Researchers from Oxford University Hospitals NHS Foundation Trust has recommended that tranexamic acid should not be used for the treatment of gastrointestinal bleeding.
Tranexamic acid is routinely used to treat excessive blood loss.
The results of the study named the HALT-IT trial, which was a six-year international collaboration, were published in the Lancet on the 20th June.
The trail found that tranexamic acid did not reduce deaths from gastrointestinal bleeding, where the acid has been used to counteract excessive bleeding during surgery or after major trauma of childbirth.
The study found that the acid actually increased the risk of deep vein thrombosis or a pulmonary embolism.
Seizures were also more common in those given tranexamic acid.
Prof Simon Stanworth, Consultant Haematologist at the John Radcliffe Hospital for NHS Blood & Transplant and OUH, who was a member of the trial team said:
“It is often assumed that a treatment that works in one bleeding situation will probably work in another, but this trial has shown otherwise.
“Our results highlight the need for clinical trials that target specific causes of bleeding.”
“On the basis of these results, we recommend that tranexamic acid should not be used to treat gastrointestinal bleeding, although it has an important role in other patients with major bleeding, such as trauma and post-partum haemorrhage.”
More than 12,000 patients with acute gastrointestinal bleeding took part in this randomised, double-blind trial, half receiving a high-dose 24-hour infusion of tranexamic acid, and the rest a placebo.
The trial was conducted 164 hospitals in 15 countries.
The number of patients who died due to bleeding within five days of randomisation was the same, 4%, in both groups, as was the rate of heart attack and stroke (0.7% versus 0.8%).
The level of deep vein thrombosis or pulmonary embolism in the tranexamic acid group was double that in the placebo group (0·8% against 0·4%).
In the UK, the trial was funded by the NIHR’s Health Technology Assessment Programme.
