Fraser and Viner reported grants from the National Institute for Health Research outside the submitted work. No other disclosures were reported.

Funding/Support: Dr Harwood is in receipt of a fellowship from Kidney Research UK (grant TF_010_20171124). Dr Hargreaves is supported by the National Institute for Health Research through the National School for Public Health Research Programme and the Applied Health Research programme for North-West London. Dr Fraser is in receipt of funding from Martin House Children’s Hospice. Dr Viner is in receipt of a grant from the National Institute of Health Research to support this work (grant NIHR202322).

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Data Sharing Statement: See Supplement 2.

Additional Contributions: We would like to thank the National Child Mortality Database; Paediatric Intensive Care Audit Network; UK Health Security Agency; National Health Service Digital, National Health Service England, and National Health Service Improvement Children and Young People Team (particularly Sophie Solti, Tiffany Watson-Koszel, and Muhammad Ambia) for their support in identifying, linking and making the data used in this study available for analysis. We would like to further thank the UK Health Security Agency for providing estimates for number of new monthly SARS-CoV-2 infections by age group.