ABSTRACT
OBJECTIVES: Diagnostic strategies in patients with suspected pulmonary embolism have been extensively studied in outpatients; their value in hospitalized patients has not been well established. Our aim was to determine the safety and clinical utility of a simple diagnostic strategy in hospitalized patients with suspected pulmonary embolism. DESIGN: Prospective management study. SETTING: Twelve teaching hospitals (five academic, seven general hospitals). SUBJECT: A total of 605 hospitalized patients with clinically suspected pulmonary embolism. All patients completed the study. INTERVENTIONS: First the clinical decision rule (CDR)-score was calculated. An unlikely CDR-score in combination with a normal D-dimer excluded pulmonary embolism. All other patients underwent helical computed tomography (CT). CT either diagnosed or excluded pulmonary embolism, in which case anticoagulants were started or withheld. All patients were instructed to report symptoms of venous thrombosis. Objective tests were performed to confirm venous thromboembolism. The primary outcome was the incidence of symptomatic venous thrombosis during 3-month follow-up. RESULTS: The combination of an unlikely CDR-score and a normal D-dimer excluded pulmonary embolism in 60 patients (10% of all patients); no venous thromboembolic event occurred during follow-up (0%; 95% CI 0-6.7%). CT excluded pulmonary embolism in 380 patients; during follow-up venous thromboembolism occurred in five patients (1.4%; 95% CI 0.4-3.1%). CONCLUSIONS: An unlikely CDR-score in combination with a normal D-dimer appears to exclude pulmonary embolism safely in hospitalized patients. Before clinical implementation it is important this safety is confirmed by others. CT testing was obviated in only 10% of patients. CT can safely exclude pulmonary embolism in hospitalized patients.
