Influence of specialty affiliation on physicians' pattern of use of CEA test

Review of 437 patients' charts, having CEA determinations at the University of Kentucky Medical Center from January 1 to April 30, 1976, revealed several different patterns of CEA test use which appeared to be related to the specialty affiliations of attending physicians. A majority of patients having nonneoplastic disease diagnoses (69%) were from general medicine and medical specialty services. Also, most patients with benign disease diagnoses (95%) had single CEA determinations which were performed for diagnosis. No change in patient management resulted from CEA test use in patients with nonneoplastic disease. In contrast, a majority (82%) of patients having neoplastic disease diagnoses were on the gynecologic and general surgical services and two‐thirds of patients with neoplastic disease diagnoses had multiple determinations. Since patients with neoplastic disease from the gynecologic service were included in ongoing studies conducted by that service in which clinical decisions were not made on the basis of CEA test results, they were excluded from further analysis. In the remaining patients with malignancy, the CEA test was mostly used for purposes of follow‐up. In this group, three patients from the surgery services and one from the medicine services had alterations in patient management resulting from CEA test use. Prognosis was altered as a result of CEA test results leading to diagnosis and therapy in one patient with lung cancer and re‐exploration of two patients during follow‐up of previously resected colon cancer. Maximal cost effectiveness of CEA test appeared to result from the pattern of use most commonly employed by the general surgical services, i.e., follow‐up of patients following curative resection of colon cancer.