The Efficacy and Utility of Acute Normovolemic Hemodilution

Abstract

Acute normovolemic hemodilution (ANH) is a blood conservation technique that was first described in the early 1970s in the setting of cardiac surgery.1,2 The principle behind this procedure is to reduce the patient’s hematocrit by phlebotomy along with infusing crystalloid and/or colloid before the onset of surgical blood loss such that for a given amount of bleeding, a smaller red blood cell (RBC) mass will be lost. The maximum benefit from ANH is achieved when a low, but physiologically adequate, hematocrit is maintained during the blood loss phase of the surgical procedure, after which the fresh whole blood that was removed is given back to the patient near the end of the surgical procedure. Ideally, for ANH to be effective in reducing allogeneic transfusion requirements, the patient should: (1) have a relatively high preoperative hematocrit; (2) undergo the maximum allowable phlebotomy; and (3) lose a substantial amount of blood during surgery. If any of these 3 parameters is not present, the potential benefit of ANH in reducing allogeneic transfusions will be limited.3 Whether ANH is effective in reducing transfusion requirements has been a point of controversy. Multiple controlled trials evaluating ANH in patients who underwent a variety of surgical procedures have been published, which allows for meta-analysis to determine efficacy. In fact, 2 previous meta-analyses, including 1 in 1998 with 24 trials4 and 1 in 2004 with 42 trials,5 did not show conclusive evidence to support the widespread use of ANH. Now 11 years later, in the current issue of Anesthesia & Analgesia, Zhou et al.6 report an updated systematic review and meta-analysis that includes additional recent studies (63 studies in total) in an attempt to answer the question of whether ANH is efficacious in reducing allogeneic transfusion. The authors should be commended for conducting the largest and most rigorous analysis of ANH studies to date. Despite their comprehensive meta-analysis, high-quality evidence to support the routine use of ANH is still lacking. Although ANH decreased both the likelihood of any allogeneic transfusion (by 26%) and the volume of allogeneic transfusion (by approximately 1 unit), there was evidence for publication bias leading to an overestimation of benefit from ANH. The primary limitations of this study were the inclusion of small trials, some with as few as 10 patients per group, and the absence of transfusion protocols with a designated hemoglobin trigger or target, which introduces bias when blinding is not possible in such studies. As expected, the type of surgery, the volume of blood withdrawn, the presence or absence of other blood management methods, and even the year of publication all influenced the beneficial impact of ANH. The secondary outcomes were adverse (morbid) events. The only morbid event for which the incidence differed between groups was “any infection,” which unfortunately was reported in only 10 of the 63 studies but occurred with less frequency in the ANH group (relative risk, 0.64; P = 0.037). Because allogeneic transfusion is known to be associated with hospital-acquired infection,7 this finding is plausible, interesting, and perhaps understated in the current study. In addition, the inability to assess such outcomes in a blinded fashion is problematic. The authors go on to conclude that based on their findings, in combination with previous mathematical modeling studies,8 surgical procedures with blood loss of 1 L or greater are the cases in which the benefits of ANH are most likely to be recognized. Assessing the efficacy of ANH is no simple task. For example, what is the primary outcome that should be measured? Is it the percentage of patients exposed to allogeneic transfusion or the volume of blood transfused? What about other important outcomes that are often not reported in these studies such as length of stay, morbid events, mortality, and overall costs? Is ANH being compared with no blood conservati