It is estimated that of 50 million transfusions performed each year, 20% result in adverse effects. Although numerous studies have evaluated other modalities to decrease the allogeneic transfusion rate following total knee arthroplasty (TKA), allogeneic transfusions have not been eliminated. This paper discusses the evolution of transfusion management at our institution.
Blood loss and transfusion risks following TKA have been evaluated since the early 1970s. In 1973, Patterson et al1 examined the blood loss in 44 rheumatoid arthritis patients and concluded that blood loss was secondary to a hemolytic or redistribution process that resulted in a decrease in the patient's hemoglobin (Hb). This paper was limited in that osteoarthritic patients were not considered and the presumed hemolytic or redistribution process was not defined. In 1981, Erskine et al2 evaluated this issue, noting blood loss from a variety of sources and concluding that the majority of blood loss occurred intraoperatively. This was a small study and, once again, the osteoarthritic patient was not considered.
In a study that evaluated 112 consecutive primary TKAs that followed a standard protocol of immediate continuous passive motion, intraoperative tourniquet deflation, and closed suction drainage,3 a major predictor of transfusion risk was preoperative hematocrit (HCT). The rate of decrease in HCT levels was similar in both the transfused and nontransfused groups. Thus, increased transfusion rates in women, as well as in rheumatoid arthritis patients, were thought to be secondary to lower preoperative HCT levels. Burman et al came to a similar conclusion regarding the importance of preoperative HCT.4
Although the importance of preoperative HCT was noted, there was little that could be done to improve preoperative anemia. Although one could treat anemia of iron deficiency in theory, in practice preoperative anemia was multifactorial and not correctable merely with routine iron therapy. Allogeneic transfusion rates of 38% to 50% were quoted in the literature and the solution to preventing these allogeneic transfusions was not yet available.
Continuing concern about the safety of allogeneic transfusions has led to the development of new methods to decrease allogeneic transfusion rates in primary unilateral TKAs. At our institution, this change in protocol included the initiation of both preoperative autologous donation (PAD) and the placement of drains to salvage blood from the surgical site postoperatively. Other modalities to decrease blood loss include cementing of all prostheses, intraoperative tourniquet deflation with hemostasis, and bone plugging of the femoral intramedullary canal. We no longer transfuse blood based on rigid transfusion trigger criteria, but rather transfuse only when clinical symptoms warrant it.
PRIMARY UNILATERAL TOTAL KNEE RESULTS
Between 1993 and 1995, it was standard practice at our institution to request 2 units of autologous blood from patients prior to TKA surgery. Review of data includes only those patients for whom predonation HCT values could be obtained. Before donation, patients had a mean HCT level of 42.1%, yet the mean HCT had decreased to 35.8% by preadmission testing - a decrease of 3% per unit of autologous blood donated. As a general trend, most patients were not anemic before donation, but many were so after donation. While these PAD patients did experience an allogeneic transfusion rate reduction to 1 .2%, over 53% of predonated autologous blood was wasted. Based on these results, a decision was made to request only 1 unit of PAD blood for TKA.
In the cohort of patients who donated only 1 unit of blood, the mean baseline (predonation) HCT level was 41.2%. At preadmission testing, the mean HCT level was 38.2% - again representing a drop of 3% per unit donated. Nevertheless, the allogeneic transfusion rate remained low, at 2.9%, and 72% of patients in this group did not require a return of the donated blood. Therefore, although the allogeneic transfusion rates were similar between patients who donated 1 unit versus 2 units of blood, patients who donated 2 units were more anemic preoperatively and had a higher autologous transfusion rate than patients who donated only 1 unit.
A third group of patients, those who were unable to donate blood for TKA, was also evaluated. Hematocrit levels in this group were 40.6% at baseline and 40.2% at preadmission testing. Further, the allogeneic transfusion rate was 30%, which was similar to historical transfusion rates in patients lacking predonated blood.
Several trends emerged from these data. Although the practice of autologous donation helps to decrease allogeneic transfusion rates, PAD tends to induce anemia before surgery, as others have reported.5'6 Consequently, the anemia induced by PAD actually serves to increase the patient's risk of transfusion. Contrary to popular belief, PAD patients do not have a substantial increase in erythropoietin levels from donating 1 to 2 units before surgery and generally do not recover to their predonation HCT level by the time of the indicated procedure.
Based on these trends, we have altered our blood management strategy for TKA, but not our goals. We strive to maintain a low allogeneic transfusion rate, to decrease the wasting of PAD blood, and to focus on the importance of Hb values for blood management decisions. It is necessary to identify and separate those patients who will not need to donate blood from those who will require 1 unit of autologous blood during surgery, and to identify those patients who may benefit from other modalities, such as erythropoietin supplementation.
In our current procedure, an estimation of a patient's surgical blood need is based on a complete blood count obtained at the time of surgical scheduling. We are currently performing a prospective study to evaluate a blood management treatment algorithm revolving around baseline preoperative HCT levels and the select role of autologous blood donation and erythropoietin. The algorithm is based on the following criteria: HCT > 42%, no PAD required; HCT > 39% and < 42%, 1 unit of autologous blood; HCT < 39%, erythropoietin therapy only. This algorithm should raise Hb in anemic patients, reduce PAD wastage and limit preoperative anemia secondary to PAD. The threshold HCT of 42% was determined by considering that most patients have a 10% decrease in HCT with total knee replacement surgery, such that some patients would have postoperative HCTs low enough to require a transfusion. The threshold HCT of 39% was chosen because patients below that value are anemic and are therefore candidates for erythropoietin therapy.
By employing a patient-specific approach to blood management, we hope to maintain a low allogeneic transfusion rate, to decrease wasting of autologous blood, and to eliminate unnecessary autologous transfusions.
Infected Total Knee Revision Arthroplasty. The current plan for surgical blood management is not suitable for all patients undergoing TKA. For example, patients who have TKA in the presence of infection are no longer candidates for reinfusion drains or autologous donations because these can cause reinfection. The alternative procedure for TKA in the presence of infection requires two surgeries. During the first, the existing prosthesis is replaced with an antibiotic spacer that remains for 6 weeks while the infection is controlled. During the second surgery, a new prosthesis is implanted. Preoperative treatment with erythropoietin before the initial implant is removed is not feasible because there is insufficient time to administer the course of therapy. Because these patients have a higher transfusion risk due to a very short time span between the two major procedures, we decided to evaluate, retrospectively, their transfusion needs.
In a current study, Pagnano and colleagues have reviewed 70 patients treated for infected TKA at two major medical centers. The transfusion rates were 80% following prosthesis removal and 82% at the time of reimplantation. Only 12% of patients avoided allogeneic transfusion in either procedure; the majority of patients received allogeneic blood at both procedures. Additionally, a transfusion after prosthesis removal seldom protected the patient from receiving a transfusion for the reimplantation procedure. Because of concerns of immunosuppression associated with allogeneic transfusions, these patients may be more at risk for reinfection of the prosthesis.
Based on our review of these cases, we have begun to include erythropoietin therapy as part of the postoperative treatment following prosthesis removal in an attempt to reduce die transfusion rate during the second procedure. Patients are administered 40,000 International Units of erythropoietin subcutaneously and oral iron supplementation after their HCT stabihzes postoperatively. Erythropoietin treatment is continued after discharge with weekly monitoring of complete blood counts. Since initiating this procedure, the allogenic transfusion rate following reimplantation has been reduced to 35%.
Bilateral TKA. Preliminary review of bilateral TKAs performed at our institution has been completed, sbowing that numerous patients have donated 4 units of PAD blood in the past. Recently, we have changed our protocol to include only 1 PAD unit per knee, consistent with our unilateral TKA data. Allogeneic transfusion rates remain low, and we have noted a decrease in autologous wastage. Final results of this study are pending.
Blood loss management following TKA at our institution has evolved and continues to evolve. We feel that, with this patient-baséd protocol based on a predonation complete blood count, allogeneic transfusions as well as autologous donations can be reduced. Blood loss and transfusion management has improved and will continue to improve as these protocols are refined.
1. Patterson FP, Brown CS. Complications of total hip replacement arthroplasty. Orthop Clin North Am. 1 973; 4:503-5 12.
2. Erskine JG, Fraser C, Simpson R. Protheroe K, Walker ID. Blood loss with knee joint replacement. / R Coll Surg Edinb. 1 98 1 ; 26:295-297.
3. Cushner FD, Friedman RJ. Blood loss in total knee arthroplasty. Clin Orthop. 1991; 269:98101.
4. Griffiths EM, Kaplan DK, Goldstraw P, Burman JF. Review of blood transfusion practices in thoracic surgery. Ann Thorac Surg. 1994; 57:736-739.
5. Goodnough LT, Brittenham GM. Limitations of the erythropoietic response to serial phlebotomy: implications for autologous blood donor programs. J Lab Clin Med. 1990; 1 15:28-35.
6. Kickler TS, Spivak JL. Effect o/ repeated whole blood donations on serum imiminoreactive erymropoietin levels in autologous donors. JAMA. 1988; 260:65-67.