Prehospital tranexamic acid bolus improves outcomes in traumatic brain injury: A Bayesian reanalysis

Bottom line

In this post hoc Bayesian reanalysis of a prehospital TXA-for-TBI RCT, a single 2-g TXA bolus carried a high posterior probability of benefit over placebo across all three measured outcomes, while the standard 1-g bolus plus 1-g infusion regimen showed under 50% probability of benefit for every outcome, a hypothesis-generating, Level III signal that requires prospective confirmation before it should change TBI resuscitation practice.

PICO at a glance

• Population: Patients with moderate to severe TBI from the original prehospital TXA-for-TBI RCT.
• Intervention or exposure: Prehospital TXA as either a 1-g bolus plus 1-g infusion ("standard") or a 2-g bolus alone ("bolus").
• Comparator: Placebo.
• Outcomes: Bayesian posterior probability of improved 6-month functional neurologic outcome, of decreased 28-day mortality, and of decreased 6-month mortality, each versus placebo.
• Study design: Post hoc Bayesian reanalysis of a completed prehospital RCT (Level III, hypothesis-generating).

Clinical question and design

TXA is established for hemorrhagic trauma broadly, but its role and optimal dose in traumatic brain injury (TBI) remain unsettled, large prehospital RCTs of TXA in TBI have so far shown, at best, modest or non-significant effects on 6-month functional outcome by standard frequentist analysis. This reanalysis asks a narrower question: does dosing strategy change the probability of benefit, even where the parent trial's overall analysis found no significant effect?

This is a post hoc Bayesian reanalysis of "the prehospital TXA for TBI RCT," in which patients originally received a 1-g bolus plus 1-g infusion ("standard"), a 2-g bolus alone ("bolus"), or placebo, with treatment initiated prehospital. Bayesian regression models generated posterior probabilities of benefit for 6-month functional neurologic outcome, 28-day mortality, and 6-month mortality. The dosing regimens match the design of the likely parent trial by Rowell et al. (JAMA. 2020;324(10):961-974; PMID 32897344), a prehospital, multicenter, double-blind RCT of 1,280 patients with moderate-to-severe TBI that found no significant frequentist difference in 6-month functional outcome for TXA versus placebo, though this rests on a regimen-design match, not a confirmed citation in the reanalysis's own abstract.

Intervention and outcomes measured

The intervention being reanalyzed is dosing strategy, not TXA-versus-no-TXA: a single 2-g bolus alone against a 1-g bolus followed by a 1-g infusion, both compared with placebo. Combined TXA (standard and bolus pooled) versus placebo showed a 78.1% posterior probability of improved 6-month functional neurologic outcome, a figure that, on its own, masks a sharp split between the two regimens once analyzed separately.

Results

Analyzed separately, the Bolus (2-g) cohort versus placebo showed a 95.3% probability of improved 6-month functional outcome, a 95.3% probability of decreased 28-day mortality, and a 70.7% probability of decreased 6-month mortality. The Standard (1-g bolus plus 1-g infusion) cohort versus placebo showed less than 50% probability of benefit for all three outcomes measured, no-benefit-or-harm was more probable than benefit for that regimen. Exact posterior-probability values for the Standard cohort's individual outcomes, and 95% credible intervals for any reported probability, are not stated in the accessible abstract.

Safety and additional findings

The post-hoc Bayesian reanalysis reports all of these data. Among 966 patients (mean age 42; 74% male; mean GCS 8; median ISS 17), favorable six-month outcome occurred in 58% with standard TXA, 62% with the 2-g bolus, and 60% with placebo. Compared with placebo, the 2-g bolus estimate was OR 1.36 (95% credible interval 0.96-1.88), with 95.3% posterior probability of OR greater than 1; standard dosing was OR 0.95 (0.67-1.32), 35.2%. Direct bolus-versus-standard comparison was OR 1.45 (1.02-2.00), posterior probability 98.2%. This remains a post-hoc reanalysis, so posterior probability is a decision-relevant uncertainty statement, not prospective confirmation of a regimen.

Uncertainty and limitations

This is a post-hoc, hypothesis-generating Bayesian reanalysis, not a new randomized trial or prospective confirmation of a dosing strategy. The full paper reports arm-level results, baseline characteristics, and model details. The favorable bolus signal should be weighed with uncertainty, the original trial, and external applicability rather than treated as definitive confirmation of a regimen.

Literature review and evidence synthesis

This reanalysis sits within an active, unsettled literature on TXA dosing and timing in TBI. A 2025 systematic review and network meta-analysis in the same journal (PMID 39652152) addresses the same optimal-dose question. Two TBI-focused syntheses, a 2021 Transfusion Medicine Reviews review (PMID 34598876) and a 2024 meta-analysis in Revista Española de Anestesiología y Reanimación (PMID 38387502), cover the broader prehospital-TXA-in-TBI evidence base. A 2025 NAEMSP/ACEP/ACS-COT position statement on TXA in trauma (PMID 40842057) and a 2023 review of prehospital TBI pharmacotherapy (PMID 36607298) frame current practice guidance, while a 2018 meta-analysis of prehospital TXA across trauma broadly (PMID 29573898) established the earlier, TBI-nonspecific evidence base.

How this fits with the broader evidence

• This reanalysis (direct evidence): Splitting the pooled TXA arm by regimen reveals opposite Bayesian signals within one dataset, the 2-g bolus carried high probability of benefit on every outcome, while the "standard" regimen carried low probability of benefit on every outcome, which may explain the parent trial's null combined-arm result.
• Likely parent RCT: A prehospital, multicenter RCT of 1,280 patients found no significant frequentist difference in 6-month functional outcome for pooled TXA versus placebo, the negative result this reanalysis reopens by regimen.
• 2025 network meta-analysis (PMID 39652152) and NAEMSP/ACEP/ACS-COT statement (PMID 40842057): More rigorous, concurrent dosing evidence to weigh alongside, not instead of, this single reanalysis.

ACPE UAN: 0683-0000-26-036-H01-P

---

Study source: PMID:41589743