Heightened clinical suspicion of pulmonary embolism and disregard of the D-dimer assay: a contemporary trend in an era of increased access to computed tomography pulmonary angiogram?Intern Med J. 2013 Nov; 43(11):1231-6.IM
Prospective studies have shown that utilising qualitative D-dimers in those with a low Wells pre-test probability (PTP) of pulmonary embolism (PE) have significantly reduced the number of computed tomography pulmonary angiograms (CTPA) being performed. These studies have been based on a PE prevalence of approximately 6% in the low PTP group.
This study compares the diagnostic approach to PE in the study institution to well-established guidelines. The study also re-examines the cost-benefit analyses of qualitative d-dimers and CTPA in the low PTP group.
A retrospective study of 169 consecutive CTPA requested in the emergency department of a major teaching hospital during a 12-month period.
The prevalence of PE was 0% (0/65), 11.7% (9/77) and 0% (0/2) in the low, moderate and high Wells PTP groups respectively, and 6.3% (9/144) overall. PTP was documented in 10 (6.9%) cases, and the qualitative Clearview Simplify D-dimer was only ordered in 33.8% (22/65) of low PTP subjects. The false positive D-dimer rate was 90.2% (37/41). Cost-benefit analysis and assay performance defines a narrow range of low PTP PE prevalence between 1% and 5% for the utilisation of the qualitative D-dimer assay.
The overall prevalence of PE in subjects undergoing CTPA was significantly lower compared with data in the literature. The authors recommend warranted clinical suspicion of PE should be confirmed by a senior physician prior to placing a patient in the PE work-up pathway. In such patients, the qualitative D-dimer assay should be utilised if PTP is low, and the exclusionary efficiency of the D-dimer will be improved in the setting of higher PE prevalence in this subgroup. Hospitals should audit local PE prevalence, as cost-benefit analyses raises questions about the effectiveness of D-dimers when PE prevalence is very low in the low PTP subgroup.