Attending Radiologist Variability and Its Effect on Radiology Resident Discrepancy Rates

Area of Science:

• Radiology
• Medical Education
• Quality Improvement

Background:

• Resident performance evaluation relies on discrepancy rates.
• Existing benchmarks may not account for attending radiologist variability.

Purpose of the Study:

• Investigate the impact of attending variability on resident discrepancy rates.
• Determine accuracy of current benchmarks and propose updates.

Main Methods:

• Reviewed 5 years of resident interpretations (chest, MSK radiographs; chest, abdomen/pelvis, head CTs).
• Simulated expected discrepancy distributions using Gibbs sampling.
• Compared actual resident distributions to simulated benchmarks.

Main Results:

• Found inverse correlation between resident volume and discrepancy rates.
• Demonstrated significant attending variability in overread volume and discrepancy issuance.
• Simulations indicated current benchmarks are appropriate for chest radiographs, abdomen/pelvis CTs, and head CTs, but too strict for MSK radiographs and chest CTs.

Conclusions:

• Attending practice patterns heavily influence resident discrepancy rates.
• Current benchmarks for MSK radiographs and chest CTs require revision.
• Updated benchmarks are needed for accurate resident performance assessment.