Abstract

We studied the effect of the initiation of a five part program of blood component conservation on blood product utilization in patients undergoing major cardiovascular surgery. We encouraged physicians and nurses to adhere to a set of guidelines that would limit the exposure of these patients to complications associated with blood product administration. Through this effort the amounts of packed red cells, fresh frozen plasma, platelets, and cryoprecipitate used were markedly reduced. After the start of the conservation program, overall blood product usage was reduced 29% despite a 46% increase in the surgical caseload, and there was a 53% decrease in the amount of donor exposure (units of individual blood products administered) per patient.

Methods

We studied the blood product utilization of all patients at our hospital who were undergoing CABG, single valve replacement or repair, multiple valve operations, reoperations for valve repair or replacement, closure of atrial septal defect in patients above and below 16 years of age, and ventriculoseptal defect closure. All patients undergoing these operations from February 1 to August 31, 1988, were placed in the control group (Group I). A second group of patients operated on from May 1 to October 31, 1989 (Group II), was studied after the introduction of a new set of procedures and guidelines governing blood product administration.

The following conservation techniques were utilized whenever possible for the patients in Group II:

1. Withdrawal of 450 to 900 ml of the patient's blood from the sideport of an 8.5 F percutaneous sheath introducer kit (Arrow AK 09801) after insertion of the Swan-Ganz catheter or from a 16ga × 8-inch central venous catheter, both inserted through an internal jugular or subclavian vein. Blood was collected by gravity into a bag containing 63 ml of citrate phosphate dextrose adenine solution (CPDA-1) over a period of 5 to 10 minutes for each 450 ml. This autologous blood was reinfused after neutralization of heparin. Pre-bypass blood withdrawal was not used when the patient's weight was less than 60 kg or the hemoglobin was less than 110 gm/L.

2. Reinfusion of all possible blood from the pump oxygenator after discontinuation of bypass, either through the aortic cannula or by intravenous infusion.

3. Acceptance of normovolemic anemia in patients who were not actively bleeding before the transfusion of packed cells.

4. Reinfusion of mediastinal chest tube drainage in the ICU using a described technique [2].

5. Administration of platelets, cryoprecipitate, and fresh frozen plasma only after specific coagulation deficiences could be demonstrated by coagulation tests. If the patient was not actively bleeding postoperatively, transfusions of platelets were withheld for platelet counts over 80,000; cryoprecipitate was given if the fibrinogen level was less than 1.0 gm/L; and FFP was reserved for patients with elevated prothrombin time (PT), partial thromboplastin time (PTT), thrombin time, and reptilase time.

All of these conservation methods represent a significant departure from our previous transfusion practice [3], which was used by Group I.

All patients in both groups were anesthetized with fentanyl (50 μg/kg) or sufentanil (5–10 μg/kg) supplemented with isoflurane or additional narcotics. Patients were anticoagulated with 300 units/kg of heparin before initiation of cardiopulmonary bypass. Adequate heparinization was assured by giving additional heparin to maintain the activated coagulation time (ACT) above 480 seconds. A disposable membrane oxygenator with a crystalloid prime was employed and bypass was carried out at 24 to 28°C. Cold potassium cardioplegic ischemic arrest was used in all cases.

At the conclusion of bypass, heparin was neutralized with protamine sulfate 1.5 times the initial heparin dose. Additional protamine was given if the ACT did not return to baseline levels. Platelet count and fibrinogen levels were monitored during bypass and the following tests were performed when the patient arrived in the ICU: hemoglobin, hematocrit, fibrinogen, PT, PTT, thrombin time, reptilase time, and fibrin degradation products. Transfusion of packed cells, platelets, cryoprecipitate, and FFP was recorded as units per patient, and accurate records of blood product usage were kept by the blood bank for each patient. Efforts were made to follow these blood product conservation guidelines for each patient during surgery and in the postoperative period. Whenever possible, the administration of blood products was discussed at daily morning rounds in the ICU. Transfusions were deemed inappropriate if they were not justified by the results of hemoglobin levels and coagulation tests.

Data were analyzed using a Student's two-sample t test.

Results

There were 177 patients in Group I and 259 in Group II. The number of patients undergoing the different surgical procedures is shown in Table 1. No significant differences in the height, weight, preoperative hemoglobin level, or mortality were found between Group I and Group II patients for each type of operation. The amounts of blood products utilized for the two patient groups are shown in Table 2 and demonstrate reduced blood product usage in Group II patients for all four types of blood products. These results were statistically significant (P <0.001). In addition, patients in Group II received less packed cells, FFP, platelets, and cryoprecipitate for every type of operation. Overall utilization of blood bank products decreased by 29%, with 1880 individual products administered in Group I patients compared with 1336 products given to Group II patients, even though there were 46% more patients in Group II.

Table 1. Summary of patients according to surgical procedures.

CABC = coronary artery bypass graft; ASD = atrial septal defect; VSD = ventriculoseptal defect.

Table 2. Summary of blood product usage.

The number of donors required to supply the blood products for each group of patients is estimated from the numbers of units of blood products administered, considering each unit of packed cells, platelets, cryoprecipitate, and FFP to have come from an individual donor. The number of donors for each patient group is presented in Figure 1 and represents the relative amount of potential exposure to the acquisition of infectious diseases for each operative group. There was an average reduction in donor exposure of 53% for Group II patients.

There were four patients in Group II with hemoglobin greater than 130 gm/L in the ICU. Two of these patients were overtransfused in the operating room and the two remaining received additional packed cells in the ICU despite a hemoglobin of greater than 110 gm/L.

Discussion

The results of this study demonstrate that a significant reduction in blood bank products can be achieved by the establishment of a blood product conservation program for cardiac surgical patients. It was not possible to assess the relative contributions of the several blood conservation methods used.

We use autologous prebypass withdrawal of 1 to 2 units in patients of suitable size and hemoglobin levels. Withdrawal of autologous blood prior to bypass has been shown to result in a 20% savings in blood products, although this was accomplished in a study which aimed for a postoperative hemoglobin level of 120 to 130 gm/L and did not use other conservation methods [4]. Other objections to this method are a questionable improvement of the platelet count after retransfusion [4], and lack of applicability in small or anemic patients.

The use of blood conservation during cardiac surgery has been used with great success at the Cleveland Clinic [1]. There, patients did not receive packed red cells until the hematocrit fell below 22%. Ninety percent of their patients did not receive any blood products during hospitalization and the mean transfusion rate was 0.3 units of packed cells per patient. The major predictor of the need for transfusion was age (above 60 years) and preoperative red cell volume (below 1600 vol.%) which is related to patient weight (below 65 kg) and hematocrit (below 37.5%). We did not use a cell saver because of the effort and expense involved in saving a relatively small blood loss prior to patient heparinization. This agrees with the experience of others [5]. There are other reasons why our experience does not match the excellent results of the Cleveland Clinic. Their CABG patients required less packed cell transfusions because they were heavier with higher red cell volumes, and we did not want to withhold transfusions until a hematocrit of 22% was reached.

There is strong clinical evidence to support a program of perioperative blood product conservation, both because of the complications of transfusions as well as revised guidelines regarding the indications for transfusions. The concern about the number of donor exposures in Group I patients was a major argument in convincing staff members to reevaluate their transfusion practices. In the Western world, the fear of posttransfusion acquired immune deficiency syndrome (AIDS) has probably done more to force physicians to reevaluate and redefine safe limits in transfusion practice than have all the educational programs and research in the past decade, combined [6]. In a recent prospective study of 4163 cardiac surgical patients who received 36,282 HIV-screened components, there was one HIV seroconversion [7]. The first cases of transfusion-associated AIDS in Saudi Arabia were reported in two patients who had received blood transfusions in 1981 [8].

Postransfusion hepatitis (PTH) represents a greater viral risk from blood product transfusion. In one study [9], the incidence of PTH was 7.7% when diagnosed by elevation of serum ALT, but was only 0.07% (1% as many patients) when the diagnosis was based on clinical symptoms. Chronic hepatitis, cirrhosis, and death will occur in an even smaller percentage of patients with clinical signs of PTH. There is a high incidence of hepatitis B in Saudi Arabia, but the incidence of PTH is not known and was not examined in our study.

Other hazards from transfusion include cytomegalovirus in immunocompromised patients [10], noncardiogenic pulmonary edema in post-bypass cardiac surgical patients [11], and posttransfusion myelopathy [12].

Hemorrhage following cardiopulmonary bypass is a dangerous complication with many causes, including lack of surgical hemostasis and inadequate reversal of heparinization. The alterations in coagulation factors related to open heart surgery and cardiopulmonary bypass have been well studied. Immediately after the start of cardiopulmonary bypass there is a 50% reduction in circulating levels of platelets, fibrinogen, and factors II, VII, IX, and X due to hemodilution by the nonblood prime in the bypass machine. Factor V falls to less than 50% while factor VIII does not fall. None of these components are reduced below hemostatic levels [13].

Until recently, it was acceptable to administer blood products empirically to cardiac surgical patients to prevent or treat postoperative bleeding. The indications and limitations for the perioperative administration of blood components in surgical patients are now better defined [14–16]. The use of fewer red cell transfusions in cardiac surgery is supported by the lack of morbidity in postoperative patients who tolerate hematocrits of 25 to 30% [1,17,18]. There is considerable experience with Jehovah's witnesses who receive no blood during cardiac surgery and tolerate hemoglobin levels as low as 50 gm/L without suffering increased morbidity [19]. The safe limits for normovolemic anemia in cardiac patients have not been defined, but there is a suggestion that this may be poorly tolerated prior to anesthesia and surgery in patients with coronary artery disease [20].

No benefit has been demonstrated for prophylactic administration of FFP in cardiac surgical patients [21,22]. In fact, FFP administration may be detrimental in patients with inadequate protamine reversal of heparin and may lead to reheparinization through the re-formation of heparin–ATIII complexes [3,23].

Prophylactic platelet administration has also been shown to be of no benefit in reducing transfusion requirements in cardiac surgical patients [24]. Disorders of platelet activation and aggregation occur during cardiopulmonary bypass. In uncomplicated cases, platelet function becomes normal within one hour of the completion of bypass [13].

We recognize that our approach to improved blood conservation is a matter of educating many surgeons, residents, nurses, and anesthesiologists. This educational process will take more time before it is fully accepted by all members of the team [25,26]. In addition, all facets of the regimen of blood product conservation were not or could not be applied to all patients in Group II. Not all Group II patients underwent postoperative autotransfusion of mediastinal drainage using a cardiotomy reservoir, and it was not always possible to adhere strictly to guidelines for administration of clotting factors in patients who were actively bleeding in the ICU. Achieving normovolemic anemia has been the most difficult part of our conservation program, as shown by the fact that 70% of Group II patients had a hemoglobin level greater than 100 gm/L upon discharge from the ICU. There is certainly room for improvement in our transfusion practices for these patients, but the results reported here are still encouraging because they have led to significantly reduced exposure of these patients to unwanted complications. Future efforts will be exerted to assure more complete acceptance of our established guidelines and compliance with the principles of reducing usage of blood bank products. In addition, the use of other modalities which reduce postoperative bleeding in cardiac surgical patients, such as desmopressin and aprotinin, will be encouraged [27,28].

Acknowledgments

We thank Dr. Carlos Duran, Chairman, Department of Cardiovascular Diseases, for his guidance, and Dr. Stener Bernvil and Mr. Joseph Schifano, Blood Bank, Department of Pathology and Laboratory Medicine, for their kind assistance with this project and paper.

References