Diagnostic strategies in venous thromboembolism.

BACKGROUND AND OBJECTIVE Diagnosis of acute deep vein thrombosis (DVT) and of pulmonary embolism (PE) is often difficult: symptomatic patients are usually investigated employing several diagnostic tests, which should be appropriately selected and sequenced, taking into account their sensitivity, specificity, safety and cost. The objective of this paper is to evaluate the performance of the new diagnostic tests and their combination in rational diagnostic strategies. DESIGN AND METHODS A literature review was made using a Medline(R) database search for the period 1988-1998 on the following key words in various combinations: diagnosis, diagnostic strategy, venous thrombosis, pulmonary embolism, venous thromboembolism. Results of a new study by our group on diagnosis of DVT in hospitalized patients are also discussed. RESULTS In patients with symptoms or signs suggestive of DVT, compression ultrasound (CUS) appears to be the diagnostic test of first choice, since it is a noninvasive test with high specificity and sensitivity for proximal DVT (about 97%). When CUS gives a negative result it is usually recommended that the test is repeated after one week, since its sensitivity for calf DVT is poor. The positive and negative predictive values (PPV and NPV) of CUS in symptomatic outpatients can be improved if adequate consideration is given to clinical diagnosis, using a standardized model (ref. #9), which allows symptomatic outpatients to be categorized as having a high, moderate or low probability of DVT. In case of agreement between clinical diagnosis and CUS results, no further testing is needed: patients with high or intermediate clinical probability and positive CUS results are treated, while in patients with low clinical probability and negative CUS results the diagnosis of DVT is excluded. In the case of discrepancy between clinical diagnosis and CUS results, D-dimer test and/or venography are requested. However in patients who develop signs or symptoms of DVT in the hospital the clinical model does not work, and diagnosis should be based on an appropriate mix of CUS, D-dimer (DD) test and venography. In patients presenting with signs or symptoms of pulmonary embolism, the ventilation/perfusion (V/P) lung scan remains a pivotal diagnostic test, and pulmonary angiography the reference standard, but both methods have limitations and in recent years other diagnostic tests such as echocardiography, helical (or spiral) computerized tomography, and magnetic resonance imaging have been introduced into clinical practice. Moreover, all four diagnostic tools mentioned for DVT diagnosis can be considered. Several diagnostic strategies have been proposed and evaluated in comparative studies but there is still debate over the most efficient test combination or sequence. INTERPRETATION AND CONCLUSIONS Diagnostic strategies which include adequate consideration of clinical diagnosis using standardized models have the potential of being more efficient for outpatients (but not for inpatients) with symptoms or signs suggesting DVT of lower limbs. For patients with suspected PE, several diagnostic strategies have been assessed: V/P lung scan remains a pivotal diagnostic test, but its limitations have been increasingly recognized and newer non-invasive techniques are gaining credit. A consensus is still to be reached over the most appropriate combination of diagnostic tests.