Abstract
Background: Management of body fluid volumes and adequate prescription of ultrafiltration (UF) remain key issues in the treatment of chronic kidney disease patients.Objective: This study aims to estimate the magnitude as well as the precision of absolute blood volume (Vb) modeled during regular hemodialysis (HD) using standard data available with modern dialysis machines.Methods: The estimation utilizes a two-compartment fluid model and a mathematical optimization technique to predict UF-induced changes in hematocrit measured by available on-line techniques. The method does not rely on a specific hematocrit sensor or a specific UF or volume infusion protocol and uses modeling and prediction tools to quantify the error in Vb estimation.Results: The method was applied to 21 treatments (pre-UF body mass: 65.57±13.44 kg, UF-volume: 3.99±1.14 L) obtained in ten patients (4 female). Pre-HD Vb was 5.4±0.53 L with an average coefficient of variation of 9.8% (range 1 to 22%). A significant moderate correlation was obtained when Vb was compared to a different method applied to the same data set (r = 0.5). Specific blood volumes remained above the critical level of 65 mL/kg in 17 treatments (80.9%).Conclusion: The method offers the opportunity to detect critical blood volumes during HD and to judge the quality and reliability of that information based on the precision of the Vb estimate.