Paediatric safety assessment of BNT162b2 vaccination in a multistate hospital-based electronic health record system in the USA: a retrospective analysis

Background: The emergency use authorisation of BNT162b2 (tozinameran; Comirnaty, Pfizer–BioNTech) for children aged 5–17 years has resulted in rapid vaccination in the paediatric population. However, there are few studies of adverse events associated with vaccination in children. The aim of this study was to systematically assess the adverse events of two-dose BNT162b2 vaccination in the paediatric population. Methods: We conducted a retrospective analysis of patient electronic health records (EHRs) of children aged 5–17 years who received the primary two-dose series of the BNT162b2 vaccine between Jan 5, 2021, and Aug 5, 2022, at the Mayo Clinic Health System (MN, FL, AZ, IA, and WI), USA. Using natural language processing, we automatically curated adverse events reported by physicians in EHR clinical notes before and after vaccination. To determine significant adverse events after BNT162b2 vaccination, we calculated risk differences, which was defined as the percentage difference between the rate of children with an adverse event after a vaccine dose and the baseline rate of children with an adverse event before vaccination. 95% CIs and p values were calculated using the Miettinen and Nurminen score method. Findings: 56 436 individuals aged 5–17 years (20 227 aged 5–11 years and 36 209 aged 12–17 years) with EHRs in the Mayo Clinic Health Systems were included in the study. Overall, the reporting of adverse events remained low in passive surveillance. Serious adverse events were rare after the first and second doses of BNT162b2, with rates of anaphylaxis (six [0·01%] of 56 436), myocarditis (five [0·01%]), and pericarditis (three [0·01%]) consistent with previous studies. Among the 20 227 5–11-year-olds, there were increased risks of fatigue (58 after second dose vs 41 before first dose; risk difference [RD]_dose2 0·08% [95% CI –0·01 to 0·18], p=0·044) and fever (104 after second dose vs 77 before first dose; RD_dose2 0·13% [0·00 to 0·27], p=0·022) after the second dose. Among the 36 209 12–17-year-olds, there were increased risks of arthralgia (69 after second dose vs 48 before first dose; RD_dose2 0·06% [–0·00 to 0·12], p=0·026), chills (58 after second dose vs 40 before first dose; RD_dose2 0·05% [–0·00 to 0·11], p=0·034), and myalgia (96 after second dose vs 73 before first dose; RD_dose2 0·06% [–0·01 to 0·14], p=0·038) after the second dose. Although the overall incidence was low, there was an increased risk of myocarditis in males aged 12–17 years after the second dose (five after second dose vs zero before first dose; RD_dose2 0·03% [0·01 to 0·07], p=0·013), with median age being 15 years (IQR 14 to 16). Interpretation: Overall, this data suggests that vaccination with BNT162b2 in the paediatric population is generally safe and well-tolerated. Further research is warranted to investigate the basis for the increased risk of myocarditis in adolescent males. Additionally, further studies are needed to confirm whether the findings in our study population apply to the whole vaccinated paediatric population. Funding: nference.