Influenza-Associated Excess Mortality and Hospitalization in Germany from 1996 to 2018

Introduction: Influenza-associated excess mortality and morbidity is commonly estimated using statistical methods. In Germany, the Robert Koch Institute (RKI) uses the relative mortality distribution method (RMDM) to estimate influenza-associated excess mortality without reporting age-specific values. In order to better differentiate the distribution of the disease burden, a distinction by age is of high relevance. Therefore, we aimed to revise the existing excess mortality model and provide age-specific excess mortality estimates over multiple seasons. We also used the model to determine influenza-associated excess hospitalizations, since the RKI excess hospitalization model is currently based on another approach (i.e., combination of excess physician visits and hospitalized proportion).

Methods: This study was a retrospective data analysis based on secondary data of the German population from 1996-2018. We adapted the RKI's method of estimating influenza-associated excess mortality with the RMDM and also applied this approach to excess hospitalizations. We calculated the number of excess deaths/hospitalizations using weekly and age-specific data.

Results: Data available in Germany are suitable for addressing the restrictions of the RKI's mortality model. In total, we estimated 175,858 (176,482 with age stratification) influenza-associated excess all cause deaths between 1995-1996 and 2017-2018 ranging from 0 (17 with age stratification) in 2005-2006 to 25,599 (25,527 with age stratification) in 2017-2018. Total influenza-associated excess deaths were comparable to RKI's estimates in most seasons. Most excess deaths/hospitalizations occurred in patients aged ≥ 60 years (95.42%/57.49%) followed by those aged 35-59 years (3,80%/24,98%). Compared with our model, the RKI hospitalization model implies a substantial underestimation of excess hospitalizations (828,090 vs. 374,200 over all seasons).

Conclusion: This is the first study that provides age-specific estimates of influenza-associated excess mortality in Germany. The results clearly show that the main burden of influenza is in the elderly, for whom prevention and control measures should be prioritized.