Am Fam Physician. 2002;65(4):677-678
Ventilation-perfusion (V/Q) lung scanning, the method most commonly used to identify pulmonary embolism (PE), may not provide a definitive diagnosis. Rapid identification of pulmonary emboli is essential so that appropriate anticoagulation can be initiated. An alternative diagnostic method could include a clinical model with whole blood agglutinin d-dimer testing. Wells and associates looked at a combination of a clinical algorithm and d-dimer test results as a way to manage patients presenting to an emergency department with recent acute symptoms suspicious for PE.
| Clinical signs and symptoms of DVT | Probability (points)* |
|---|---|
| Evidence of DVT (leg swelling and pain with palpation) | 3.0 |
| Heart rate higher than 100 beats per minute | 1.5 |
| Previous objectively diagnosed DVT or pulmonary embolism | 1.5 |
| Immobilization for three or more consecutive days or surgery in previous four weeks | 1.5 |
| Hemoptysis | 1.0 |
| Malignancy | 1.0 |
| Pulmonary embolism as a highly likely diagnosis | 3.0 |
Four participating Canadian medical centers evaluated 930 patients for clinical signs (see accompanying table). All patients had d-dimer testing after the clinical score was determined. Patients who had a low clinical probability for PE and a negative d-dimer test did not undergo imaging studies. These patients were not given anticoagulants but were told to return for further assessment if they developed any suspicious symptoms. All other patients underwent V/Q lung scanning. Scans were interpreted as normal, high probability, or nondiagnostic.
When indicated, compression ultrasonography was done to look for deep venous thrombosis. Patients who had an abnormal ultrasonogram, a high-probability result on V/Q scan or a venous thromboembolic event during the three-month follow-up period were considered to have PE and were given anticoagulation therapy. Of the 86 patients found to have PE during the entire study, seven of them were in the group with low clinical probability, and one had a negative d-dimer test. Among patients with moderate or high embolism probability and those with a positive d-dimer test in whom PE was not found on diagnostic evaluation, only one patient was diagnosed with PE on follow-up. The negative predictive value of d-dimer testing was 99.5 percent in the low-probability group. Fewer imaging studies were obtained among the participating patients than in previous studies in which d-dimer was not considered in the evaluation algorithm.
The authors conclude that including d-dimer testing in the clinical pretest algorithm for PE diagnosis is reliable and reduces the number of imaging studies needed. In patients with a low clinical probability of embolism and a negative d-dimer test, further imaging is not needed.