Tranexamic Acid Use

The molecule. Tranexamic acid (TXA) is a synthetic derivative of the amino acid lysine: trans-4-(aminomethyl)cyclohexanecarboxylic acid, C8H15NO2. Its structural likeness to lysine is what allows it to bind competitively, and it acts as a competitive inhibitor of fibrinolysis.

How it works. TXA binds reversibly to the lysine-binding sites on plasminogen. Plasminogen can then no longer bind fibrin, and tissue plasminogen activator (tPA) cannot convert it to plasmin. At higher concentrations TXA also inhibits plasmin directly, preventing degradation of the fibrin clot into fibrin degradation products (FDPs).

What that achieves. Formed clot is protected from premature breakdown, so surgical haemostasis is more effective and more durable. TXA inhibits fibrinolysis, not coagulation initiation: it protects formed clot rather than promoting new clot formation.

View larger

Why surgery switches fibrinolysis on. Tissue damage releases tPA, and contact with artificial surfaces, especially cardiopulmonary bypass, adds to it. This physiological response can be excessive and produce microvascular oozing, and surgical fibrinolysis peaks during and immediately after surgery. TXA counteracts it by stabilising clot at the surgical site.

Pharmacokinetics. The figures worth knowing:

• Peak plasma level immediately after an IV bolus and 1 hour after an oral dose; oral bioavailability approximately 45%
• Low protein binding (3%); distributes into the extravascular space, including joint fluid
• Half-life 2-3 hours, prolonged in renal impairment
• 90% excreted unchanged in the urine by glomerular filtration
• Antifibrinolytic effect at 10-15 micrograms/mL, lasting 3-6 hours after a single dose

Dose and effect. Low-to-moderate doses (10-25mg/kg) give an effective antifibrinolytic action without significant adverse effects, and 15-20mg/kg IV or a 1-2g flat dose provides therapeutic levels for typical orthopaedic procedures. Doses over 100mg/kg add no haemostatic benefit and raise the seizure risk (see Complications).

TXA is best understood against the other two antifibrinolytics. Aminocaproic acid shares its mechanism at lower potency and is the fallback; aprotinin works differently and is the cautionary tale.

Before giving TXA. Routine use in TKA or THA needs nothing beyond the standard preoperative workup, and routine coagulation testing is not required unless the patient is anticoagulated. The screen is mostly history, with eGFR the one value that changes the dose:

• Renal function (eGFR)
• Previous thromboembolic disease: DVT, PE, stroke, MI
• Current anticoagulant or antiplatelet therapy
• Seizure history
• Pregnancy or breastfeeding

Most orthopaedic patients are candidates.

Renal function and dose. Because TXA is renally excreted, impairment prolongs its half-life, to 4-8 hours, and the higher plasma levels increase the seizure risk. The dose follows the eGFR, and in a patient on dialysis it is given after dialysis.

Thrombophilia screening. No routine screening is recommended. Surgical meta-analyses show no clear increase in VTE with TXA at therapeutic doses, testing is expensive and rarely changes management, and a patient with known thrombophilia can still receive TXA with appropriate VTE prophylaxis. Individualised discussion is reserved for severe hereditary thrombophilia.

Monitoring. TXA needs no coagulation studies, D-dimer or fibrinogen monitoring, during or after surgery, and PT and aPTT are unaffected. Routine observations cover what little it can do:

Anything further is directed at a specific suspected adverse event:

The decision. Using TXA well is mostly about who should not have it, and when. Screen for contraindications, then choose a route, time the dose, decide whether a repeat is needed, and document the plan.

Algorithm

View larger

What the evidence settles. The Fillingham network meta-analyses underpinning the combined AAOS/AAHKS/Hip and Knee Society/ASRA guideline found topical, IV and oral TXA all superior to placebo, with no formulation, dose, dosing frequency or timing clearly superior in either TKA or THA. Moderate evidence supported pre-incision IV administration in TKA. Pre-incision IV is therefore a reasonable default, and the confident protocols in circulation are convention rather than demonstrated optimum.

Intravenous TXA gives a systemic effect that reaches every surgical site, with predictable plasma levels and the most extensive RCT data, and it is the standard approach in most institutions.

Topical TXA gives a high local concentration with minimal systemic absorption, although even topical use achieves measurable plasma levels. Xie et al. (Thrombosis Research 2017) found similar transfusion requirements with the two routes in THA and TKA, with a marginally smaller maximum haemoglobin drop with IV; there is no clinically significant difference between routes.

Choosing the route. Choose on patient factors and surgeon preference. In a standard patient either route is acceptable. Renal impairment favours topical or a reduced IV dose, a seizure history favours topical, and topical is preferred for any relative contraindication to systemic use because it reduces systemic exposure.

Combined IV and topical. IV pre-incision plus topical at closure has an attractive rationale, systemic plus high local concentration, but multiple RCTs show no significant additional benefit over a single route. It is safe and can be reserved for high-bleeding-risk cases such as revision or a complex primary.

Oral TXA, 1-2g preoperatively (for example 2g about 2 hours before surgery), is an emerging, cost-effective and less-studied alternative, with efficacy comparable to IV in TJA network meta-analysis.

Repeat dosing. Consider a repeat dose when the procedure runs over 3 hours (the half-life is 2-3 hours), with ongoing significant or unexpected blood loss, or in revision and complex reconstruction with a large surface area. Give half the initial dose, or the full dose at 3 hours, and do not exceed about 3g total IV. Routine postoperative dosing is not required; a single 1g IV dose at around 3 hours may be considered only for ongoing oozing or concern about late fibrinolysis.

Special populations. The relative contraindications, and pregnancy, translate into these plans:

Trauma. The regimen is the CRASH-2 protocol, 1g IV over 10 minutes, then 1g over 8 hours, given as soon as possible. CRASH-2 reduced all-cause mortality in bleeding trauma patients, with an NNT of roughly 67 to save one life, but the benefit is confined to treatment within 3 hours of injury. That window comes from the trial's time-to-treatment analysis, not from the primary report, which randomised within 8 hours.

Head injury. CRASH-3 extended the finding to traumatic brain injury, reducing head-injury-related death in mild-to-moderate TBI treated within 3 hours, with no increase in seizures or vascular occlusive events.

Bleeding despite TXA. Work through the causes in order:

• Verify delivery: pre-incision timing, adequate dose, and a repeat if over 3 hours have elapsed
• Surgical control: electrocautery, bone wax, topical haemostatics, direct pressure
• Resuscitation: point-of-care Hb, transfusion by restrictive thresholds, correction of hypothermia and acidosis, FFP and platelets if coagulopathic
• Pharmacological: further TXA up to about 3g total; aminocaproic acid is the alternative antifibrinolytic, aprotinin having been withdrawn

Intravenous administration. A slow infusion, finished before the knife goes in:

1. Confirm there are no contraindications during the WHO checklist
2. Prepare 15-20mg/kg or 1-2g in 100mL saline
3. Give it as a slow IV infusion over 10 minutes; slow infusion reduces the risk of hypotension
4. Complete it 10-15 minutes before skin incision
5. Document dose, timing and route in the anaesthetic or operative record

TXA is compatible with most IV fluids.

Topical administration. Haemostasis first, then the soak, at the end of the procedure before closure:

1. Prepare 2-3g TXA in 50-100mL normal saline
2. Achieve mechanical surgical haemostasis
3. Instil into the joint cavity or wound, in contact with the bleeding surfaces
4. Leave a dwell time of 5-10 minutes
5. If a drain is used, clamp it during the dwell, then release

An absorbable gelatin sponge can be soaked in TXA for additional local effect, the solution should not be mixed with blood products, and some surgeons atomise it for even distribution.

Drains. If a drain is used, instil TXA, close the wound layers and clamp the drain for the dwell period, 5-10 minutes or up to 1-2 hours in some protocols, then unclamp, allow free drainage, monitor output as usual and remove it per standard protocol, typically at 24-48 hours. Extended clamping has mixed evidence and risks haematoma if bleeding is brisk. Many surgeons now use drainless protocols, as the blood-loss reduction from TXA makes drains less necessary; otherwise drain management is not significantly altered.

Tourniquet in TKA. Tourniquet release increases bleeding, which is where TXA is particularly beneficial. Timing is pre-incision regardless of tourniquet strategy: given systemically before inflation, therapeutic plasma levels are present to control the reactive bleeding on release, whether the tourniquet is used throughout, for cementation only, or not at all.

By procedure. Points specific to each operation:

• THA: IV or topical, effective with all approaches (posterior, anterior, lateral)
• Revision arthroplasty: higher blood loss, so consider the combined IV and topical approach and a higher dose
• Spine: IV loading dose and infusion; topical is less studied but likely effective
• Trauma fixation: IV by the CRASH-2 protocol, time-critical within 3 hours of injury

Thromboembolism. No clear increase in VTE has been detected at therapeutic doses. CRASH-2 (20,211 patients) found no difference in vascular occlusive events (DVT, PE, stroke, MI) even in a high-risk trauma population without routine chemical prophylaxis. In the Ker surgical meta-analysis (129 trials, 10,488 patients) the estimates showed no clear increase, but every interval is wide:

• DVT RR 0.86 (95% CI 0.53-1.39)
• PE RR 0.61 (0.25-1.47)
• MI RR 0.68 (0.43-1.09)
• Stroke RR 1.14 (0.65-2.00)

POISE-3. In noncardiac surgery the cardiovascular composite was near-identical, 14.2% vs 13.9%, but formal noninferiority was not met, so vigilance and standard prophylaxis remain essential.

Seizures. The risk is dose-dependent and very low at orthopaedic doses.

At plasma concentrations over about 80-100 micrograms/mL TXA acts as a GABA-A receptor antagonist, lowering the seizure threshold, and orthopaedic doses rarely reach these levels. If a TXA-related seizure occurs, stop the infusion, protect the airway, give a benzodiazepine, check glucose, document it and avoid TXA in future.

Other adverse effects. Most are minor or rare, serious adverse events are rare at orthopaedic doses, and the risk rises with dose and renal impairment:

• Nausea, vomiting and diarrhoea, especially with rapid IV or oral dosing
• Hypotension with rapid IV administration
• Visual disturbance and retinal toxicity: extremely rare, reported with prolonged oral use (e.g. menorrhagia), not with short-term perioperative use
• Allergic reactions: very rare, and anaphylaxis has not been reported at standard doses

Interactions. Concurrent antiplatelets, anticoagulants and other drugs are handled as follows:

The balance sheet. What TXA does to complications:

• Reduced by TXA: transfusion requirements, wound haematoma, reoperation for bleeding
• Not increased by TXA: DVT, PE, wound infection, periprosthetic joint infection
• Uncertain or dose-dependent: seizures (only at very high doses or in renal impairment), retinal toxicity (not with short-term use)

Routine care. TXA needs no additional postoperative monitoring: standard wound observation, a routine day-1 haemoglobin and VTE prophylaxis per protocol. TXA typically reduces wound haematoma, and rehabilitation is unchanged, with early mobilisation on day 0 or day 1.

Transfusion. Standard restrictive thresholds apply: transfuse below 70 g/L for most patients, or at 70-80 g/L if symptomatic or with cardiac disease. TXA reduces transfusion need but does not eliminate it.

Ongoing bleeding. Any of these is cause for concern:

• Drain output over 500mL in the first 4 hours
• An expanding wound haematoma
• Hb drop greater than 30 g/L
• Haemodynamic instability

Managing it. Start conservatively: compression dressing and ice, reversal of excessive anticoagulation, and transfusion if indicated. By the time excessive drainage or haematoma appears the TXA effect has usually worn off, so the priority is to assess for surgical bleeding rather than routinely re-dose, although an additional 1g IV may be considered within about 6 hours of surgery.

Return to theatre is indicated for:

• Expanding haematoma compromising the wound
• Haemodynamic instability despite resuscitation
• Drain output greater than 1500mL in 6-8 hours
• Suspected arterial bleeding

Enhanced recovery. TXA is the core blood-conservation component of an ERAS pathway for joint replacement, with a transfusion rate under 5% that reduces transfusion-related morbidity. Less anaemia enables early mobilisation, an indirect benefit, and TXA contributes to reduced length of stay as part of a multimodal strategy; the VTE prophylaxis protocol does not change.

Blood loss. TXA reduces total blood loss by 30-50% in TKA and 30-40% in THA, an absolute reduction of approximately 300-500 mL per procedure. That reduces anaemia, transfusion need and the complications of both.

Transfusion. In arthroplasty the transfusion rate falls from 15-25% without TXA to 2-5% with it, a reduction of roughly one third to two thirds. Across surgery the Ker meta-analysis found transfusion reduced by about one third (RR 0.62). Beyond arthroplasty, POISE-3 (9535 noncardiac surgery patients) reduced the composite bleeding outcome, 9.1% vs 11.7% (HR 0.76).

Cost. TXA is one of the most cost-effective interventions in orthopaedic surgery and is generally cost-saving, a dominant strategy that both saves money and improves outcomes. A single bolus is simple to give and effective in many settings.

• TXA costs of the order of a few units of currency per gram, roughly 10-30 per patient
• A single unit of blood costs hundreds, including testing, administration and monitoring, plus the cost of transfusion-related complications
• Transfusion falls by the order of 15-20% absolute, an NNT of roughly 5-10 to prevent one transfusion, with net savings of several hundred per patient
• Indirect benefits: shorter length of stay, fewer transfusion-related complications, less burden on the blood bank

Other settings and the long term. The evidence outside primary joint replacement, and over time:

• Spine surgery: 20-40% blood-loss reduction, most beneficial in multilevel fusions and deformity correction; no increase in epidural haematoma or neurological complications, and no effect on MEP/SSEP monitoring
• Hip fracture and periarticular trauma: reduced transfusion; give as early as possible, ideally within 3 hours
• Shoulder arthroplasty: limited but consistent evidence of blood-loss reduction; topical use is feasible
• Long term: no delayed VTE signal, no increase in implant failure or revision, no delayed wound complications

The trauma sections explain that TXA given after 3 hours may increase mortality and attribute this to a shift toward fibrinolytic shutdown. That shift is part of the trauma-induced-coagulopathy fibrinolysis spectrum, and understanding the phenotypes explains why TXA timing matters.

Global epidemiology and rationale

Perioperative bleeding and allogeneic transfusion are common across orthopaedic surgery worldwide. Untreated, transfusion rates reach 15-25% in primary TKA/THA and are higher in revision, bilateral and trauma cases. TXA is on the WHO Model List of Essential Medicines, reflecting its global relevance and low cost, and is a cornerstone of patient blood management (PBM) programmes internationally.

Major society guidance (side by side)

Registry and blood-management context

National joint registries (NJR-UK, AJRR-US, AOANJRR-Australia, Swedish/Norwegian registries) do not all capture TXA use directly, but transfusion rates following primary TKA/THA have fallen markedly over the past decade in regions where TXA and PBM programmes have been adopted (e.g. TKA transfusion rates declining from roughly 20% to under 10% nationally in several registries). This temporal correlation supports TXA as a key driver of reduced perioperative transfusion. No registry-level signal of increased VTE or revision has emerged.

Patient blood management (PBM) - three pillars

1. Optimise erythropoiesis preoperatively - treat anaemia (iron, erythropoietin if indicated), manage anticoagulation
2. Minimise perioperative blood loss - TXA as the key pharmacological intervention, cell salvage, meticulous haemostasis, hypotensive anaesthesia where safe
3. Harness tolerance of anaemia - restrictive transfusion thresholds (Hb less than 70-80 g/L unless symptomatic), continued iron supplementation

High- vs limited-resource variation

• High-resource settings: TXA is often embedded in pre-printed order sets and surgical checklists for TJA and trauma, with pharmacy-driven automatic dispensing; cell salvage and restrictive transfusion are routinely available
• Limited-resource settings: TXA is especially valuable because it is cheap, stable and easy to administer, reducing dependence on a constrained donor blood supply; CRASH-2 deliberately recruited across 40 countries to establish global applicability

Barriers and strategies for adoption

Despite strong evidence, uptake varies due to knowledge gaps, outdated VTE-risk concerns and protocol inertia. Effective strategies include institutional protocols, default order sets, education for surgeons/anaesthetists/nurses, and audit-and-feedback on transfusion and TXA utilisation.