Diagnosis of pulmonary embolism: have we reached our goal?

Accessible online at: www.karger.com/res Pulmonary embolism (PE) is a common disease that may be fatal if untreated, while it may show a favorable course when promptly recognized and treated. Although clinicians have searched for objective tests to confirm or rule out the diagnosis, PE remains a challenging diagnostic problem since it is detected only at autopsy in about two thirds of cases, even in most recent published series. In recent years, several diagnostic strategies have been introduced in clinical practice. In outpatients, the sequential use of D-dimer measurement, lower limb venous compression ultrasonography, clinical evaluation of the likelihood of PE and a correct interpretation of lung scan results make it possible to correctly treat more than 90% of patients with suspected PE and to safely not treat more than 99% [1]. In both outand inpatients, the evaluation of clinical likelihood combined with a simple, practical reading of perfusion lung scintigraphy helps to restrict the need for angiography to a minority of patients [2]. Nevertheless, the demand for a single test that might accurately and noninvasively allow diagnosis of PE has remained high, more than for other diseases; lately, the hope of such a test seems finally to have become reality with the helical CT scan. Indeed, after the skeptical conclusion of Rathbun et al. [3] in their systematic review reported in 2000, evidence has been published that in patients with suspected PE, helical CT scan may be used efficaciously and safely as the primary test to detect and rule out PE [4]. A valid contribution to the exclusion of PE by helical CT scan is also given by Friera et al. [5] in their paper published in this issue of Respiration. In a prospective study including more than 200 patients, the authors give an elegant demonstration of the utility of a negative CT scan (when used as the only imaging method) in excluding clinically suspected PE. In their report, patients with clinically suspected PE, not otherwise anticoagulated and followed for a 3-month period of time, did not show venous thromboembolic events during the entire follow-up. Though the aforementioned studies have raised widespread enthusiasm, we believe that the old issue concerning the entire diagnostic process and the following management of PE still remains far from having a definitive solution, and it deserves further consideration. Firstly, the strategy for diagnosis varies among different centers according to the availability of one or more techniques that may integrate with each other. For instance, strong evidence has existed for a long time that perfusion lung scan, when normal, reliably excludes the diagnosis of PE without the necessity of further tests. Perfusion lung scan is objective, has a low cost and is easily repeatable, while helical CT is perhaps not cost-effective as a single test. Thanks to the same characteristics, perfusion lung scan may be used to control the efficacy of therapy after an adequate lapse of time, usually before hospital discharge. Is it cost-effective to perform such a control by helical CT? Is it feasible to use an instrument designed for emergencies for routine controls? As a matter of fact, it