Related Experiment Videos
Radiologist uncertainty and the interpretation of screening
Patricia A Carney1, Joann G Elmore, Linn A Abraham
1Department of Community and Family Medicine, Dartmouth Medical School, Hanover and Lebanon, New Hampshire, USA. patricia.a.carney@dartmouth.edu
Summary
Radiologists
Area of Science:
- Radiology
- Medical Imaging
- Diagnostic Procedures
Background:
- Mammography interpretation involves inherent uncertainty.
- Understanding factors influencing radiologist uncertainty is crucial for improving diagnostic accuracy.
- Variability in mammography interpretation performance is a significant concern in breast cancer screening.
Purpose of the Study:
- To investigate radiologists' responses to uncertainty during mammography interpretation.
- To assess the impact of radiologist uncertainty on interpretive performance metrics.
- To identify practice-related factors associated with varying levels of uncertainty.
Main Methods:
- A mailed survey was used to collect data on radiologist demographics, practice characteristics, and reactions to uncertainty.
- Radiologist responses were linked to actual performance data from mammography registries.
- Interpretive performance was evaluated using metrics such as positive predictive value, recall rates, and specificity.
Main Results:
- A response rate of 76.8% was achieved with 139 participating radiologists.
- Factors such as female gender, increased years of experience, and higher interpretation volume were associated with lower uncertainty scores.
- While certain performance metrics showed trends of being affected by uncertainty, these relationships were not statistically significant.
Conclusions:
- Radiologist uncertainty scores are influenced by practice factors including gender and interpretive experience.
- The study found no statistically significant evidence that radiologists' reactions to uncertainty impact their mammography interpretive performance.
- Further research may be needed to fully elucidate the complex relationship between uncertainty and diagnostic accuracy in mammography.