CHAPTER 30 A Multimodal Approach to Transfusion Avoidance and Blood Loss Management in Partial Knee Arthroplasty
Introduction
Modern surgical techniques have reduced the amount of blood loss during total knee arthroplasty (TKA) procedures. Despite numerous advances, allogeneic transfusion rates still remain high. Transfusion rates following unilateral TKA range from 4% to 46%, while bilateral TKA results in transfusion rates between 31% and 72%.1,2 Acute postoperative anemia, risks of allogeneic transfusion, and wound complications remain of great concern to the patient and surgeon. Allogeneic transfusions traditionally have been used to ameliorate the occurrence of anemia, but complications following allogeneic transfusion remain. Incorrect blood component transfusion, disease transmission, allergic reactions, fluid overload, transfusion reactions, and immunosuppression from allogeneic transfusions are well described.3–8 Few data currently exist with regard to blood loss and transfusion rates in patients undergoing partial knee arthroplasty (PKA). In available studies regarding unicompartmental arthroplasty, average blood loss ranges from less than 200 to 240 ml, transfusion rates appear low, and postoperative hemoglobin drop ranges from 1.8 to 2.73 g/dl.9–12 While patient risks for transfusion are lower with PKA, our practice follows a similar protocol with TKA and PKA with regard to minimizing perioperative blood loss and transfusion avoidance. The protocol focuses on a multimodal approach. This approach needs to be utilized in the preoperative, intraoperative, and postoperative periods to optimize the patient prior to surgery and minimize blood loss during and following arthroplasty. Conservation techniques for the preoperative, intraoperative, and postoperative periods are reviewed, and blood management protocols are presented.
Why We Should Try To Avoid Transfusions
Historically, 50% of patients undergoing total joint arthroplasty receive an allogeneic blood transfusion. The risks of allogeneic transfusions are well described in the literature for TKA, but data following PKA are sparse.9,13–19 The main risks for direct transfusion-related morbidity and mortality include incorrect blood component transfusion (70%) and immunologic risks (28%), while transfusion-transmissible infection (2%) is of less significance.20 With regard to transmissible infection, West Nile, TT, parvovirus B19, and SEN virus transmission has been discussed as a potential source of infection after allogeneic transfusions.20,21 On a global scale, protozoa infections, including malaria and toxoplasmosis, remain some of the most common transfusion-transmitted infections.20 In addition, some cases of the blood-borne transmission of prions have been described.20
Perhaps more concerning to the surgeon, allogeneic transfusions have been implicated in the increased incidence of infection after surgery. Tang et al.22 performed a prospective series of 2809 consecutive colon resections. In this report, transfusion was the single most powerful risk factor for postoperative infection, with an odds ratio greater than 5. Kendall et al.23 described immunosuppression secondary to allogeneic transfusions in 34 patients undergoing total hip arthroplasty (THA). According to Kendall et al., the lymphocyte function is impaired, which may be the etiology of the increased risk of deep prosthesis infection.23 The effect of allogeneic transfusion on the immune system is now well documented in the literature.20 Authors believe that the immunomodulating effect of red blood cell transfusion may be responsible for the better outcome in transplantation patients receiving transfusion, the higher risk of recurrence following malignancy resection in patients receiving transfusion, and the higher risk of postoperative infection following transfusion.20 The incriminated pathomechanism of these effects of blood transfusion is called transfusion-related immunomodulation.20
The increased risk of infection following allogeneic transfusion is fairly well established in the joint arthroplasty literature. Pulido et al.24 looked at predisposing factors for developing periprosthetic infection in a series of 9245 arthroplasty patients. They demonstrated a 2.1-fold increase in the rate of periprosthetic infection following allogeneic transfusion. Similarly, Shaunder et al.25 demonstrated a 3.6 times greater relative risk for developing postoperative infection after cardiac and orthopaedic surgery following allogeneic blood transfusion. Similarly, Murphy et al.26 showed an increased rate of confirmed or suspected infections in a series of 84 patients undergoing THA. They compared patients receiving autologous blood transfusion with those receiving allogeneic blood. Although the numbers were small in this series, the infection rate in the group receiving allogeneic transfusion was 32% compared to 3% in the autologous group. Other complications have also been linked to transfusion. In a recent population-based review of 28,087 THA patients, Pedersen et al.27 compared 2254 patients receiving allogeneic transfusion with 2254 nontransfused, matched patients. In this series, transfused patients demonstrated a higher 90-day mortality and increased odds of pneumonia (odds ratio 2.2 and 2.1, respectively). Bierbaum et al.1 evaluated 9482 patients undergoing major orthopedic surgery. This study demonstrated an increase in overall complication rate in patients requiring transfusion. Complications associated with transfusion in this series included an increase in infection, fluid overload, and a longer hospital stay.
Risk Factors For Transfusion
Preoperative blood values remain the best way to predict which patients will require perioperative allogeneic transfusion.28–31 Checking the hemoglobin and hematocrit before scheduling the indicated procedure identifies those patients at risk for requiring transfusion. In one of Cushner et al.’s original papers, they described factors that influence transfusion rates following TKA.32 In this study, they found preoperative hematocrit values were the best predictor for transfusion needs. Nuttall et al.33 as well as Boettnner et al.28 had similar findings, also noting the importance of preoperative hemoglobin. To minimize postoperative transfusion requirements, the patient’s hemoglobin must be maximized during the preoperative period. Patients average a 10% hematocrit loss routinely after total joint arthroplasty. Therefore, those who begin with higher preoperative hemoglobin concentrations are better able to tolerate the loss. Patients who are anemic preoperatively will be anemic during the postoperative period and may require transfusion. Guerin et al.34 performed a prospective review of 162 consecutive hip and knee arthroplasties. In this series, patients with preoperative hemoglobin levels less than 13 g/dl were four times more likely to receive a transfusion than those patients with preoperative hemoglobin levels greater than 15 g/dl.
Nuttall et al.33 evaluated 299 patients who underwent primary or revision THA in an attempt to predict the risk factors for allogeneic transfusion. In this study, risk factors for transfusion included preoperative hemoglobin, weight, age, anticipated blood loss, and aspirin use. Interestingly, they noted that predonated blood was often not transfused, leading to blood wastage and an increase in cost. Nuttall and colleagues33 concluded that identifying patients who were unlikely to require transfusion could avoid the unnecessary autologous predonation. Similar results were recently published by Boettner et al.28 In their study of 283 patients undergoing THA, not only was preoperative autologous donation (PAD) not beneficial in nonanemic patients, but PAD increased the overall transfusion rate, as our practice has demonstrated previously.
The preoperative hemoglobin concentrate has been shown to be the best predictor for postoperative transfusion in several other studies as well. Cabibbo et al.35 recently reviewed their autologous blood donation program in 1198 orthopaedic patients over the past 10 years. In this series, a preoperative hemoglobin level greater than 14 g/dl was a strong predictor for not requiring postoperative transfusion. Sculco and Gallina36 evaluated 1405 patients who underwent total joint arthroplasty. Again, preoperative hemoglobin level was inversely related to the frequency of perioperative allogeneic transfusions. In a large multicenter study, Bierbaum et al.1 evaluated 9482 patients undergoing major orthopaedic surgery. An increased transfusion rate for patients with a preoperative hemoglobin value of less than 13 g/dl was demonstrated. These studies suggest that the patient’s hemoglobin should be checked and optimized prior to surgery to minimize postoperative transfusion and complication rates. In our practice, we feel that preoperative optimization of the patient’s hemoglobin/hematocrit is the most valuable step in reducing perioperative transfusion rates.
Preoperative Blood Management
Preoperative Autologous Donations
In the late 1980s, the standard of care for a patient undergoing TKA was the preoperative donation of autologous blood. Although commonplace in the United States, PAD is uncommon in other areas of the world. PAD has several limitations. As discussed previously, patients who are most anemic in the preoperative period are at greatest risk for postoperative transfusion. These same patients are often excluded from PAD due to the preoperative anemia. This alone inhibits the ability of a PAD program to helping those patients with the greatest need. In addition, PAD may actually increase the risk of transfusion in the remaining patients by reducing their preoperative hemoglobin level, making a previously low-risk patient now at higher risk for transfusion.32,33,37 The assumption cannot be made that, after donating blood, the patients return to their predonation status. Several studies have shown that PAD will lower the patient’s preoperative blood levels.28,38–40 The patient donates the blood, does not return to baseline, and arrives for the surgery in an anemic state. The patient is now more at risk for allogeneic transfusion because of the autologous donation.
In our experience, too many patients donated blood too close in time to the scheduled surgery to recover the blood lost with PAD. In addition, a 1- to 2-unit PAD program does not cause a significant erythropoietic response. Because no erythropoietic response occurs, patients do not return to their baseline level. This is demonstrated by the literature. For example, Hatzidakis and coworkers41 performed a retrospective analysis of 489 consecutive patients undergoing total joint arthroplasty. A decrease in hemoglobin concentration from the time of the donation to the time of surgery was reported; the average decrease was 1.22 g/dl. The authors did not recommend PAD for patients with predonation status greater than 13 g/dl. We evaluated our PAD program previously with regard to preoperative hemoglobin levels. Between 1993 and 1995, 2 units of PAD were obtained on average compared with 1 unit of PAD obtained in the years between 1995 and 1997. Our studies showed a 3% decrease in hematocrit values for every unit donated before surgery.42 When 1 unit was donated, a 3% decrease in hematocrit was noted before surgery. When 2 units were donated, a 6% decrease from baseline was noted; 2 units of PAD resulted in more anemia before the surgical procedure.39
Our practice tried to address the issue of wastage with automatic infusion of the autologous blood. We reviewed our results of a 1-unit PAD program with automatic infusion of the donated blood. All patients were given their PAD immediately after surgery, resulting in 0% wastage. No numerical transfusion triggers were utilized and subsequent allogeneic transfusion was based on symptoms. Despite ordering the PAD unit 1 month before surgery, significant preoperative anemia was noted. In this retrospective review of 148 patients undergoing unilateral TKA, a 1.3-g/dl decrease in hemoglobin was noted between predonation and presurgical testing.43 We refer to this occurrence as “orthopaedic-induced anemia.” Whereas only 26.2% of patients were in the high-transfusion-risk group (hemoglobin >10 g/dl and ≤13 g/dl) before surgery, 55.7% of patients were in this high-risk category after PAD. 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. As documented by others, the use of PAD resulted in anemia, and the patients did not return to the predonation hemoglobin and hematocrit values. Although the allogeneic infusion rate was low in this series, we believe this reflects our lower trigger point for transfusion in the immediate postoperative period. Had historical transfusion triggers been followed (hemoglobin concentration < 10 g/dl), a transfusion rate of 30% would have been found. Fifty percent of our patients were discharged with hemoglobin concentrations less than 9 g/dl.
The low allogeneic transfusion rate should not be misconstrued as efficacy of a 1-unit PAD program, but rather a change in our transfusion practice. PAD programs are limited by significant wastage if automatic transfusion is not followed. In a review of 1198 patients enrolled in an autologous blood donation program, Cabibbo et al.35 demonstrated a high degree of autologous blood wastage in patients with an optimized preoperative hemoglobin. In this series, a preoperative hemoglobin level greater than 14 g/dl was a strong predictor for not requiring postoperative transfusion. In this subgroup of patients, autologous blood wastage was over 90%. Even in the subgroup of patients with preoperative hemoglobin levels between 13 and 14 g/dl, autologous blood wastage was still over 50%. Even in a recent study that concluded autologous blood donation was a cost-effective measure in reducing allogeneic transfusion rates following arthroplasty, Green et al.44 demonstrated overall autologous blood wastage of 38% in 356 patients. When looking at TKA alone, the autologous blood wastage was 51%. These findings are similar to those of Bierbaum et al.1 In their large prospective series including 9482 patients, the autologous blood wastage was 45%. Our institution has abandoned the aforementioned protocol based on the apparent lack of efficacy. In addition, this type of protocol places patients at an additional risk. A protocol with a 100% autologous rate exposes patients to the added risk of donation error. Goldman et al. reviewed autologous error rates in Canada and found an error rate of 6 in 149.45 The majority of these errors were related to labeling error (48%) or error in component preparation (25%). One patient received the wrong unit of donated blood, which is not an uncommon occurrence. According to the College of American Pathologists,46 0.9% of the 3852 institutions studied had at least 1 unit of PAD given to the wrong patient.
Cost is the final issue of PAD, for this is not an inexpensive process. The costs of a PAD program are related to procurement and the cost connected with giving the blood. Billote et al.38 evaluated PAD in patients who were receiving THA and found no benefit in PAD for nonanemic patients undergoing primary hip replacement. Each patient donated 2 units of autologous blood, with an additional cost of $758 per patient despite frequent blood wastage. Etchason et al.42 studied the cost of the PAD program and concluded that increased protection afforded by autologous blood is limited and may not justify the increased cost. However, a more recent study comparing the cost of erythropoietin versus autologous and allogenic blood donation in total joint arthroplasty showed that autologous blood donation might be cost effective.44 In 356 unilateral total joint arthroplasties performed during an 11-month period, a combination of autologous blood donation and allogeneic blood was the least costly approach at $856 and $892 per patient for THA and TKA, respectively. The most costly strategy was allogeneic blood only at $1769 and $1352 per THA and TKA patient, respectively.
Use of Erythropoietins
The importance of preoperative hemoglobin concentrations was discussed earlier. The surgeon is often limited in his or her ability to maximize the patient’s predonation hemoglobin values. In the past, we participated in a PAD program. Unhappy with the anemia caused by donations, we put our patients on a new protocol. We began incorporating a patient-specific protocol to utilize epoetin alfa (EPO) (Procrit; Ortho Biotech, PA) into our busy knee practice.39 We obtain a preoperative hemoglobin and hematocrit prior to surgical booking. Patients with a hemoglobin level greater than 10 and less than 13 g/dl are indicated to receive EPO injections prior to surgery. Patients receive 40,000 units at 3 weeks, 2 weeks, and 1 week prior to surgery, with an average rise in hemoglobin of 1.5 g.39 We compared 50 patients who received EPO injections with 50 patients participating in the autologous program with automatic reinfusion.43 The patients receiving EPO had higher blood parameters preoperatively, postoperatively, and on discharge than the patients who participated in the autologous program. Additionally, our overall cost was reduced because the autologous program was used in 25% of the patients with EPO compared with 100% in previous protocols. Due to the success of this protocol, our indications have expanded. We now employ the same protocol for all arthroplasty patients, including those undergoing PKA, unilateral arthroplasty, bilateral arthroplasty, revision arthroplasty, and two-stage revisions for infections.
The success of this protocol is best demonstrated with discussion of our more complicated cases. By utilizing EPO between stages during the two-stage treatment of an infected total joint arthroplasty, our allogeneic transfusion rates were decreased from a high of 88% to 33%.47 On evaluation of blood loss and transfusion rates following revision TKA, we found that 75% of female patients scheduled for revision were in the high-risk group (hemoglobin > 10 and <13 g/dl).48 In our practice, approximately 25–30% of patients fall into the high-risk group and require EPO injections. This means that 75% required no preoperative intervention. This saves our office the time and inconvenience of routinely scheduling autologous blood donations for all patients, for many of whom it would be unnecessary. Treating 25% of patients with EPO injections is significantly less costly than using an autologous donation program for 100% of the patients, especially in light of potential risks with an automatic reinfusion protocol. A similar protocol was recently studied by Moonen et al.49 and compared to re-transfusion of autologous shed blood. All patients in this study had a preoperative hemoglobin level between 10 and 13 g/dl. In 100 arthroplasty patients randomized to preoperative erythropoietin injection or re-transfusion of autologous shed blood, the allogeneic transfusion rate for the group receiving preoperative injections of erythropoietin was 4% compared to 28% for the group with a reinfusion drain. A second recent study compared the use of erythropoietin to a PAD program.50 In this series, 121 arthroplasty patients with a preoperative hemoglobin between 11 and 14 g/dl either received weekly EPO injections or underwent PAD. The EPO group demonstrated higher hemoglobin levels, lower transfusion rates, and a significant increase in postoperative vigor.50
