Dr Hourlier is from the Department of Orthopedic Surgery, Polyclinique de la Thiérache, Wignehies, France; and Mr Fennema
is from the Department of Clinical Research, Smith & Nephew Orthopaedics AG, Baar, Switzerland.
Dr Hourlier has no relevant financial relationships to disclose. Mr Fennema is employed by Smith & Nephew Orthopaedics AG,
This study was sponsored by Smith & Nephew, Courbevoie, France, which had no involvement in the study design; the collection,
analysis, or interpretation of the data; the writing of the article; or the decision to submit for publication.
Correspondence should be addressed to: Hervé Hourlier, MD, Department of Orthopedic Surgery, Polyclinique de la Thiérache,
59212 Wignehies, France (email@example.com).
Blood loss and transfusion rates remain high during total knee arthroplasty (TKA), despite the availability of a number of
strategies to prevent or lessen the likelihood of these outcomes.
Several blood conservation regimens are currently available, yet the best method of decreasing the risk of postoperative
blood transfusion is not clear.
The effect of tranexamic acid on blood loss and blood transfusion rate was determined by Benoni and Fredin
in a placebo-controlled randomized study. A recent national survey determined that blood salvage systems and reinfusion drains
are routinely used for reducing allogeneic exposure in patients undergoing TKA.
At our institution, the use of autotransfusion systems in primary unilateral total hip arthroplasty (THA) was abandoned following
an internal audit in 2005. Since 2006, we have adopted a blood-saving program that uses chemoprophylaxis support to lessen
blood loss and allogeneic blood requirements. The program served a series of 215 consecutive patients who underwent primary
THA and remained blood transfusion free.
In the same period, we used a similar blood transfusion-sparing plan for unilateral primary TKA.
The purpose of the current study was to compare blood loss rates and blood transfusion requirements after unilateral TKA in
2 comparable groups that differed only with respect to the blood-saving plan used. We hypothesized that chemoprophylaxis would
lead to less blood loss and fewer blood transfusion, without having an adverse effect on patients.
Materials and Methods
The study comprised 2 groups. Group A included 90 prospective patients (91 knees) undergoing consecutive cemented TKA. This
cohort was operated on between January 2008 and February 2009 under a global blood transfusion-sparing plan configured with
chemoprophylaxis to reduce blood loss and allogeneic exposure. Neither perioperative autotransfusion systems nor postoperative
intra-articular wound drainage at surgery were used in this TKA group. Group B was used as an historical comparator. This
group consisted of 44 TKAs (44 knees) operated on between January 2004 and September 2005 using a blood plan that used a reinfusion
drain (Constavac CB II; Stryker, Kalamazoo, Michigan). All TKAs were performed by the senior author (H.H.) at the Polyclinique
de la Thiérache in Wignehies, France. An identical cemented knee prosthesis design was implanted in each group (TC-SB implants;
Smith & Nephew Orthopaedics AG, Baar, Switzerland).
All TKAs were primary and unilateral. Bilateral TKA under a single anesthesia and revision TKAs were excluded. One male patient
in group A underwent bilateral TKA, which was performed separately during different hospital stays at a 9-month interval between
operations. Each knee was therefore considered as a separate case. The proportion of patients who underwent a previous operation
on their knee was 13% (12 of 90) in group A (4 tibial osteotomies and 8 arthroscopic debridements) and 10% (4 of 44) in group
B (2 osteotomies and 2 debridements), while the remaining patients (78 in group A and 40 in group B) had no previous knee
surgery. Indications for surgery were degenerative primary or secondary osteoarthritis that was severe enough to warrant TKA.
No statistically significant differences were observed between the 2 study groups with respect to sex, body mass index (BMI),
preoperative hemoglobin level, American Society of Anaesthesiologists physical status score, incision length, or operating
time. However, a significant difference was observed in mean patient age between the 2 groups (Table ).
Table 1. Patient Demographics
All surgeries were performed in an operating room with horizontal laminar airflow. All patients received prophylactic antibiotics
for 24 hours, starting 1 hour before skin incision. A straight longitudinal midline skin incision was made for patients who
had not undergone previous knee surgery. A medial parapatellar approach was taken, with patellar eversion and tibial translation
in all patients. Both cruciate ligaments were excised. Synovial tissue was removed if patients had an inflammatory knee condition.
The tibia was prepared for TKA with an extramedullary alignment jig. The distal femur was cut by opening the femoral intramedullary
canal. The procedures followed a standardized technique, including necessary soft tissue release to obtain adequate ligament
balance. The ligaments were balanced with spacers. The patella was resurfaced at the surgeon’s discretion with an all-polyethylene
patellar component. All parts were cemented (CMW Genta 3 cement; DePuy, Blackpool, United Kingdom) without pressurization.
The lumen of the femur was plugged by autologous bone before femoral component cementation.
All but 1 patient underwent TKA under general anesthesia. A femoral catheter was inserted for local anesthetic administration
before general anesthesia. Following an initial bolus of 20 mL ropivacaine 0.2%, continuous infusion was instituted and prolonged
for 2 to 4 days postoperatively.
General anesthesia was induced by propofol, sufentanyl, and atracurium. The patients were intubated, and ventilation was controlled
with nitrous oxide/oxygen mixture. Anesthesia was maintained with sufentanyl reinjections and sevoflurane inhalation. A warming
blanket set at 37°C was applied on the upper body throughout the procedure. Mean body temperature was recorded at entry into
the recovery room.
Common Blood-Sparing Practices
Patients were submitted to common blood-sparing practices in both groups. No patients were included in a predeposit autologous
blood donation program before TKA. Oral nonsteroidal anti-inflammatory drugs were discontinued at least 48 hours preoperatively.
Hemostasis was achieved with standard electrocoagulation during knee surgery. Knees were operated routinely with a pneumatic
tourniquet under 320 mm Hg pressures. The tourniquet was released to secure hemostasis during surgery before closing the wound
in 86 of 91 knees in group A and in 40 of 44 knees in group B. For the remaining knees, the tourniquet was either contraindicated
(1 case in group A and no cases in group B) or released after the wound was closed and wrapped with occlusive dressings and
elastic bandages. Apart from these common procedures, the 2 groups were investigated under radically different blood-sparing
strategies (Table ).
Table 2. Blood Plans
In group A, suitable patients received chemoprophylaxis to reduce blood losses and transfusion requirements: oral iron was
prescribed with posology of 3 pills daily for 20 days preoperatively, with information given to the patient regarding the
gastrointestinal outcome of this treatment.
Men and women with baseline hemoglobin levels <13 g/dL and 12 g/dL, respectively, received 1 preoperative weekly subcutaneous
dose of 40,000 IU epoetin-alpha. The epoetin-alpha regimen varied from 2 to 4 injections. The decision regarding the number
of epoetin-alpha injections was made in relation to the patient’s medical status and the expected postoperative hemoglobin
level at discharge. During surgery, all patients were administered 1 intravenous low infusion of 30 mg/kg of tranexamic acid,
unless contraindications were present. These included renal insufficiency, arterial disease, myocardial infarction antecedents,
coronary disease, history of deep venous thrombosis, and pulmonary emboli.
In group B, the abovementioned chemoprophylaxis was not administered.
Wound drainage was radically different in the 2 groups. In group A, a single superficial subcutaneous vacuum drain was placed
at the time of closure with the intent of anticipating benefit for the skin wound-healing conditions. No drains were installed
deep in the joint cavity at surgery. In group B, TKAs were performed in a blood plan using reinfusion drains (Constavac CB
II) as the central blood-sparing strategy. Two suction drains, both external and internal to the joint, were Y-connected to
the Constavac reservoir. The blood from fluid recovered by drainage was reinfused intravenously during the first 6 hours after
skin closure when >100 mL of drained fluid was collected in the reservoir.
Postoperative care incorporated a multimodal pain control program. Ice packs were applied during the first postoperative days.
The transfusion triggers were 8 g/dL of hemoglobin in patients with irrelevant comorbidities and 10 g/dL of hemoglobin in
patients with preexisting cardiac insufficiency or coronary disease.
The rehabilitation protocol consisted of immediate knee mobilization. Full weight bearing was allowed postoperatively. Antithromboembolic
prophylaxis was started the day of surgery, specifically 8 hours after skin closure, with 2.5 mg fondaparinux and continued
once daily for 10 days. Thereafter, thrombosis prophylaxis was prolonged subcutaneously for 30 days with 1 daily dose of 4500
IU tinzaparine, except in cases in which vitamin K antagonists were reestablished. All patients underwent Doppler ultrasound
of both lower limbs at postoperative day 7 or earlier if there was any clinical suspicion of deep vein thrombosis.
Assessment of Blood Losses, Complications, and Treatment Costs
Total blood loss (TBL) was calculated indirectly from the hematological data using the formula: TBL=estimated blood volume×(hematocrit
(Ht) reduction/mean Ht), where Ht reduction was the difference between preoperative and postoperative day 7 values.
were applied to estimate patient blood volume.
Compensated blood loss was taken into account, as 1 unit of homologous blood contained 150 mL of red blood cells, and every
100 mL of blood from reinfusion corresponded to 54 mL of red blood cells.
Laboratory measurements, including hematocrit and hemoglobin in venous blood samples, were recorded at the assessment office
visit when surgery was planned (30±7 days before admission), at admission (day -1) in patients treated preoperatively with
epoetin-alpha, on the morning of postoperative day 1 (specifically between 15 and 21 hours postoperatively), and at postoperative
day 7 (±1 day).
Length of hospital stay and number and type of postoperative complications were recorded for each patient. Major complications
were defined as death within 90 days postoperatively, perioperative myocardial infarction, cerebrovascular accident, and symptomatic
Global costs of the blood plan per TKA were computed with the same index prices as those used in a previously reported algorithm,
which involved the variables epoetin-alpha, blood-giving sets, and blood cell units. Our clinic financed tranexamic acid,
blood cell units, and postoperative reinfusion systems. Epoetin-alpha is covered by national health insurance in France.
Data on the series were compared using the 2-sided unpaired parametric (
t test) and nonparametric (Mann-Whitney, chi-square, or Fisher exact) tests.
P<.05 was considered statistically significant.
Blood Loss and Blood Administration
Mean total blood loss throughout the 7-day perioperative period was 1490±636 mL in group A and 1828±559 mL in group B. Total
blood loss was 1252±605 mL in the subgroup of 21 patients who had received preoperative epoetin-alpha and 1570±615 mL for
the remaining 69 patients who did not receive epoetin-alpha. Postoperatively, only 1 patient received an allogeneic blood
transfusion in group A. This patient (aged 89 years) was operated on without the aid of a tourniquet or tranexamic acid because
of severe pre-existing arterial disease. The remaining 89 patients did not receive any blood administration. In group B, most
of patients (41 of 44) received autologous blood administration from fluid collected by reinfusion drain. The mean reinfused
blood volume from drainage was 315 mL (range, 0–800 mL). Blood volume recovered by drain was insufficient to be reinfused
in the 3 remaining patients. Of 41 patients who received autologous blood from postoperative drainage, 2 received an additional
allogeneic blood transfusion. Finally, only 3 patients in group B did not receive blood administration (Table ).
Table 3. Comparison of Blood Loss and Transfusion Rates
Baseline hemoglobin values at screening (30±7 days before admission) were 13.6±1.2 g/dL in group A and 13.7±1.2 g/dL in group
B. In group A, mean hemoglobin values were 14.17 g/dL (SD 0.95) on the day before surgery due to the preoperative management
of anemia in patients with mild to moderate anemia.
Postoperative hemoglobin levels were significantly higher in patients operated on under the blood transfusion-sparing plan
than those under reinfusion drainage. The lowest hemoglobin levels recorded during the postoperative period were 8.7 g/dL
in group A and 7.4 g/dL in group B. Average hemoglobin reduction was significantly less in group A than in group B at day
1 (12.4±1.0 g/dL vs 11.7±1.2 g/dL;
P=.002) and day 7 (11.0±1.2 g/dL vs 10.4±1.1 g/dL;
Figure:. Graph Showing Hemoglobin Concentrations in the Preoperative Period and on Postoperative Days 1 and 7 in Groups A and B. Boxes
Are Joined by Lines Through Their Means. Day 1 Is the First Postoperative Day and Day 7 Is the Seventh Postoperative Day After
the Day of Surgery.
No major bleeding complications emerged in the immediate postoperative period in either group. No patient returned to the
operating room for evacuation and debridement of hematoma. In group B, 1 patient developed a knee infection 3 months postoperatively.
No knee infection was identified at 1-year follow-up in group A.
The cost of the blood plan per TKA was higher in group A than in group B, due to the use of epoetin-alpha in group A (Table
). However, the length of hospital stay was shorter in group A than in group B (8.7±2.3 days [range, 6–17 days] vs 11.5±3.8
days [range, 7–20 days], respectively).
Table 4. Cost Comparison
Total knee arthroplasty is associated with significant perioperative blood loss. Various techniques can be deployed to minimize
this loss and consequently reduce the need for allogeneic transfusion. Most techniques for minimizing perioperative allogeneic
transfusion involve the reinfusion of autologous blood either from preoperative deposit, cell salvage, or reinfusion drains.
However, these techniques likely only have a modest blood-sparing effect when a rigorous transfusion protocol is used.
Data from several randomized trials have provided convincing evidence of the efficacy of intraoperatively administering the
antifibrinolytic drug tranexamic acid to reduce perioperative bleeding in patients treated with TKA.
In the current study, we compared 2 consecutive series of unilateral cemented TKAs undertaken with radically different blood
plans. The results showed that routine TKA can be performed with a marginal risk of blood transfusion by using chemoprophylaxis
to diminish blood loss and transfusion requirements. The rate of blood transfusion was nil in patients who had received either
preoperative epoetin-alpha or intraoperative tranexamic acid during TKA. No cases of deleterious anemia emerged postoperatively,
despite the absence of autotransfusion.
It is recommended that preoperative anemia be detected, evaluated, and managed to improve outcomes in the elective orthopedic
Interestingly, postoperative hemoglobin levels were significantly higher in patients operated on under the blood transfusion-sparing
strategy than in those receiving reinfusion drains. We prescribed preoperative epoetin-alpha in patients with mild to moderate
anemia and administrated a single intraoperative dose of tranexamic acid in all patients, unless contraindications were present.
The proportion of patients who received epoetin-alpha in group A (21 of 90) is concordant with the prevalence of anemia (20%)
in patients undergoing total joint arthroplasty.
The administration of epoetin-alpha was associated with an increase in the preoperative blood hemoglobin concentration to
almost 2 g/dL between the assessment visit and the day before surgery (12.27±0.77 g/dL and 14.14±0.82 g/dL, respectively).
This value is similar to that reported in the European Epoetin Alfa Surgery Trial.
The difference in the average hemoglobin at day 7 between groups A and B was minor but statistically significant (11.0±1.2
P=.003). However, this finding is nonetheless clinically relevant for patients with preexisting cardiovascular disease, given
that the trigger for transfusion in these patients is not 8 g but rather 10 g.
Therefore, the use of chemoprophylaxis can potentially increase the safety gap in these patients.
In the current study, we administrated intraoperatively a single dose of tranexamic acid in all patients in group A, unless
contraindications were present. We did not repeat the bolus or continuous infusion of tranexamic acid because we preferred
to use a simple clinical protocol when we planned our study. The mean total blood loss calculated throughout the 7-day perioperative
period was 1437±582 mL for the 83 patients who received tranexamic acid in group A, and 1935±496 mL for the remaining 9 patients
who did not receive tranexamic acid due to contraindications. These values are similar to those reported elsewhere in the
There were no epoetin-alpha- nor tranexamic acid-related complications in our study. Although concerns exist about the risk
of thrombotic events with tranexamic acid, large meta-analyses indicate that tranexamic acid can be safely and efficaciously
used to decrease perioperative blood loss and transfusion requirements.
Since we adopted tranexamic acid at our institution in 2005, we have exercised caution with respect to its contraindications
by administering potent anticoagulants and using echo-Doppler to provide systematic postoperative venous screening of the
lower limbs. We feel these precautions are necessary to avoid compromising the future development of tranexamic acid in orthopedic
Wound drainage and the use of a reinfusion drain with unilateral primary TKA remains controversial in the orthopedic community.
Prior to adopting tranexamic acid in our unit, we used reinfusion drains as the main blood-sparing strategy for patients
undergoing TKA. After implementing tranexamic acid, we discerned that the volume collected by drainage was extensively reduced
and insufficient for autotransfusion, an observation that has also been made elsewhere in literature.
The decision was therefore made to abandon reinfusion drains in primary TKA, as it was in a separate analysis,
given that this device represents an unnecessary expense. Additionally, we had already ceased intra-articular wound drainage
in TKA. Our study design benefits from comparing 2 uniform series of TKAs in which identical prostheses and anesthetic practices
were used. One surgeon was responsible for all the operations, which is also an advantage considering the influence of potential
confounding variables such as patient selection, surgical technique, incision length, duration of ischemia, and postoperative
course management. Remarkably, the same anticoagulant was exclusively used in both study groups.
Conversely, our investigation also has several notable potential limitations. An important weakness of the study was the nonrandomized
and partially retrospective study design that may have introduced bias to the results. Relevant differences in baseline characteristics
(ie, patient age) between the 2 groups also may have confounded outcomes. In addition, our results were obtained in a population
of patients with a relatively elevated mean BMI. It was reported by Salido et al
that patient weight represents a significant factor in the need for blood transfusion in total joint replacement. Therefore,
it is unlikely that our results will be replicated in a patient population with a lower BMI.
Our study was not intended to deliver a specific cost analysis. As we wished to evaluate cost differences in TKA undertaken
with and without autotransfusion, we compared the cost differences between our series of TKA undertaken without autotransfusion
and the series by Martinez et al
undertaken with a variety of perioperative autotransfusion systems. By using the same nonactualized prices of epoetin-alpha,
blood-giving sets, and red blood cell units, we found a cost-saving difference of approximately €200 for the mean global cost
per TKA in favor of our blood plan.
In our study, hospitalization duration was reduced by approximately 3 days in patients who did not receive venous blood during
TKA. However, the length of hospital stay is a complex parameter and is known not to be purely determined by clinical factors.
The present study demonstrated that a global blood-sparing transfusion strategy is an efficient way of decreasing blood transfusion
requirements in the routine daily setting of unilateral cemented primary TKA.
Our blood-sparing transfusion plan comprised 2 main elements:
Detection, evaluation, and management of preoperative anemia in the elective orthopedic patient, and
Minimization of perioperative erythrocyte loss by administering intraoperative tranexamic acid and by avoiding intra-articular
drainage at surgery.
To date, accumulating evidence that tranexamic acid is not a supplementary risk factor for thrombotic events contrasts with
growing proof that blood transfusion increases the risk for poorer clinical outcomes. The potential savings associated with
keeping patients transfusion free is considerable when set against the cumulative risks of blood transfusion and the contemporary
shortfalls in allogeneic blood availability.
A strategy that properly screens patients and deploys chemoprophylaxis may limit blood exposure, reduce postoperative anemia,
and decrease the financial costs of blood management in routine TKA.
Future studies are needed to confirm whether this blood strategy confers a benefit in conventional TKA (cemented and uncemented)
in both experienced and less-experienced surgeons and in populations with a lower BMI.
The authors thank Bernard Liné for the computerized calculation of blood losses and Ovid Da Silva for revising the manuscript.
- 1. Lemaire R. Strategies for blood management in orthopaedic and trauma surgery.
J Bone Joint Surg Br. 2008; 90(9):1128–1136. doi: 10.1302/0301-620X.90B9.21115
- 2. Evans RPAmerican Academy of Orthopaedic Surgeons Patient Safety Committee. Surgical site infection prevention and control: an emerging paradigm.
J Bone Joint Surg Am. 2009; 91(Suppl 6):2–9. doi: 10.2106/JBJS.I.00549
- 3. Benoni G, Fredin H. Fibrinolytic inhibition with tranexamic acid reduces blood loss and blood transfusion after knee arthroplasty: a prospective,
randomised, double-blind study of 86 patients.
J Bone Joint Surg Br. 1996; 78(3):434–440.
- 4. Vuillaume C, Fuzier R, Magues JP, Richez AS, Bataille B, Bonnavialle P. Blood conservation practices in primary total hip and total knee arthroplasty: a French survey [in French].
Rev Chir Orthop Trauma. 2010; (96):242–247.
- 5. Hourlier H, Fennema P, Liné B. A novel blood-saving plan for less-invasive primary total hip replacement.
Orthopedics. 2008; 31(12 Suppl 2). pii:
- 6. Gross JB. Estimating allowable blood loss: corrected for dilution.
Anesthesiology. 1983; 58(3):277–280. doi: 10.1097/00000542-198303000-00016
- 7. Gilcher RO, Belcher AJ. Predeposit programs. In: Sander G, Silvergleid AJ, eds.
Autologous Transfusion. Arlington, VA: American Association of Blood Banks; 1983:11–22.
- 8. Camarasa MA, Ollé G, Serra-Prat M, et al. Efficacy of aminocaproic, tranexamic acids in the control of bleeding during total knee replacement: a randomized clinical
trial [published online ahead of print March 10, 2006].
Br J Anaesth. 2006; 96(5):576–582. doi: 10.1093/bja/ael057
- 9. Martinez V, Monsaingeon-Lion A, Cherif K, Judet T, Chauvin M, Fletcher D. Transfusion strategy for primary knee and hip arthroplasty: impact of an algorithm to lower transfusion rates and hospital
costs [published online ahead of print October 9, 2007].
Br J Anaesth. 2007; 99(6):794–800. doi: 10.1093/bja/aem266
- 10. Henry DA, Carless PA, Moxey AJ, et al. Antifibrinolytic use for minimising perioperative allogeneic blood transfusion.
Cochrane Database Syst Rev. 2011; (1):CD001886.
- 11. Eubanks JD. Antifibrinolytics in major orthopaedic surgery.
J Am Acad Orthop Surg. 2010; 18(3):132–138.
- 12. Goodnough LT, Maniatis A, Earnshaw P, et al. Detection, evaluation, and management of preoperative anaemia in the elective orthopaedic surgical patient: NATA guidelines.
Br J Anaesth. 2011; 106(1):13–22. doi: 10.1093/bja/aeq361
- 13. Saleh E, McClelland DB, Hay A, Semple D, Walsh TS. Prevalence of anaemia before major joint arthroplasty and the potential impact of preoperative investigation and correction
on perioperative blood transfusions [published online ahead of print October 23, 2007].
Br J Anaesth. 2007; 99(6):801–808. doi: 10.1093/bja/aem299
- 14. Weber EW, Slappendel R, Hémon Y, et al. Effects of epoetin alfa on blood transfusions and postoperative recovery in orthopaedic surgery: the European Epoetin Alfa
Surgery Trial (EEST).
Eur J Anaesthesiol. 2005; 22(4):249–257. doi: 10.1017/S0265021505000426
- 15. Carson JL, Terrin ML, Barton FB, et al. A pilot randomized trial comparing symptomatic vs. hemoglobin-level-driven red blood cell transfusions following hip fracture.
Transfusion. 1998; 38(6):522–529. doi: 10.1046/j.1537-2995.1998.38698326331.x
- 16. Foss NB, Kristensen MT, Jensen PS, Palm H, Krasheninnikoff M, Kehlet H. The effects of liberal versus restrictive transfusion thresholds on ambulation after hip fracture surgery.
Transfusion. 2009; 49(2):227–234. doi: 10.1111/j.1537-2995.2008.01967.x
- 17. Alvarez JC, Santiveri FX, Ramos I, Vela E, Puig L, Escolano F. Tranexamic acid reduces blood transfusion in total knee arthroplasty even when a blood conservation program is applied [published
online ahead of print December 7, 2007].
Transfusion. 2008; 48(3):519–525. doi: 10.1111/j.1537-2995.2007.01564.x
- 18. Hazarika S, Bhattacharya R, Bhavikatti M, Dawson M. A comparison of post-op haemoglobin levels and allogeneic blood transfusion rates following total knee arthroplasty without
drainage or with reinfusion drains.
Acta Orthop Belg. 2010; 76(1):74–78.
- 19. Parker MJ, Roberts CP, Hay D. Closed suction drainage for hip and knee arthroplasty. A meta-analysis.
J Bone Joint Surg Am. 2004; 86(6):1146–1152.
- 20. Salido JA, Marín LA, Gómez LA, Zorrilla P, Martínez C. Preoperative hemoglobin levels and the need for transfusion after prosthetic hip and knee surgery: analysis of predictive
J Bone Joint Surg Am. 2002; 84(2):216–220.
|No. of TKAs
|Mean age, y
|Mean BMI, kg/m
|Mean baseline hemoglobin, g/dL
|Mean ASA score
|Mean operation time, min
|Mean incision length, cm
||Postoperative salvage and reinfusion
| Ferrous sulfate
|Intraoperative tranexamic acid administration
||All except in cases of contraindications
Comparison of Blood Loss and Transfusion Rates
|Mean total blood loss in 7-day period, mL
|No. of patients receiving allogeneic and/or autologous blood after TKA
|No. of patients receiving allogeneic blood units
|Mean volume reinfused blood, mL
|No. of TKAs
|Red blood cell units (a)
|Tranexamic acid injections (c)
|Epoetin-alpha injections (d)
|Mean hospital cost per TKA (a+b+c), €
|Mean out-of-hospital costs per TKA (d), €
|Mean global cost of TKA (a+b+c+d), €