Sensitivity and Specificity of Helical Computed Tomography in the Diagnosis of Pulmonary Embolism

PURPOSE: To determine the sensitivity and specificity of helical computed tomography (CT) for the diagnosis of pulmonary embolism and to determine the safety of withholding anticoagulant therapy in patients who have clinically suspected pulmonary embolism and negative results on helical CT. DATA SOURCES: The MEDLINE database was searched for all reports published from 1986 to October 1999 that evaluated the use of helical CT for the diagnosis of pulmonary embolism. Bibliographies of the retrieved articles were cross-checked to identify additional studies. STUDY SELECTION: All prospective English-language studies were selected. Retrospective studies, review articles, and case reports were excluded, and 5 of the 20 identified articles were excluded. The scientific validity of the remaining 15 articles was assessed. DATA EXTRACTION: Two of the authors used a priori, pre-defined criteria to independently assess each study. A third author resolved disagreements by adjudication. The pre-defined criteria were inclusion of a consecutive series of all patients with suspected pulmonary embolism, inclusion of patients with and those without pulmonary embolism, a broad spectrum of patient characteristics, performance of helical CT and pulmonary angiography (or an appropriate reference test) in all patients, and independent interpretation of the CT scan and pulmonary angiogram (or reference test). Specific data on sensitivity and specificity and the associated 95% CIs were recorded when available. DATA SYNTHESIS: No study met all of the predefined criteria for adequately evaluating sensitivity and specificity. The reported sensitivity of helical CT ranged from 53% to 100%, and specificity ranged from 81% to 100%. In no prospective study was anticoagulant therapy withheld without further testing for venous thromboembolism in consecutive patients with suspected pulmonary embolism. One prospective study reported the outcome of selected patients with negative results on helical CT who did not receive anticoagulant therapy. CONCLUSIONS: Use of helical CT in the diagnosis of pulmonary embolism has not been adequately evaluated. The safety of withholding anticoagulant treatment in patients with negative results on helical CT is uncertain. Definitive large, prospective studies should be done to evaluate the sensitivity, specificity, and safety of helical CT for diagnosis of suspected pulmonary embolism.