Radiologic interpretation by family physicians in an office practice setting

Abstract

Background: Radiology is an integral part of the office practice of many family physicians. Nevertheless, data are sparse on the performance of family physicians in this endeavor. This study investigated the performance of family physicians at interpreting radiographs ordered in a free-standing family practice office.

Methods: A consecutive series of radiographic studies performed at a family practice office during a 3-year period was surveyed. All radiographic studies included in this analysis (N = 1674) were separately interpreted by the family physician ordering the study and an overreading radiologist. If the interpretations agreed, the studies were accepted as having been correctly interpreted. Cases in which the interpretations disagreed were reexamined.

Results: Family physicians correctly interpreted 92.4% of the radiographic studies (95% confidence interval, 91.0 to 93.6). Their accuracy with extremity films (96.0%) was significantly higher than their accuracy with chest films (89.3%, P < .001). Family physicians were more likely to correctly interpret normal films (95.2%) than abnormal ones (85.9%, P < .001). Thirty-five percent of the cases in which there were differences between family physician and radiologist interpretations were correctly interpreted by family physicians.

Conclusions: Family physicians showed a high degree of accuracy in radiologic interpretation in an office setting. Chest films were inherently more difficult to interpret than extremity films. Because correct interpretation depends on body part examined and the prevalence of disease, the performance of family physicians will probably vary in different practice settings.