Safety outcomes following COVID-19 vaccination and infection in 5.1 million children in England – Nature.com
June 2, 2024
Number of children receiving COVID-19 vaccines
A total of 5,197,925 young people aged 517 years, comprising 1,842,159 children aged 511 years and 3,355,766 adolescents aged 1217 years were included in the study. 4,347,781 young adults aged 1824 years, were included as a comparison. The characteristics of the young people included in the study are detailed in Table1 and Supplementary Table1.
In children aged 511 years, 32% (n=581,545) received at least one dose of COVID-19 vaccine and 16% (n=303,118) received a second dose within the study period. Over 99.9% of 511-year-olds who received at least one COVID-19 vaccine dose received the BNT162b2 vaccine (Table2). 82% (n=1,508,661) of children in this age group had a positive SARS-CoV-2 test recorded between 8th December 2020 and 7th August 2022, 0.4% (n=5665) of whom received their first COVID-19 vaccine dose prior to their first recorded positive SARS-CoV-2 test (Table2).
In adolescents aged 1217 years, 86% (n=2,882,229) received a first dose of COVID-19 vaccine, 67% (n=2,254,214) received a second dose and 14% (n=454,868) received a third dose (Table2).
The characteristics of the population excluded from the self-controlled case series analysis (i.e. those who did not receive any COVID-19 vaccine and did not have a positiveSARS-CoV-2 test recorded during the study period) are presented in Supplementary Table2.
In children aged 511 years, we did not observe an increased risk of any of the pre-specified outcomes in the 142 days following any dose COVID-19 vaccine with BNT162b2, mRNA-1273 or ChAdOX1 (Table3). However, given that less than 0.1% of vaccinated 5-11-year-olds received a ChAdOX1 or mRNA-1273 vaccine, the probability of type II errors was high as the sample size was too small to detect statistically significant associations for these vaccines.
The clinical characteristics of all children hospitalised with a pre-specified safety event are shown in Supplementary Table3. Supplementary Tables47 show the incidence rate ratios (IRR) and 95% confidence intervals (CI) for all outcomes 142 days and in weekly risk periods following each vaccine dose in 511-year-olds in males and females separately, and the effect of ethnicity on the risk of each outcome.
In the 142 days after the first and second doses of BNT162b2, we observed an increased risk of myocarditis in adolescents aged 1217 years (IRR 1.92, 95%CI 1.083.43 and IRR 2.96, 95%CI 1.655.32 for first and second dose, respectively) (Table4). We estimated that an additional 3 (95%CI 05) cases per million exposed would be anticipated after the first dose and 5 (95%CI 36) after the second dose (Fig.1). When we split the risk period into weekly blocks, the increased risk was restricted to 114 days following each dose (Supplementary Table8). There was also an increased risk of hospitalisation with epilepsy (IRR 1.17, 95%CI 1.001.37; excess events per million: 12, 95%CI 023) in the 142 days following the second dose of BNT162b2 (Table4, Fig.1).
Estimated number of excess events per million (95% CI) based on incidence rate ratios of each outcome in the 142 days following vaccination or SARS-CoV-2 positive test compared to baseline period are presented where there were at least five events during the exposure period and when number of excess events is greater than zero. Data available from 8th December 2020 and 7th August 2022. Table4 contains the data presented in this figure. MIS-C Multisystem inflammatory syndrome; ITP Idiopathic or immune thrombocytopenic purpura, ADEM Acute disseminated encephalomyelitis.
In the sex-stratified analysis, the increased risk of myocarditis after the first dose of BNT162b2 was only observed in females (IRR 4.01, 95%CI 1.3312.09; excess events per million: 3, 95%CI 14), while the increased risk following the second dose was observed in males only (IRR 2.87, 95%CI 1.505.51; excess events per million: 9, 95%CI 411) (Supplementary Figs.1 & 2, Supplementary Tables4 & 5). We additionally observed an increased risk of demyelinating disease, restricted to females (IRR 2.41, 95%CI 1.06-5.48; excess events per million: 4, 95%CI 06) following the second dose of BNT162b2. Of the eight female adolescents who experienced demyelinating disease in the 142 days following a second dose of BNT162b2, five were coded as optic neuritis.
In a post hoc analysis investigating differences in risk between children of different ethnic backgrounds, we found that the risk of anaphylaxis following a second dose of BNT162b2 in adolescents with non-white ethnicity was higher relative to the risk in adolescents with white ethnicity (relative IRR 2.55, 95%CI 1.006.46) (Supplementary Table6). However, when the analysis was restricted to the subgroup of adolescents from non-white ethnic backgrounds, the risk of anaphylaxis in the 142 days following a second dose of BNT162b2 was not significantly increased compared to the baseline period (IRR 1.69, 95%CI 0.803.54) (Supplementary Table6). We did not identify any differences in vaccine safety between white and non-white ethnicity for any of the other pre-specified outcomesin under-18s.
We found no evidence for significantly increased risks for any of the pre-specified outcomes in the 142 days following a first, second or third dose of mRNA-1273 vaccine in 1217-year-olds (Table4). However, this analysis lacked power to detect statistically significant associations, except for very large effect sizes, as less than 0.1% of adolescents received a first or second dose of mRNA-1273 vaccine.
There was an increased risk of hospitalisation with epilepsy 142 days after a first dose of ChAdOX1 (IRR 1.93, 95%CI 1.103.39), with an additional 705 (95%CI 1291033) cases expected per million exposed (Table4). This increased risk was restricted to females in the sex-stratified analysis (IRR 2.26, 95%CI 1.034.94) with an additional 813 (95%CI 441164) hospitalisations with epilepsy expected per million female adolescents exposed (Supplementary Table4).
We also observed an increased risk of appendicitis in the 142 days following the second dose of ChAdOX1 (IRR 4.64, 95%CI 1.7712.17; excess events per million: 512, 95%CI 283599) (Table4).
The IRRs and 95% CIs for all outcomes 142 days and in weekly risk periods following each vaccine dose in 1217-year-olds in males and females separately, and the effect of ethnicity on the risk of each outcome, are presented in Supplementary Figs.1 & 2, Supplementary Tables46 & 8.
In children aged 5-11 years who had received at least one dose of COVID-19 vaccine before the date that their positive SARS-CoV-2 test was recorded, we did not observe increased risks of any of the pre-specified outcomes in the 142 days following SARS-CoV-2 infection. In children who had not been vaccinated against COVID-19 prior to infection, there was an increased risk of hospital admission with MIS-C following a SARS-CoV-2 positive test (IRR 11.52, 95%CI 9.2514.36), with an additional 137 (95%CI 134140) cases expected per million exposed (Table3, Fig.2). In the sex-stratified analysis, the risk of MIS-C was slightly greater in male children (IRR 12.00, 95%CI 8.9216.12; excess events per million: 162, 95%CI 157-166) compared to female children (IRR 11.13, 95%CI 7.9615.57; excess events per million: 124, 95%CI 119127) (Supplementary Figs.3 & 4, Supplementary Tables4 & 5). The increased risk was mainly observed in the 2242 days following the date that the positive SARS-CoV-2 test was recorded (Supplementary Table7).
Estimated number of excess events per million (95% CI) based on incidence rate ratios of each outcome in the 142 days following SARS-CoV-2 positive test compared to baseline period are presented where there were at least five events during the exposure period and when number of excess events is greater than zero. Data available from 8th December 2020 and 7th August 2022. Table3 contains the data presented in this figure. MIS-C Multisystem inflammatory syndrome, ITP Idiopathic or immune thrombocytopenic purpura, ADEM Acute disseminated encephalomyelitis.
We also observed increased risks of hospital admission for myositis, myocarditis, acute pancreatitis and ADEM following SARS-CoV-2 infection before vaccination(Table3, Fig.2). In the sex-stratified analyses, we additionally identified increased risks of ITP (in both sexes) and anaphylaxis (in females only) (Supplementary Figs.3 & 4, Supplementary Tables4 & 5).
The IRRs and 95% CIs for all outcomes 142 days following SARS-CoV-2 infection in 511-year-olds in males and females separately, and the effect of ethnicity on the risk of each outcome, are presented in Supplementary Figs.3 & 4, Supplementary Tables4, 5, 6 & 7.
In adolescents aged 1217 years, we observed an increased risk of MIS-C (IRR 12.38, 95%CI 8.8817.28; excess events per million: 84, 95%CI 8186) in the 142 days following a SARS-CoV-2 infection in those who had not been vaccinated prior to SARS-CoV-2 infection (Table4, Fig.1). In the sex-stratified analysis, male adolescents were at higher risk of MIS-C following infection compared to females (IRR 12.33, 95%CI 8.3118.31and IRR 13.11, 95%CI 6.9024.91 in males and females, respectively), with an additional 131 (95%CI 126135) cases expected per million males exposed compared to 48 (95%CI 44-50) in females (Supplementary Figs.1 & 2, Supplementary Tables4 & 5).
We also observed increased risks ofhospitalisation with myocarditis, ITP and epilepsy in the 1-42 days following SARS-CoV-2 infection in adolescents who had not been vaccinated against COVID-19 prior to infection as well as an increased risk of hospitalisation with epilepsy in those who had received at least one vaccine dose prior to infection (Table4, Fig.1). In the sex-stratified analysis, the increased risks ofhospitalisation with myocarditis and epilepsy were restricted to males while the increased risk of ITP following infection was only observed in females (Supplementary Figs.1 & 2, Supplementary Tables4 & 5). We additionally identified increased risks of appendicitis (in females only) and anaphylaxis (in males only) following SARS-CoV-2 infection.
The IRRs and 95% CIs for all outcomes 142 days following SARS-CoV-2 infection in 12-17-year-olds in males and females separately, and the effect of ethnicity on the risk of each outcome, are presented in Supplementary Figs.1 & 2, Supplementary Tables4, 5, 6 & 8.
The results for all analyses in young adults aged 18-24 years are presented in Supplementary Tables6 & 9.
The robustness of the results of the self-controlled case series analyses were assessed by (1) checking that the risk of outcomes during the pre-vaccination period (month prior to vaccination to account for potential bias of people with recent hospitalisation being less likely to get vaccinated) was lower than the baseline period and (2) checking that the risk of the positive control outcome (anaphylaxis) was higherfollowing vaccination or SARS-CoV-2 infection than the baseline period. In the vast majority of analyses the estimates of the pre-vaccination period and the risk of anaphylaxis following vaccination or SARS-CoV-2 agreed with what was expected (SupplementaryTables8 & 9).
Our matched cohort analysis included 1,580,869 children aged 511 years and 1,535,341 adolescents aged 1217 years. Characteristics of the cohort are detailed in Supplementary Table10.
Incidence rates of vaccine safety outcomes in the 142 days following each vaccine dose and following SARS-CoV-2 infection in vaccinated and unvaccinated children are presented in Supplementary Table11. Incidence rates for all outcomes were significantly higher following SARS-CoV-2 infection compared to COVID-19 vaccination.
We matched 160,262 children aged 511 years and 848,186 adolescents aged 1217 years who had received at least one dose of COVID-19 vaccine to a child of the same age and sex who had not received any COVID-19 vaccine doses by the date of the vaccinated childs first vaccine dose (characteristics of matched cohort reported in Supplementary Table12).
As in the self-controlled case series analysis, we identified an increased risk of hospitalisation with epilepsy in the 142 days following a second dose of COVID-19 vaccine with BNT162b2 in 12-17-year-olds (unadjusted IRR 1.77, 95%CI 1.052.99, adjusted IRR 3.88, 95%CI 1.2711.86), but did not find significantly increased risks of appendicitis or myocarditis with BNT162b2 vaccination in adolescents (Supplementary Table13).
We identified additional increased risks of anaphylaxis and appendicitis in 12-17-year-olds following a first dose of BNT162b2 (unadjusted IRR 3.71, 95%CI 1.2311.14 and unadjusted IRR 1.37, 95%CI 1.051.80, respectively) and an increased risk of hospitalisation with epilepsy following a first dose with BNT162b2 in 511-year-olds, although the confidence interval was very wide reflecting the uncertainty of the estimate (unadjusted IRR 16.00, 95%CI 2.12120.65) (Supplementary Table13).
In general, the estimates from the matched cohort study were in agreement with the results from the self-controlled case series analysis in under-18s.
Unadjusted IRRs and IRRs adjusted for self-reported ethnicity (white, non-white, missing), quintile of deprivation (based on Townsend score) and presence of comorbidity (yes/no) for each outcome are reported in Supplementary Table13.
View post:
Safety outcomes following COVID-19 vaccination and infection in 5.1 million children in England - Nature.com