The value of thoracic computed tomography scans in clinical diagnosis: a prospective study.

BACKGROUND Computed tomography (CT) scans are used extensively to investigate chest disease because of their cross-sectional perspective and superior contrast resolution compared with chest radiographs. These advantages lead to a more accurate imaging assessment of thoracic disease. The actual use and evaluation of the clinical impact of thoracic CT has not been assessed since scanners became widely available. OBJECTIVE To identify patterns of utilization, waiting times and the impact of CT scan results on clinical diagnoses. DESIGN A before and after survey of physicians who had ordered thoracic CT scans. SETTING Vancouver General Hospital--a tertiary care teaching centre in Vancouver, British Columbia. SUBJECTS Physicians who had ordered CT scans. INTERVENTION Physicians completed a standard questionnaire before and after the CT scan result was available. MEASUREMENTS Changes in the clinical diagnosis, estimates of the probabilities for the diagnosis both before and after the CT scan, and waiting times. RESULTS Four hundred fifty-four thoracic CT cases had completed questionnaires, of whom 80% were outpatients. A change in diagnosis was made in 48% of cases (25% with a normal CT scan and 23% with CT scan findings that indicated a different diagnosis). The largest change in probability scores for the clinical diagnosis before and after the CT scan was 43.9% for normal scans, while it was 36.3% for a different diagnosis and 26.3% for the same diagnosis. High-priority scans were associated with decreased waiting time (--7.89 days for each unit increase in priority). CONCLUSIONS The CT scan results were associated with a change in diagnosis in 48% of cases. Normal scans constituted 25% of the total and had the greatest impact scores. Waiting times were highly correlated with increased urgency of the presenting problem.