
Detecting Clinician Bias in Psychiatric Notes
AI-powered analysis of sentiment and perspective in healthcare documentation
This research develops techniques to identify negative sentiment and clinician bias in psychiatric notes that could harm patient trust and treatment outcomes.
Key Findings:
- Analyzes psychiatric clinical notes to detect stigmatizing language and negative patient descriptions
- Identifies how documentation bias contributes to healthcare disparities in two ways: eroding patient trust and influencing future clinician perspectives
- Focuses on psychiatry where therapeutic alliance is particularly critical for clinical outcomes
Business Impact: This work provides healthcare systems with tools to improve documentation quality, enhance patient-clinician relationships, reduce healthcare disparities, and potentially improve treatment outcomes through more objective clinical documentation.
The Point of View of a Sentiment: Towards Clinician Bias Detection in Psychiatric Notes