Is Hemodilution Effective?

In Response: Dr. Harris contends that the red blood cell (RBC) volume removed by hemodilution and available for reinfusion (585 mL) in the case illustrated in Figure 1of our report [1] should be credited as blood conserved or "saved." Following this line of reasoning, the same volume (585 mL) that was removed would have to be credited as intraoperative blood loss. Intraoperative hemodilution cannot be perceived as the magical generation of RBC volume because it does not allow time for the patient to replenish RBC volume lost [2,3]. For the patient in Figure 1, the infusion of the 585 mL of RBC at the end of the procedure replaced the RBC mass that was removed during hemodilution. Hemodilution conserves blood only because the patient's hematocrit level is lowered before surgical blood loss, so that, for a given volume of whole blood loss, less RBC volume is lost. Thus, the true savings from hemodilution only was the 204 mL of RBC volume that was not lost during the period of surgical blood loss because the patient's hematocrit level was decreased by hemodilution. Dr. Harris questions why all autologous blood units were not returned to the patients. All blood removed by hemodilution was returned in the operating room; we did not include this hemodiluted blood in the data on blood transfusions. In this study and our previous one [4], the predeposited blood was transfused only when medically necessary. In our previous report, approximately 50% of the autologous blood predonated was transfused to the patients who did not require allogeneic blood [4]. In the present series, the mean of 0.38 autologous blood units transfused (about 6 units total) represented two-thirds of the predeposited autologous blood units. Five patients who did not predeposit blood required allogeneic blood. We are well aware of the inherent limitations of blood volume and blood loss estimates, but these limitations are the same for all clinical studies and yet do not detract from the usefulness of comparison of clinical outcomes in patients retrospectively. One important advantage of retrospective analysis is that the audit does not bias physician transfusion behavior, as would be the case for any prospective study of transfusion outcomes [5]. Finally, Dr. Harris points out that the standard deviations exceed the means of blood units transfused. The values listed in Table 4are not based on any statistical test but are simply the means and standard deviations of positively skewed samples. As with any variable having finite limits, one can add or subtract sufficient standard deviation units from the mean to exceed those limits. Lawrence T. Goodnough, MD Division of Laboratory Medicine, Departments of Medicine and Pathology Terri G. Monk, MD Department of Anesthesiology, Washington University School of Medicine, St. Louis, MO 63110-1092