Profound normovolemic hemodilution: hemostatic effects in patients and in a porcine model

Previous systematic investigations of the hemostatic effects of normovolemic hemodilution (NHD) have not explored the influence of hematocrits less than 20% in humans or animals. However, clinical interest in maximizing the perioperative conservation of erythrocytes may involve profound NHD beyond traditionally accepted empiric end points. We report here on coagulation data in eight healthy adolescent patients undergoing profound NHD in concert with surgical correction of idiopathic scoliosis, and in 29 swine undergoing experimental stepwise NHD until death. Blood was replaced with 5% albumin in 0.9% saline in our patients, and with 5% albumin in lactated Ringer's solution in our pigs. A 75% blood volume exchange in our patients yielded a platelet count (PLT) of 158 +/- 26 x 10(3)/microL, fibrinogen concentration (FIB), 50 +/- 7 mg/dL, prothrombin time (PT), 25.4 +/- 2.6 s, activated partial thromboplastin time (aPTT), 87 +/- 15s, and a nadir hemoglobin of 2.8 +/- 0.2 g/dL; however, global oxygen delivery as assessed by body oxygen consumption remained adequate. Coagulation during the experimental porcine hemodilution was assessed by measuring PLT, FIB, PT, and aPTT, as well as by measurement of coagulation factor activities. In neither species did clinically significant thrombocytopenia (PLT < 100 x 10(3)/ microL) become manifest prior to clinical or other laboratory evidence of coagulopathy. Rather, a combined deficiency of coagulation factors explains the coagulopathy developing during NHD in both patients and swine. Abnormal hemostasis develops prior to compromise of global tissue oxygenation, assessed by mixed venous oxygen saturation and total body oxygen consumption, during NHD in healthy patients anesthetized as described. Therefore, NHD may be more limited by preservation of normal coagulation than of global oxygen delivery and consumption.