Imputation des Erkrankungs-Datums bei unvollständigen Daten im Rahmen der COVID19 Epidemie, Österreich

Lukas Richter, Daniela Schmid*, Ernst Stadlober

*Corresponding author for this work

Research output: Working paper

Abstract

The date of disease onset in case reports of notifiable diseases is often missing or reported with delay.Then, a surrogate date, such as the date of the diagnostic laboratory report, has to be used for time seriesanalyses. We introduce a statistical model, which allows the imputation of the missing onset dates based onthe known difference between date of lab report and of disease onset, obtained from complete case reports.The model was applied on the Austrian COVID19 surveillance data. The mean difference between lab reportdate and disease onset was 5.4 days with variation by calendar week of the report: the more case reportsper week, the higher the difference. We found the Austrian COVID19 epidemic peaked on March 26, 2020,using lab report dates, compared to an epidemic peak on March 16, 2020, when using known and imputeddisease onset dates
Original languageGerman
PublisherAGES- Österreichische Agentur für Gesundheit und Ernährungssicherheit
Number of pages8
Publication statusPublished - 4 Jun 2020

ASJC Scopus subject areas

  • Epidemiology
  • Statistics and Probability

Fields of Expertise

  • Information, Communication & Computing

Treatment code (Nähere Zuordnung)

  • Experimental

Cite this

Richter, L., Schmid, D., & Stadlober, E. (2020). Imputation des Erkrankungs-Datums bei unvollständigen Daten im Rahmen der COVID19 Epidemie, Österreich. AGES- Österreichische Agentur für Gesundheit und Ernährungssicherheit.