From the aInstitute of Public Health, Charité – Universitätsmedizin Berlin, Berlin, Germany

bDigital Health—Machine Learning Research Group, Hasso Plattner Institute for Digital Engineering, Potsdam, Germany

cDigital Engineering Faculty, University of Potsdam, Potsdam, Germany

dDepartment of Mathematics, Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland.

Submitted August 21, 2023; accepted January 28, 2024

M.J.S. is funded by the Swiss National Science Foundation, Grant 200021_207436.

M.P. reports having received partial funding from Novartis Pharma, and being awarded a research grant from the Center for Stroke Research Berlin (private donations), both outside of the submitted work.

Supplemental digital content is available through direct URL citations in the HTML and PDF versions of this article (www.epidem.com).

All data used in this article consisted in aggregated data previously published in Grotta et al. 2021 (https://doi.org/10.1056/NEJMoa2103879). The R code used to load the data and run the statistical analyses can be found in the Supplemental Digital Content (eAppendix; https://links.lww.com/EDE/C121).

Correspondence: Mats Julius Stensrud, Department of Mathematics, Ecole Polytechnique Fédérale de Lausanne, EPFL SB MATH BIOSTAT, MA B2 477 (Bâtiment MA), Station 8, CH-1015, Lausanne, Switzerland. E-mail: [email protected].