Quantitative high D-dimer value is predictive of pulmonary embolism occurrence independently of clinical score in a...

We performed a prospective study to assess whether positive quantitative D-dimer (DD) levels could be integrated for a selected population in a defined strategy to accurately diagnose pulmonary embolism (PE). For this purpose, 1528 in- or outpatients with clinically suspected PE were investigated according to our prescription rules. Clinical probability was defined as low, intermediate or high. Patients in whom DD levels were measured met criteria defined by our previously described decision-making algorithm: in- and outpatients, < 80 years, without surgery in the previous 30 days or active cancer. Nine hundred and twenty-three patients (60.4%) had quantitative DD measurement using automated latex DD assay (STA-Liatest D-Di). According to our decision-making algorithm, DD measurement was applied to 70.5% of out-, and 55.7% of inpatients, and PE diagnosis was ruled out in 49.5% of the 923 patients. This allowed us to confirm prospectively that our specific rules greatly improve the DD testing efficiency. PE was diagnosed in 115 (12.5%) patients. For a 0.5 mg L(-1) cut-off, the test sensitivity was 97.4%, but its specificity was only 56.7%. However, PE prevalence increased gradually with DD levels. The true observed PE prevalence, according to the quantitative assessment of DD levels, differed from that predicted with pretest clinical probability only. Moreover, in this well-defined patient group, a quantitative DD level > 2 mg L(-1) was predictive of PE occurrence independently of the clinical score (odds ratio 6.9, 95% confidence interval 3.7, 12.8). As part of a defined strategy, knowledge of positive DD quantitative value, together with the clinical probability score, improves the PE predictive model. A clinical validation of these results in a follow-up study would now be necessary before considering the implementation of this strategy into clinical practice.