Reducing Physicians' Cognitive Load During Chart Review: A Problem-Oriented Summary of the Patient Electronic Record

Area of Science:

• Medical Informatics
• Natural Language Processing
• Clinical Decision Support

Background:

• Electronic health records (EHRs) contain vast amounts of data, increasing physician cognitive load.
• Existing EHR systems present challenges for efficient information retrieval and synthesis.
• Automated summarization tools are needed to mitigate physician burden and enhance clinical workflow.

Purpose of the Study:

• To introduce ProSPER (Problem-oriented Summary of the Patient Electronic Record), an automated system for summarizing EHR data.
• To reduce the cognitive burden on physicians by providing problem-centered patient summaries.
• To improve clinical care through more efficient review of patient electronic records.

Main Methods:

• Developed ProSPER using 1,500 longitudinal patient records from two US medical groups.
• Employed multiple natural language processing (NLP) components for tasks including syntactic analysis, adverse drug event extraction, and problem list generation.
• Focused on a problem-oriented approach with disease-specific views for chronic conditions.

Main Results:

• Evaluated the performance of individual NLP components within the ProSPER system.
• Demonstrated how ProSPER components address specific physician challenges in EHR data review.
• Showcased the system's ability to generate auto-summaries and problem lists.

Conclusions:

• Leveraging holistic EHR information is crucial for developing comprehensive summarization applications.
• NLP-based applications like ProSPER show significant potential to support physicians.
• Automated EHR summarization can contribute to improved clinical care and physician efficiency.