Preoperative Acute Normovolemic Hemodilution Is an Alternative to Hypervolemic Hemodilution-In Case of Proper Use

Abstract

In Response: Kreimeier and Finsterer emphasize that our method of hypervolemic hemodilution (HHD) proved to be as (in)effective as acute normovolemic hemodilution (ANH) because the target hematocrit (Hct) of 33% +/- 5% was higher than recommended. This is exactly the message we intend to emphasize. ANH is of limited clinical value in reducing homologous blood saving, and it should only be used if target Hct values <20% are accepted. However, in our experience and as published, only 2-3 units of red blood cells (RBC) are typically removed for ANH [2]. As we mentioned, it was first shown by Messmer [3] that it is necessary to remove at least approximately 2000 mL whole blood and let the intraoperative Hct not increase to more than 19% to reach a saving effect of 298 mL [3]. These 298 mL (approximately 1 single RCB unit) seem to be the maximal saving effect that may be reached using ANH. It requires the preoperative removal of at least 4 units of homologous blood. In addition, even authors presenting several models of the blood-sparing effect of ANH question any reasonable effect of ANH [2,4,5]. The aim of our study was only to facilitate the use of hemodilution in clinical practice (removal of 2-3 RBC units), proposing a less cost- and time-consuming, but still safe, method [1]. Our study was preceded by an investigation looking for intravascular volume status and cardiac compensation, which we quoted in the text [6]. The speculations of Kreimeier and Finsterer concerning the intravascular volume status after HHD are disproved by the results of this study, showing no detrimental side effects with our infusion regimen. These data are confirmed by Trouwborst et al. [7]. In this context, it is a common experience that rapid intravascular volume infusion before or immediately after the induction of anesthesia is necessary to stabilize hemodynamics, especially in elderly patients. We strongly refute the allegation that we put our patients at risk with an infusion rate of 100 mL/min for the experimental purpose of showing efficacy in terms of the time-saving effect of HHD. The problem of an intra- and postoperative transfusion trigger depends on the patients' individual physiological compensation possibilities. We do not consider ANH with hemoglobin levels below 7.5 g/dL a safe method for every patient. As shown by Doss et al. [8], hemodilution with Hct values lower than 28% exerts detrimental effects in patients with compromised myocardial function [8]. The median age of our patients was approximately 60 yr, so we considered a transfusion trigger of 9 g/dL to be more justifiable. We doubt that the extreme hemodilution (Hct <20%), which is necessary to achieve an acceptable blood saving effect, is a safe procedure for most patients in daily anesthetic routine. This and the amount of time and financial effort are probably the reasons that ANH is very often used only in a limited manner and not in the recommended way (Hct <20%). For reasons mentioned above, we are not convinced that a more profound hemodilution would have been more effective. However, the use of extreme ANH with HCT <20% compared with HHD has not been studied either by us or by others, and all conclusions drawn by Kreimeier and Finsterer (including the title of the letter) are of a highly speculative manner. We want to point out again that we only compared ANH and HHD in orthopedic patients undergoing hip replacement with an estimated blood loss of approximately 1000 mL and concluded that in this patient population, HHD seems to be a good alternative to ANH in routine practice by offering a gain in time and effort. Lars L. Mielke, MD Elmar K. Entholzner, MD Michael Kling, MD Barbara E. M. Breinbauer, MD Rainer Burgkart, MD Stefan R. Hargasser, MD Rudolf F. J. Hipp, MD Institut fur Anaesthesiologie; Klinikum rechts der Isar; Technische Universitat Munchen; D-81675 Munchen, Germany