(Q4748)
Jump to navigation
Jump to search
Statements
From a statistical and methodological perspective, the use of smart devices suffers from under-coverage and non response, which are rarely addressed by proponents of digital epidemiology. [..]. For COVID-19 surveillance, we recommend that instead of digital epidemiology, available routine data and random samples should be used
0 references
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:
0 references
Older people, children, persons without smartphones due to lower-income and members of vulnerable populations.
0 references
June 2020
0 references