Using Natural Language Processing to Classify Serious Illness Communication with Oncology Patients

Area of Science:

• Healthcare Quality
• Oncology
• Medical Informatics

Background:

• High-quality serious illness communication is essential for patient-centered care decisions.
• Current methods struggle to measure communication quality and alignment with patient priorities.
• Objective evaluation of serious illness communication is needed for quality improvement.

Purpose of the Study:

• To train natural language processing (NLP) algorithms for identifying and characterizing serious illness communication.
• To explore the potential of NLP for developing quality metrics in serious illness care.
• To assess the feasibility of using NLP to evaluate communication in oncology settings.

Main Methods:

• Development and training of NLP algorithms.
• Application of NLP to analyze documented serious illness communication in oncology patients.
• Characterization of communication patterns and content using NLP techniques.

Main Results:

• Successfully trained NLP algorithms to identify and characterize serious illness communication.
• Demonstrated NLP's efficiency in analyzing large volumes of clinical text.
• Provided a foundation for objective measurement of serious illness communication quality.

Conclusions:

• Natural language processing offers a promising approach to measure serious illness communication quality.
• NLP can support the development of new quality metrics for patient-centered care.
• This methodology has potential applications in oncology and other serious illness contexts.