Standard and computed tomography in the evaluation of neoplasms of the chest. A comparative efficacy assessment.

Although in practice computed tomography (CT) has nearly replaced standard tomography in the evaluation of chest diseases, an analysis of the literature shows that standard tomography may be preferred in some settings. After a detailed review of studies on test performance, we calculated overall sensitivity and specificity values. Using Bayesian analysis, we then developed guidelines for choosing between computed and standard tomography. A test is recommended if its results might obviate the need for an invasive procedure. Standard tomography is recommended in the detection of mediastinal metastases from lung cancer. In the detection of hilar metastases, standard tomography is recommended for patients who have low pretest probabilities of metastases to this site, and computed tomography is recommended for patients who have high pretest probabilities. Computed tomography is recommended for the detection of calcification in solitary pulmonary nodules. Tomographic procedures are not recommended when the pretest probability is intermediate. The discrepancy between current practice and these guidelines highlights the need for carefully designed, multi-institutional comparative studies of radiographic procedures.