Background Overuse of imaging for people with musculoskeletal conditions, including low back pain, is a widespread and costly healthcare problem with potential for adverse health outcomes. The radiology report plays an important role in determining decisions about further investigation and management. Understanding the imaging report and how findings are communicated may identify ways in which to modify radiology reports to address the problem of imaging overuse.

Aim To develop and pilot a template to assess the content of imaging reports for musculoskeletal conditions.

Methods This pilot study is a retrospective audit of lumbar spine plain radiograph (X-ray) reports of patients attending a private hospital emergency department in Melbourne, Australia with a primary complaint of low back pain during the 2015 calendar year. Descriptive and thematic analysis of the terms used in each report was conducted. Each report was examined according to whether or not it adhered to items identified from the Royal Australian and New Zealand College of Radiologist’s reporting guidelines including whether: the clinical question was addressed; normal or likely nonsignificant findings were explicitly stated; and whether the terminology was appropriate to the readers of the report. Terms related to specific findings were classified according to whether they are likely or unlikely to be of clinical importance based on expert opinion and previous research.

Results The content of 25 lumbar spine X-ray reports written by 13 different radiologists was analysed. Patients were predominantly female (80%) with a mean age of 76 years. There was a mean of 78 ± 24 words in each report. 48% of reports included clinical information provided by the requestor but only approximately half of these (20% of total) explicitly stated a clinical question. Of the reports that included a clear clinical question, 80% explicitly answered the question. Terms used were classified into three themes representing anatomic parts, specific findings and descriptors such as severity and degree of uncertainty. 128 terms relating to specific findings were extracted but there were only 24 unique meanings. 83% of normal or likely nonsignificant terms referred to the absence of a fracture or misalignment. There was a high frequency of terms relating to disc and facet joint degeneration considered unlikely to be of clinical importance.

Conclusion Our template for assessing the content of lumbar spine X-ray reports extracted useful information that can be used to identify ways of improving these reports to aid with their interpretation by the referrer and patient. The reports generally conformed to reporting guidelines but may be improved by ensuring an explicit clinical question is posed and considered. Our findings suggest it may be possible to develop a consistent template language to describe findings of lumbar spine X-ray reports requested by emergency physicians. Qualifying that degenerative terms may be normal for age may also aid interpretation. Further research is required to apply this template to different types of radiology reports (e.g. CT and MRI scans), for different musculoskeletal conditions (e.g. shoulder pain), and in different settings (e.g. primary care).