Gaining Insights Into Patient Satisfaction Through Interpretable Machine Learning.

Area of Science:

• Healthcare Management
• Health Informatics
• Machine Learning in Healthcare

Background:

• Patient satisfaction is crucial for quality care and hospital reimbursement.
• Understanding factors influencing patient experience is vital for improvement.
• Healthcare journeys involve multiple professionals and generate valuable data.

Purpose of the Study:

• To develop an interpretable machine learning framework for analyzing patient satisfaction.
• To identify the most influential factors affecting patient experiences in hospitals.
• To integrate data transformation, variable selection, and coefficient learning for actionable insights.

Main Methods:

• Formulated patient satisfaction as a supervised learning task.
• Utilized a mixed-integer programming model for factor identification.
• Developed a framework transforming heterogeneous data into understandable features.

Main Results:

• Achieved desirable model performance in predicting patient satisfaction.
• Maintained excellent model interpretability for practical application.
• Identified key influential factors impacting patient experiences.

Conclusions:

• The proposed interpretable machine learning framework offers a robust method for analyzing patient satisfaction.
• This approach facilitates the extraction of actionable insights from complex healthcare data.
• Enables data-driven strategies to enhance patient-centered care and hospital performance.