Statistics from Altmetric.com
Hip and knee arthroplasties are routine orthopedic procedures commonly associated with acute postoperative anemia, and in many cases require an allogeneic or autologous blood transfusion. Several techniques have been used to limit postoperative blood loss and risk of transfusion. Tranexamic acid (TXA) is an antifibrinolytic agent that has fundamentally changed blood management in total joint arthroplasty (TJA) by making transfusion an infrequent event.
Although TXA is widely used in total hip arthroplasty (THA) and total knee arthroplasty (TKA), it has not yet become the standard of care. A significant body of literature has been compiled on the use of TXA in hip and knee arthroplasty but a comprehensive review and analysis of the existing evidence to provide clinical guidance is lacking. Therefore, the American Association of Hip and Knee Surgeons (AAHKS), the American Academy of Orthopaedic Surgeons (AAOS), the Hip Society, the Knee Society and the American Society of Regional Anesthesia and Pain Medicine (ASRA) have worked together to develop evidence-based guidelines on the use of TXA in primary TJA. The purpose of these guidelines is to improve the treatment of orthopedic surgical patients and reduce practice variation by promoting a multidisciplinary evidence-based approach on the use of TXA.
The combined clinical practice guidelines are meant to address common and important questions related to the efficacy and safety of TXA in primary TJA. Using the AAOS Clinical Practice Guidelines and Systematic Review Methodology, the committee members completed a series of direct meta-analyses and network meta-analyses to support the clinical practice guidelines.1 For each question, we have provided a recommendation, assessed the strength of the recommendation and elaborated on the rationale of the recommendation, which should be interpreted in accordance with the AAOS Clinical Practice Guidelines and Systematic Review Methodology.1 The current clinical practice guidelines were based on the available evidence, so future updates may become necessary as additional literature becomes available with future research.
Guideline question 1
For patients undergoing primary TJA, what method of administration of TXA, compared with placebo, reduces the risk of transfusion and/or reduces blood loss?
Administration of intravenous (, topical, and oral TXA as well as combinations of individual formulations of TXA are all effective strategies when compared with placebo for reducing calculated blood loss and the need for transfusion during the perioperative episode of a primary TJA.
Strength of recommendation
The direct meta-analysis of one moderate-quality and 82 high-quality studies provided significant evidence for the ability of TXA to reduce the risk of blood loss and need for transfusion during the perioperative episode of primary hip and knee arthroplasties.2 Subsequent network meta-analysis supported the blood-sparing properties of TXA.2
Total hip arthroplasty
Intravenous and topical TXA have been shown with limited heterogeneity in direct meta-analysis to reduce blood loss.2 Similarly, intravenous and topical TXA were found to reduce the risk of transfusion compared with placebo by 60% and 71%, respectively.2 Network meta-analysis of low-dose intravenous (<20 mg/kg or ≤1 g), high-dose intravenous (≥20 mg/kg or >1 g), high-dose topical (>1.5 g), oral and combined intravenous/topical TXA reduced the risk for blood loss compared with placebo.2 Correspondingly, network meta-analysis demonstrated low-dose intravenous, high-dose intravenous, high-dose topical, low-dose topical and combined intravenous/topical TXA to significantly reduce the risk of transfusion.2
Due to a lack of studies directly comparing oral TXA with placebo, no conclusions could be derived in the direct meta-analysis.2 Network meta-analysis was performed to provide an indirect comparison of oral TXA, which demonstrated significantly reduced blood loss compared with placebo.2 Although oral TXA was shown to be equivalent regarding risk of transfusion to all other formulations of TXA in the network meta-analysis, oral TXA did not reduce the risk of transfusion compared with placebo. The inconsistency is the result of the network relying on indirect comparisons and a limited number of studies. Similar limitations were observed for lower doses (≤1.5 g) of topical TXA in the network meta-analysis.2 In the analysis, low-dose topical TXA did not significantly reduce blood loss but did reduce the risk of transfusion compared with placebo.2
Total knee arthroplasty
Intravenous and topical TXA consistently demonstrated in direct meta-analysis the ability to reduce blood loss during the perioperative episode of a primary TKA.2 Topical and oral TXA reduced the risk of transfusion by 66% and 61% for each respective formulation of TXA when compared with placebo.2 Intravenous TXA administered as a single dose either before or after incision reduced the risk of transfusion by either 81% or 55% compared with placebo.2 When multiple doses of intravenous TXA were administered, the observed reduction in transfusion was 75% compared with placebo.2 Subsequent network meta-analysis demonstrated low-dose intravenous, high-dose intravenous, low-dose topical, high-dose topical and oral TXA as well as combinations of intravenous/topical and intravenous/oral TXA to reduce blood loss and risk of transfusion during the perioperative episode of a TKA.2
Guideline question 2
For patients undergoing primary TJA, what method of administration of TXA, compared with a different method of administration, reduces the risk of transfusion and/or reduces blood loss?
The analysis of studies did not identify a clearly superior method, or combinations of methods, for the administration of TXA. All methods of administration effectively demonstrate equivalent efficacy at reducing calculated blood loss and the risk of transfusion during the perioperative episode of a primary TJA.
Strength of recommendation
The direct meta-analysis of 31 high-quality studies provided no evidence to favor any specific method of TXA to reduce the risk of blood loss and need for transfusion during the perioperative episode of primary hip and knee arthroplasties.2 Subsequent network meta-analysis included a more expansive comparison between the methods of TXA administration with no evidence to clearly support a superior method of administration.2
Total hip arthroplasty
Intravenous and topical TXA have been compared together in multiple randomized clinical trials, which through direct meta-analysis showed no difference in the risk of transfusion.2 Network meta-analysis provided the opportunity to perform direct and indirect comparisons between low-dose intravenous (<20 mg/kg or ≤1 g), high-dose intravenous (≥20 mg/kg or >1 g), low-dose topical (≤1.5 g), high-dose topical (>1.5 g), oral and combined intravenous/topical TXA.2 In terms of the ability to reduce blood loss, no method of TXA administration was found to provide a significantly different outcome.2 Similar results in the network meta-analysis were observed for risk of transfusion with the exception that a combination of intravenous and topical TXA was equivalent to oral TXA but superior to low-dose intravenous, high-dose intravenous, low-dose topical and high-dose topical TXA.2 However, the inconsistent result regarding combined intravenous/topical TXA likely represents bias from a limited number of studies and not superiority to other methods of TXA administration.
Total knee arthroplasty
Direct comparisons were performed between intravenous TXA and topical, oral or combined intravenous/topical TXA, which found no difference in the risk of transfusion.2 Similar to the network meta-analysis of THA, direct and indirect comparisons were performed between low-dose intravenous, high-dose intravenous, low-dose topical, high-dose topical and oral TXA as well as combinations of intravenous/topical and intravenous/oral TXA that resulted in no difference in their blood-sparing properties.2 The significant differences between methods of TXA administration were observed in respect to the risk of transfusion being higher for low-dose intravenous TXA compared with high-dose intravenous or combined intravenous/topical TXA, which could represent a dose response or the limited number of studies.2
Guideline question 3
For patients undergoing primary TJA, does the dose amount of intravenous or topical TXA affect the risk of transfusion and/or reduction in blood loss?
Within the context of the TXA doses used in primary TJA, the dose amount of TXA was not found to significantly affect its reduction of calculated blood loss or the need for transfusion during the perioperative episode of a primary TJA.
Strength of recommendation
Due to a limited number of studies, direct meta-analysis could not be performed on the six high-quality publications that solely investigated the dose effect of either intravenous or topical TXA.2 Only two of the published studies observed a difference in favor of higher doses of intravenous or topical TXA.3 4 However, two different studies investigating the same comparative doses of topical TXA (1.5 and 3 g) in primary TKA did not observe a difference in calculated blood loss or risk of transfusion with higher doses of topical TXA.5 6 Additionally, two publications of intravenous TXA in THA or TKA did not favor higher doses of intravenous TXA.7 8 When a network meta-analysis was performed regarding the dose effect of intravenous or topical TXA, it demonstrated limited evidence for a reduction in either transfusion risk or calculated blood loss with higher doses of TXA.2
Although a dose–response has only been observed in the analysis of TKA for intravenous TXA with the risk of transfusion, it does not preclude the presence of a dose–response for intravenous or topical TXA. The anticipated dose–response is likely not present at the levels of blood loss or the doses of TXA used for primary THA or TKA. In the network meta-analysis of primary THA and TKA, the range of intravenous TXA doses was 10 mg/kg and three doses of 15 mg/kg while for topical TXA doses the range was between 0.5 and 3 g.2
Network meta-analysis demonstrates no additional reduction in blood loss following a hip or knee arthroplasty with high-dose intravenous (≥20 mg/kg or >1 g) TXA compared with low-dose intravenous (<20 mg/kg or ≤1 g) TXA.2 A dose–response was observed in the network meta-analysis with a reduced risk of transfusion for higher doses of TXA in primary TKA, but similar results were not found for primary THA.2 Additional evidence shows no significant difference comparing single-dose and multiple doses of intravenous TXA, which further supports the notion that higher doses of intravenous TXA are not necessarily clinically needed to improve the blood-sparing effects in the setting of primary hip or knee arthroplasties.2
Direct meta-analysis and network meta-analysis consistently reported no statistical difference between low-dose (≤1.5 g) and high-dose (>1.5 g) topical TXA.2 A dose–response was not observed for either blood loss or risk of transfusion regarding low-dose or high-dose topical TXA following hip or knee arthroplasty.2
Guideline question 4
For patients undergoing primary TJA, do multiple doses of or oral TXA, compared with a single dose, reduce the risk of transfusion and/or reduce blood loss?
Administration of multiple doses of intravenous or oral TXA compared with a single dose of intravenous or oral TXA does not significantly alter the amount of calculated blood loss and need for transfusion during the perioperative episode of a primary TJA.
Strength of recommendation
Direct meta-analysis was performed using six high-quality studies to compare between a single dose and multiple doses of intravenous TXA during primary TKA, which demonstrated no additional benefit for multiple doses of intravenous TXA.2 Due to a limited number of studies, direct meta-analysis could not be performed for primary THA. However, the results of the available individual studies consistently demonstrate no additional benefit for multiple doses of intravenous TXA in primary THA.7 9 10
An expanded comparison was performed through a network meta-analysis on primary hip and knee arthroplasties. Multiple doses of oral TXA have not been studied in primary THA, thus analysis was only available for intravenous TXA in primary THA. Network meta-analysis showed no additional benefit in terms of blood loss or risk of transfusion for multiple doses of intravenous TXA compared with a single dose.2 Similarly, the results for primary TKA supported no benefit for multiple doses of intravenous or oral TXA compared with a single dose of either intravenous or oral TXA.2
Guideline question 5
For patients undergoing primary TJA, does the timing of the administration of TXA in relation to the surgical incision affect the ability of TXA to reduce the risk of transfusion and/or blood loss?
In primary TJA, administration of intravenous TXA before the incision potentially reduces blood loss and the need for transfusion compared with its administration after incision.
Strength of recommendation
Direct meta-analysis investigating the effect on timing of TXA administration was available in five high-quality studies.2 Due to a limited number of studies, direct meta-analysis was only performed for primary TKA.2 The result showed no significant difference between preincision and postincision administration. However, trends towards significance were identified.2 Network meta-analysis for primary TKA demonstrated that a single dose of intravenous TXA before incision reduced the risk of transfusion compared with administration after incision; however, there was no significant difference regarding blood loss.2 In the setting of a primary THA, network meta-analysis failed to show any significant difference in blood loss or risk of transfusion between preincision and postincision administration of TXA.2
Due to the inconsistency in the results of high-quality studies, we are only able to provide ‘moderate’ support to our recommendation. Despite the inconsistencies, we still recommend preincision TXA administration because the results demonstrate no potential benefit for postincision administration. In contrast, preincision administration does demonstrate benefit in some of the analyses.
Guideline question 6
For patients undergoing primary TJA without a history of a venous thromboembolic event (VTE), does the treatment with TXA affect the risk of VTE?
Administration of intravenous, topical and oral TXA in patients without a known history of a VTE does not increase the risk of developing a VTE compared with placebo during the perioperative episode of a primary TJA.
Strength of recommendation
Direct meta-analysis investigating the impact of TXA administration on the risk of VTE was performed using 77 high-quality and one moderate-quality randomized clinical trial. Of those, 92% of the studies used history of a thromboembolic event as an exclusion criterion.2 Individual results for hip and knee arthroplasty demonstrated no significant difference between all methods of TXA administration compared with placebo.2 Since individual studies included were noted to be comprising small sample sizes, the hip and knee arthroplasty populations were combined to analyze for differences between intravenous and topical application of TXA in order to comment on the incidence of rare complications like VTE.2 This combined analysis further supports the group’s conclusion that there is no evidence of increased risk of VTE with TXA administration (combined results had a relative risk closer to ‘no difference’ than the subgroup analysis).2 Given this overwhelming evidence, we provide ‘strong’ support to our recommendation that TXA administration at doses typically used in hip and knee arthroplasty is not associated with an increased risk of VTE for patients without a known history of VTE.
Guideline question 7
For patients undergoing primary TJA with a history of a VTE, myocardial infarction (MI), cerebrovascular accident (CVA), transient ischemic attack (TIA) and/or vascular stent placement, does the treatment with TXA affect the risk of VTE?
There is a paucity of randomized clinical trials on the risk of adverse effects of intravenous, topical and oral TXA in patients with known history of a VTE, MI, CVA, TIA and/or vascular stent placement. The existing high-quality literature regarding administration of TXA in patients of generally higher comorbidity burden does not suggest increased risk of adverse thromboembolic events during the perioperative episode of a primary TJA.
Strength of recommendation
Despite an established proposed mechanism of action for TXA as a fibrin clot stabilizer, clinician concerns remain over the use of any antifibrinolytic medication in patients considered at ‘high-risk’ for thromboembolic events (eg, history of VTE, MI with vascular stents, cerebral vascular occlusive disease), which continues to limit the widespread adoption of TXA use in hip and knee arthroplasty.11 Since no clinical trials have investigated specific risk factors, the American Society of Anesthesiologists (ASA) physical status classification system was used as a proxy to identify ‘high-risk’ patients among the literature available. In a metaregression analysis comparing a population of patients with greater than 50% ASA status ≥3 to another population with patients of greater than 50% ASA status I or II, the results demonstrated no increase in the risk of VTE for patients undergoing a primary hip or knee arthroplasty.2 Due to the absence of experimental evidence, we also reviewed observational studies on the topic of TXA administration in patients with specific risk factors. Large database studies suggest TXA administration in patients with a history of VTE or ASA status of ≥3 does not experience an increased risk of VTE.12–15
In the ‘high-risk’ patient population, we must consider the summation of the benefits and potential risks of administering TXA. Despite limited data on the safety of TXA, we wish to highlight a parallel lack of evidence for harm. Additionally, evidence has demonstrated that postoperative cardiovascular complications are associated with anemia and high blood transfusion rates.16–18 Therefore, we wish to highlight the overall positive summation of data on TXA efficacy in the context of limited evidence for added risk with the use of TXA. In sum, the calculation of the number needed to harm for VTE among the total joint population was 983 patients, but the number needed to treated with intravenous TXA in THA and TKA to prevent a transfusion was only four and three patients, respectively.2 Although the available data limit for stronger advocacy and more widespread use in those at ‘high-risk’, each patient with known risks factors should be considered individually and we advocate for a multidisciplinary approach when deciding whether to withhold or administer TXA.
Guideline question 8
For patients undergoing primary TJA, does the treatment with TXA affect the risk of arterial thromboembolic events (ATE)?
There is a paucity of randomized clinical trials on the risk of ATE due to the administration of TXA intravenously, topically and orally. However, the existing evidence does not suggest that TXA increases the risk of developing an ATE compared with placebo during the perioperative episode of a primary TJA.
Strength of recommendation
ATEs following primary hip or knee arthroplasty are exceptionally infrequent complications and are rarely reported in the randomized clinical trials as most patients who have known risk factors for these events are often excluded from these types of studies. In a direct meta-analysis investigating arterial and VTEs, only nine studies reported the incidence of ATE.2 Due to the limited number of studies, the authors performed combined hip and knee arthroplasty direct meta-analysis only. The results demonstrated no statistical difference in the risk of ATE between intravenous or topical TXA and placebo.11
Although the direct meta-analysis was limited to high-quality studies, we are only able to provide ‘moderate’ support to our recommendation. The combination of paucity of data, lack of studies specifically designed to investigate the complication of ATE and exclusion of patients with known risk factors among those studies that were available diminished our ability to provide for a stronger recommendation and an individual risk benefit assessment will be necessary.
As highlighted in the individual network meta-analyses for hip and knee arthroplasties, there exists a relatively small amount of literature on the use of oral TXA and low-dose topical TXA. Therefore, we believe there still exists an opportunity to meaningfully contribute to the current literature on TXA by investigating those formulations of TXA. However, if oral TXA is used, we suggest the use of 2 g of oral TXA approximately 2 hours before the desired effect of the medication to ensure appropriate pharmacokinetics is obtained.
The most substantial shortcoming of the current literature on TXA involves the lack of high-level evidence to support the use of TXA in patients with a history of a VTE, MI, CVA, TIA and/or vascular stent placement. Although well-designed retrospective comparison studies have been published, we lack the ability to provide stronger support on the use of TXA in these patients considered to be high risk. Therefore, we would encourage future research to determine the safety of TXA in the high-risk patient. Additionally, TJA is being performed more frequently in patients with a history of ‘chronic’ cancers such as prostate cancer. Similar to other patient factors, we lack the ability to comment on the safety of TXA in this particular subset of patients, and we encourage further research on these patients.
Peer review process
Following the committee’s formulation of the Clinical Practice Guideline draft, it underwent a peer review by the board of directors from the AAHKS, the ASRA, the Hip Society and the Knee Society. The AAOS Evidence-Based Quality and Value Committee reviewed the Clinical Practice Guideline draft for endorsement. Additionally, the three publications of the direct and network meta-analyses on the efficacy and safety of TXA in primary hip and knee arthroplasties that supported the formulation of the Clinical Practice Guideline have undergone peer review for publication.
We thank AAHKS for providing the funding and administrative support. We thank Jayson Murray, Peter Shores and Kyle Mullen from the AAOS Department of Research, Quality, and Scientific Affairs for their assistance with the analysis and guidance. Lastly, we thank the leadership of the AAHKS, AAOS, ASRA, the Hip Society and the Knee Society for help with organizational support.
Presented at This work was presented in part at the 2017 annual meeting of the American Association of Hip and Knee Surgeons, Dallas, TX, 2–5 November 2017; and the 2018 annual meeting of the American Academy of Orthopaedic Surgeons, New Orleans, 6–10 March 2018.
Funding The American Association of Hip and Knee Surgeons provided financial support for the cost of the statistical analysis.
Disclaimer All authors or contributors to the Clinical Practice Guideline have provided a disclosure statement in accordance with the publicly available AAOS Orthopaedic Disclosure Program. All authors and contributors attest none of the disclosures present are relevant to the Clinical Practice Guidelines.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Author note Tranexamic acid is a drug described in this Clinical Practice Guideline that has only been approved by the Food and Drug Administration (FDA) for dental bleeding prophylaxis in patients with hemophilia and menorrhagia. According to the FDA, it is the prescribing physician's responsibility to ascertain the FDA clearance status for all medications prior to use in a clinical setting.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.