Prophylactic placement of inferior vena cava filters and the risk of death or venous thromboembolism in severe trauma patients: a retrospective study comparing two hospitals with different approaches

BACKGROUND: Prophylactic use of inferior vena cava filters to prevent pulmonary embolism in trauma is controversial. The practice varies between hospitals and countries, in part due to conflicting evidence and guidelines. PURPOSE: To compare the effects of pulmonary embolism, deep venous thrombosis and mortality in two hospitals using prophylactic inferior vena cava filter placement or prophylactic anticoagulation alone. MATERIAL AND METHODS: Patients presenting with severe trauma were recruited from two level-1 trauma centres between January 2008 and December 2013. Recruited patients from an US hospital having prophylactic inferior vena cava filter inserted were compared to a Scandinavian hospital using prophylactic anticoagulation alone. Inclusion criteria were age >15 years, Injury Severity Score >15 and survival >24 h after hospital admission. Patients with venous thromboembolism diagnosed prior to inferior vena cava filter placement were excluded. A Cox proportional hazard regression model was used with adjustment for immortal time bias and predictor variables. RESULTS: In total, 951 patients were reviewed, 282 from an US hospital having inferior vena cava filters placed and 669 from a Scandinavian hospital without inferior vena cava filters. The mean age was 45.9 vs. 47.4 years and the mean Injury Severity Score was 29.8 vs. 25.9, respectively. Inferior vena cava filter placement was not associated with the hazard of pulmonary embolism (Hazard ratio=0.43; 95% confidence interval (CI) 0.12, 1.45; P=0.17) or mortality (Hazard ratio=1.16; 95% CI 0.70, 1.95; P=0.56). However, an increased rate of deep venous thrombosis was observed with inferior vena cava filters in place (Hazard ratio=3.75; 95% CI 1.68, 8.36; P=0.001). CONCLUSION: In severely injured trauma patients, prophylactic inferior vena cava filter placement was not associated with pulmonary embolism or mortality. However, inferior vena cava filters were associated with increased rate of deep venous thrombosis.