Abstract
BackgroundIn the United Kingdom radiographers with a qualification in image interpretation have interpreted mammograms since 1995. Radiographers work under the title of, radiography advanced practitioner (RAP), with many achieving consultant radiographer status, yet no definitive data are available on what determines RAP performance. This PhD examines this evidence gap through investigating clinical, non-clinical and experiential factors with potential to impact the reporting performance of RAPs.
Aim and Objectives
The aim of this research is to optimise radiographic interpretation by RAPs through an advanced novel assessment platform to evaluate performance levels and reader characteristics. The objectives of this study are to:
1. Identify reader-based factors that impact upon RAP image interpretation with a particular focus on:
a. Agents less relevant to clinical practice
b. Clinically relevant agents
2. Identify image-based mammographic appearances that challenge RAPs as opposed to Radiologists.
Methods
This PhD is comprised of a series of investigations including a systematic review and two research investigations. To provide a comprehensive synthesis of the associated literature regarding the performance of RAPs, a systematic literature review was undertaken exploring factors associated with diagnostic performance variation in reporting of mammograms. PubMed, ScienceDirect and Google Scholar databases were searched using the following terms: ‘Radiology’, ‘Radiologist’, ‘Radiographer’, ‘Radiography’, ‘Mammography’, ‘Interpret’, ‘read’, ‘observe’ ‘report’, ‘screen’, ‘image’, ‘performance’ and ‘characteristics.’ Exclusion criteria included articles published prior to 2000 as digital mammography was introduced at this time. Non-English articles language were also excluded. 38 of 2542 studies identified were analysed.
Study 1 identified factors that impact RAP performance. To identify what factors impact RAP performance, the performance of 18 UK based RAPs interpreting a cloud-based test set of 60 mammograms with known truth were assessed in Study 1. Performance values were established for each RAP and statistical tests were used to explore the impact of varying factors on image interpretation accuracy.
In Study 2, to determine what appearances challenge RAPs, two cohorts of readers, radiologists and radiographers were compared, each of whom read the same set of 60 cases, some of which demonstrated cancer, whilst the remaining cases were normal. Difficulty indices were calculated for each case and were analysed for each reader cohort.
Results
A systematic literature review demonstrated the paucity of information related to RAP performance. It demonstrated that data on radiologist performance variability is widely available but data on radiographers interpreting mammograms is non-specific.
Key findings from Study 1 included: lower specificity values (p=0.0337) for RAPs who stated that their emotional mind set impacted their image reading (88) compared to those who did not (66.5); higher ROC values (p=0.0042) in those readers who: had less than 10 years ’experience (0.93), compared to readers with greater than 10 years of experience(0.84); read greater than 100 cases a week (0.93), compared to those who read less 100 (0.87) (p=0.0358)and those who consistently relied on prior images (0.94), compared to those who only occasionally relied on prior projections (0.89) (p=0.0231). Further, higher sensitivity values of 90.7% were seen in those RAPs who had an eye test in the last year compared to those who had not, 82% (p=0.021).
Key findings for Study 2 included strong correlations were shown between radiologist and radiographer groups for cancer and normal cases difficulty indices respectively (r=0.83CI:0.61-0.93) and (r=0.73; CI:0.54 -0.85). Specific to RAPs, greatest difficulty scores were identified for cancer cases with soft tissue appearances as opposed calcifications (p=0.003) and for normal cases without prior images, compared to those with (p=0.03).
Conclusion
The systematic literature review identified the need to address the paucity of information on RAP performance. Both studies demonstrate agents and challenges that may impact upon the diagnostic performance of RAPs when reading mammograms. This included volume of reads, the use of prior mammograms, eye testing, emotional mindset and the mammographic challenge of interpreting soft tissue presentations. This data has given rise to a variety of analyses, specifically in the case of the image interpretation style of the radiographer which has not yet been investigated.
Advances in knowledge
This is the first study to implement an image learning system of assessment and feedback that instantly is available to RAPs, which allow the paucity of research to be addressed. The new knowledge provided by the findings of this research elevate the work of RAPs and provide insights that may enhance image interpretation. This information will support the future endeavours of RAPs, not only in mammography interpretation but in all radiography reporting activities
| Date of Award | Aug 2025 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Prof Chantal Ski (Supervisor) |