This evaluation of the use of enzyme-linked immunosorbent assay (ELISA) D-dimer test in routine clinical practice supports other evidence that the assay has a high sensitivity to exclude pulmonary embolism in patient populations in which there is clinical suspicion. Nevertheless, the assay incorrectly excluded the diagnosis of pulmonary embolism in 2 cases.
Other examples of clinical decision-making exist for which the acceptable negative predictive value for screening is set at 100%—eg, the diagnosis of phenylketonuria in newborns.
Physicians who do not want to miss cases of acute pulmonary embolism when they clinically suspect the diagnosis should not rely solely on negative D-dimer assay results when the value to rule out the diagnosis is set at 500 ng/mL. If a lower value is used to define normal—eg, 250 ng/mL, as used in other studies—no cases of acute pulmonary embolism would have been missed in this group of patients. Regardless of the cutoff used, the assay will yield many false-positive results.