Early Vena Cava Filter Placement Fails to Reduce Embolic Risk in Severely Injured Patients

Placement of an inferior vena cava (IVC) filter in patients with a contraindication to prophylactic anticoagulation who have recently been admitted to the hospital for a major trauma did not reduce risk of pulmonary embolism (PE) or death at 90 days, compared with no filter placement, according to a study published in the New England Journal of Medicine. However, certain patients who have a prolonged contraindication after injury may derive some benefit from this approach, the researchers wrote.

“For an invasive, expensive, and potentially risky preventive measure such as IVC filter, selecting the most appropriate patients is the key to making a treatment strategy the most cost-effective,” study author Kwok M. Ho, PhD, of the University of Western Australia in Perth, told ASH Clinical News. “Having considered this important issue in the design of our trial, we have confirmed that an IVC filter should be reserved for those who are perceived to have a prolonged contraindication beyond seven days after injury to anticoagulant prophylaxis. A more restrictive approach to using the filters in trauma patients will reduce potential unnecessary complications, health-care cost, and medical litigation.”

In this multicenter, randomized trial, Dr. Ho and researchers enrolled 240 severely injured patients (median age = 39 years; interquartile range [IQR] = 27-57 years) who met the following criteria:

  • Injury Severity Score >15 (scores range from 0-75, with higher scores indicating more severe injury)
  • contraindication to prophylactic anticoagulant agents (e.g., stroke, history of bleeding)

Participants were randomized in an unblinded fashion to having no IVC filter placed (n=118) or to having a retrievable vena cava filter placed within 72 hours after hospital admission for injury (n=122).

The primary endpoint was a composite of symptomatic PE or death from any cause at 90 days. Secondary endpoints included filter-related complications, mortality, and bleeding events at 90 days, as well as symptomatic PE between day eight and day 90 in patients who did not receive prophylactic anticoagulation within seven days of injury.

At two weeks following study enrollment, patients underwent ultrasonography of the legs as well as CT pulmonary angiography.

The overall cohort had a median Injury Severity Score of 27 (IQR=22-34), mainly of traumatic intracranial hematoma or contusions (57.5%). Most patients who were assigned to the IVC filter group (89%) had the filter placed within 24 hours following enrollment (median = 15.6 hours [IQR=3.0-22.3 hours]).

There was no difference between the IVC-filter group and the no-filter group in the proportion of patients who experienced symptomatic PE or death at 90 days: 13.9% versus 14.4%, respectively (hazard ratio [HR] = 0.99; 95% CI 0.51-1.94; p=0.98).

Only two patient characteristics were significantly associated with the primary endpoint: higher age (HR=1.3 per 10-year increment) and higher Injury Severity Score (HR=1.5 per 10-point increment; p values not reported).

In a subgroup of patients in the IVC-filter group (n=46) and the no-filter group (n=34) who had survived up to seven days and did not receive prophylactic anticoagulation within seven days of injury, no patients in the filter group had symptomatic PE between days 8 and day 90, compared with five patients (14.7%) in the control group (relative risk [RR] with the filter = 0; 95% CI 0.00-0.55; p value not reported). “This patient cohort may benefit from the use of a vena cava filter as a temporizing measure to prevent symptomatic pulmonary embolism,” the authors reported.

Safety outcomes were similar between the two groups, as well. In the filter group, 11.4% of patients experienced a deep vein thrombosis (DVT), compared with 10.1% in the no-filter group (RR=1.1; 95% CI 0.6-2.3; p value not reported). There also were no differences in the incidence of major bleeding (70.5% vs. 66.1%; RR=1.07; 95% CI 0.90-1.27) and nonmajor bleeding (23.8% vs. 17.8%; RR=1.34; 95% CI 0.81-2.20; p values not reported).

Among the 122 patients in the IVC-filter group, the researchers observed entrapped thrombus within the filter at first removal attempt in six patients (4.9%). Only one patient (0.8%) had a filter that was adherent to the caval wall and was subsequently removed via surgery.

“The number of patients with thrombus entrapped in the filters was identical to the number of patients with symptomatic PE in the control group, while the lower-limb DVT rates were similar between the two groups,” Dr. Ho added. “It is tempting to speculate that filters were achieving precisely the desired effect – preventing embolization of lower-limb thrombi to the pulmonary circulation.”

Limitations of the study include the relatively small sample size and the lack of blinding of both patients and investigators. In addition, Dr. Ho explained that the study was underpowered to identify the primary endpoint of symptomatic PE or mortality, noting that the incidence of symptomatic PE in study participants who received anticoagulant prophylaxis between three and seven days of injury without a filter is low (1.2%). “To confirm the benefit of the filter in such patients would require a sample size of 9,712 per group and a number-needed-to-treat of approximately 250, assuming a power of 80% and an alpha error of 5% as acceptable, which is realistically not feasible nor clinically relevant,” he added.

The authors reported relationships with Medtronic and Cardinal Health.


Ho KM, Rao S, Honeybul S, et al. A multicenter trial of vena cava filters in severely injured patients. N Engl J Med. 2019;381:328-37.