Long COVID in children and adolescents: results from three cross-sectional school-based cohorts with adjudication

STUDY AIMS: The prevalence of Long COVID in children and adolescents is heterogeneous, ranging from 1% to 51%, depending on the population studied. The lack of a standardised approach for establishing a Long COVID diagnosis in children and adolescents complicates the accurate assessment of prevalence, risk factors and outcomes. The present study aimed to examine the value of standardised interviews and an adjudication process to better understand self- or proxy-reported symptoms lasting longer than 12 weeks compatible with Long COVID in children and adolescents during the COVID-19 pandemic. METHODS:We conducted a school-based, prospective cohort study (Ciao Corona) from March 2020 to July 2022 in the Canton of Zurich, Switzerland. Of 156 invited schools, 55 agreed to participate. Primary schools were randomly selected across all 12 districts of the Canton of Zurich, with nearby secondary schools subsequently invited. Within participating schools, classes were randomly selected, stratified by school level, and all students aged 6–17 years in selected classes were eligible. At three different time points (March/April 2021, November/December 2021 and June/July 2022), school-aged children and adolescents underwent serology testing and completed online questionnaires, including questions on symptoms lasting ≥12 weeks compatible with Long COVID. We invited those with persisting symptoms and who were seropositive for SARS-CoV-2 – whether “infected” i.e. infection and no vaccination or having “hybrid immunity” i.e. infection and vaccination – to participate in interviews to allow us to better understand the pattern, severity and timing of the reported symptoms. An adjudication process with experts then followed to assess the probability of Long COVID. RESULTS: 39/1120 (3.5%) seropositive children and adolescents (i.e. infected or with hybrid immunity) reported persisting symptoms (≥12 weeks). The most frequently reported symptoms were headache, tiredness and stomach ache. In 20/39 (51%) with persisting symptoms who agreed to be interviewed, the adjudication committee concluded that Long COVID was unlikely in 13 (65%), possible in 7 (35%) and likely in 0 participants. CONCLUSIONS: Relying exclusively on self- or proxy-reported questionnaire data, without more detailed information may overestimate Long COVID in children and adolescents. Implementing standardised interviews and an adjudication process helps to contextualise self- or proxy-reported symptoms compatible with Long COVID. Trial registration: https://clinicaltrials.gov NCT04448717.