Russell L. Gruen, Biswadev Mitra, Stephen A. Bernard, Colin J. McArthur, Brian Burns, Dashiell C. Gantner, Marc Maegele, Peter A. Cameron, Bridget Dicker, Andrew B. Forbes, Sally Hurford, Catherine A. Martin, Stefan M. Mazur, Robert L. Medcalf, Lynnette J. Murray, Paul S. Myles, Sze J. Ng, Veronica Pitt, Stephen Rashford, Michael C. Reade, Andrew H. Swain, Tony Trapani, Paul J. Young
8 min
Trauma is a leading cause of death, often complicated by trauma-induced coagulopathy—a condition where bleeding is worsened by the body's over-active fibrinolytic system. Tranexamic acid is an antifibrinolytic drug known to reduce mortality in settings where trauma care systems are still developing. However, its effectiveness in advanced trauma systems—which provide immediate access to specialized surgical and critical care—remained uncertain. The PATCH-Trauma trial was a large-scale, international, double-blind, randomized controlled trial conducted in Australia, New Zealand, and Germany. Researchers evaluated whether prehospital administration of tranexamic acid (a 1g bolus followed by a 1g infusion over 8 hours) improved functional survival at 6 months compared to a placebo in patients at high risk for coagulopathy.
The trial enrolled 1,310 patients across 15 emergency medical services. The researchers found no statistically significant difference in the primary outcome: 53.7% of the tranexamic acid group achieved a favorable functional outcome at 6 months, compared to 53.5% in the placebo group. While the treatment was associated with a decrease in early mortality (at 28 days), this benefit did not translate into better long-term functional recovery. Furthermore, there was no meaningful increase in serious adverse events, such as vascular occlusive events, between the two groups.
This study provides critical evidence for clinicians working in highly resourced trauma centers. While earlier landmark trials (like CRASH-2) established that tranexamic acid saves lives in more austere, less-resourced environments, this trial suggests those findings might not directly translate into improved long-term functional quality of life for patients in advanced systems. It suggests that while the drug may reduce bleeding-related deaths, its impact on the final functional status of survivors is negligible in modern, high-functioning trauma networks.
Sam: That means the trial tested treatment based on a practical risk assessment, not a laboratory-confirmed disorder.
Alex: Yes. Only a minority subsequently had laboratory evidence of early clotting dysfunction. Eligible patients had severe injuries and a high assessed risk, and treatment had to begin soon after injury, before admission.
Sam: What was actually randomized? Was the question just whether to give a dose in the ambulance?
Alex: It was a treatment course, not just an ambulance dose. Patients received intravenous drug or matched placebo before arrival, followed by a hospital infusion. Everyone also received usual care, and patients, clinicians, and follow-up assessors were unaware of the assignment.
Sam: Then the distinctive choice was the endpoint. What did the authors mean by favorable function?
Alex: They assessed survival with favorable function six months after injury. They used the Glasgow Outcome Scale–Extended, a scale spanning death, severe disability, moderate disability, and good recovery. Trained interviewers administered a standardized questionnaire by telephone.
Sam: “Favorable” could sound like returning to normal. Did patients have to recover completely to count?
Alex: No. Moderate disability still counted as favorable. Severe disability did not. That boundary matters: the trial asked whether more people were alive above that functional threshold, not simply whether more were alive.
Sam: With that definition in place, how close were the treatment groups?
Alex: About fifty-four percent in each group survived with favorable function. The estimated difference was essentially zero. The uncertainty still allowed a modest benefit or harm, so this is not proof that the treatments are equivalent.
Sam: A near-zero average can hide patients who respond differently. Did the authors find a subgroup where treatment clearly helped recovery?
Alex: They found no evidence of different effects in the prespecified subgroups, including age, blood pressure, injury mechanism, and treatment timing. But small subgroups had limited statistical power. So this does not establish that every kind of trauma responds identically.
Sam: Now the tension from the opening: did fewer people die even though favorable recovery didn’t improve?
Alex: At twenty-eight days, mortality was about seventeen percent with tranexamic acid versus twenty-two percent with placebo. Early mortality also favored treatment. But these were secondary outcomes, and the analyses were not corrected for multiple comparisons.
Sam: So the mortality pattern deserves attention, but it shouldn’t become an unqualified claim that the trial proved a survival benefit.
Alex: That is how I would read it. The authors describe an apparent association with lower early mortality. Later mortality remained numerically lower, but its uncertainty included no difference. The primary functional-survival result stayed neutral.
Sam: Does that pattern mean the drug saved people who then had severe disability?
Alex: The authors describe more survivors alongside a similar additional number of people categorized with severe disability. But the trial cannot identify which individuals would otherwise have died. Nor does that pattern establish that the drug caused disability.
Sam: And the recovery assessment is a checkpoint, not necessarily the end of recovery.
Alex: The authors note that comparable trauma populations can improve beyond that checkpoint. Their findings do not rule out preventing early bleeding deaths in people who later recover well. They simply did not demonstrate more favorable functional survival at the measured time.
Sam: What about the clotting concern? Limiting clot breakdown could carry its own risks.
Alex: The researchers screened hospitalized patients for leg-vein clots. Blood-vessel blockages were numerically more common with treatment, but the comparison was uncertain. The authors found little evidence of increased risk; that is not the same as ruling it out.
Sam: What is the biggest reason to be cautious about the main recovery result?
Alex: Thirteen percent of patients lacked primary outcome data, mostly because follow-up was lost. Analyses estimating the missing outcomes gave similar results. Still, the authors could not exclude differences between groups among patients whose outcomes were missing.
Sam: There were also treatment deviations. Could those have blurred the comparison?
Alex: Some patients missed intended doses, and some received tranexamic acid outside the assigned treatment. Analyses restricted to patients following the protocol gave similar findings. That is reassuring, though it does not remove the missing-data limitation.
Sam: Who should read the full document, and where should they start?
Alex: Trauma clinicians, prehospital researchers, and investigators choosing recovery endpoints should read it closely. Start with Outcomes and Primary Outcome, especially the definition of favorable function. Then compare Secondary Outcomes with the Discussion’s cautions about missing follow-up and generalizability.
Sam: And what should everyone else carry away, without turning this into a blanket verdict on the drug?
Alex: This result applies to the tested regimen in severely injured adults at risk of clotting dysfunction within advanced trauma systems. It does not settle other doses or settings. The line to remember is: fewer early deaths do not automatically mean more survivors with favorable recovery.
Sam: Thanks for listening.