Advancing Motivational Interviewing Training with Artificial Intelligence: ReadMI

Area of Science:

• Medical Education
• Behavioral Science
• Natural Language Processing

Background:

• Motivational interviewing (MI) is effective for behavior change but underutilized due to training gaps.
• Effective feedback is crucial for improving MI skills in clinical practice.
• A novel tool, Real-time Assessment of Dialogue in Motivational Interviewing (ReadMI), was developed to address these training needs.

Purpose of the Study:

• To evaluate the Real-time Assessment of Dialogue in Motivational Interviewing (ReadMI) tool.
• To examine the relationship between physician speaking time and MI-consistent behaviors.
• To assess the accuracy of ReadMI's automated analysis.

Main Methods:

• Analysis of 48 simulated patient interviews using the ReadMI tool.
• Examination of correlations between physician speaking time and MI metrics (e.g., open-ended questions).
• Interrater reliability testing to validate ReadMI's analysis accuracy.

Main Results:

• Increased physician speaking time correlated with decreased MI-consistent behaviors (r = -0.403, p = 0.007).
• This included fewer open-ended questions, reflective statements, and change ruler uses.
• ReadMI demonstrated accurate analysis of physician responses.

Conclusions:

• ReadMI provides objective, immediate metrics for MI training feedback.
• The tool can enhance skill development for students, residents, and clinicians.
• ReadMI offers an efficient alternative to subjective feedback and time-consuming video reviews.