We present a statistical model to estimate, i.e. impute, the missing data for the disease date based on the calculated difference between laboratory reporting date and disease date of case reports with complete data. The model is applied to the COVID19 surveillance dataset for Austria. The difference between disease date and laboratory reporting date averaged 5.4 days, with variability by calendar week of epidemic: the difference increased with case number per calendar week. Based on the laboratory reporting date, the case number peak was on 26.03.2020 and based on the disease date, including cases with imputed disease date, already on 16.03.2020.

Lukas Richter, Daniela Schmid, Department of Infection Epidemiology & Surveillance, AGES Ernst Stadlober, Institute of Statistics, Graz University of Technology