A 26-year-old man involved in a high-speed motor vehicle collision sustained traumatic brain injury, pulmonary contusions, a grade 2 liver laceration, extraperitoneal bladder rupture, pelvic fracture, and lower extremity fractures. He required mechanical ventilation for acute respiratory distress syndrome (ARDS) as well as skeletal traction, which prevented the placement of sequential compression devices. Prophylactic doses of heparin were contraindicated because of the patient's injuries and an unstable hemoglobin level. A vena cava filter (VCF) was inserted at the bedside in the intensive care unit on the second day postinjury. A computed tomographic scan (Figure 1) was obtained 5 days after VCF insertion to evaluate the patient for possible abdominal sepsis. The scan revealed an embolus trapped in the cone of the filter. The clot was nonocclusive and asymptomatic. 

A 14-year-old boy sustained a lateral compression fracture of his pelvis, bilateral femur fractures, and a lumbar spine fracture with resultant paraplegia when he was ejected from his seat during a motor vehicle collision. Acute respiratory distress syndrome required that he be given high-frequency oscillatory ventilation. A retrievable VCF was inserted on the third day postinjury. Sixty days after VCF insertion and 14 days after hospital discharge, the patient returned for removal of the VCF. A venogram obtained before removal of the filter showed the presence of two small clots in the VCF (Figure 2). The filter was removed, and a regimen of therapeutic anticoagulation was initiated. 

A 26-year-old man involved in a motor vehicle accident sustained a traumatic subarachnoid hemorrhage and a spinal cord injury, resulting in incomplete tetraplegia. He required mechanical ventilation for ARDS caused by aspiration during the trauma. Vena cava filter placement was done at the bedside in the intensive care unit on the second day postinjury. Sixty days after insertion, the patient returned for VCF removal. A venogram obtained before VCF removal identified a large thrombus in the VCF (Figure 3). Therapeutic anticoagulation was started and, due to the size of the thrombus, the VCF was left in place as a permanent implant. 

Despite adequate prophylaxis with subcutaneous heparin and lower extremity compression devices, venous thromboembolism, which includes deep vein thrombosis (DVT) and pulmonary embolism (PE), is a frequent cause of morbidity and mortality in injured patients.^1–4 Unfortunately, PE occurs without warning, and in severely injured patients, just one occurrence may be fatal. Although pharmacologic prophylaxis has been shown to decrease the incidence of both DVT and PE, it is not 100% protective, particularly in high-risk patients.⁵ To complicate matters, high-risk patients frequently have contraindications to even prophylactic doses of subcutaneous heparin because of concomitant injuries. 

Rogers and colleagues⁶ identified clinical factors that contribute to high-risk stratification in a retrospective review of PE in patients with severe injuries. They reported that 92% of PEs occurred in patients with specific injuries, such as spinal cord injuries, head injuries, and long bone and pelvic fractures. An analysis from the American College of Surgeons' National Trauma Data Bank⁷ identified similar risk factors in addition to age (>40 years), venous injury, and mechanical ventilation for longer than 3 days, confirming the report by Rogers and colleagues,⁸ who found significant protection against PE in patients with prophylactic VCFs. It should be noted that all VCFs are considered prophylactic in that they prevent thrombi from reaching the pulmonary circulation; they do not treat DVT or PE. Although there are a relatively small number of patients in the study by Rogers and colleagues,⁸ their observations have been corroborated by those of other investigators.^9–12 Based on the findings of these investigators, the Eastern Association for the Surgery of Trauma (East Northport, NY) developed the Practice Management Guidelines for the Prevention of Venous Thromboembolism in Trauma Patients,¹³ with a level 3 recommendation (ie, recommendation based on retrospective studies or expert opinion) for VCF insertion in very-high-risk patients. 

Most prospective studies of injured patients with PE have been limited to symptomatic PE. The true incidence of PE may be greatly underestimated, however, as the illness is asymptomatic in most patients. Schultz and colleagues,¹⁴ who performed a prospective study of patients with Injury Severity Scores of 9 or higher, identified a 24% incidence of asymptomatic PE, including four patients with massive PE detected on contrast-enhanced computed tomography of the chest. Combes and colleagues,¹⁵ using autopsy-identified PE in patients who died in an intensive care unit, reported a 5.9% incidence of PE. 

No method has been developed to determine when a VCF actually performs its function. It is our opinion that an acute vena caval occlusion is likely the entrapment of a large, possibly fatal PE, whereas the intimal overgrowth caused by the VCF itself leads to a more chronic occlusion. Large, randomized, prospective studies evaluating prophylactic insertion of VCFs are lacking at this time. 

Although it is arguable that the clots described in the present case series were caused by the VCFs themselves, we contend that a thrombus trapped in the cone of the filter makes an embolic source more likely than a primary thrombus arising de novo within the filter. The laminar flow of blood is highest in the center of the vena cava and slowest at the periphery, where a primary thrombus would more likely form on the struts of the filter as a result of endothelial damage or stagnation of blood flow.¹⁶ 

The controversy over prophylactic VCF insertions continues.¹⁷ The present case series demonstrates that the filters inserted in injured patients are performing their intended task, thus lending support to the use of prophylactic VCFs. Further study is needed to define the role of VCFs in the absence of documented venous thromboembolism.