Looking beyond the D-dimer
The D-dimer test ruled out a pulmonary embolism in our patient, but her signs and symptoms suggested that we take another look.
This case report illustrates a key problem with the Wells criteria—the somewhat subjective nature of the scoring. Some physicians find it questionable to award 3 points for “alternative diagnosis less likely than PE,” for example.4 Similarly, with respect to immobilization, some clinicians might have awarded our patient 1.5 points for her recent car trip to New York. We did not think that riding in a car for 2 uninterrupted hours for each leg of the trip was significant enough. However, awarding this patient 1.5 points could have made an important difference in her clinical management if the alternative classification scheme was used. Instead of having a score of 3, the patient would have had a score of 4.5, placing her in the “likely to have a PE” group and prompting us to perform a CTA sooner (FIGURE).
FIGURE
Diagnostic algorithm for pulmonary embolism6,7,10
CTA, computed tomography angiogram; PE, pulmonary embolism.
Inappropriate work-ups are common
Some physicians ignore algorithms when working up a PE and simply order a CTA. In fact, a large multicenter trial showed that 43% of patients suspected of having a PE were inappropriately managed diagnostically.9 Similarly, a meta-analysis of 4 studies including 1660 patients found that only 58% of those with a positive D-dimer had the requisite CTA, as did 7% of patients with a negative D-dimer.2
Physicians should not be concerned about ruling out a PE in the setting of a negative D-dimer, as a meta-analysis found that this diagnostic approach has a negative predictive value (NPV) of 99.7%.2 It is important to note that the NPV is significantly affected by the sensitivity of the D-dimer assay used. If the D-dimer assay is highly sensitive, a negative result in combination with a low, moderate, or unlikely probability Wells score rules out the diagnosis of PE. If the assay is moderately sensitive, however, only a low or unlikely probability Wells score rules out PE.10
The inappropriate work-up of this group of patients is significant and extends beyond the ultimate goal of preventing morbidity and mortality. The unnecessary use of pulmonary CTA is extremely expensive, exposes patients to unnecessary radiation, and results in contrast nephrotoxicity in about 4% of patients.9 Although pulmonary CTA is the standard diagnostic test for PE, other imaging modalities are more appropriate in some cases (TABLE 2).
TABLE 2
Alternative imaging modalities for diagnosing PE1,4,7,11
| Modality | Indication |
|---|---|
| Ventilation-perfusion scanning | Patients with contrast allergies or renal failure; test of choice for diagnosing chronic PE due to limited sensitivity of CT |
| Venous compression ultrasonography | Patients with symptoms of PE and signs/symptoms of DVT |
| Pulmonary angiography | Most invasive test. Should be used only in patients with high probability of PE who may need vascular intervention |
| CT, computed tomography; DVT, deep venous thrombosis; PE, pulmonary embolism. | |
The bottom line
This case report illustrates the importance of using sound clinical judgment when diagnosing a PE. Although our patient initially had a moderate probability Wells score and a negative D-dimer, her symptoms persisted. Her history of OCP use, persistent dyspnea, and new symptoms of a DVT prompted us to reinitiate the diagnostic algorithm and eventually diagnose a PE.
It is always essential to treat the patient and not simply react to laboratory values. To avoid unnecessary testing, however, adhering to the algorithm is equally important.
CORRESPONDENCE
Michael S. Kelleher, MD, University of Connecticut School of Medicine, 263 Farmington Avenue, Farmington, CT 06030; mkelleher@student.uchc.edu
