Postoperative autologous retransfusion of collected shed blood after total knee arthroplasty with the cell saver

AIM: The purpose of this study was to examine if it is possible to reduce transfusion of blood units by collecting shed blood with the Cell Saver for autologous retransfusion in total knee arthroplasty (TKA). METHOD: In 186 patients drainage blood was collected over a 6-h period after total knee arthroplasty with a Cell Saver system in order to make retransfusions if necessary. A tourniquet was used routinely throughout the operation. No preoperative blood donation was performed. In 19 patients preoperative haemoglobin levels were below 12 g/dL (group A, anaemic patients). In the other 167 patients (group B) the preoperative haemoglobin levels were higher. RESULTS: 4 patients (21 %) in group A received a homologous blood transfusion. Only 1 patient (0.6 %) in group B received one unit of erythrocyte concentrate (difference statistically significant, P < 0.001). In group A 8 patients (42 %) received 284 ml (145-621 ml) Cell-Saver concentrate on average, 38 patients (23 %) in group B received 358 mL (147-776 ml) Cell-Saver concentrate on average. CONCLUSION: With a risk lower than 1 % for patients without anaemia to get a homologous blood transfusion one can do without the more expensive preoperative blood donation in total knee arthroplasty if a tourniquet is used for the operation and a Cell Saver is used for facultative retransfusion of drainage blood.     

The effects of preoperative autologous donations on perioperative blood levels

The purpose of this study was to evaluate the blood levels of patients preparing for total knee arthroplasty (TKA) who were enrolled in a preoperative autologous donation program. The charts and hospital records of 70 consecutive patients who underwent primary unilateral TKA between 2000 and 2002 were retrospectively reviewed. Study participants were instructed to donate one unit of blood approximately 4 weeks prior to surgery. Predonation and preoperative hemoglobin levels were assessed throughout the study and transfusion requirements were recorded. Transfusions were administered only when warranted by clinical symptoms. The mean initial (predonation) hemoglobin concentration was 14.1 g/dL. The mean number of days donations were made prior to surgery was 13 +/- 3.3 days. Prior to surgery, the average hemoglobin concentration dropped to 12.8 g/dL. Fifty (71%) patients had a hemoglobin value > 13.0 g/dL prior to their autologous donation, but only 30 (43%) patients had blood levels > or = 13.0 g/dL following blood donation. Postoperatively, the mean hemoglobin concentration in the recovery room was 11.6 g/dL and dropped to a nadir of 10.8 g/dL on postoperative day 3. Overall, 91% of patients required autologous blood transfusion following TKA but no patients required allogeneic blood transfusions. Preoperative autologous donation was associated with a decrease in preoperative hemoglobin levels and with a high rate of autologous transfusion based on clinical symptoms of postoperative anemia.     

Relevance of wound drainage in total knee arthroplasty--a prospective comparative study

AIM: The benefits of postoperative wound drainage in patients with total knee arthroplasty (TKA) with regards to mobilisation and wound healing were studied. We wanted to determine the efficacy of an autologous blood retransfusion system. METHOD: 150 patients with TKA were divided into three groups of 50 patients: A) three wound drainages with an autotransfusion system and suction; B) no wound drainage; C) one intraarticular wound drainage without suction. Hemoglobin values, blood transfusion requirements, blood loss, postoperative range of motion, Insall knee score and rate of complications were observed and recorded. All patients were operated without tourniquets for lower blood loss. RESULTS: In the group of patients with wound drainage and a retransfusion system the requirement of postoperative additional blood transfusion was not significantly less than in the group without wound drainage. Group A had the highest blood loss of all. The group without wound drainage had more hematomas and wound healing complications. Best results were observed within the group with one intraarticular drainage without suction. The rate of complications was not increased and the blood transfusion requirements were the lowest. CONCLUSION: This study shows that total knee replacement involving one intraarticular wound drainage without suction attains the best results.     

Blood loss and transfusion rates in bilateral total knee arthroplasty

From 1994 to 1998, we performed 170 bilateral total knee arthroplasties (TKAs) with cemented, posterior cruciate-substituting prostheses. Blood management included preoperative autologous donation, symptom-based transfusion, and autoreinfusion devices. Perioperative allogeneic transfusion rates for patients who donated 0, 1, 2, 3, or 4 units of blood were 40.00%, 0.00%, 3.70%, 0.00%, and 3.23%, respectively. Preoperative autologous donation >2 units also resulted in lower preoperative hemoglobin levels. For bilateral TKA, a protocol of 2 preoperative autologous donation units and reinfusion of postoperative drainage reduces anemia during the preoperative and postoperative periods.     

Blood loss management in high-risk patients undergoing total knee arthroplasty: a comparison of two techniques

Although it is well known that patients with preoperative hemoglobin levels <13.0 g/dL are at a higher risk for requiring postoperative transfusions, the ideal blood management strategy for this group of patients remains unclear. This study compared preoperative autologous donation with preoperative administration of epoetin alfa as a method to maximize perioperative hemoglobin levels and minimize blood transfusions in these high-risk patients undergoing total knee arthroplasty (TKA). Results show that both preoperative autologous donation and epoetin alfa were successful in decreasing the need for allogeneic blood transfusions following TKA in high-risk patients. Epoetin alfa was more effective in maximizing perioperative hemoglobin levels.     

Who needs autologous blood donation in joint replacement?

Conflicting results are available about the efficacy of routine preoperative autologous donation (PAD) in reducing allogenic blood transfusion during total joint arthroplasty (TJA). This study aimed to determine if PAD is effective in reducing the need for allogenic transfusion after TJA. For this retrospective study, data on 409 patients who received total knee arthroplasty (TKA) and 513 who underwent total hip arthroplasty (THA) from January to June 2005 were evaluated. It is our institutional policy to offer preoperative donation to all patients. Based on patient decision, preoperative hemoglobin, and other factors, PAD may or may not take place. Five hundred forty-six (61%) patients donated on average 1.3 units (range, 1 to 2) of blood. Autologous and allogenic transfusions were respectively performed in 91% (514 patients) and 24% of our cohort. The rate of allogenic transfusion after TKA was lower in the PAD group at 21% versus 27% among the nondonors, although it was not statistically significant (p = 0.10). The allogenic transfusion rate after THA was significantly lower among autologous donors (16% versus 34%, p = 0.003). Advanced age, lower body mass index, simultaneous bilateral arthroplasty, and lower preoperative hemoglobin were independently associated with increased allogenic blood transfusion. PAD seems to be effective in reducing allogenic transfusions after THA but not TKA.     

Postoperative autotransfusion in primary knee replacement surgery

OBJECTIVE: To assess the efficacy of postoperative autologous transfusion to reduce homologous blood transfusion needs in primary knee replacement surgery. PATIENTS AND METHODS: A prospective study was carried out in 33 consecutive patients with diagnoses of arthrosis scheduled for primary knee replacement surgery with postoperative autotransfusion using a CBCII Constavac-Stryker (Stryker Instruments, Michigan, USA) recovery system from June through October 2002. We analyzed patient age, sex, preoperative and postoperative (24 hours) hemoglobin and hematocrit values, autologous blood reinfused and homologous blood transfusion incidence rate (if hematocrit was below 25%). RESULTS: Of the 33 patients receiving postoperative autotransfusion, one also needed homologous blood transfusion (3%). The mean volume of filtered whole blood reinfused was 538.63+/-261.23 mL, 1100 mL being the largest volume reinfused. We observed no complications related to use of autotransfusion devices during the perioperative period. CONCLUSIONS: Postoperative autotransfusion as the only blood salvage technique in primary knee prosthesis surgery nearly eliminates homologous transfusion needs. In addition, it is a safe, simple procedure and has replaced our hospital's preoperative autologous transfusion procedure.     

Preoperative use of recombinant human erythropoietin before total joint arthroplasty

BACKGROUND: Previous reports have suggested that the use of recombinant human erythropoietin is effective for decreasing the need for perioperative allogeneic blood transfusion. The purpose of this study was to evaluate the efficacy of erythropoietin in combination with, and compared with, preoperative autologous donation for reducing allogeneic blood requirements for total joint arthroplasty. METHODS: Two hundred and forty patients undergoing primary and revision total hip or knee arthroplasty were enrolled into three groups with different treatment regimens: (1) erythropoietin and preoperative autologous donation (Group 1), (2) erythropoietin alone (Group 2), and (3) preoperative autologous donation alone (Group 3). Patients were evaluated with regard to requirements for allogeneic transfusion, change from the baseline to the lowest postoperative hemoglobin value, postoperative complications, and adverse reactions. RESULTS: The rate of allogeneic transfusion was 11% in Group 1 (erythropoietin and preoperative autologous donation) compared with 28% in Group 2 (erythropoietin alone) and 33% in Group 3 (preoperative autologous donation alone). Within Group 1, patients who had a unilateral primary arthroplasty had an allogeneic transfusion rate of 4% and those who had a bilateral or revision arthroplasty had an allogeneic transfusion rate of 17%. In Groups 2 and 3, the allogeneic transfusion rates were 14% and 15%, respectively, for the patients who had a unilateral primary arthroplasty and 35% and 47%, respectively, for those who had a bilateral or revision arthroplasty. CONCLUSIONS: Preoperative use of erythropoietin in conjunction with preoperative autologous donation reduces the need for allogeneic blood transfusion associated with total joint arthroplasty more effectively than does either erythropoietin or preoperative autologous donation alone.     

Perioperative blood salvage as an alternative to predonating blood for primary total knee and hip arthroplasty

A total of 200 consecutive patients who underwent primary total knee or hip arthroplasty were reviewed to assess the efficacy of perioperative blood salvage and retransfusion. Five of 132 (3.8%) patients undergoing total knee arthroplasty and 3 of 68 (4.4%) patients undergoing total hip arthroplasty required allogeneic transfusion in addition to retransfusion of salvaged autologous blood. The risk of receiving allogeneic transfusion in addition to retransfusion of salvaged blood was 1.2% (2 of 173) in patients with a preoperative hematocrit of > or=37%. The risk of requiring allogeneic transfusion was 22% (6 of 27) in patients with a preoperative hematocrit of 37%.     

Orthopaedic-induced anemia: the fallacy of autologous donation programs

Total knee arthroplasty is associated with significant blood loss. Despite the initiation of various blood conservation modalities, allogeneic transfusion has yet to be eliminated. One hundred forty-eight patients who had unilateral primary total knee arthroplasties during a 3-year period were evaluated retrospectively for blood loss and transfusion rates. The patients were prescribed one unit of preoperative autologous donation that was to be transfused automatically on postoperative Day 1. Allogeneic transfusion was based on symptoms, and no numerical transfusion triggers were used. The preoperative autologous donation program resulted in increased preoperative anemia. Whereas only 26.2% of patients were in the high transfusion-risk group (hemoglobin >10 g/dL and < or = 13 g/dL) before surgery, 55.7% of patients were in this high-risk category after preoperative autologous donation. The patients did not recover from the autologous donations that occurred 4 weeks before surgery. A mean hemoglobin level of 14.0 g/dL was seen before donation, whereas the mean preoperative hemoglobin level decreased to 12.6 g/dL. We think that a preoperative autologous donation program leads to an increased risk of anemia before surgery.     

Efficiency of Autologous Blood Donation in Combination with a Cell Saver in Bilateral Total Knee Arthroplasty

The increased blood loss and resulting need for allogenic blood has been a major concern of one-stage bilateral total knee arthroplasty (TKA). One hundred eighteen consecutive patients donating either 2 units (87 patients) or 3 units (31 patients) of autologous blood prior to one-stage bilateral TKA were retrospectively evaluated to determine: (1) how many patients received allogenic transfusion; (2) what percentage of autologous blood was wasted; and (3) whether donating 2 or 3 units of autologous blood before surgery is more cost-effective. Fifteen patients in the 2-units donation group (17.2%) and one patient in the 3-units donation group (3.2%) required allogenic blood transfusions. In the 2-units group, 37.9% of the patients wasted 21.8% of predonated autologous blood, and in the 3-units group, 64.5% of the patients wasted 32.3% of predonated autologous blood. The estimated cost for patients donating 2 or 3 units of blood was $1,814.17 and $1,996.10, respectively. Donating 2 units of autologous blood is more cost-effective; however, patients donating 3 units of blood required less allogenic blood. Ethical Board Review statement: Institutional Review Board Approval IRB 25053 Level of evidence: level III: retrospective, case-controlled study     

Safety and efficacy of autologous blood donation before elective off-pump coronary artery bypass grafting

BACKGROUND: Preoperative autologous blood donation reduces exposure to homologous blood transfusions in cardiac surgery. The purpose of this study was to ascertain, how much volume of predonated autologous blood needed to avoid of homologous blood transfusion in scheduled off-pump coronary artery bypass grafting (OPCAB). METHOD: Fifty patients underwent scheduled OPCAB. These patients donated 400 ml (group A, n = 30) or 800 ml (group B, n = 20) of autologous blood before operation. These patients donated at a rate of 400 ml per week. All patients were given an equal volume of saline solution at the time of autologous donation. RESULT: There were no significant differences mean age, mean body weight, mean preoperative hematocrit values, mean graft number or mean volume of intraoperative blood loss between groups A and B. There was significant difference the mean postoperative day-7 hematocrit value (33.4 +/- 1.5% vs 38.7 +/- 1.5%, p < 0.05). The rates of avoiding homologous blood transfusion were 63.3% in group A and 100% in group B (p < 0.05). CONCLUSIONS: Autologous blood transfusion was effective for reducing the homologous blood requirement. We believe that 800 ml predonation is sufficient to avoid homologous blood transfusion in scheduled OPCAB, further patients with cardiovascular disease including severe coronary artery should be donated with the administration of saline.     

Autologous blood transfusion after total knee arthroplasty: a randomized,prospective study comparing predonated and postoperative salvage blood

A randomized, prospective study of the use of allogeneic blood was performed in a consecutive series of patients who underwent primary total knee arthroplasty (TKA) and had autologous transfusion either from one unit of predonated autologous blood or from postoperative unwashed blood salvage. In this study, 83 patients (88 knees) were included, with 47 knees in the salvage group and 41 in the predonation group. There were no differences between groups in average age, height, and weight, or gender, diagnoses, or anesthesia type. No significant difference was seen between the groups in the prevalence of allogeneic blood transfusion (5% for the predonation group and 0% for salvage group). Postoperative blood salvage was as effective as predonated autologous blood in preventing the risk associated with allogeneic blood after TKA.     

Blood transfusion strategies for total knee arthroplasty: Minimizing autologous blood wastage,risk of homologous blood transfusion,and transfusion cost

In this nonrandomized study, alternative strategies were suggested to 10 orthopaedic surgeons to minimize autologous blood wastage, the risk of homologous blood transfusion, and cost associated with blood product usage after total knee arthroplasty (TKA). One hundred fifty-five patients with 177 consecutive TKAs over a 2-year period were studied. Group 1 patients had undergone unilateral TKA and did not predonate; 1A patients (n = 19) were drained with a Hemovac, and 1B patients (n = 28) with a postoperative blood recovery system. Group 2 patients (n = 47) predonated one packed red blood cell (pRBC) unit. Group 3 patients (n = 20) predonated 2 pRBC units. Group 4 patients had undergone bilateral sequential TKAs (n = 21) and had predonated 2 pRBC units. Group 5 patients (n = 14) had undergone revision TKA procedures and their blood requirements were individualized. Group 6 patients (n = 6) had preexisting anemia and were excluded from the study. There was no significant difference in total blood loss (909 mL) between groups. Female sex was associated with significantly lower admission hematocrit. Homologous blood was required for 4% of patients in the entire study and the percentage was not statistically different between groups. Twenty-five percent of patients who predonated autologous pRBCs did not use all or some of it. In group 1, the postoperative blood recovery system had a significant effect on reducing postoperative hematocrit drop (P = .0001), but it was not a significant factor if autologous pRBCs were available. The costs associated with group 1A were significantly less (P = .0001) compared with the other groups; group 1A had the highest admission hematocrit (43.2). Transfusion with autologous pRBCs was related to lower admission hematocrit rather than to increased postoperative blood loss. An algorithm is presented to provide cost-effective management of blood products after TKA.     

Blood management for patients with hemoglobin level lower than 130 g/l in total knee arthroplasty

Introduction The risk of requiring allogeneic blood transfusion after total knee arthroplasty has been recently decreased with several methods such as blood donation, blood salvage, and hematinic. For patients with a low baseline hemoglobin level, however, the rate of allogeneic transfusion is still high, and an effective method for avoiding allogeneic blood transfusion has not been established. We introduced intra- and postoperative blood salvage with the Cell Saver for patients with a baseline hemoglobin level lower than 130 g/l and analyzed the frequency of allogeneic blood transfusion.Materials and methods From 1993 to 1997, 218 consecutive total knee arthroplasties were performed, and 155 knees with baseline hemoglobin lower than 130 g/l were included in this study. Baseline hemoglobin ranged from 62 to 129 g/l, with a mean of 110 g/l. All patients were managed with intra- and postoperative blood salvage with the Cell Saver. Preoperative autologous blood donation and/or use of hematinic was performed for 129 knees (group 1) and was not performed for 26 knees (group 2). There were no significant differences between the two groups with regard to preoperative factors. To examine the role of perioperative factors (age, sex, baseline hemoglobin level, revision procedure, preoperative blood donation, and use of hematinic) in determining the requirements for allogeneic transfusion, backward elimination logistic regression analysis was used.Results Seven knees (4.5%) required allogeneic transfusion. Group 1 (2.3%) exhibited a lower rate of allogeneic blood transfusion than group 2 (15.4%) (p=0.016). Hemoglobin levels on the day (p=0.016), 1 week (p=0.0001), and 2 weeks (p=0.007) after surgery were lower in group 1 than in group 2. Backward elimination logistic regression analysis showed that preoperative blood donation (p=0.048) and use of hematinic (p=0.040) were significantly associated with a requirement for allogeneic blood transfusion.Conclusion Preoperative blood donation and use of hematinic were associated with a low incidence of allogeneic blood transfusion after total knee arthloplasty with intra- and postoperative blood salvage, even for patients with a baseline hemoglobin level below 130 g/l.     

Disminución de las transfusiones alogénicas por la generalización del uso de recuperadores de sangre postoperatorios en cirugía de prótesis de rodilla

ObjectivesSurgical teams have several tools in order to reduce the need for postoperative allogenic transfusion. Postoperative autotransfusion of unwashed shed blood has become common practice for total knee replacement surgery since 2006 in our hospital. This study was designed to evaluate if this practice has reduced allogenic blood transfusions.Material and methodsA retrospective study comparing two cohorts, group 2004 with patients operated on for total knee replacement during the year 2004, before the use of the retransfusion system, and group 2008, patients operated on in the year 2008, with regular use of the retransfusion system. Gender, preoperative and postoperative haemoglobin levels, total amount of calculated erythrocytes lost, reinfusion of shed blood and allogenic blood transfusion during hospital stay were recorded.ResultsBoth groups were similar as regards gender, preoperative and postoperative hemoglobin levels, and total amount of erythrocytes lost. The proportion of transfused patients was significantly lower in group 2008 versus group 2004 (20.18% versus 42.19%), with a relative risk of being transfused of 0.47 and a NNT of 4.54. P = .0017.ConclusionsIn our hospital the use of postoperative retransfusion systems has reduced the proportion of transfused patients during hospitalization for total knee replacement surgery, although this result cannot be generalized due to the lack of a fixed transfusion trigger.     

Cost Minimization Analysis of Preoperative Erythropoietin vs Autologous and Allogeneic Blood Donation in Total Joint Arthroplasty

Autologous blood donation and erythropoietin (EPO) have been shown to be effective in reducing allogeneic blood transfusion, but the cost-effectiveness of these interventions remains unclear. A cost minimization analysis was performed, comparing the total costs of allogeneic blood transfusion strategy and autologous and allogeneic blood transfusion strategy for 161 primary total hip arthroplasty (THA) and 195 total knee arthroplasty (TKA) patients. An EPO cost minimization model was constructed using a previously published algorithm for blood management after total joint arthroplasty. The least costly strategy was autologous blood donation in combination with allogeneic blood for THA and TKA patients at $856 and $892 per patient, respectively. The most costly strategy was allogeneic only at $1769 and $1352 per THA and TKA patient, respectively. The EPO strategy model predicted costs similar to the autologous and allogeneic. A strategy that combines autologous blood donation with EPO for patients who cannot donate autologous blood may provide the greatest cost savings and minimize allogeneic blood transfusion.     

Acute normovolemic hemodilution for radical prostatectomy: Can it replace preoperative autologous blood transfusion?

BACKGROUND: Although preoperative autologous blood donation (PAD) is accepted as a standard of care for radical prostatectomy, it is costly, time-consuming and has risks associated with blood storage. Acute normovolemic hemodilution (ANH) is reported to be less expensive and to preserve blood components more effectively than PAD. In the present study, the efficacy and safety of these two autologous blood-collection techniques were compared. METHODS: The study included 16 consecutive patients scheduled for radical prostatectomy. The first eight patients underwent conventional preoperative autologous blood donation of 400 mL 1 week before the operation (PAD group) and the second eight patients underwent acute normovolemic hemodilution followed by immediate operation (ANH group). All blood collected was transfused in the perioperative period. Preoperative and postoperative hematocrit levels in these two groups were compared. RESULTS: There were no differences in preoperative hematocrit, time of operation or operative blood loss between the two groups. In the ANH group, 1080 +/- 160 mL of blood were collected. The postoperative hematocrit level did not differ significantly between the groups. No patient in either group received allogeneic blood transfusion or experienced an adverse event directly related to blood transfusion. CONCLUSION: The two blood-conservation strategies resulted in similar postoperative hematologic outcomes. Given its advantages, which include lower cost, lower risk and higher convenience, ANH is one of the procedures that may replace conventional PAD for use in radical prostatectomy.     

The reduction of preoperative autologous blood donation for primary total hip or knee arthroplasty: the effect on subsequent transfusion rates

We conducted this quality assurance observational study to examine the effects of a change in policy regarding preoperative autologous blood donation (PABD) and indications for perioperative blood transfusion in patients undergoing primary total hip or knee arthroplasty. Two successive time periods, each including 182 successive patients treated by the same medical team and with standardized anesthesia, were compared. The first study had the following standard transfusion policy: 3 U of PABD collected (n = 119) and liberal autologous transfusion (AT). The second study introduced a specific indication for PABD, on the basis of estimated red blood cell reserve and a life expectancy of more than 10 years; 2 U of PABD was collected (n = 81), and criteria were identical for AT and allogeneic transfusion. We mainly compared the incidence of AT; allogeneic and overall transfusions; the inclusion, admission, and discharge hematocrit values; and the wastage of PABD units. This novel policy increased the number of untransfused patients by a factor of 10 (5.5% vs 56.6%) (P < 0.0001), decreased the number of PABD patients by 30% with a 2.4-fold reduction in AT (30% vs 80%) (P < 0.0001), and did not change allogeneic requirements (13% vs 15%). Although fewer autologous units were collected (172 vs 426), the wastage was higher in Study 2 (46% vs 12%) (P < 0.0001). We conclude that incorporation of patients' individual factors improves the efficiency of transfusion for total hip and total knee arthroplasty surgeries. IMPLICATIONS: We compared two transfusion policies for primary total hip or knee arthroplasties: first, a standard preoperative autologous donation with a liberal autologous transfusion policy; and second, a more restrictive indication for autologous donation that was based on patients' individual factors, with identical criteria for autologous and allogeneic transfusion. We found that this change of policy reduced autologous donation and transfusion with no increase in allogeneic transfusion.     

One hundred consecutive hepatobiliary resections for biliary hilar malignancy: preoperative blood donation, blood loss, transfusion, and outcome

BACKGROUND: Many reports on blood loss and transfusion requirements during hepatectomy for metastatic liver cancer or hepatocellular carcinoma have been published; however, there are no reports on these issues in hepatectomy for biliary hilar malignancy. The aim of this study was to review our experience with blood loss and perioperative blood requirements in 100 consecutive hepatectomies for biliary hilar malignancy. METHODS: One hundred consecutive hepatectomies with en bloc resection of the caudate lobe and extrahepatic bile duct for hilar malignancies were performed, including 81 perihilar cholangiocarcinomas and 19 advanced gallbladder carcinomas involving the hepatic hilus. Fifty-eight hilar resections were combined with other organ and/or vascular resection. Data on preoperative blood donation, intraoperative blood loss, and perioperative transfusion were collected and analyzed. RESULTS: Preoperative autologous blood donation was possible in 73 patients (3.4 +/- 1.2 U). Intraoperative blood loss was 1850 +/- 1000 mL (range, 677-5900 mL), and it was < 2000 mL in 62 patients. Intraoperatively, only 7 of the 73 patients (10%) who donated blood received transfusion of unheated, homologous blood products (packed red blood cells or fresh frozen plasma), whereas 18 the 23 patients (67%) without donation received homologous transfusions. Only 16 patients received transfusion postoperatively, and overall, 35 patients received unheated homologous blood products. Total serum bilirubin concentrations after hepatectomy in patients receiving autologous blood transfusion only was similar to those in patients who did not receive transfusion. The incidence of postoperative complications was higher in the 35 patients who received perioperative homologous transfusion than in 65 patients who did not (94% vs 52%; P <.0001). The mortality rate (including all deaths) was 3% (myocardial infarction, intra-abdominal bleeding, and liver failure, 1 patient each). CONCLUSIONS: Despite the technical difficulties arising from hepatectomy for biliary hilar malignancy, approximately two thirds of hepatectomies can be performed in an experienced center without perioperative homologous blood transfusion using preoperative blood donation.