Addressing inequities in cardiovascular health is a critical public health objective, particularly for persons with serious mental illness (SMI) who experience alarmingly high rates of cardiovascular disease (CVD) risk factors (obesity, hypertension, dyslipidemia, and diabetes mellitus) and risk behaviors (tobacco smoking, physical inactivity, and unhealthy diet) [1-4]. Lifestyle interventions, especially those that adopt a patient-centered approach, have been shown to be effective in helping persons with SMI to improve a wide range of health behaviors, including weight loss management, smoking cessation, and diabetes management [5,6]. A foundational approach for many of these lifestyle interventions is motivational interviewing (MI), which is an evidence-based, patient-centered communication method aimed at evoking patients’ motivations for change [7]. MI practitioners adopt a collaborative and empathic attitude, empower patients to act as an equal partner in their disease management, and use specific and core patient-centered communication skills to guide them toward change.

Patients’ CVD risk profiles improved significantly in clinical trials in which the study staff incorporated an MI approach to conduct lifestyle interventions among participants with SMI [8-12]. To lessen the disparity of CVD risk experienced by this population on a wider scale, evidence-based programs to improve cardiovascular health need to be implemented in community mental health settings where patients routinely receive their care [13]. Real-world practitioners in these settings need the training not only to deliver the CVD health intervention itself (the requirements of which will likely differ based on education and prior experience) but also for the specific preparation necessary to incorporate an MI approach. Although some mental health practitioners may have had previous exposure to MI, the method involves a complex skill set such that attending a seminar or a traditional 2- to 3-day in-person workshop alone is usually insufficient to build competency [14-16]. A meta-analysis of MI training studies suggested that for learners to gain and sustain MI skills after a workshop training, an average of 3 to 4 coaching sessions with performance feedback from practice sessions over several months were needed [17]. Yet, when scaling up evidence-based practices (EBP) for implementation research, the stated amount of training to deliver patient-centered behavior change lifestyle interventions may be harder to accomplish among practitioners in community mental health settings who have demanding caseloads and limited time for additional training activities [18]. Furthermore, because of the diverse occupational backgrounds of the health care team in these real-world community settings (which often include physicians, nurses, therapists, and frontline day program staff) [19], it is essential to develop MI training that is both tailored and acceptable to a variety of practitioners.

There are a few promising formats to efficiently train busy mental health care practitioners in MI [20]. Synchronous or real-time trainings over videoconferencing can be used to provide learners with a basic introduction to MI. Live videoconferencing, although still requiring coordinating schedules of learners and trainers, can simulate an in-person event. Videoconference trainings may be shorter than standard multiday trainings and can be broken into smaller chunks of time; this may be more appealing to managers and allows learners time to meet their other work demands. This format also allows practitioners the opportunity to reflect on and integrate the concepts they are learning between training sessions.

Web-based self-study programs that include video and workbook activities can be entirely asynchronous or may integrate a live component in which learners meet with trainers during scheduled times [21-23]. Both real-time synchronous and web-based training options have the advantage of being able to reach a wider audience of practitioners, including those working in rural settings or where skilled trainers of MI are not available, than in-person events. Both approaches reduce the travel time and cost for trainers and limit the need for learners’ transportation, workshop space, and other expenses associated with in-person workshops. When implementing the intervention at multiple locations, videoconference training provides a practical way to reach several sites simultaneously. Constraints around social distancing imposed during the COVID-19 pandemic have created the need for effective options other than traditional in-person trainings for teaching different counseling approaches. Although these modalities have been used more frequently in the past few years, research on the impact of web-based [22] and video trainings in MI remains limited [24].

Another scalable remote training format for MI involves the use of computer simulations [20,25,26]. SIMmersion [27] offers video-interface trainings; videotaped role-play conversations with human actors give learners the ability to practice and receive coaching and feedback on their recorded responses [20,28]. Avatar-interface programs, by contrast, provide learners with a virtual reality training experience to learn and practice their skills with simulated avatar patients [29,30]. Avatar trainings may involve a human trainer [31] or consist of fully automated, preprogrammed simulations; some platforms also offer coaching and feedback features to learners. Kognito [32] offers several avatar simulations in which the users enter an interactive virtual environment and engage in role-play conversations with emotionally responsive virtual humans on a variety of topics in health care, mental health, and patient-provider communication. Their approach combines the science of learning, the art of conversation, and the power of gaming technology. Avatar-interface programs can be used as stand-alone or as follow-up MI practice opportunities; they have the advantage of allowing learners to practice their skills in a convenient and private environment.

Despite the many different approaches to training and the demonstrated effectiveness of MI [6], the amount and type of training necessary for real-world practitioners to develop and sustain proficiency in the method is uncertain [24]. In the context of 3 pilot implementation research projects that address weight loss, smoking cessation, and CVD risk reduction among persons with SMI, we aimed to develop scalable training in MI. This paper will provide a brief description of the 3 pilot projects; their corresponding setting and intervention; the design and development of the 2 MI training approaches, real-time brief workshops and follow-up asynchronous avatar training used to implement the EBP; and how we propose to evaluate the impact of the training on practitioners’ skills, attitudes, and beliefs about MI.

The Johns Hopkins ALACRITY Center for Health and Longevity in Mental Illness is a research-to-practice translation center funded by the National Institute of Mental Health that aims to reduce premature mortality from CVD in people with SMI [33,34]. There is an untapped opportunity to develop scalable and generalizable solutions that integrate physical health care into mental health service sectors, which have not historically provided such services. Through the ALACRITY Center, we are conducting 3 pilot implementation projects in community mental health settings to study strategies for the adoption and integration of evidence-based CVD risk reduction interventions related to (1) weight loss; (2) tobacco smoking cessation treatment; and (3) hypertension, dyslipidemia, and diabetes care for people with SMI. This study is a substudy within the ALACRITY Center methods core and focuses on the 2 types of innovative MI training: real-time brief workshop training over videoconferencing and asynchronous avatar training, which span the 3 projects described in the following paragraphs.

ACHIEVE-D (Adapting an Evidenced-based Weight Management Intervention and Testing Strategies to Increase Implementation in Community Mental Health Programs) is a 6-month, 2-arm, weight loss pilot cluster randomized controlled trial (RCT), which is designed to assess the degree to which the implementation strategy, that is, performance coaching, impacts psychiatric rehabilitation program (PRP) coaches’ knowledge, self-efficacy, and fidelity to the delivery of the adapted ACHIEVE-D weight loss program curriculum when compared with a standard control group. ACHIEVE-D is a behavioral weight loss program designed for persons with SMI that provides basic weight loss health education using an MI approach, which was adapted from an intervention tested in a previous RCT [10,12,35]. Coaches in PRPs will be provided with the materials, trained on all aspects of the ACHIEVE-D weight loss curriculum, and encouraged to use MI skills in their interactions with the participants.

IMPACT (Promoting Evidenced-Based Tobacco Smoking Cessation Treatment in Community Mental Health Clinics) is a pretest-posttest design–based 12-month observational study that aims to examine the effects of an implementation intervention on mental health providers’ delivery of four EBP related to smoking cessation treatment for clients with SMI: (1) assessment of smoking status, (2) assessment of willingness to quit, (3) behavioral counseling, and (4) pharmacotherapy prescribing. Psychiatrists and mental health counselors at 5 community mental health clinics will receive training to implement smoking cessation practices, incorporating an MI style in their interactions with the patients. Counselors and prescribers will be encouraged to use MI to understand their patients’ smoking behavior and engage them in smoking cessation treatment.

RHYTHM (A Quality Improvement Process to Support Delivery of Cardiovascular Care in Community Mental Health Organizations) is a pretest-posttest design–based 12-month pilot study examining the effects of an adapted 5-step Comprehensive Unit–Based Safety Program implemented during care coordination aimed at reducing CVD risk for patients participating in behavioral health home programs [36]. Behavioral health home staff will receive the training necessary to implement the program, which will include the use of MI as a way of interacting with patients as they discuss various aspects of their patients’ CVD care.

In these pilot implementation projects, we are assessing the use of MI among clinician and staff participants who have been trained to deliver the lifestyle interventions to mental health clients. Practitioners recruited from community mental health settings across Maryland will implement the 3 projects. Only English-speaking adults (aged ≥18 years) who provide informed consent are enrolled. Additional inclusion criteria include willingness to provide baseline data, attend training, implement the study intervention, and participate in follow-up data collection. Exclusion criteria across the 3 projects are minimal and exclude any staff who express an intention to leave the facility before project completion. ACHIEVE-D enrolled interested residential staff and employees at the PRP program as coaches. IMPACT participants include community mental health clinic staff (eg, licensed counselors, psychologists, and social workers) and prescribers (eg, nurse practitioners and psychiatrists). RHYTHM participants include PRP staff and health home nurses who are currently employed at the study sites.

Approval to conduct the studies and the associated MI trainings was granted by the Johns Hopkins Institutional Review Board (IRB) protocols. ACHIEVE-D study received IRB approval in March 2021 (IRB00247344). IMPACT study received IRB approval in January 2021 (IRB00231836), and RHYTHM study received IRB approval in December 2020 (IRB00269855).

MI Training for the ALACRITY Center studies consists of 2 distinct approaches: real-time brief workshop training over videoconferencing and asynchronous avatar training.

The treatment of CVD risk factors is a public health priority for persons with SMI who experience symptoms disproportionally owing to these chronic health conditions. In community mental health practices where patients regularly receive their care, providing physical health interventions using an MI approach has untapped potential [4]. However, practitioners in these facilities already have excessive demands on their time, a challenge exacerbated by staff shortages and turnover, and they may not have the time, energy, or organizational support to learn how to implement patient-centered CVD risk reduction interventions [18,19]. To overcome some of the barriers to wide-scale adoption of MI in these contexts, we have adapted the traditional MI training format and created brief videoconference workshops supplemented by postworkshop avatar-based practice simulations to prepare the practitioners to deliver 3 evidence-based health interventions in community mental health settings. We hope that these formats will be acceptable and lead to an uptake and improvement of practitioners’ patient-centered skills.

There are several strengths in this study. We created a tailored training that has shorter duration than traditional training as a way to expose busy practitioners to MI. Brief trainings (<8 hours) are more time efficient and have been shown to make some difference in physician consultation behavior [51]. We adapted the real-time training to be delivered remotely in shorter bouts of time. We aim to identify whether the brief trainings provided are adequate to prepare real-world clinicians to perform the EBP and subsequently assess their skill acquisition with a validated measure.

Although technology-based virtual reality training methods have been used in medical and other health professionals’ education, the use of avatars and other computer programs that provide interactive learning opportunities is not as common in behavioral health skills training [25,26,30,52-54]. Virtual platforms are a relatively new, potentially promising mode for training in MI and allow busy practitioners to learn the patient-centered counseling approach from expert trainers without having to travel to an off-site location. We incorporated 3 novel avatar conversation simulations tailored for practitioners to practice and receive feedback on their use of MI in simulated discussions about weight loss, smoking cessation, and cholesterol management. We will assess their acceptability to a wide range of practitioners at the end of each project.

We describe the study’s training content and development of the avatar simulations as well as 2 methods of measuring adherence to MI and assessing change in practice that results from the trainings provided. Research has shown that the immediate skill gains from a workshop training diminish within a few months and that postworkshop follow-up performance feedback with coaching is needed for practitioners to sustain or further improve MI skills [14,16,17,55]. Although our initial training is brief, the automated coaching and feedback offered during the avatar simulations should mirror the role of coaching as a way of providing follow-up training to improve learners’ skills. Simulation technology offers the capability to move beyond traditional in-person training and provide skill training through repeated practice and feedback [56].

One limitation of the study is that none of the training methods have a comparator condition. The pretest-posttest outcomes are used here to assess the acceptability, feasibility, and effectiveness of the 2 training methods for improving practitioners’ MI skills.

Avatar practice requires practitioners to set aside time during their busy workdays. The extra workload and responsibilities of staff during the COVID-19 pandemic may leave little time or motivation for avatar skills practice after their initial exposure to MI, thus potentially decreasing the impact of this component of the training.

Another limitation is that because of the nature of the avatar platform and how the 3 conversations simulations are built, there are limited choice options. Therefore, after a learner has used the platform a few times, they might find it redundant and be less willing to practice. The didactic module and the conversation simulations may be a great opportunity for novice learners to practice their skills; practitioners with more experience in patient-centered counseling may be less interested in using a platform that offers few multiple-choice conversation options.

To overcome the challenges of assessing MI skills in implementation settings, we are using multiple measures, namely, MITI and the automated feedback scores from the avatar simulations. The feedback feature of the avatar simulations was originally designed to assist learners in improving MI skills. The mapping approach we used to create the automated scoring system for each of the global scores may limit the precision of the measure to capture how well the clinician is using MI. In addition, because participants complete the same simulations for training and assessment purposes, scoring from the simulations may not be a valid outcome measure of their MI skills.

This study design was in the context of implementation research and does not allow us to examine the effect of the MI workshop training component alone. However, examining MI in real-world settings is an innovative aspect of the study. Although there are limitations to asking about experiences of the training over a 6- or 12-month period, our qualitative interviews at the end of the study will add detail to the quantitative results.

Although the development of our avatar conversation simulations included multiple iterations between our study experts and the developers of the platform, further advances in virtual training have become available since the study was first conceptualized. Other programs that allow for additional branching conversation scripts could be integrated into future versions of this training to better support virtual learning [57]. Virtual reality programs that create endless iterations of interactive experiences such as those offered by SIMmersion are another option to promote learning. Platforms that use natural language processing with actual recorded patient encounters also have the potential to offer feedback to learners on their MI skills and could be an addition to MI training for busy mental health practitioners [56,58].

This study will help us understand the effectiveness of 2 scalable training strategies on practitioners’ use of MI and how these approaches impact practitioners’ beliefs about, the importance of, and their confidence to use MI when delivering 3 interventions focused on CVD risk reduction to patients with SMI. The degree of tailoring to specific lifestyle change targets and the amount of real-time interaction needed to garner training effects in MI are not known; this research will contribute to knowledge about the effectiveness of brief videoconferences for training community mental health practitioners in MI. Testing MI skill uptake from videoconference training sessions has not been documented in the literature; this project will be a step forward in filling this gap. The study will also provide information about the perceived usefulness of the avatar as a modality for learning MI. On the basis of our results of this study, a logical next step to determine the most effective scalable training approach would be an RCT that compares different MI training modalities and includes a comparator condition.