THE DEVELOPMENT OF EXPERTISE IN NON-MEDICAL PERSONNEL INTERPRETING DIAGNOSTIC MAMMOGRAMS

Purpose: To explain how radiographers (non-medical allied health professionals) develop expertise in mammography image interpretation and reporting (MIIR) in order to task substitute for medically qualified radiologists.

Method(s): Realist evaluation and qualitative methods were used to articulate and test the hypothesis that MIIR ‘expertise’ is characterised by decision speed-accuracy. In a two-stage study design, firstly unstructured interviews were conducted with 31 radiographers and 8 radiologists at 8 NHS sites in the UK. Secondly, non-participant observation of 10 diagnostic clinics (91 patient cases) and 5 multidisciplinary meetings (104 patient cases) and post-observation semi-structured ‘teacher-learner’ interviews with 11 radiographers were conducted at 2 NHS sites. Interview and observation data were analysed thematically using three theoretically derived coding categories: expertise takes time, expertise involves intuitive and analytical reasoning, expertise is characterised by decision speed and diagnostic accuracy.

Result(s): All radiographers task-substituting for radiologists had a formal qualification in MIIR that conferred competence to practice and had several years’ experience. Prior to this the radiographers had undertaken mammography image acquisition during which they had learned experientially to recognise some common and typical normal and abnormal appearances. 

Irrespective of experience all radiographers appeared to use a combination of analytic and intuitive reasoning (dual processing) for all cases. They had an initial quick ‘global look’ at the images (thinking fast) followed by a systematic and analytical (thinking slow) two-stage process - careful, systematic and sequential interrogation of images and careful correlation of image appearances with clinical information and physical examination findings.  

Intrinsic features of each case appeared to determine if it was intuitive, quick and easy to interpret or inherently more difficult, requiring longer reasoning. Over time as radiographers learned to interpret and report difficult cases, detect small, subtle abnormalities and recognise rare and unusual pathology, they became more confident decision makers. Following up cases developed their ability to recognise their own (cognitive) limitations and the (technical) limitations of their discipline and enabled them to invoke decision ‘stopping rules’ with greater confidence.  

Conclusion(s): The path to ‘expertise’ for the radiographers in this study was an extended learning journey of instruction, continued practice (repetition), feedback and reflection. ‘Expert’ radiographers continued to use intuitive and analytical reasoning strategies, with improved diagnostic speed occurring as they became more confident decision makers.