One hundred fourteen patients undergoing partial hepatectomy were randomized to a clinical arm (ANH used for resection of ¡Ý3 liver segments) or a nomogram arm (ANH used for predicted probability of transfusion ¡Ý50 % based on a previously validated nomogram). The primary end point was appropriate management, defined as avoidance of ANH in patients at low risk or use of ANH in patients at high risk for allogeneic red blood cell transfusions.
Between September 2009 and May 2011, 58 patients were randomized to the clinical arm and 56 to the nomogram arm. Demographics, diagnoses, extent of resection, blood loss, and incidence and grade of complications did not differ between the 2 groups. There were no differences in perioperative transfusions or laboratory values. Nomogram-based allocation did not change appropriate management overall (80 % vs 76 % in the clinical arm; p?=?0.65), but did result in comparable perioperative outcomes and a trend toward decreased ANH use (30 % vs 47 % ; p?= 0.09), particularly in low blood loss (estimated blood loss ¡Ü400?mL) cases (12 % vs 25 % ; p?= 0.04).
Although allocation of intraoperative management using a transfusion nomogram did not improve appropriate management overall, it more effectively identified low blood loss cases and reduced ANH use in patients least likely to benefit.