Response to: Incidence of stroke in Switzerland

The almost simultaneous release of new epidemiological stroke data from different databases in Switzerland is a lucky fortuity [1, 2]. In a Letter to the Editor, Dirren et al. discussed their main finding: a discrepancy between the results of their stroke incidence and our stroke discharge findings. Comparing results of the Geneva Stroke Study (GSS) with findings of the Federal Hospital Discharge Statistics (HOST) or Cause of Death (CoD) statistics in Switzerland is challenging [1, 2]. The goal of our study was to show trends in stroke epidemiology between 2004 and 2017. Therefore, we used the same methodical approach and data sources as Meyer et al., revealing increasing stroke events and discharge rates, but an approximately halving of the overall case-fatality and mortality, likely reflecting better diagnostic standards and therapeutic approaches [2, 3]. However, when using the HOST and CoD statistics, calculating the true incidences is impossible due to the lack of linkage of both databases, which makes a comparison with the population-based true incidence of the GSS difficult. However, both studies indicate a similar number: the annual number of strokes in Dirren et al. for the years 2018–19 is 13 612 cases, while we report 26 032 stroke discharges [1, 2]. The difference would dissipate if similar inclusion criteria were applied: (i) in contrast to the GSS and in accordance with Meyer et al., we included multiple hospitalizations (at least two) of one patient (n = 5890); (ii) beside similar included ICD-10 codes (International Statistical Classification of Diseases and Related Health Problems 10th Revision: I60, I61, I63), we also included I62 (n = 1761) and I64 (n = 1227). With the subtraction of these non-ischaemic stroke patients from our number, only 17 154 stroke discharges in 2017 remain. (iii) Considering the fact, as discussed before in the Letter to the Editor, that we could not separate first-ever strokes from recurrent strokes, this would have further reduced the absolute number derived from stroke discharges to