Total knee arthroplasty (TKA) has the potential for major blood loss. To decrease perioperative bleeding and transfusions surgeons are not only examining technique, but also considering pharmacologic agents that minimize bleeding through the inhibition of blood clot degradation. Antifibrinolytics function to decrease perioperative blood loss by inhibiting clot degradation. Tranexamic acid (TXA) is an antifibrinolytic that saturates the plasminogen-binding sites to fibrin, thereby inhibiting plasminogen from dissolving the clot. Previous studies have examined the effect of TXA on blood loss and transfusion requirements for unilateral TKA patients.1-4 Dhillon and colleagues evaluated the use of TXA in single-stage, bilateral, simultaneous TKAs and found a reduction in blood loss and decreased blood transfusion requirements. Our purpose was to evaluate the effect of TXA on blood loss, hemoglobin (Hb) changes, and transfusion in patients undergoing bilateral TKA staged 3 days apart.
- Top of page
- Materials and Methods
- Conflict of Interest
There were significant differences between subjects and controls in terms of blood loss, postoperative Hb levels, and transfusion requirements. Subjects (those receiving TXA) had a significantly lower (p < 0.001) mean (±SD) total drain output than did controls: 373.8 (±264.6) and 871.6 (±457.7) mL, respectively (Table 2). Subjects also had significantly fewer (p < 0.001) mean allogenic blood units transfused: 0.60 (±0.84) and 1.53 (±1.30) units, respectively (Table 2). Significantly fewer (p = 0.003) subjects than controls required an allogenic blood transfusion: 22 of 51 (43%) and 50 of 70 (71%), respectively. Subjects also had significantly higher Hb levels than did controls on Days 1 (11.13 ± 1.22 g/dL vs. 10.49 ± 1.15 g/dL, p = 0.005) and 2 (11.05 ± 1.30 g/dL vs. 10.16 ± 0.95 g/dL, p = 0.002) after the first TKA and also on Days 1 (9.74 ± 1.26 g/dL vs. 8.53 ± 1.00 g/dL, p < 0.001) and 2 (9.90 ± 1.05 g/dL vs. 9.30 ± 0.93 g/dL, p = 0.001) after the second TKA. When adjusting for differences in preoperative Hb levels and Hct, the difference in mean total blood units between subjects and controls was still significant (mean difference, −0.82; SE, 0.19; p < 0.001).
Table 2. Operative characteristics*
|Characteristic||Control group (n = 70)||Study group (n = 51)||p value|
|Total drain output (mL)||871.6 (±457.7)||373.8 (±264.6)||<0.001|
|Blood transfusion required||50 (71.4)||22 (43.1)||0.003|
|Amount allogenic blood transfused (units)||1.53 (±1.30)||0.60 (±0.84)||<0.001|
Post hoc sample size estimates were calculated. Based on the pooled SD of total blood loss (SD, 615.8), approximately 46 participants would be required per group to detect as significant the observed difference in total blood loss between subjects and controls. Similarly, using the pooled SD of transfused allogenic blood units (SD, 1.10), approximately 42 participants would be required per group to detect as significant the observed difference in transfused allogenic blood units between subjects and controls.
Patient comorbidities were reviewed and Charlson comorbidity score was calculated for each group. No statistical difference was found between groups (p = 0.878).
Postoperative complications (Table 3), Knee Society Scores, and range-of-motion scores (Table 4) were recorded for each group.
Table 3. Postoperative complications
|Complication||Control group (n = 70)||Study group (n = 51)|
|Reoperation||1 (manipulation for arthrofibrosis)||1 (revision for acute infection)|
Table 4. Preoperative and postoperative Knee Society Scores and range of motion*
|Group||Knee Society Scores||Range of motion (°)|
|Control (n = 70)|| || || || || |
|Preoperative||8.2 (0-20)||54.5 (15-70)||44.6 (22-60)||3.5 (0-15)||123.6 (90-135)|
|Postoperative (6 weeks)||44.2 (10-50)||72.7 (30-100)||88.7 (46-100)||2.4 (0-15)||114 (70-135)|
|Study (n = 51)|| || || || || |
|Preoperative||10.4 (0-30)||57.2 (20-70)||42 (21-60)||4.8 (0-15)||121 (50-135)|
|Postoperative (6 weeks)||45.2 (30-50)||69.6 (30-100)||89.4 (59-100)||4.7 (0-20)||113.9 (87-138)|
- Top of page
- Materials and Methods
- Conflict of Interest
Blood loss is expected to be high for bilateral TKA, and the likelihood of a transfusion is higher for patients undergoing bilateral TKA than for those undergoing unilateral TKA. TXA has been proven to reduce blood loss1-4 and transfusion rates1-4 for TKA. Our study examined the effectiveness of TXA on reducing blood loss and reducing the transfusion rates for patients undergoing bilateral TKAs staged 3 days apart.
Our results confirm previously published studies of single TKAs with TXA showing a reduction in blood loss, higher postoperative Hb levels, and a decrease in the transfusion rate1-4 compared with no use of TXA. However, in addition, our study of patients undergoing bilateral TKAs has also shown that patients receiving TXA for each surgery had a significant decrease in total drain output compared with those who did not receive TXA.
Our study had some limitations. First, it was not a randomized, blinded study and the numbers were limited. However, even with our sample size, we were able to show significant differences in drain output, postoperative Hb levels, and transfusion rates. A larger, multicenter, randomized, blinded study would provide a better level of evidence but, because bilateral staged TKA is not very common, it may take several more years to complete a larger study. Although our sample size is limited, our data help support the use of TXA for bilateral TKA.
Second, our dosing of TXA was not weight based. Per our protocol, for ease of administration and because each vial is a 1-g dose, we used a dose of 1 g of TXA at the start of the case and 1 g when the tourniquet was deflated. Other studies have evaluated the dose of TXA and found that 10 mg/kg or 20 mg/kg TXA are effective doses. Although our dose resulted in a decrease in drain output and transfusion rate, our study may not be comparable with other studies that base TXA dosing on patient weight.
Third, we did not routinely perform venous ultrasonography, venogram, or lung perfusion scan to rule out DVT or pulmonary embolism. Therefore, it is possible that we missed a subclinical DVT or pulmonary embolism in these patients. We recognize that a much larger sample size would be required to determine any increased risk of DVT or pulmonary embolism with this protocol. We also do not have documentation of patient compliance with DVT prophylaxis (325 mg of aspirin twice a day). Nevertheless, symptomatic DVTs were not increased in our study group, which is consistent with findings in earlier studies.4,10-12 There was no adverse reaction to the TXA that we could identify. Whether TXA would increase thromboembolic events in patients with preexisting conditions has not been established. Our study excluded patients with existing conditions such as hypercoagulable states or a history of DVT.
Fourth, we did not attempt to calculate total blood loss by incorporating intraoperative blood loss estimates. We compared only postoperative blood loss through the hemovac drain output. Also, we did not standardize the exact timing of drain removal. All drains were routinely pulled on the morning of Postoperative Day 1, when there had been less than 100 mL of output per 8-hour shift, but we did not standardize the exact number of hours. However, all drains remained for at least 20 hours, with no drain remaining in place for more than 28 hours. Although the exact number of hours was not standardized, the protocol was the same for both groups, and we still found a significant decrease in drain output for patients who received TXA. The decreases in blood loss and transfusion rate also correlate with findings in previous studies.1-4
A relative strength of our study is that we evaluated the use of TXA in patients having bilateral staged TKAs. Previous studies evaluating the effectiveness of TXA have focused on single TKAs or bilateral concurrent TKAs.[5, 8, 14] Dhillon and colleagues found that TXA administered to patients undergoing single-stage bilateral TKAs reduced blood loss and decreased allogenic blood transfusion requirements, findings similar to ours. MacGillivray and coworkers evaluated different TXA dosing regimens for patients undergoing concurrent bilateral TKAs and found that a lower dose of TXA (10 mg/kg) was sufficient to provide a lower allogenic blood transfusion requirement when used in combination with an autologous blood transfusion strategy. Our study did not use an autologous reinfusion strategy, our dosing of TXA was not weight based, and we staged our TKAs apart by 3 days; however, similar to MacGillivray and colleagues, we also showed a decrease in allogenic blood transfusion requirement for patients undergoing bilateral TKAs. In a double-blind study, Kakar and colleagues found decreased drain output and transfusion rates when patients were given TXA for unilateral or bilateral TKAs.
In conclusion, we have shown that TXA reduces postoperative drain output, reduces postoperative transfusion requirements, and increases postoperative Hb for patients undergoing bilateral TKAs staged 3 days apart.