BACKGROUND: COVID-19 (coronavirus disease 2019) vaccines are authorized for use in children in the United States (US). Real-world assessment is needed regarding the effectiveness of these vaccines in children.
OBJECTIVES: To estimate the real-world effectiveness of receiving a complete primary series of BNT162b2 (Pfizer-BioNTech) COVID-19 Vaccine in US children.
METHODS: A retrospective cohort of children aged 5 to 17 years was identified in Optum pre-adjudicated insurance claims linked with Immunization Information System (IIS) records from 10 US jurisdictions between 11 Dec 2020 and 31 May 2022 (end dates varied based on IIS availability). Vaccines were identified using IIS records and claims data. Vaccinated children were followed from their first BNT162b2 dose. Unvaccinated children were matched to vaccinated children on calendar date, US county, and demographic and clinical factors. Censoring occurred if vaccinated children failed to receive dose 2 or if unvaccinated children received any dose. Tw o outcome definitions were evaluated: COVID-19 diagnosis in any medical care setting and specifically in the hospital/emergency department (ED) setting. Propensity score–weighted Cox proportional hazards models estimated hazards ratios (HRs) and 95% confidence intervals (CIs); vaccine effectiveness (VE) was estimated as 1 minus the HR. VE was estimated overall, within age groups, and within variant-specific eras.
RESULTS: This study matched 91,947 eligible vaccinated children to unvaccinated comparators (mean age 12 years; 50% female). Rates of hospital/ED visits and hospitalizations were low among children, regardless of vaccination status (40.5 per 10,000 person-years). Overall, vaccination was associated with reduced incidence of COVID-19 in any medical care setting (VE = 36%; 95% CI, 31%-40%) and hospital/ED-diagnosed COVID-19 (VE = 53%; 95% CI, 37%-64%) . VE estimates were highest among children aged 16-17 years and lowest among those aged 5-11 years and during the Omicron variant era.
CONCLUSIONS: Vaccination with a complete primary series of BNT162b2 was associated with lower medically diagnosed COVID-19 and hospital/ED-diagnosed COVID-19 risk in these children, although the observed VE estimates varied by age group and variant era. Because pandemic conditions and vaccine authorizations have evolved rapidly over time, further research in children is needed.
