 |
 |
 |
|
| home | CME | help | feedback | subscriptions | archive |
|
© 2004 The Journal of Bone and Joint Surgery, Inc. A Blood-Conservation Algorithm to Reduce Blood Transfusions After Total Hip and Knee Arthroplasty
1 St. Vincent Center for Joint Replacement, 8402 Harcourt Road, Suite 128,
Indianapolis, IN 46260. E-mail address for J.L. Pierson:
jlpierso{at}stvincent.org
Investigation performed at St. Vincent Center for Joint Replacement,
Indianapolis, Indiana
Background: Donation of autologous blood before total joint arthroplasty is inconvenient and costly, causes a phlebotomy-induced anemia, and may be wasteful and unnecessary for the nonanemic patient. We developed a blood-conservation algorithm that does not require predonation of autologous blood, employs selective use of epoetin alfa, and uses evidence-based transfusion criteria. Our hypothesis was that use of this algorithm would reduce the rate of transfusion after unilateral total hip and knee arthroplasty as compared with the rates described in previous reports. Methods: We retrospectively reviewed the records of 500 consecutive patients in whom unilateral primary total hip or knee arthroplasty had been performed by a single surgeon. The same blood-conservation algorithm was recommended to all patients. Two groups of patients were identified: the first group consisted of 433 patients in whom the algorithm was followed, and the second group consisted of sixty-seven patients in whom the algorithm was not followed. Results: In the group in which the algorithm was followed, the rates of allogeneic transfusion after total knee and total hip arthroplasty were 1.4% (three of 220) and 2.8% (six of 213), respectively. The overall rate of transfusion in this group was only 2.1% (nine of 433). The prevalence of transfusion in the group in which the algorithm was not followed was 16.4% (eleven of sixty-seven). This difference was significant (p = 0.0001). Conclusions: The use of this blood-conservation algorithm resulted in a significant reduction in the need for allogeneic blood transfusions after unilateral total hip and knee arthroplasty, and the results compare favorably with the rates of transfusion described in previous reports. Level of Evidence: Therapeutic study, Level III-2 (retrospective cohort study). See Instructions to Authors for a complete description of levels of evidence.
Total hip arthroplasty and total knee arthroplasty are associated with substantial blood loss1-4. Bierbaum et al.5 reported a 46% rate of allogeneic and autologous transfusion in a series of 9482 patients managed with total joint arthroplasty. A reduction in the use of blood transfusions is desirable as it would reduce the potential for disease transmission6,7, it would reduce costs8-10, and it would not strain the limited availability of allogeneic blood11. Since the 1980s, the most common blood-conservation strategy associated with total joint arthroplasty has been the predonation of autologous blood5,12-18. Although emotionally appealing to patients and physicians5,19,20, this strategy is associated with a number of problems, including inconvenience21, wastefulness5,12,22,23, cost6,12,22,24, the creation of a phlebotomy-induced anemia25-29, and uncertainty regarding the indications for predonation by nonanemic patients3,5,22,27,30. We believed that use of evidence-based transfusion criteria would likely result in a decreased rate of transfusions, but such criteria have been infrequently adopted by physicians11,25,26,31-35. The purpose of this study was to evaluate the efficacy of a blood-conservation algorithm in a series of patients managed with primary total hip and total knee arthroplasty.
The records of all 500 patients in whom a primary total hip or total knee arthroplasty had been performed by the senior author (J.L.P.) from July 1999 through May 2002 were reviewed retrospectively. The institutional review board at our institution approved the study. The inclusion criterion was primary unilateral total hip or knee arthroplasty. The exclusion criteria included simultaneous bilateral total joint arthroplasty, revision total joint arthroplasty, and primary total hip arthroplasty for fracture. No patients were excluded for medical reasons. All total hip arthroplasties had been performed through a posterolateral approach, and all total knee arthroplasties had been performed through a median parapatellar approach. Drains had not been used in any patient. No blood-salvage techniques, such as intraoperative blood salvage or postoperative reinfusion of blood, had been used. Thromboembolic prophylaxis had consisted of low-dose warfarin, with a target international normalized ratio of 1.8 to 2.2, for a duration of four weeks postoperatively. The height, weight, body-mass index, estimated red blood-cell volume, preoperative hemoglobin level, lowest postoperative hemoglobin level, estimated preoperative blood volume, estimated blood loss, postoperative complications, length of stay, and number and type of transfusions were recorded for each patient (Table I). Major complications were defined as death within ninety days after the procedure, perioperative myocardial infarction, cerebrovascular accident, and symptomatic pulmonary embolism.
The same blood-conservation algorithm had been suggested to all patients. The algorithm consisted of a mathematical formula that was designed to predict which patients would be likely to need an allogeneic blood transfusion (Fig. 1). The senior author used previously reported data3,36 as well as unpublished data on the expected decline in hemoglobin levels after primary unilateral total joint arthroplasty to develop the formula. The formula identifies patients with substantial anemia, in whom the procedure-specific expected loss of blood (that is, the expected drop in the hemoglobin level plus one standard deviation) places them at risk for the need of transfusion (a predicted lowest hemoglobin level of <7.0 g/dL)4,26,37,38. Use of the algorithm led to one of two recommendations to patients undergoing unilateral total joint arthroplasty: (1) treatment with observation only or (2) treatment with preoperative epoetin alfa. Predonation of autologous blood was discouraged.
Two groups were retrospectively identified (Table II). The first group consisted of 433 patients in whom the algorithm was implemented (the "on-algorithm" group). The second group consisted of sixty-seven patients in whom the algorithm was not followed (the "off-algorithm" group). The reasons for nonimplementation of the algorithm included patient insistence on the predonation of autologous blood (four patients), patient or family insistence on directed-donor predonation (five patients), or nonimplementation of treatment with preoperative epoetin alfa despite an algorithmic recommendation to do so (fifty-eight patients).
There were several reasons that epoetin alfa was not implemented despite a recommendation to do so, including logistical difficulties in arranging the series of injections, patient refusal, and/or the fact that epoetin alfa was not covered by the patient's insurance. The senior author made all decisions regarding whether to order a transfusion on the basis of evidence-based criteria that had been established by the medical director of our institution's blood-conservation program39. According to these criteria, an allogeneic transfusion should be given when the hemoglobin level is <7 g/dL (<4.3 mmol/L) or when a patient requires additional oxygen-carrying capacity because of symptomatic anemia. Signs of tachycardia, hypotension, and inadequate urinary output were deemed to be responses to hypovolemia rather than anemia and were treated with restoration of the patient's circulating blood volume. In the event that these signs persisted after volume replacement, the patient was then considered to have symptomatic anemia and a transfusion was given.
Statistical Methods
The estimated blood loss was significantly greater after total hip arthroplasty than after total knee arthroplasty (1428 compared with 1353 mL) (p = 0.048), and the mean decline in the hemoglobin level was greater after total hip arthroplasty than after total knee arthroplasty (4.0 compared with 3.8 g/dL [2.5 compared with 2.4 mmol/L]). Only four (0.8%) of the 500 patients predonated autologous blood. The overall rate of transfusion (both autologous and allogeneic) for the entire study group was 4.0% (twenty of 500). The transfusion rates associated with total knee and total hip arthroplasty were 3.3% (eight of 240) and 4.6% (twelve of 260), respectively. Pearson product-moment correlation coefficients revealed that the rate of transfusion was correlated with the baseline hemoglobin level (r = –0.26, p = 0.0001), the lowest hemoglobin level (r = –0.39, p = 0.0001), the length of stay (r = –0.25, p = 0.001), and the estimated blood loss (r = 0.15, p = 0.001). Although the demographic characteristics of the two groups were similar, the off-algorithm group had a significantly lower baseline hemoglobin level than the on-algorithm group did (11.6 compared with 14.0 g/dL [7.2 compared with 8.7 mmol/L]; p = 0.0001) (Table II). The off-algorithm group consisted predominantly of anemic patients in whom preoperative epoetin alfa had been recommended but had not been implemented for the reasons previously described. This group had a significantly higher transfusion rate than the on-algorithm group did (16.4% compared with 2.1%; p = 0.0001). Only nine (2.1%) of the 433 patients in the on-algorithm group received a transfusion. In this group, the transfusion rates after total knee and total hip arthroplasty were 1.4% (three of 220) and 2.8% (six of 213), respectively. This group included eighteen patients who had been managed with epoetin alfa for the treatment of preoperative anemia. None of these patients received a transfusion.
Complications
Anemia secondary to preoperative autologous blood donation and a conventionally low transfusion threshold leads to substantially higher overall rates of transfusion (both allogeneic and autologous) among autologous blood donors as compared with nondonors25. Recent evidence11,25,26,31-34,40 has challenged the frequently followed practice of the "10/30" rule and other arbitrarily defined transfusion triggers. The "10/30" transfusion trigger suggests that a patient receive a transfusion when the hemoglobin level falls below 10 g/dL (6.2 mmol/L) or the hematocrit falls below 30%41. Orthopaedic surgeons have been slow to adopt evidence-based transfusion guidelines35. A randomized, controlled trial of 838 patients in critical care units demonstrated that a transfusion threshold based on a hemoglobin concentration of as low as 7 g/dL (4.3 mmol/L) (combined with the maintenance of hemoglobin concentrations in the range of 7 to 9 g/dL [4.3 to 5.6 mmol/L]) was at least as effective as or superior to a threshold of 10 g/dL [6.2 mmol/L] (combined with the maintenance of hemoglobin concentrations in the range of 10 to 12 g/dL [6.2 to 7.4 mmol/L])42. Our transfusion criteria require a patient-demonstrated need for increased oxygen-carrying capacity. Our findings suggest that a patient should rarely receive a transfusion for a hemoglobin level of >7 g/dL (>4.3 mmol/L) in the absence of symptoms that are attributable to the anemia. It is important to note that most hemodynamic parameters, such as pulse, blood pressure, and urine output, are a reflection of the circulating blood volume status. A transfusion should rarely be performed for tachycardia, hypotension, or low urinary output until the circulating blood volume has been normalized. In most cases, doing so will result in normalization of these parameters. The key element of our blood-conservation strategy is the preoperative identification of patients who are at substantial risk of needing an allogeneic blood transfusion after total joint arthroplasty. As the expected blood loss after total knee and total hip arthroplasty is reasonably predictable, this strategy requires the identification of patients with substantial preoperative anemia. Previous studies have demonstrated that improving the preoperative hemoglobin level is the best strategy for these patients28,40,43,44. We believe that the most effective method of achieving this goal is for appropriately selected patients to be treated with epoetin alfa preoperatively. In order for us to recommend treatment with epoetin alfa to patients undergoing primary unilateral total knee or hip arthroplasty, the preoperative hemoglobin level must be <12 g/dL (<7.4 mmol/L). Epoetin alfa was utilized for eighteen patients in this study, and none of them required a transfusion. We believe that our data demonstrate the effectiveness of an algorithm-based blood-conservation program for patients undergoing unilateral total joint arthroplasty. In the group of 433 patients for whom this algorithm was followed, the transfusion rate was only 2.1% (1.4% for patients managed with total knee arthroplasty and 2.8% for those managed with total hip arthroplasty). These transfusion rates are among the lowest that have been reported in the literature. For example, in the study by Bierbaum et al.5, in which preoperative autologous donation was the dominant blood-conservation strategy and in which poorly defined transfusion criteria were utilized, the overall rate of transfusion (both autologous and allogeneic) was 46%. Furthermore, the overall rate of allogeneic transfusion was 16%. Feagan et al.30 reported allogeneic transfusion rates of 25% and 30% after total knee and hip arthroplasty in a cohort of patients who did not predonate blood. They also reported allogeneic transfusion rates of 11% and 15% after total hip and knee arthroplasty in patients who did predonate blood. Preoperatively, the on-algorithm cohort had a higher baseline hemoglobin level than the off-algorithm cohort did (14.0 compared with 11.6 g/dL [8.7 compared with 7.2 mmol/L]; p = 0.0001). This difference is to be expected as the algorithm is designed to identify patients with substantial preoperative anemia. The rates of major postoperative complications and ninety-day mortality were not different from those reported in the literature45-49. Similarly, the length of stay was similar to that in previous reports. We believe euvolemic patients (including elderly patients) tolerate anemia well and are capable of participating in a typical total joint arthroplasty rehabilitation program. In conclusion, an algorithm-based blood-conservation strategy reduces the rate of allogeneic blood transfusions after unilateral total joint arthroplasty. We continue to utilize the algorithm, including its key elements. We discourage the predonation of autologous blood, selectively utilize preoperative epoetin alfa for the treatment of anemic patients who are at high risk for perioperative allogeneic transfusion, and employ evidence-based transfusion criteria.
  CiteULike  Connotea  Del.icio.us  Facebook  Technorati  Twitter What's this?
This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
| home | CME | help | feedback | subscriptions | archive | site index |
|
Stanford University Libraries' HighWire Press (TM) assists in the publication of JBJS Online. |