The clinical value of body computed tomography over time and technologic change

A clinical study of body computed tomography (CT) at Massachusetts General Hospital evaluated 2,619 patients who were prospectively assigned to one of 12 examination protocols. Data obtained from referring physicians and patient records just before CT examination and later in the course of care served as a basis for judging the contribution of CT to diagnostic understanding, use of other tests, and choice of therapy. Fifty-three percent of examinations produced a substantial or unique contribution to diagnostic understanding, and 15% contributed to a change in treatment. Performance in different protocols varied greatly: lymphoma, pancreas, retroperitoneum, lung, and liver ranked in the top half for both diagnostic and therapeutic efficacy; pelvis, urology, and colon fell in the bottom third. Overall, CT reduced surgery by an estimated 14% and angiography by an estimated 11% in the study population. Availability of CT was also associated over time with significant declines in the frequency of sonographic examinations and of lymphangiography, though not of endoscopic retrograde cholangiopancreatography. Compared with an 18-sec scanner, examinations on a 3-sec unit more frequently contributed to improved diagnostic understanding (p less than 0.05) and to increased physician confidence in previously chosen treatment (p less than 0.001). Studies of the diagnostic and therapeutic efficacy of devices like CT can guide clinical expectations and provide a basis for evaluating new imaging methods.