P1090DETERMINATION OF ABSOLUTE BLOOD VOLUME USING ONLINE DIALYSATE DILUTION: WHEN SHOULD BE MEASURED?

Abstract Background and Aims Current on-line haemodiafiltration (HDF) machines equipped with a blood volume monitor (BVM) and an on-line bolus function have the potential for measuring absolute blood volume (aBV). Recently, we developed a simple method to determine absolute BV in everyday dialysis sessions. The aim of the present study was to evaluate the reproducibility of measurements. Method Intra-individual reproducibility was studied in 10 patients during a single dialysis session by 4 measurements of absolute BV: immediately after beginning before ultrafiltration (UF) was started, and after one, two and three hours. ABV was determined by indicator dilution. A defined volume bolus of 240 mL dialysate was infused into the venous blood line by pressing the emergency button of the HDF machine 5008 (FMC). For this reason, total UF volume was increased by 1L. UF was automatically stopped during and after the infusion. The resulting increase in relative blood volume (RBVpost-RBVpre) was measured by the ultrasonic relative BVM incorporated in the dialysis machine. ABV was measured in hourly intervals and for assessment of reproducibility the volume at treatment start (t=0) where RBV is 100% was calculated for all measurements as: aBV in mL = bolus volume 240 mL x 100% / increase RBV in % ABV data were normalized for body mass at dry weight (in mL/kg). Additionally, in 5 patients the RBV graph was monitored immediately at the beginning of dialysis without UF in a separate dialysis session. Results ABV at t=0 were consistently larger when calculated from measurements done immediately after the beginning compared to measurements obtained after 1 h (6.52 ± 1.40 L or 80.6 ± 14.5 mL/kg vs. 5.16 ± 1.40 L or 63.9 ± 14.3 mL/kg). Specific BV derived from 2 and 3 h measurements did not significantly differ from the measured volumes after 1 hour (61.4 ± 13.8 mL/kg, and 60.9 ± 13.9 mL/kg). The standard deviations of the 3 examinations in the same study patient during a further course of dialysis were between 0.6 and 5.3 ml/kg (ø 2.6 ml/kg). In a separate session, RBV decreases without UF at the beginning of dialysis in the first 3 minutes by 0.5 % and in 5 minutes by 0.6 %. Conclusion If BV is diluted by additional priming volume and bolus volume, a part of this volume will leave the circulation. This represented the time frame where the bolus was initially infused and the measurements were carried out. This loss is caused by the reduction in plasma colloid osmotic pressure induced by the dilution thereby changing the microvascular filtration equilibrium. The increase in RBV display is not solely caused by the bolus volume in this time and, and therefore, calculated BV would be overestimated by about 17 mL/kg. If measurement is performed at a later time, UF will take place and, consequently, refilling. This inward drive matches the outward bolus escape as a counterforce. BV measurement during a further course of dialysis is well reproducible with a deviation of only ± 2.6 ml/kg. The method would therefore be sufficiently precise in clinical practice. Therefore, we propose the determination of aBV only after 1 hour dialysis when a sufficient refilling takes place. With a software modification, the BV measurement could be routinely automated during each dialysis treatment. Manufacturers are asked to implement this technology in their devices.