LAUSR

Background Temporary lower limb immobilisation following injury is a risk factor for symptomatic venous thromboembolism (VTE). Pharmacological thromboprophylaxis can mitigate this risk but it is unclear which patients benefit from this intervention. The Aberdeen VTE risk tool was developed to tailor thromboprophylaxis decisions in these patients and this evaluation aimed to describe its performance in clinical practice. Secondarily, diagnostic metrics were compared with other risk assessment methods (RAMs). Methods A prospective cohort service evaluation was conducted. Adult patients (≥16 years) managed with lower limb immobilisation for injury who were evaluated with the Aberdeen VTE risk tool prior to discharge from the ED were identified contemporaneously between February 2014 and December 2020. Electronic patient records were scrutinised up to 3 months after removal of immobilisation for the development of symptomatic VTE or sudden death due to pulmonary embolism (PE). Other RAMs, including the Thrombosis Risk Prediction for Patients with cast immobilisation (TRiP(cast)) and Plymouth scores, were assimilated retrospectively and diagnostic performance compared. Results Of 1763 patients (mean age 46 (SD 18) years, 51% women), 15 (0.85%, 95% CI 0.52% to 1.40%) suffered a symptomatic VTE or death due to PE. The Aberdeen VTE tool identified 1053 (59.7%) patients for thromboprophylaxis with a sensitivity of 80.0% (95% CI 54.8% to 93.0%) and specificity of 40.4% (95% CI 38.1% to 42.6%) for the primary outcome. In 1695 patients, fewer were identified as high risk by the TRiP(cast) (33.3%) and Plymouth (24.4%) scores, but with greater specificity, 67.0% and 75.6%, respectively, than dichotomous RAMs, including the Aberdeen VTE tool. Conclusion Routine use of the Aberdeen VTE tool in our population resulted in an incidence of symptomatic VTE of less than 1%. Ordinal RAMs, such as the TRiP(cast) score, may more accurately reflect VTE risk and permit more individually tailored thromboprophylaxis decisions but prospective comparison is needed.