Venous thromboembolic risk assessment models should not solely guide prophylaxis and surveillance in trauma patients

Venous thromboembolism (VTE) risk assessment models exist to stratify patients at risk for VTE and guide surveillance and prophylaxis. We evaluated the only two models developed specifically for trauma patients: the Trauma Embolic Scoring System (TESS) and the Risk Assessment Profile (RAP). Clinical and demographic data on patients admitted from July 2006 to December 2011 who underwent surveillance lower extremity duplex ultrasound were recorded. Patients were excluded if they were missing one or more of the variables required to calculate either TESS or RAP. Patients received prophylaxis according to American College of Chest Physicians guidelines. TESS and RAP scores were calculated retrospectively and compared between patients with VTE and patients without VTE. High risk was defined by the models as TESS score of 7 or greater and RAP score of 5 or greater. A total of 2,868 patients received surveillance lower extremity duplex ultrasound. TESS score was calculated for 2,140 patients; 215 developed VTE, 110 (51%) of whom had TESS score less than 7. The sensitivity and specificity at a cutoff point of 7 were 49% and 72%, respectively. RAP score was calculated for 1,505 patients; 152 developed VTE, 26 (17%) of whom had RAP score of less than 5. The sensitivity and specificity at a cutoff point of 5 were 83% and 37%, respectively. The area under the receiver operating characteristic curve for each model was 0.66. A clinically significant number of patients who developed VTE were classified as low risk by both TESS and RAP. The indications for VTE surveillance and chemoprophylaxis should not be based exclusively on these scores. These results suggest that additional variables should be sought to improve risk assessment for VTE following trauma. Care management study, level III.