Citizen Science for Infectious Disease Surveillance in Belgium – COVID-19

Abstract

In this study, data from the survey ”de Grote Corona-Studie” are used to derive suspected COVID-19 cases and compare these to the official laboratory-confirmed incidence in Belgium. There are two goals to this: Firstly, to assess the general feasibility of using citizen science data to estimate incidence, and secondly, to compare the performance of different case definitions as given by national and supranational health agencies, namely Sciensano, CDC, RKI, ECDC and WHO. The comparison is performed using graphical means, raw correlations between the different time series, and correlations between the time series after pre-whitening using ARIMA models. Furthermore, different lags between the survey derived times series and the laboratory confirmed cases are tested and additional correlation analyses are per- formed for different gender and age groups. Lastly, sensitivity and specificity of the case definitions are compared based on a small subset of the survey data for which PCR test results were available.