Nomograms for the prediction of changes in patient chemistries during hemodialysis

Prediction of changes in patient chemistries as a result of a hemodialysis has been difficult due to 1) the different physical characteristics of the many types of hemodialyzers, 2) intra-patient and interpatient variations of blood flow through the devices, 3) differing body weights among the patient population, 4) variability of the predialysis body chemical composition of the patients, 5) unknown mass transfer characteristics of chemical substances between the blood and other body fluid compartments in the patients, and 6) variations in the duration of dialysis. One or more of these factors has hindered attempts by Blackmore and Elder [1], Wolf et al [2] and others to accurately predict the results of dialysis in vivo. Recently Wolf, Watson, and Barbour [3] have shown that one can use mathematical models to predict the chemical changes of urea and creatinine in the body fluids of the uremic patient during dialysis with a Kiil type dialyzer. All that needs to be specified is the weight of the patient, the duration (hours) of dialysis, and the average rate of blood flow into the device. This paper uses the computer results of Wolf to construct nomograms for the prediction of changes in plasma urea and creatinine concentrations during hemodialysis with a Kiil dialyzer.