Fig. 1
Kinetic model and goodness-of-fit. a Schematic drawing of the volume kinetic model. b Plasma dilution as predicted by the base model versus the measured plasma dilution for all data points in the 111 experiments. Hence, the predictions are based only on the mean values of V c, k 12, k 21, k 10, and k b for all experiments. c Same plot as b but after correction of the predicted plasma dilution for covariates, which gives different values to k 21 and k 10 depending on gender. Random distribution around the solid line of unity indicates a good goodness-of-fit. d Predicted versus measured urinary excretion for all 111 infusions according to the base model. Each point represents one patient. e Same plot as d but after correction for covariates (i.e., gender). Comparing d with e shows how important gender is when predicting the diuretic response to infused fluid