Fig. 4
From: COVID-19 outbreaks surveillance through text mining applied to electronic health records
Stratified time series of suspected and confirmed cases of COVID-19. The bars represent the daily numbers of suspected (in blue) and confirmed (in orange) cases of COVID-19. The solid lines represent moving averages with windows of 7 days for both quantities. We show the time series of suspected cases of COVID-19 stratified by the following age groups: patients with ages (a) 18-24 years, (b) 25-34 years, (c) 35-44 years, (d) 45-54 years, (e) 55-64 years, and (f) 65 years. We leave the age stratum of patients with age <18 years out of the main analysis due to its particular behavior likely related to seasonal respiratory diseases in children. We also emphasize that, in all remaining strata, the time series of suspected cases of COVID-19 are delayed in comparison to the time series of confirmed cases of COVID-19, endorsing, even more, the idea that it is possible to early detect COVID-19 waves using only the patient's reported symptoms