Requirement in Coronary Bypass Surgery

Blood Loss and Bank Blood Requirement in Coronary Bypass Surgery Thomas Yeh, Jr., Larry Shelton, C.C.P., and Thomas J. Yeh, M.D. ABSTRACT With the use of nonblood prime and refinement in perfusion and surgical techniques, blood requirement for coronary bypass operations has been reduced to a minimum. Of 240 patients (average number of bypasses, 3.07; average pump time, two hours and 22 minutes), no blood was used in pump prime or before perfusion. During perfusion, 29 patients (12%) received 34 units of blood in the pump-oxygenator, and after bypass 64 patients (27%) received 65 units of blood in the operating room (average intraoperative use, 203 ml per patient). For the total hospital stay, the blood requirement was 728 ml per patient. For the last 60 patients operated on, the figure was 328 ml. There were no surgical deaths, and only 1 reexploration for postoperative hemorrhage (0.4%). Discharge hemoglobin level averaged 11.8 gm, whereas the admission hemoglobin level had averaged 13.8 gm. Autotransfusion, avoidance of entry into the pleural space, shorter perfusion time, postoperative platelet count of more than 150,000, and normal partial thromboplastin time tend to reduce blood requirement, but not to a striking degree. Bank blood requirement for the coronary bypass program accounted for 3.7% of the hospital need and 2% of the community need. It was estimated that approximately 70,000 patients would undergo coronary bypass operations in the United States in 1977 [2], and the number increases every year. The bank blood requirement for these procedures can conceivably strain the capacity of the local blood bank resources [ll]. This study was undertaken to determine the bank blood requirement of the coronary bypass surgery program in Savannah, GA, to measure its impact on the local blood bank system, and to find the effects of various factors From the Department of Surgery, Memorial Medical Center, Savannah, GA. Presented at the Twenty-fourth Annual Meeting of the Southern Thoracic Surgical Association, Nov 3-5, 1977, Marco Island, FL. Address reprint requests to Dr. Thomas J. Yeh, 5112 Paulsen St, Savannah, GA 31405. upon the blood loss and replacement in coronary bypass operations. Material and Methods Two hundred forty consecutive coronary bypass operations done by one surgeon since May 15, 1975, were analyzed. Excluded from this series were reoperations and those patients in whom ventricular aneurysmectomy, prosthetic valve replacement, or both were done concomitantly. There were 206 men and 34 women (male-tofemale ratio, 6:l). Ages ranged from 27 to 75 years (average 56.5 years). Weight ranged from 53 to 140 kg (average, 79.1 kg). The saphenous vein was harvested from the ankle up toward the knee. Ascending aortic cannulation was used in all instances. Venous drainage was by a single catheter in the right atrium. The left ventricle was vented through the apex. The patient was heparinized (3 mglkg body weight) intravenously just prior to cannulation. A disposable bubble oxygenator was primed with Ringer s lactate with serum albumin, sodium bicarbonate, potassium chloride, and methylprednisolone sodium succinate added. Total prime volume was 1,500 to 2,000 ml. Throughout the perfusion period, heparin was added to the perfusate at a rate of 0.5 mg per kilogram of body weight per hour. Following completion of cardiopulmonary bypass, protamine was given in a dosage equal to 80 to 90% of the initial heparin dose. Heparin added to the perfusate during perfusion was not neutralized. Clotting time or activated partial thromboplastin time (PTT) was not determined as a guide to protamine dosage. Postoperative chest drainage was recorded daily for three days or until the chest tubes were removed. Blood transfusion was given to replace excessive loss or to stabilize hemodynamics and to keep the hemoglobin level at 11 gm or more. Fresh-frozen plasma and platelet concentrates were rarely given (less than 5% of patients). Epsilon-aminocaproic acid was not used in any patient [6]. Blood transfusion 11 0003-497517810026-0104$01.25 @ 1978 by Thomas Yeh, Jr,

12 The Annals of Thoracic Surgery Vol 26 No 1 July 1978 records were meticulously kept. Hemoglobin and hematocrit values were obtained on admission, on arrival at the intensive care unit, on the first and second postoperative days, and if needed any time thereafter. The platelet count and PTT were obtained on admission and on arrival at the intensive care unit in 80 patients operated on during 1977. The blood loss and replacement were analyzed and were correlated to various factors such as the number of grafts, pump time, autotransfusion, method of pericardial drainage, postoperative platelet count, and PTT. Results The average number of bypasses per patient was 3.07. The total perfusion time ranged from 49 to 300 minutes (average, 142 minutes). The average admission hemoglobin was 13.8 gm, and the average discharge hemoglobin was 11.8 gm. There was no operative or hospital death in this series, The average hospital stay was eight days. One patient required a thoracotomy for hemorrhage, and 1 patient had posttransfusion hepatitis. lntraoperative Blood Requirement Blood replacement was not required in any patient before going on bypass, even for the patients from whom l unit of blood was withdrawn for later autotransfusion. No blood was used in priming the pump. In 29 patients (12%) bank blood was added in the pump-oxygenator during prolonged perfusion to maintain the reservoir level. In 64 patients (27%) bank blood was given after perfusion to stabilize the circulation and to elevate excessively low hemoglobin levels. One hundred fifty-six patients (659'0) re- ceived no blood at all in the operating room. Thus the total intraoperative bank blood requirement for 240 patients was 97 units, an average of 203 ml per patient. Postoperative Blood Loss and Requirement The chest drainage averaged 428 ml for the first twenty-four hours and 736 ml for the total period. In 9 patients (4%) the drainage exceeded 1,000 ml in the first twenty-four hours and in only 1 patient (0.4%) was it more than 2,000 ml, necessitating reexploration for the bleeding. Fifty-four patients received no blood for the entire hospital stay, and 73 patients received 1 unit each. The maximum amount of blood used was 7 units for the patient who was reexplored. The total blood used for the 240 patients was 350 units, or 728 ml per patient. No patient was discharged with a hemoglobin level under 10 gm. Effect of Number of Bypasses on Blood Loss and Replacement As may be expected, the greater the number of bypass grafts, the longer the pump time and the greater the blood requirement intraoperatively and postoperatively (Table 1). Effect of Autotransfusion In 116 patients, 1 unit of blood was collected from the arterial pressure monitoring line in a citrate-phosphate-dextrose bag before heparinization. This blood was returned to the patient after the completion of cardiopulmonary bypass and after the administration of protamine. The autotransfusion had no statistically demonstrable effect upon postoperative platelet count, PTT, or blood loss (Table 2). The blood Table 1. Pump Time, Blood Loss, and Blood Replacement According to the Number of Bypasses No. of Bypasses Variable Single Double Triple Quadruple Quintuple (N = 19) (N = 52) (N = 99) (N = 53) (N = 17) Pump time (min) 72 110 145 175 186 Total blood loss (ml) 636 695 720 870 839 Total blood use (ml) 534 510 716 958 1,038 Intraoperative blood use (ml) 158 311 336 407 441

~~~ 13 Yeh, Shelton, and Yeh: Blood Loss in Coronary Bypass Surgery Table 2. Baseline Variables, Blood Loss, and Replacement Data in 116 Autotransfusion and 124 Control Patients Variable Autotransfusion Control Average age (yr) 54 58a Average weight (kg) 82 77b No. of bypasses 3.07 3.06 Pump time (min) 143 141 Preoperative hemoglobin (gm1100 ml) 14.4 13.3b Discharge hemoglobin (gm1100 ml) 11.9 11.6 Postoperative platelet count 116,000 124,000 Postoperative PTT (sec) 28.7127.2 26.8125.5 Total blood loss (ml) 740 732 Total blood used (ml) 627 829b Blood in pump (ml) 34 104b Transfusion in operating room after perfusion (ml) 77 189b Intraoperative blood use 111 293b "Difference significant to < 0.05. bdifference significant to < 0.01. requirement was reduced by an average of 200 ml for the entire hospital stay. This reduction is completely accounted for by decreased intraoperative use (70 ml less added to the pump and 112 ml less given after perfusion). Effect of Venting the Pericardium into the Pleural Space In the first patients seen in the series, the pericardium was vented widely into the right pleural space for prevention of cardiac tamponade and possible reduction in the incidence of postpericardiotomy syndrome. A random study was done of the last 80 patients in the series; 40 patients had the pleural space opened and the remaining 40 served as controls. In the group in whom the pericardial space was not vented into the pleural space, the total blood loss was less (498 vs 641 ml) and blood replacement was also less than in the control group (280 vs 410 ml) (Table 3). Effect of Experience of the Team The total of 240 patients was divided into four groups of 60 patients each in chronological sequences. Although the average number of bypasses done per patient in each period did not differ greatly, the pump time decreased slightly and blood losses and replacement decreased noticeably (Table 4). In Period IV, the blood loss Table 3. Baseline Variables, Blood Loss, and Replacement Data as lnfluenced by Entry into the Pleural Space Variable Age (yr) 59 Weight (kg) 79.8 No. of bypasses 2.8 Pump time (min) 120 Preoperative hemoglobin 13.9 (gm1100 ml) Discharge hemoglobin 11.9 (gm1100 ml) Total blood loss (ml) 641 Total blood use (ml) 410 adifference significant to < 0.05. bdifference significant to < 0.01. Pleural Space Pleura Entered Intact (N = 40) (N = 40) 56 79.0 3.2a 140b 14.0 11.9 498a 280a was reduced to 50% and replacement to 33% of those values in Period I. Sixty percent of the patients in Period IV required no bank blood at all. Effect of Postoperative Platelet Count In the last 80 patients seen, platelet counts were done routinely. The postoperative platelet count averaged 120,000 (52,000 to 278,000) compared with the preoperative level of 228,000 (101,000 to

14 The Annals of Thoracic Surgery Vol 26 No 1 July 1978 Table 4. Baseline Variables, Blood Loss, and Replacement Data According to Periods When Operations Were Done Period I I1 111 IV (5115175 to (3104176 to (9124176 to (3114177 to Variable 2127176) 9/21/76] 3111177) 8122177) No. of patients 60 60 60 60 No. of bypasses 3.15 2.8 2.9 3.1 Pump time (min) 155 143 133 135 Total blood loss (ml) 1,009 677 772 544 Total blood use (ml) 1,001 776 803 32ga Blood in pump (ml) 200 58 8 17 Transfusion in operating 183 150 125 125 room after perfusion (mu Intraoperative blood use (ml) 383 208 133 142a adifference from Period I significant to < 0.01. bdifference from Period I significant to < 0.05. Table 5. Blood Requirement for Coronary Bypass Surgery in Relation to Hospital and Community Requirements Institution No. of Cross-matches (units) No. of Units Used Savannah, GA (3 Data 11,568 hospitals), 1976 unavailable Memorial Medical Center, 14,750 6,203 1976 Coronary bypass surgery 840 232 program, 1976 (140 patients) Coronary bypass surgery 320 55 program, first half of 1977 (80 patients) 393,000). The blood loss did not differ significantly according to the postoperative platelet counts. The correlation between the blood loss and platelet count was poor, and some patients with a platelet count in the range of 50,000 to 70,000 had a blood loss of only 250 ml during the total postoperative period and required no blood transfusion, platelet concentrates, or fresh-frozen plasma. If the platelet count was higher than 150,000 postoperatively, the blood requirement was definitely less (178 vs 428 ml, p < 0.01) even though blood loss was not signifi- cantly different from that in other patients (523 vs 610 ml, p > 0.05). PTT Of 134 patients in whom postoperative PTT was determined, 111 had normal or slightly elevated PTT (up to 9 seconds over control) and 23 patients had markedly prolonged PTT (10 to 60 seconds over control). The average blood loss was approximately 200 ml greater in the latter group, but the difference was not statistically significant. The blood requirement for the latter group was significantly greater than for the former (853 vs 559 ml, p < 0.01). Blood Requirement in Coronary Bypass Operations in Relation to Total Hospital and Community Need In 1976,140 coronary bypasses were done at this center. Six units of blood were requested for each patient, or a total of 840 units for crossmatching. Of these, 232 units were actually used (Table 5). This usage amounted to 5.7% of total cross-matches and 3.7% of total blood given at the hospital, or 2% of all blood used in the city of Savannah, GA. Based on this figure, since January, 1977, only 4 units of blood were crossmatched and only 55 units of blood were used, an outcome indicating that further reduction in cross-matching requests is in order.

15 Yeh, Shelton, and Yeh: Blood Loss in Coronary Bypass Surgery Comment Cardiac operation with the use of a pumpoxygenator has traditionally been associated with a requirement for a large amount of donor blood [4, 81. In the early days of open-heart surgery, it was not unusual to have 12 to 20 units of donor blood on hand for each patient. The use of nonblood prime [l, 3,7,9] and the development of more efficient and less traumatic oxygenators with smaller priming volume have reduced the adverse effect of extracorporeal circulation. The refinement of both the cardiopulmonary bypass technique and surgical technique also have reduced blood loss and requirement. Conservation of all blood spilled in the operative field after heparinization, avoidance of unnecessary blood trauma, avoidance of discarding cardiotomy blood needlessly or allowing it to mix with irrigating fluid, and, finally, returning all blood-fluid mixture in the extracorporeal circuit to the patient, with occasional use of diuretic to eliminate fluid overload, are some of the helpful measures that have been adopted. The economic benefit of reduction of blood transfusion is obvious. More important, however, is the reduction of transfusion reaction, pulmonary complications, and incidence of posttransfusion hepatitis [81. Paradoxical though it may seem, the use of a higher initial dose of heparin and the addition of heparin in the system during perfusion have reduced blood loss and requirement. By fully heparinizing the blood, it is possible to prevent so-called intravascular coagulation in the extracorporeal circuit with depletion of various clotting factors. The dosage of protamine, on the other hand, has been reduced. With a protamine dosage equivalent to the total amount of heparin used (initial amount + dosage added during perfusion), we have noticed that the relatively dry operative fields after perfusion often become bloody after the administration of protamine. Because of this, we administered protamine in increments of 50 to 100 mg, and when visible clots began to appear in the wound, we judged the dosage to be adequate. Based on this experience, we now administer about 80 to 85 mg of protamine per 100 mg of heparin used initially and disregard the heparin given subsequently in the pump-oxygenator. We have not used any laboratory test to modify protamine dosage in the past eight years. The platelet count and PTT were determined in some patients for the purpose of clinical investigation. The postoperative platelet count was invariably greatly reduced from the preoperative count, but the effect of the postoperative platelet count on blood loss was inconsequential. In fact, no statistically significant difference could be detected at all. The postoperative PTT level likewise correlated poorly with the postoperative blood loss. The blood loss and replacement predictably rose with the increased number of bypass grafts inserted. The differences, however, were not all that great. The effect of autotransfusion on blood loss and replacement as reported in the literature varies [5,10,12]. Our studies on this subject are inconclusive. There was no difference in the blood loss between autotransfusion and control groups. Blood replacement was moderately lower (627 vs 829 ml, p < 0.01). This seemingly contradictory result may be due to the fact that preoperative hemoglobin was about 1 gm higher in the autotransfusion group compared with the control group, a situation enabling the patients in the former group to withstand the same amount of loss with less replacement. It is interesting to note that postoperative platelet count and PTT were not significantly different between the two groups. One unit of autotransfusion is equivalent to 1 unit of platelet concentrate and 1 unit of fresh-frozen plasma, a quantity hardly sufficient as a therapeutic dose for any coagulation anomalies. The statistical analysis confirmed the clinical impression that blood loss and replacement can be reduced by keeping pleural spaces intact. The difference, however, again was rather modest. The most significant factor influencing blood loss and replacement was the increasing experience of the team. Thus for the last 60 patients of the series, the blood loss and replacement were only 50% and 33% of the figures for the first 60 patients seen. It is reassuring to know that in our hospital the coronary bypass program accounts for no more than 2% of total community bank requirement. The entire open-heart surgery program probably will not account for 4%. Compared with the expected 10% waste caused by outdating of

16 The Annals of Thoracic Surgery Vol 26 No 1 July 1978 bank blood in the course of normal operation, the impact of the coronary surgery program upon local blood bank resources can be said to be negligible. References 1. Bailey CP, Hirose T, Gollub S, et al: Open heart surgery without blood transfusion. Vasc Dis 5:179, 1968 2. Braunwald E: Coronary artery surgery at the crossroads. N Engl J Med 297:661, 1977 3. Cooley DA, Crawford ES, Howell JF, et al: Open heart surgery in Jehovah's Witnesses. Am J Cardiol 13:779, 1964 4. Heimbecker RO: Progress in extracorporeal circulation. J Thorac Cardiovasc Surg 74:157, 1977 5. Kaplan JA, Vannarella C, Jones EL, et al: Autologous blood transfusion during cardiac surgery: a re-evaluation of three methods. J Thorac Cardiovasc Surg 74:4, 1977 6. Kevy 5V, Glickman RM, Bernhard WF, et al: The pathogenesis and control of the hemorrhagic defect in open heart surgery. 5urg Gynecol Obstet 123:313, 1966 7. Litwak R5, Jurado RA, Lukban 5B, et al: Perfusion without donor blood. J Thorac Cardiovasc 5urg 64:714, 1972 8. National Transfusion Hepatitis Study: Risk of posttransfusion hepatitis in the United States: a prospective cooperative study. JAMA 220:692, 1972 9. Ott DA, Cooley DA: Cardiovascular surgery in Jehovah's Witnesses: report of 542 operations without blood transfusion. JAMA 238:1256, 1977 10. Pliam MB, McGoon DC, Tarhan S: Failure of transfusion of autologous whole blood to reduce banked-blood requirements in open-heart surgical patients. J Thorac Cardiovasc Surg 70:338,1975 11. Roche JK, Stengle JM: Open-heart surgery and the demand for blood. JAMA 225:1516, 1973 12. Sherman MM, Dobnik DB, Dennis RC, et al: Autologous blood transfusion during cardiopulmonary bypass. Chest 70:592, 1976 Notice from the Southern Thoracic Surgical Association The Twenty-fifth Annual Meeting of the Southern Thoracic Surgical Association will be held at the Marco Beach Hotel, Marco Island, FL, on November 2-4,1978. There will be a $100 registration fee for nonmember physicians except for guest speakers, authors and coauthors on the program, and residents. There will be a Postgraduate Program on Difficult Problems in Surgical Management preceding the regular program. This meeting has been approved for Category I CME credit. The Scientific Program will be published by title and authors in the September issue of The Annals of Thoracic Surgery. Manuscripts of pa- pers accepted for the program must be submitted to The Annals of Thoracic Surgery by October 15, 1978. Applications for membership should be completed by September 1, 1978, and forwarded to Ronald F. Galloway, M.D., 1467 Harper St, Augusta, GE 30902. Hotel reservations may be made by writing to the Marco Beach Hotel, 400 South Collier Blvd, Marco Island, FL 33937. (Phone: 813-394-2511). J. Kent Trinkle, M.D. Secreta y-treasurer