Radiologists’ practice patterns vary considerably, with substantial deviation from evidence-based approaches in medical literature, when interpreting incidental findings on chest CT scans, according to the results of a survey published in the December issue of Academic Radiology .
Leslie Eisenbud Quint, MD, of the University of Michigan Health System in Ann Arbor, and colleagues wanted to study patterns in the evaluation of incidental findings on chest CT scans to determine areas requiring further education. They noted that incidental findings are encountered more frequently due to advances in scanning technology and an increase in the number of exams performed.
“Therefore, it is highly important for radiologists to be educated with regard to the optimal interpretation and management of incidental findings, so that extra, inappropriate examinations and procedures may be avoided,” wrote the authors.
Surveys were administered to 1,600 radiologists, largely from the University of Michigan Health System alumni society and the Society of Computed Body Tomography and Magnetic Resonance, but also some radiologists who were participating in continuing education courses.
The surveys presented four clinical case questions involving the incidental findings of a thyroid lesion; enlarged mediastinal lymph nodes; asymptomatic, small pulmonary embolus; and small lung nodules. Respondents’ answers were compared with the best practices as defined in medical literature. Additional questions were used to determine demographics and the respondents’ comfort level in addressing the findings.
Overall survey response was 28 percent, and correct response rates ranged from 26 percent to 79 percent. Six percent of respondents had answers that corresponded with evidence-based practice for all cases. Eighty percent felt comfortable in addressing findings, while 57 percent believed they needed more training.
As far as factors that were associated with high proportions of correct answers, respondents who were working in a teaching practice, had fellowship training in cardiothoracic radiology or subspecialized in abdominal or cardiothoracic radiology tended to have higher correct response rates. Comfort level was predictive of correct response rate only in the case question involving enlarged mediastinal lymph nodes.
“The results of this survey reveal that there is considerable disagreement among practicing radiologists with regard to the significance and any recommendations for further evaluation of various types of incidental findings on chest CT scans,” wrote the authors, adding that many respondents “overestimated their ability to make evidence-based diagnostic recommendations.”
The authors noted that only one of the scenarios, incidental finding of lung nodules, has well defined, evidence-based guidelines that have been widely disseminated in the radiology community. However, despite the more highly publicized guidelines for this scenario, approximately 60 percent of respondents still incorrectly answered that question, a higher incorrect response rate than two of the other three scenarios.
Quint et al concluded that the findings of the survey demonstrated a significant need for further education in these areas.
A separate study published in the Journal of the American College of Radiology in November found that radiologists often disagreed with one another on interpretations for a number of incidental finding scenarios on body CT scans.
The importance of being able to identify incidental findings during CT exams is underscored by the completion of the National Lung Screening Trial , which found a 20 percent reduction in mortality among heavy smokers receiving annual CT screening compared with conventional x-ray screening and fueled a debate on whether CT screening should be more commonplace.