Relationship of model parameters to clinical risk factors
We analysed the relationship between clinical risk factors (age, sex,
weight, height, body surface area, blood volume, WHO infection grade and
LDH) and the derived individual model parameters by formal regression
analysis. Individual parameter estimates resulting from fitting all
cycles of the validation data set were used for that purpose. Details of
regression analysis are explained in the Supplement Material S.5. To
check the performance of these regression models, we applied them to our
primary data set adding a new prediction scenario, which was only based
on the clinical risk factors of the patient.