The purpose of this small randomized pilot study is to evaluate adherence to strict transfusion protocols in both the control and ANH group examining compliant and noncompliant transfusion rates and reasons for violation of the protocol if present. Data on transfusion requirements for both groups (with and without ANH) will be evaluated to determine sample size and feasibility of performing a larger study on ANH in our patient population. The ultimate purpose of a larger study would be to determine if acute normovolemic hemodilution results in a reduction in red blood cell units transfused in patients undergoing cardiac surgery involving cardiopulmonary bypass (CPB) and to carefully determine provider adherence to specific transfusion triggers for the administration of each unit of harvested and banked blood transfused. Secondary endpoints that will be evaluated in a larger study include any difference in the number of other blood products transfused including fresh frozen plasma, platelets, and cryoprecipitate as well as differences in coagulation status as measured by thromboelastography (TEG), INR/PTT, platelet count, and fibrinogen level, and ICU/hospital length of stay.

Cardiac surgery patients are typically at higher risk of receiving blood transfusions during surgery than patients presenting for most other surgical procedures. Acute normovolemic hemodilution (ANH) is a method of blood conservation used to reduce the amount of donor blood transfused. It involves removing blood from appropriate patients immediately after the induction of anesthesia, replacing the volume lost with colloids (e.g. 5% albumin), then storing and returning this blood to the patient at the completion of the surgical procedures. By reducing the patient's hematocrit during surgery, the blood that is shed during the procedure has a lower red blood cell volume and therefore fewer red blood cells are lost during major blood loss. In addition, the autologous stored blood contains coagulation factors that are spared and returned following the procedure. This process is already performed here and at other institutions as a potential means of reducing transfusions in high risk patients. However, in spite of multiple publications, the actual value of this procedure is unclear, in part because many of the published studies in cardiac surgical patients failed to establish firm protocol- specific guidelines for either returning harvested blood to the patient or for using banked blood, or because it is unclear whether these guidelines were actually followed during surgery (rather than left to the discretion of anesthesiologists and surgeons).

The primary goal of this pilot study is to implement rigorous transfusion protocols for both ANH and routine transfusion practice and to examine compliant and noncompliant transfusion rates in both the ANH and control groups, evaluating reasons for violation of the transfusion protocol if applicable and to determine the feasibility of performing a larger study on ANH. Secondary goals for this pilot study include gathering data on transfusion rates in ANH versus control patients to determine the sample size for a larger, definitive study in which the primary goal would be to evaluate the effectiveness of ANH in reducing transfusion of banked blood units during cardiac surgery as compared to control patients in which ANH is not used. Secondary endpoints of a larger study would also include evaluating coagulation status in patients undergoing ANH versus control patients.

Inclusion Criteria:

- Patients undergoing cardiac surgery involving one of the following procedures at the University of Minnesota: A) combined coronary artery bypass graft and valve repair/replacement B) redo cardiac surgery C) left ventricular assist device placement

Exclusion Criteria:

• Hemoglobin concentration < 12 g/dl
• Patients undergoing cardiac surgery without the use of cardiopulmonary bypass