COVID-19 epidemic trends from a Self-Assessment Web Application in France: Observational Study (Preprint)
BACKGROUND We developed a self-assessment and participatory surveillance web-application for coronavirus disease (COVID-19), which was launched in France in March 2020. OBJECTIVE We compared daily large-scale RT-PCR tests results to daily connections to a self-triage application by anosmic users to assess dynamics of emergency visits, hospitalizations and ICU admissions for COVID-19 positive patients in France. METHODS Between 3/21/2020 and 11/18//2020, users of maladiecoronavirus.fr self-triage application were asked questions about COVID-19 symptoms. Data of daily hospitalizations, large-scale RT-PCR positive tests, emergency visits and ICU entrances for COVID-19 patients were compared to data of anosmic users of the applications. RESULTS As of November 18, 2020, 575,214 users reported recent anosmia on near 13 Million responders. Daily anosmia reported during the peak of connections to the application on September 16 were spatially correlated with daily COVID-19–related hospitalizations peak which occurred in November (spearman correlation coefficients=0.77, p<.001). Decrease of the connections of anosmic users after the main peaks of connections preceded the decrease of daily hospitalizations by 10 and 9 days during the first and the second outbreak waves respectively although the decrease of RT-PCR positive tests occurred only 2 days before daily hospitalizations during the second wave. CONCLUSIONS A peak of daily reported anosmia in young adults was observed 49 days before the peak of the hospitalizations corresponding to the first phase of a large-scale contamination of young population followed by older people leading to hospitalization’s peak in November. Data of anosmic users of a national widespread self-assessment application can be a relevant tool to anticipate outbreak surge, and hospitalizations and ICU decrease of COVID-19 patients. CLINICALTRIAL ClinicalTrials.gov NCT04331171