Reconstruction of SARS-CoV-2 outbreaks in a primary school using epidemiological and genomic data

Abstract

Mathematicalmodelling studies have shown that repetitive screening canbeused tomitigate SARS-CoV2 transmission inprimary schoolswhile keeping schools open. However, notmuch is knownabout how transmissionprogresseswithin schools andwhether there is a riskof importation tohouseholds. During theacademicyear2020–2021, aprospectivesurveillancestudyusingrepetitivescreeningwasconductedin aprimary school andassociatedhouseholds inLiège (Belgium). SARS-CoV-2 screeningwas performedvia throatwashingeitheronceor twiceaweek.Weusedgenomicandepidemiological data toreconstruct the observedschooloutbreaksusingtwodifferentmodels.Theoutbreaker2model combines informationonthe generationtimeandcontactpatternswithamodelofsequenceevolution.ForcomparisonwealsousedSCOTTI, aphylogeneticmodel basedon the structuredcoalescent. Inaddition,weperformeda simulationstudy to investigatehowtheaccuracyofestimatedpositivityratesinaschooldependsontheproportionofaschoolthat issampledinarepetitivescreeningstrategy.WefoundnodifferenceinSARS-CoV-2positivitybetweenchildren andadults andchildrenwerenotmoreoftenasymptomaticcompared toadults. Bothmodels foroutbreak reconstruction revealed that transmissionoccurredmainlywithin the school environment. Uncertainty in outbreakreconstructionwas lowestwhenincludinggenomicaswellasepidemiologicaldata.Wefoundthat observedweeklypositivityrates areagoodapproximation to the trueweeklypositivityrate, especially in children, evenwhenonly25%of theschool populationis sampled.Theseresults indicate that, inaddition to reducing infections as shown inmodelling studies, repetitive screening inschool settings can lead toa betterunderstandingof theextentof transmissioninschoolsduringapandemicandimportationriskat the communitylevel.