REFINEMENTS IN THE DIAGNOSIS OF PULMONARY EMBOLISM

According to these studies:

* A revised Geneva score for the objective assessment of the likelihood of pulmonary embolism now is a standardized, validated tool.

* Additional testing is unnecessary in the presence of negative d-dimer test results and an assessment of low probability.

The results of two recently published studies provide clinicians with guidance in the classification of probable pulmonary embolism and in the circumstances under which extensive diagnostic testing either is warranted or can be omitted safely.

In the first study, Le Gal and colleagues examined data on patients presenting to the ED with suspected pulmonary embolism. The researchers hoped to devise a predictive rule that incorporates only objective clinical indicators into the prediction of the likelihood of pulmonary embolism before diagnostic testing is done. (This is in contrast to the Wells score, which involves subjective assessment by the physician.)

The Geneva score, another commonly used predictive rule, involves an arterial blood gas reading taken with the patient breathing room air, but such readings aren't always possible. To revise that tool, the researchers identified eight clinical variables that were independently associated with pulmonary embolism in a sample of ED patients involved in a separate study: age older than 65, a history of deep venous thrombosis or pulmonary embolism, recent bone fracture or surgery, malignancy, expectoration of blood, an accelerated heart rate, unilateral lower limb pain, and unilateral edema painful on deep palpation. Each variable was assigned a point value that, when totaled, yielded an outcome score representing clinical probability (low, intermediate, or high).

Scores obtained using the new predictive tool corresponded to the actual prevalence of pulmonary embolism in the study patients. Among the sample of 956 patients the prevalence of pulmonary embolism in the low-probability category was 9%, the prevalence in the intermediate-probability category was 27.5%, and the prevalence in the high-probability category was 71.7%. The researchers then validated the rule using an independent sample of 749 comparable patients; the prevalence of pulmonary embolism was 7.9% in the low-probability category, 28.5% in the intermediate-probability category, and 73.7% in the high-probability category. The investigators concluded that the revised Geneva score had been standardized and validated and, now based solely on objective clinical variables, could be tested for clinical usefulness and more fully validated in further outcome studies.

In a separate, randomized clinical trial conducted by Kearon and colleagues, the need for further diagnostic testing was evaluated in patients with suspected pulmonary embolism but negative d-dimer test results who had been classified into a category of either low or moderate or high probability according to the Wells score. Patients in the low-probability category and with negative d-dimer test results were randomized into either a control group or an experimental group. In the control group, ventilation-perfusion lung scans and ultrasonography of the proximal deep veins of the legs were performed successively on the same day; in the experimental group, no additional diagnostic testing was performed. Similarly, patients in the category of moderate or high probability of pulmonary embolism and with negative d-dimer test results underwent venous leg ultrasonography, repeated after seven and 14 days, and patients in the corresponding experimental group underwent no further testing.

The study revealed an overall 15.2% rate of venous thromboembolism (VTE) on follow-up after six months. No VTEs (0 in 182 patients) were found in the experimental low-probability group, one VTE (1 in 185 patients) was identified in the control low-probability group. One VTE (1 in 41 patients) was found in the experimental moderate-or-high-probability group, and no VTEs (0 in 41) were found in the control moderate-or-high-probability group. The researchers reasoned, therefore, that subsequent diagnostic testing isn't necessary in patients with negative d-dimer test results and classification of low clinical probability of pulmonary embolism. However, despite the comparable results observed in the moderate-or-high-probability group, the small number of patients evaluated prevented the authors from drawing the same conclusion regarding that population.

The findings of the two studies provide clinicians with refined clinical guidance tools and evidence that can assist in the diagnosis of pulmonary embolism. An accurate and expedited diagnosis is instrumental in the timely initiation of treatment and the prevention of complications. Also, the elimination of follow-up and costly diagnostic testing can reduce inconvenience to the patient and save money.