Skip to main content
Fig. 6 | BMC Nephrology

Fig. 6

From: Predictors of shorter- and longer-term mortality after COVID-19 presentation among dialysis patients: parallel use of machine learning models in Latin and North American countries

Fig. 6

Top 15 predictors of death 0 to 14 days, 15 to 30 days, and > 30 days after COVID-19 presentation in descending order for the Latin America (A, C, E) and North America (B, D, F) cohorts. Bar plot in the left panels show the mean absolute SHAP values that estimate the average size of each variable’s contribution to predicting the outcome on the x-axis (calculated from the average absolute value for all patients). SHAP value plots in the right panels show the size and direction (more positive = higher risk or more negative = lower risk) of each variable’s influence on the outcome for each unique patient on the x-axis, with warmer colors representing higher observed values for that measurement, cooler colors indicating lower values for that measurement, and gray representing a missing value for that measurement. SHAP values are presented in the unit of log odds (i.e. logarithm of the odds ratio)

Back to article page