Abstract
d-dimer (DD) levels are used in the diagnostic workup of suspected acute pulmonary embolism (APE), but data on DD for early risk stratification in APE are limited. In this post hoc analysis of a prospective observational study of 270 consecutive patients, we aimed to optimize the discriminant capacity of the simplified pulmonary embolism severity index (sPESI), an APE risk assessment score currently used, by combining it with DD for in-hospital adverse event prediction. We found that DD levels were higher in patients with complicated versus benign clinical course 7.2 mg/L (25th-75th percentile: 4.5-27.7 mg/L) versus 5.1 mg/L (25th-75th percentile: 2.1-11.2 mg/L), P = .004. The area under the curve of DD for serious adverse event (SAE) was 0.672, P = .003. d-dimer =1.35 mg/L showed 100% negative predictive value for SAE and identified 11 sPESI ≥1 patients with a benign clinical course, detecting the 1 patient with SAE from sPESI = 0. d-dimer >15 mg/L showed heart rate for SAE 3.04 (95% confidence interval [CI]: 1-9). A stratification model which with sPESI + DD >1.35 mg/L demonstrated improved prognostic value when compared to sPESI alone (net reclassification improvement: 0.085, P = .04). d-dimer have prognostic value, values <1.35 mg/L identify patients with a favorable outcome, improving the prognostic potential of sPESI, while DD >15 mg/L is an independent predictor of SAE.