Differential effects of serial hemodilution with hydroxyethyl starch, albumin, and 0.9% saline on whole blood coagulation.

STUDY OBJECTIVES To determine by thrombelastography assessed coagulation, the effects of progressive hemodilution with three intravascular volume expanders. DESIGN Prospective, controlled, whole blood, volumetric ex vivo hemodilution study. SETTING University of Pennsylvania Medical Center Operating Rooms. PATIENTS 60 ASA physical status I and II patients; phlebotomy prior to administration of IV fluids or medications. INTERVENTIONS Analysis of whole blood clotting determined by six thrombelastographic channels for control and five volumetric hemodilutions (11%, 25%, 33%, 50%, and 75%) with 0.9% saline, 5% albumin, and 6% hydroxyethyl starch (n = 20 for each diluent group). MEASUREMENTS AND MAIN RESULTS Thrombelastographic parameters R (minutes), angle alpha (degree), MA (mm), and lysis (%) were measured and compared to the sample control for each dilution of the same specimen. There was no significant difference between control groups in any thrombelastographic variable (R, angle alpha, MA, or lysis). No changes were seen in any variable from any diluent at 11% hemodilution. Seventy-five percent hemodilution caused significantly hypocoagulable changes from control for all thrombelastographic parameters for all three diluents. Thrombelastographic indices differed significantly from controls at intermediate hemodilutions. Both colloids caused decreases in measured angle alpha and MA at lower hemodilution than did 0.9% saline. Albumin 5% caused significant hypocoagulable changes from control values at lower hemodilution than did either 0.9% saline or 6% hydroxyethyl starch for all thrombelastographic parameters. Saline 0.9% increased angle alpha significantly at 50% hemodilution. Abnormal lysis did not occur at any dilution. Differing ex vivo effects of three different intravascular fluids thrombelastography assessed coagulation are found. CONCLUSION No differences were found after 11% hemodilution with any volume expanders. Hemodilution with up to 50% saline maintained thrombelastographic indices. Albumin produced early and profound hypocoagulable effects. Significant hypocoagulability occurred for all three diluents at 75% hemodilution. The study supports the use of albumin in patients at risk for thrombosis, and saline in patients with a need for normal hemostasis.