Entropy and expertise: assessing changes in pathologists' language over time using the UK Liver Pathology External...

Area of Science:

• Pathology
• Medical Informatics
• Quality Assurance

Background:

• External Quality Assessment (EQA) schemes are crucial for histopathologist consistency.
• Evaluating linguistic trends in pathology reports can reveal insights into diagnostic practices.

Purpose of the Study:

• To apply Shannon entropy as a novel metric for linguistic variability in the UK Liver Pathology EQA scheme.
• To quantify language trends in morphological assessments and clinicopathological diagnoses over a decade.

Main Methods:

• Analysis of free-text responses from the UK Liver Pathology EQA scheme over ten years.
• Application of Shannon entropy to assess linguistic diversity in morphological assessments and diagnoses.
• Statistical adjustments for word count and participant tenure.

Main Results:

• A significant increase in Shannon entropy for morphological assessments over time was observed.
• This growing linguistic diversity may correlate with increasing complexity in liver pathology cases.
• Entropy in clinicopathological diagnoses did not show clear evidence of convergent language.

Conclusions:

• Shannon entropy is a novel tool for analyzing pathologist language and enhancing quality assurance.
• High entropy cases indicate greater response diversity and potential diagnostic challenges.
• This method can identify areas for targeted educational interventions in pathology.