Early prophylactic use of retrievable vena cava filters after major trauma in patients with a contraindication to anticoagulation does not reduce the incidence of symptomatic pulmonary embolism (PE) or death at 90 days.
Yet the filters are widely used for the prevention of PE in patients at high risk of bleeding.
Speaking at the International Society of Thrombosis and Haemostasis (ISTH) Congress in Melbourne, Dr Kwok-Ming Ho said more than 25% of patients with severe trauma were either actively bleeding or had severely deranged clotting profiles on arrival to a trauma centre.
Dr Ho, an intensivist from the Royal Perth Hospital and the University of Western Australia, said relevant guidance ranged from the pro-filter Eastern Association for the Surgery of Trauma to the American College of Chest Physicians, which was against filters.
He said the randomised trial of filter versus no filter comprised 240 adult patients admitted to one of four major Australian hospitals with an Injury Severity Score of more than 15 and a contraindication to prophylactic anticoagulation.
Patients were mostly male (77%) with median age of 39 years.
Filter placement in the intervention group mostly occurred within 24 hours but always within 72 hours of injury. Filters were left in situ for a median of 27 days.
Anticoagulation was initiated within 7 days for 67% of patients.
The study found a composite end point of symptomatic PE or death at 90 days was similar in the intervention and control groups – 13.9% in patients with a filter compared to 14.4% in those without a filter (HR 0.99; p=0.98).
Dr Ho told the meeting that filters were clearly not a panacea.
The only apparent benefit from filters was seen in the subgroup of patients who did not receive anticoagulants within 7 days – for example, due to repeated operations.
None of the 46 patients with filters had a symptomatic PE between 8 and 90 days compared to 14.7% of patients without filters.
Patients with severe traumatic brain injury with multiple cerebral contusions were the most likely candidates for early filter use, he said.
The number needed to treat to prevent one PE was seven.
The study, published concurrently in the NEJM, found the incidence of major or non-major bleeding, DVT and transfusion requirements did not differ significantly between the two groups.
Complications with filters were largely related to their removal.