ICD-10 based syndromic surveillance enables robust estimation of burden of severe COVID-19 requiring hospitalization and intensive care treatment

Abstract

AbstractObjectiveThe emergence of coronavirus disease 2019 (COVID-19) required countries to establish COVID-19 surveillance by adapting existing systems, such as mandatory notification and syndromic surveillance systems. We estimated age-specific COVID-19 hospitalization and intensive care unit (ICU) burden from existing severe acute respiratory infections (SARI) surveillance and compared the results to COVID-19 notification data.MethodsUsing data on SARI cases with ICD-10 diagnosis codes for COVID-19 (COVID-SARI) from the ICD-10 based SARI sentinel, we estimated age-specific incidences for COVID-SARI hospitalization and ICU for the first five COVID-19 waves in Germany and compared these to incidences from notification data on COVID-19 cases using relative change Δrat the peak of each wave.FindingsThe COVID-SARI incidence from sentinel data matched the notified COVID-19 hospitalization incidence in the first wave with Δr=6% but was higher during second to fourth wave (Δr=20% to 39%). In the fifth wave, the COVID-SARI incidence was lower than the notified COVID-19 hospitalization incidence (Δr=-39%). For all waves and all age groups, the ICU incidence estimated from COVID-SARI was more than twice the estimation from notification data.ConclusionThe use of validated SARI sentinel data adds robust and important information for assessing the true disease burden of severe COVID-19. Mandatory notifications of COVID-19 for hospital and ICU admission may underestimate (work overload in local health authorities) or overestimate (hospital admission for other reasons than the laboratory-confirmed SARS-CoV-2 infection) disease burden. Syndromic ICD-10 based SARI surveillance enables sustainable cross-pathogen surveillance for seasonal epidemics and pandemic preparedness of respiratory viral diseases.