Diagnostic performance on briefly presented digital pathology images.

Area of Science:

• Digital pathology
• Histopathology expertise assessment
• Medical education and training

Background:

• Developing robust proficiency assessments in histopathology is crucial.
• Digital pathology enables novel testing methods like flash viewing.
• Traditional microscopy limits innovative assessment techniques.

Purpose of the Study:

• To evaluate the efficacy of time-restricted testing for histopathology expertise.
• To explore the use of digital images in assessing diagnostic proficiency.
• To identify potential new biomarkers of expertise in pathology.

Main Methods:

• 19 novices, 18 intermediates, and 19 experts viewed 20 histopathology cases.
• Viewing times were restricted to 1 second and 10 seconds.
• Performance differences and internal reliability (Kuder-Richardson 20) were analyzed.

Main Results:

• Significant performance differences were observed between expertise groups.
• The 10-second test demonstrated greater performance differentiation than the 1-second test.
• Both tests exhibited high internal reliability (0.905 for 1s, 0.926 for 10s).
• Expert diagnostic accuracy was 72% at 1s and 83% at 10s.

Conclusions:

• Time-restricted digital image tests are highly reliable for assessing histopathology diagnostic proficiency.
• A 10-second viewing window appears more reliable than a 1-second window.
• Caution against over-reliance on rapid diagnoses is advised to mitigate medical errors.