Enhanced Quality Measurement Event Detection: An Application to Physician Reporting

Area of Science:

• Health Informatics
• Quality Improvement
• Clinical Data Analysis

Background:

• Electronic health records (EHRs) offer vast clinical data but much is unstructured text.
• Traditional quality reporting relies on structured data, often providing incomplete care summaries.
• Automated reporting from unstructured EHR data is challenging but crucial for accurate quality measurement.

Purpose of the Study:

• To evaluate the feasibility and benefits of enhanced EHR-based physician quality measurement.
• To compare traditional (structured data) and enhanced (including unstructured notes) reporting methods.
• To assess the impact on reporting ten physician quality measures across National Quality Strategy domains.

Main Methods:

• Retrospective study comparing traditional and enhanced EHR data analysis approaches.
• Inclusion of unstructured free-text clinical notes in the enhanced measurement strategy.
• Analysis focused on ten physician quality measures from multiple National Quality Strategy domains.

Main Results:

• The enhanced approach enabled calculation of five new Physician Quality and Performance System measures.
• A five-fold increase in event detection was observed with the enhanced method.
• The enhanced approach achieved an average precision of 88% (95% CI: 83-93%).

Conclusions:

• Enhanced EHR-based quality measurement significantly increases event detection for value-based payments.
• This approach can expedite quality reporting for physician practices burdened by manual chart review.
• Leveraging unstructured EHR data is key to more accurate and comprehensive quality measurement.