AI-Augmented Motivational Interviewing Training for Primary Care

Overview

This study will examine the impact of training primary care providers (PCPs) in motivational interviewing (MI) using artificial intelligence (AI) to augment the training process. MI is a patient-centered approach to engaging patients in their own care. There will be a control group and two intervention groups, with the intervention groups receiving a different amount of MI training. The hypothesis is that the AI-augmented MI training will result in improved patient outcomes, improved clinician wellbeing, and reduced behavioral manifestation of clinician biases. This mixed-methods project will also collect qualitative data from structured interviews and focus groups with participating PCPs to examine perceived facilitators and barriers to the use of the MI approach in primary care.

Detailed description

Motivational interviewing (MI) is an evidence-based patient-centered approach that has been demonstrated to be effective in increasing patient engagement in their own care. Using MI, a clinician is compassionate curious to understand patients\' perspective on their behavior, and conversationally guides them to discover a need for behavior change. However, MI is not widely utilized in primary care practices and challenging to teach, in part because many primary care providers (PCPs) take a direct approach with their patients by educating and advising them regarding steps to improve their health. However, simply telling patient what to do tends to be ineffective, leading to frustration on the part of both patients and PCPs.

To help implement the MI approach more widely, our team has developed an artificial intelligence (AI) augmented tool for MI skill development, ReadMI™ (Real-time Assessment of Dialogue in Motivational Interviewing). The goal of this project is to: 1) examine the association of AI-measured proficiency in MI to patient outcomes, PCP wellbeing, and PCP manifestations of bias, and 2) to determine the extent to which AI-augmented MI skills training can impact the same outcome, wellbeing, and bias manifestation variables in a randomized controlled trial (RCT). This mixed-methods project will also employ structured interviews and focus groups of participating PCPs to collect qualitative data for better understanding both facilitators and barriers to the implementation of MI in primary care.

The overall hypothesis is that PCPs with the strongest MI proficiencies will have patients with better outcomes, and that those PCPs will also demonstrate less burnout and less manifestation of bias. Additionally, it is hypothesized that making use of AI in MI training will be seen by PCPs as a facilitator to the use of MI in their practices.

Interventions

• Behavioral Basic AI (artificial intelligence)-Augmented Motivational Interviewing Training
  Basic participants will participate in two motivational interviewing (MI) training sessions that involve role-play practice of MI. Artificial intelligence (AI) will be used to provide metrics on MI-skills (percentage of the time the clinician talks, use of open-ended questions, use of closed-ended questions, use of reflective statements, use of 0-10 scales) as measured by the ReadMI™ tool (Real-time Evaluation of Dialogue in Motivational Interviewing), and both the spirit of MI and presence of b
• Behavioral Plus AI (artificial intelligence)-Augmented Motivational Interviewing Training
  Plus participants will participate in four motivational interviewing (MI) training sessions that involve role-play practice of MI. Artificial intelligence (AI) will be used to provide metrics on MI-skills (percentage of the time the clinician talks, use of open-ended questions, use of closed-ended questions, use of reflective statements, use of 0-10 scales) as measured by the ReadMI™ tool, and both the spirit of MI and presence of bias as measured by Dougall GPT.

Primary outcome measures

• MI Communication Metrics [Time frame: Months 4-7, Months 16-20, Months 30-33]
• MI Spirit [Time frame: Months 4-7, Months 16-20, Months 30-33]

Secondary outcome measures (9)

• Bias Manifestation [Time frame: Months 4-7, Months 16-20, Months 30-33]
• Mini Z (Zero) Burnout Scale 2.0 [Time frame: Months 4-5, Months 16-17, Months 25-26, Months 32-33]
• Aggregated Patient Data for Each Participating PCP [Time frame: Months 4-5, Months 16-17, Months 25-26, Months 32-33]
• Perceived Stress Scale (PSS) [Time frame: Months 4-5, Months 16-17, Months 25-26, Months 32-33]
• Brief Resilience Scale (BRS) [Time frame: Months 4-5, Months 16-17, Months 25-26, Months 32-33]
• Clinician Support for Patient Activation Measure (CS-PAM) [Time frame: Months 4-5, Months 16-17, Months 25-26, Months 32-33]
• Stress Mindset Measure (SMM) [Time frame: Months 4-5, Months 16-17, Months 25-26, Months 32-33]
• MI Survey Measure [Time frame: Months 4-5, Months 16-17, Months 25-26, Months 32-33]
• Diversity Awareness Self-Reflection Tool (DASRT) [Time frame: Months 4-5, Months 16-17, Months 25-26, Months 32-33]

Eligibility criteria

Inclusion criteria

• Primary care providers (physicians, physician assistants, and nurse practitioners) employed in primary care practices of the Premier Physician Network of Premier Health (southwest Ohio) are eligible to participate.

Exclusion criteria

• Eligible PCPs who opt to not provide informed consent for participation will be excluded from the study.

Criteria are shown verbatim from the registry (in English). Final eligibility is always assessed by the study center.

Healthy volunteers: No

Study design

Allocation

Randomized

Model

Parallel assignment

Masking

Open label

Primary purpose

Health services research

Study locations

United States · 2 centers

• Premier Health — Dayton
• Wright State University — Dayton

Publications

• Lundahl B, Moleni T, Burke BL, Butters R, Tollefson D, Butler C, Rollnick S. Motivational interviewing in medical care settings: a systematic review and meta-analysis of randomized controlled trials. Patient Educ Couns. 2013 Nov;93(2):157-68. doi: 10.1016/j.pec.2013.07.012. Epub 2013 Aug 1. PMID 24001658
• Hershberger PJ, Pei Y, Bricker DA, Crawford TN, Shivakumar A, Castle A, Conway K, Medaramitta R, Rechtin M, Wilson JF. Motivational interviewing skills practice enhanced with artificial intelligence: ReadMI. BMC Med Educ. 2024 Mar 5;24(1):237. doi: 10.1186/s12909-024-05217-4. PMID 38443862
• Hershberger PJ, Bricker DA, Conway K, Torcasio MH. Turning "Lose-Lose" into "Win-Win": What Is Good for Them Is Good for Us! Med Sci Educ. 2021 Mar 30;31(3):1177-1181. doi: 10.1007/s40670-021-01280-4. eCollection 2021 Jun. PMID 34457961
• Chapman EN, Kaatz A, Carnes M. Physicians and implicit bias: how doctors may unwittingly perpetuate health care disparities. J Gen Intern Med. 2013 Nov;28(11):1504-10. doi: 10.1007/s11606-013-2441-1. Epub 2013 Apr 11. PMID 23576243
• Hershberger PJ, Martensen LS, Crawford TN, Bricker DA. Promoting Motivational Interviewing in Primary Care: More Than Intention. PRiMER. 2021 Feb 4;5:7. doi: 10.22454/PRiMER.2021.287928. eCollection 2021. PMID 33860162

Identifiers

NCT: NCT06597123 · 25-0047-P0001

Primary sources (government registries)

• ClinicalTrials.gov (US NIH)

View this study on ClinicalTrials.gov ↗