Created claim: quote(P203): The main methodological problem is the selection process described. The sample of the population actually using the app will not be a random sample. It seems likely that especially subpopulations with a higher prevalence of undetected infections will have lower coverage rates by the apps: