PURPOSE: To compare accuracy of interpretation by radiologists and emergency medicine physicians of conventional radiographs and digitized images on a workstation. MATERIALS AND METHODS: One author selected 120 radiographs from the radiology department library, including 62 musculoskeletal, 20 abdominal, and 38 chest examinations. Analog radiographs were digitized. There were 60 positive and 60 control cases. Positive cases demonstrated clinically important disease and had a high degree of diagnostic difficulty. Thirty-one cases were judged to be critical to the patient's immediate care, requiring prompt accurate interpretation. Four groups of readers were used: staff radiologists and emergency medicine physicians and second-year radiology and emergency medicine residents. RESULTS: All reader groups performed better when interpreting conventional radiographs than digitized images. Differences in favor of radiograph reading were statistically significant for overall accuracy related to all cases and to critical cases (P < .05, one-tailed test). CONCLUSION: Results with the teleradiology system were found unacceptable for primary interpretation of the spectrum of radiographs seen in an emergency department.
- Diagnostic radiology, observer performance
- Emergency medical service system
- Receiver operating characteristic curve (ROC)
- Video systems
ASJC Scopus subject areas
- Radiology Nuclear Medicine and imaging