Joannalyn Delacruz

Background: Smoking continues to be a leading cause of preventable morbidity and mortality, and more than 480,000 Americans die annually due to smoking-related illness attributable to smoking and secondhand smoke. More advanced, responsive, and tailored digital interventions using machine learning and artificial intelligence may be a valuable tool for successful smoking cessation referrals. Objective: This study used the dynamic systems development method to incorporate patient and consumer sources of conversational data to develop a technology-assisted motivational interviewing (TAMI) chatbot, a digital agent using machine learning models to deliver motivational interviewing (MI) for tobacco cessation. Methods: During the functional model iteration phase, user-centered design interviews with smokers (n=3) informed the creation of TAMI. The design and build phase involved the use of existing datasets to guide the incorporation of MI-consistent utterances, language recognition, and topic classification to guide a discussion about smoking, and providing a tailored quit plan if indicated. During the implementation phase, user experience interviews with randomly selected participants (n=9) in a pilot trial discussed their experiences with TAMI. Results: User-centered design interviews indicated a desire for a chatbot that was engaging and adaptable to personal interests in quitting smoking. Inductive analysis of user experience interviews revealed that anonymity, regular reminders, and a humanized experience facilitated engagement with TAMI, but technical glitches, chatbot misunderstandings, and issues with rapport were barriers to engagement. Conclusions: Informed by user input and patient and consumer datasets, TAMI can use MI skills to elicit change talk and/or accurately evaluate readiness for tobacco cessation. Further development will enhance TAMI’s ability to seamlessly engage with users when discussing behavior change and assist underserved populations achieve improvements in a variety of health behavior goals.

Objective: Motivational Interviewing may be an ideal communication style to use in conjunction with Collaborative Care to address opioid risk, as it can facilitate the discussion of alternative pain care strategies (APCSs) that are pharmacological (APCS-P; e.g., the use of non-opioid pain relievers) or non-pharmacological (APCS-NP; e.g., yoga). This study developed and piloted a coding system (MI Skills Code–APCS) for these discussions. Method: Sessions (n = 119) from a completed randomized controlled trial comparing Collaborative Care Motivational Interviewing (CCMI) or Attention Control Psychoeducation (ACP) delivered by care managers over 12 weeks to veterans with chronic pain and high-risk opioid use enrolled in VA primary care (N = 44). Results: Coders were able to reliably code the client utterances related to APCSs in the sessions (ICCs = 0.58–0.81). The APCS-P and APCS-NP codes were positively correlated with each other. There were two significant relationships between the MISC-APCS codes (motivational states) and the pain interference and endorsement of non-pharmacological pain care goals at 20-week follow-up. Conclusions: The MISC-APCS has promise as a coding system that can reliably record client utterances regarding different types of pain care strategies. These utterances may be associated with post-treatment reports of pain and efforts to reduce opioid risk. The rapid development of artificial intelligence applications to healthcare can utilize this coding system to assist with the assessment and treatment of chronic pain.

UNSTRUCTURED Interactions based in Motivational Interviewing (MI) [1], driven by machine learning, may provide an efficient way to evaluate readiness to quit, elicit behavior change preferences, and scalable flexibility to extend reach to more diversified target populations. This study used patient and public sources of conversational data to develop a Technology Assisted Motivational Interviewing chatbot (TAMI), a digital agent employing machine learning models to deliver MI for tobacco cessation. Consistent with the four tasks of MI (engagement, focusing, evocation, and planning), the creation of TAMI involved 1) utilization of existing de-identified datasets to identify specific smoking topics of interest; 2) incorporation of MI-Consistent utterances to enhance relational connection to TAMI and elicit language expressing interest in quitting (change talk); 3) instilling change talk recognition and topic classification to guide discussion with TAMI and 4) providing tailored treatment options if indicated. Informed by patient, provider, and public discussions about smoking, TAMI can explore motivation associated with tobacco use, and privately employ interventions to elicit change talk, or accurately evaluate readiness for tobacco cessation. TAMI is also scalable technology, which opens the possibility of having more tools for clinical assessment, delivering MI, and tailoring cessation referrals to underserved populations or those with other target behavior goals.