Respiratory Trials·org

Gastroenterology in critical care · Antifibrinolytics in gastrointestinal bleeding

HALT-IT randomised trial

High-dose tranexamic acid in acute gastrointestinal bleeding

HALT-IT trial design and results
Design International randomised, double-blind, placebo-controlled trial
Treatment High-dose 24-hour infusion of tranexamic acid
Control Placebo
Population 12,009 patients with acute gastrointestinal bleeding
Follow-up 28 days
Primary endpoint Death due to bleeding within 5 days
Result Death due to bleeding within 5 days occurred in 222 of 5956 (4%) with tranexamic acid and 226 of 5981 (4%) with placebo — risk ratio 0.99 (95% CI 0.82 to 1.18). Arterial thromboembolic events were similar, 0.7% versus 0.8%. Venous thromboembolic events were higher with tranexamic acid, 48 of 5952 (0.8%) versus 26 of 5977 (0.4%) — risk ratio 1.85 (95% CI 1.15 to 2.98).
Secondary endpoints All-cause mortality, rebleeding, thromboembolic events, seizures

We found that tranexamic acid did not reduce death from gastrointestinal bleeding. On the basis of our results, tranexamic acid should not be used for the treatment of gastrointestinal bleeding outside the context of a randomised trial.

The trial authors, in the published abstract

How it has aged

A clear, large negative that stopped extrapolation from trauma and postpartum haemorrhage, where tranexamic acid does work (WOMAN). The authors' recommendation is unusually direct, and correspondingly easy to act on.

Publications

Effects of a high-dose 24-h infusion of tranexamic acid on death and thromboembolic events in patients with acute gastrointestinal bleeding (HALT-IT): an international randomised, double-blind, placebo-controlled trial

Effects of a high-dose 24-h infusion of tranexamic acid on death and thromboembolic events in patients with acute gastrointestinal bleeding (HALT-IT): an international randomised, double-blind, placebo-controlled trial. Lancet 2020 Jun 20;395(10241):1927-1936.

  • No reduction in death due to bleeding
  • More venous thromboembolic events
  • 12,009 patients randomised