Clinical Question: Does a normald-dimer level exclude pulmonary embolism (PE)?

Setting: Emergency department

Study Design: Cohort (prospective)

Synopsis: In this study, all patients presenting to the emergency department with suspected PE had a d-dimer test (VIDAS assay, bioMerieux, an enzyme-linked immunosorbent assay [ELISA] test). A value less than or equal to 500 ng per mL was considered normal. Of 1,106 patients, 547 had negative levels of d-dimer, and only two (0.4 percent) of these had a PE. Of 559 patients with a positive level of d-dimer, only 53 actually had a PE. Overall, 55 of 1,106 (5.0 percent) had a PE. This study had a number of limitations. We do not know how the patients presented (e.g., with chest pain, shortness of breath), so it is not clear to whom the results apply, although suspected PE is probably adequate. Also, the work-up other than of d-dimer levels was not specified, so patients with a normal test were much less intensively investigated (verification bias). The authors stated that they followed up all 547 patients with a negative test at six months. This is an impressive achievement, although it is hard to believe that all of these patients had an equally thorough follow-up. The physicians making the final diagnosis of PE (the reference standard) were not blinded to the d-dimer testing and may have been influenced by the results.

Bottom Line: This flawed study provides further support for a negative d-dimer test. Given the flaws in the study design, I would only feel comfortable excluding PE in a patient with a negative d-dimer (ELISA) level if he or she also had a low probability of PE based on the clinical examination. See InfoRetriever for a validated algorithm that combines the clinical examination, d-dimer test, and other tests. (Level of Evidence: 3a)