RESUMEN
D-dimer levels significantly increase with declining renal function and hence, renal function-adjusted D-dimer cutoffs to rule out pulmonary embolism were suggested. Aim of this study was to "post hoc" validate previously defined renal function-adjusted D-dimer levels to safely rule out pulmonary embolism in patients presenting to the emergency department. In this retrospective, observational analysis, all patients with low to intermediate pre-test probability receiving D-dimer measurement and computed tomography angiography (CTA) to rule out pulmonary embolism between January 2017 and December 2020 were included. Previously defined renal function-adjusted D-dimer cutoffs (1306 µg/l for moderate and 1663 µg/l for severe renal function impairment) were applied to determine sensitivity, specificity, negative and positive predictive values. One thousand, three hundred sixty-nine patients were included of which 229 (17%) were diagnosed with pulmonary embolism. The estimated glomerular filtration rate (eGFR) was ≥ 60 ml/min in 1079 (79%), 30-59 ml/min in 266 (19%) and < 30 ml/min in 24 (2%) patients. Only three patients (1.1%) with an eGFR < 60 ml/min had a D-dimer level < 500 µg/l. There was a significant correlation between D-dimer and eGFR (R = - 0.159, p < 0.001). Calculated on the standard D-dimer cutoff value of 500 µg/l, sensitivity of D-dimer testing was 97% for patients with an eGFR ≥ 60 ml/min and 100% for those with 30-60 ml/min, while specificity decreased in patients with renal function impairment. A negative predictive value of 0.99 as a premise to safely rule out pulmonary embolism was achieved by applying a D-dimer cutoff of 1480 µg/l for eGFR 30-59 ml/min and 1351 µg/l for eGFR < 30 ml/min. The findings of this study underline that application of renal function-adapted D-dimer levels in combination with a clinical prediction rule appears feasible to rule out pulmonary embolism. Out of the current dataset, renal function-adjusted D-dimer cutoffs to rule out pulmonary embolism were slightly different compared to previously defined cutoffs. Further studies on a larger scale are needed to validate possible renal function-adjusted D-dimer cutoffs.