Skip to main content
Fig. 3 | Critical Care

Fig. 3

From: Machine learning identifies ICU outcome predictors in a multicenter COVID-19 cohort

Fig. 3

EBM prediction model showing importance of risk factors predicting need for ECMO or RRT in COVID-19 ICU patients including admission data. (a) ECMO therapy left) A significant risk factors for outcome after analysis of admission data and weighed according to their importance for outcome. Right B importance of age for outcome and distribution of age data C importance of status “intubated” on ICU admission and distribution of status D) importance of status “external transfer” on ICU admission and distribution of status E importance of Murray lung injury score and distribution of MLIS data. Green indicates patients that did not receive ECMO therapy, orange indicates patients that did receive ECMO therapy. (b) Renal Replacement Therapy (RRT). Left A significant risk factors for outcome after analysis of admission data and weighed according to their importance for outcome. Right B importance of the interaction of age and D-dimer level for outcome and distribution of data C initial creatinine values and distribution of data determined on admission D initial SOFA score w/o GCS and distribution of data determined on ICU admission. Blue indicates patients that did not receive RRT, red indicates patients that did receive RRT

Back to article page