Diagnostic performance of YEARS criteria combined with point-of-care lung and venous ultrasound in suspected pulmonary embolism

Clinical prediction rules and bedside lung and venous ultrasound are frequently used in the diagnostic workup of suspected pulmonary embolism (PE). The possibility of improving the YEARS algorithm by integrating ultrasound has never been investigated. We analyzed data from a previous study involving a total of 446 outpatients with suspected PE. Signs of deep vein thrombosis (DVT) on venous ultrasound and signs of pulmonary infarctions or alternative diagnoses on lung ultrasound were used to recalculate two items of the YEARSc: signs and symptoms of DVT and alternative diagnosis less likely than PE. Diagnostic performance of ultrasound-enhanced YEARS criteria (US-YEARSc) and YEARSc were compared after final diagnosis. 446 patients were included, PE was confirmed in 125 (28%). US-YEARSc performed significantly better than YEARSc (sensitivity 82.4% vs 64.8%, p=0.001; specificity 81% vs 48.6%, p < 0.001). US-YEARSc plus negative d-dimer compared to YEARSc and negative d-dimer showed an inferior failure rate (5.9% vs 8.9%, p=0.25) and a superior efficiency (51.8% vs 48.2%, p=0.31), without reaching statistical significance. US-YEARSc perform better than YEARSc in the diagnosis of PE.