The interpretation of office radiographs by family physicians

J. G. Halvorsen, A. Kunian, D. Gjerdingen, J. Connolly, M. Koopmeiners, J. Cesnik, C. Christenson, O. Sanderson

Research output: Contribution to journalArticlepeer-review

23 Scopus citations


This study prospectively compared the interpretations of family practice residents and faculty with those of radiologists on 532 office radiographs using a uniform protocol. A total of 136 family practice residents (44 first-year, 40 second-year, 52 third-year) and 42 full-time and part-time faculty participated in the study along with 30 radiologists. The mix of radiographs evaluated was as follows: 44 percent chest, 20 percent lower extremity, 6 percent head, 4 percent lumbosacral spine, 3 percent cervical-thoracic spine, and 3 percent abdomen-pelvis. Interpretation concordance rates between family physicians and radiologists, by level of training, were as follows: first-year residents 83.0 percent, second-year residents 84.4 percent, third-year residents 86.0 percent, and faculty 88.6 percent. Concordance after the resident and faculty preceptor discussed the film and provided a collaborative interpretation of 92.1 percent. This finding compares with previously reported error rates of 10 to 40 percent between experienced radiologists. Only 10.3 percent of the discordant readings (0.8 percent of all radiographs) contained significant discordancies that may have affected patient management or outcome. Had the family physicians been given the option to refer an x-ray film to the radiologist, all x-ray films containing significant discordancies would have been referred.

Original languageEnglish (US)
Pages (from-to)426-432
Number of pages7
JournalJournal of Family Practice
Issue number4
StatePublished - Jan 1 1989


Dive into the research topics of 'The interpretation of office radiographs by family physicians'. Together they form a unique fingerprint.

Cite this