We will first develop a semistructured qualitative interview script to assess the specific treatment needs and preferences of young adults with mTBI. We will practice and pilot the semistructured interview over a secure video platform with 2 to 3 patients. We will subsequently administer the refined semistructured qualitative interview to 20 individuals with mTBIs to gather feedback on the intervention components and gauge treatment needs, expectations, and barriers. Specifically, the interview domains will include the following: (1) case-based scenarios of perceived effects of the injury and anxiety on function; (2) areas of need for skills training; (3) impact of the injury on activity, sports, school, work, and relationships; (4) best strategies for recruitment and retention for patients who have sustained concussions; (5) difficult situations or challenges experienced by college-aged individuals recovering from concussions; (6) barriers to participation and adherence, including light sensitivity and comfort using screens, and strategies to address this (eg, night setting, education); and (7) specific topics they would like to learn about. Participants will be provided information about the proposed content of the adapted program (TOR-C; Table 1) to obtain their feedback on each intervention component. Skills will be modeled to ensure an accurate understanding.

Eligible patients who agree to participate and provide written informed consent will be assisted in downloading and installing the video software (Zoom). All qualitative interviews will be audio-recorded, transcribed, organized, and analyzed using the NVivo 12 qualitative statistical package (QSR International; see the Data Analyses section). We will use the qualitative data obtained in phase 1 to develop the TOR-C and treatment manual.

We will enroll and treat up to 5 patients and conduct live video qualitative exit interviews at the end of the treatment to gather detailed feedback on the intervention components, applicability of the study measures, general protocol issues, and perceptions about live video intervention delivery. Participants will complete baseline, postintervention, and 3-month postintervention follow-up questionnaires as well as weekly home practice. We will thus be able to explore the feasibility and acceptability of the TOR-C, study measures and procedures, adherence to sessions and home practice, and acceptability of the live video in this population. Each session of the TOR-C will last 45 minutes. Exit interviews will last 30 minutes and be audio-recorded and transcribed. The interviews will follow procedures similar to those described in phase 1 (eg, delivery via a semistructured interview script, qualitative analysis using NVivo 12) and be used to further refine the study manual and procedures as well as finalize the TOR-C before the pilot RCT. The analysis plan is described in the Data Analyses section.

After completing baseline questionnaires, eligible patients will be randomized via permuted blocks using a statistician-developed sequence to either the experimental (refined TOR-C) or control (HE-C) program in a 1:1 ratio. Participants will be blinded to intervention versus control groups. They will then be scheduled by a research assistant for their first video session. They will complete postintervention assessments after the 4-week intervention as well as at the 3-month follow-up. Both interventions will be delivered by the same mind-body clinician.

The treatment fidelity process for the RCT will follow the National Institutes of Health recommendations [40] and our previously successful clinical adherence protocol [41]. The clinicians will complete fidelity checklists after each session for both intervention and control groups and receive weekly supervision to ensure protocol adherence. To evaluate protocol fidelity for both the intervention and control, a random sample (10%) of the audio-recorded sessions will be coded by an independent coder.

The original Toolkit [34] includes 4 manualized 45-min-long weekly sessions that address mind-body skills, including eliciting the relaxation response (eg, body scan, deep breathing, mindfulness), cognitive-behavioral strategies (eg, reframing), acceptance and commitment skills (eg, acceptance), and skills for returning to activity (eg, goal setting, activity pacing). Table 1 presents specific adaptations that we propose to make to each module for patients with mTBI and anxiety. These proposed adaptations will inform the qualitative interviews (phase 1), and additional refinements to the intervention will be made based on the results of our qualitative assessments.

The HE-C will be adapted from the Health Enhancement Program [48], a manualized intervention that has been successfully used in prior live video studies [49,50]. It includes 4 modules that provide educational information on the relationship between anxiety and mTBI (session 1), return to activity (session 2), roles of nutrition and sleep (session 3), and health care self-management, including access to mental health services (session 4). This control condition will be dose-, attention-, and time-matched to the TOR-C in terms of session content and home practice. Specifically, participants in HE-C will be assigned daily readings [48] for the same amount of time as those in the intervention. Participants in the TOR-C will receive HE-C information as electronic handouts. Participants in both groups will have the option to opt in to receive text messages through EZ texting [51], a communication platform, which will remind them of dates and times for scheduled sessions to maximize attendance. Our team is using EZ texting successfully in projects with other medical populations [52,53].

This will be measured using the Post-Concussion Symptom Scale [54], a 22-item questionnaire measuring perceived presence and severity of concussion symptoms (eg, headache, nausea, balance problems, fatigue irritability, nervousness) on a 0-6 scale. Higher scores represent worse symptoms.

The GAD-7 [38] is a 7-item questionnaire measuring anxiety symptoms within the past 2 weeks on a scale of 0 to 3 and will be used to ensure participants meet the criteria for anxiety symptoms. The Hospital Anxiety and Depression Scale [55] is a 14-item questionnaire assessing anxiety and depression in the previous week on a scale of 0 to 3 and will be used to assess changes in anxiety symptoms following the intervention and at follow-up. Higher scores represent higher anxiety.

This will be measured using the World Health Organization Disability Assessment Schedule (WHODAS) 2.0 for mTBI [56,57], a 12-item questionnaire assessing functional difficulties in various life domains (eg, participation barriers, physical activity limitations, and self-care limitations) on a scale of 0 to 4. Higher scores represent lower physical function.

This will be measured via the Fear Avoidance Behavior after Traumatic Brain Injury Questionnaire [58], a 16-item questionnaire assessing beliefs about how work and physical activities affect mTBI symptoms, and whether they should be avoided, on a scale of 0 to 3. Higher scores represent higher fear avoidance.

This will be measured using the Pain Catastrophizing Scale [59], a 13-item questionnaire assessing one’s tendency to focus on pain-related thoughts and feel helpless and hopeless due to pain on a scale of 0 to 4. Higher scores indicate higher pain catastrophizing.

This will be measured via the Cognitive and Affective Mindfulness Scale-Revised, a 12-item questionnaire assessing one’s ability to pay attention to the present moment in a nonjudgmental manner [60], on a scale of 1 to 4. Higher scores represent higher self-reported mindfulness.

This will be measured using the limiting behavior (7 items) subscale of the Behavioral Response to Illness Questionnaire [61], assessing the frequency in which participants are inactive as well as the all or nothing behavior (6 items) subscale, which captures one’s tendency to overexert themselves. Both subscales are rated on a scale of 0 to 4, with higher scores indicating more limiting and all or nothing behavior.

This will be assessed using the Client Satisfaction Questionnaire [62], a 3-item questionnaire assessing the degree to which the program met the participants’ needs and their satisfaction from it, on a scale of 1 to 4. Higher scores represent higher satisfaction.

This will be assessed using the Credibility and Expectancy Questionnaire [63], a 6-item questionnaire that assesses how believable, convincing, and logical patients perceive the treatment to be and the degree to which they expect to improve. Some items are scored on a scale of 1 to 9, and others are scored on an 11-point 0%-100% scale. Higher scores represent higher credibility and expectancy.

We will additionally collect baseline data about other factors that may influence recovery and mTBI symptom persistence, such as smoking, alcohol consumption, time since injury, and posttraumatic headache. To minimize patient burden, these data will be assessed using 1 to 2 self-report questions (eg, “how many drinks do you consume a week on average?”) rather than full questionnaires. Additional information on medical history will be collected directly from patients’ electronic medical records. All these data, however, will not be included in our main analyses because of the current emphasis on feasibility. Rather, they will help us better characterize the sample and provide valuable information toward a future efficacy trial using the NCCIH UG3/UH3 mechanism. It is anticipated that patients will need 20 to 30 minutes to complete the battery of questionnaires. They will also receive a prompt at the end of each questionnaire, asking them to complete unanswered fields, in effort to reduce the possibility of missing data. A blinded research assistant who is not involved in recruitment procedures will check each questionnaire upon completion to ensure that patients do not randomly respond or acquiesce. To maintain confidentiality, patients will be identified using a specific ID. Only the unblinded research assistant and study clinicians will have access to the file connecting patients’ names and study IDs.

The individual qualitative interviews (phase 1) will be recorded and transcribed. They will then be analyzed using thematic content analyses following Miles and Huberman [64] in NVivo 12. We will use the framework method [65] to analyze our qualitative data and employ a primarily deductive approach [66], while also allowing flexibility to incorporate inductive themes derived directly from novel information collected during the interviews. We will assess the reliability (κ) of coding for themes and patterns in qualitative responses by 2 independent coders. The analyses will be conducted by trained clinicians and research assistants under the guidance of the research team’s senior psychologist. Discrepancies will be resolved through discussion until adequate reliability is obtained (κ>0.80).

The primary analyses in this project will focus on establishing key quantitative feasibility benchmarks to inform future multisite RCTs. To this end, we will use frequency and proportions to assess the feasibility of recruitment and retention procedures within each group. We will additionally use proportions of patients with scores over the midpoint on the Client Satisfaction Questionnaire [62] and the Credibility and Expectancy Questionnaire to assess satisfaction and credibility, respectively. Feasibility benchmarks for these phases are detailed in Table 2.

Participants who drop out will be counted as not meeting applicable feasibility criteria. Benchmarks that need to be met before beginning the future fully powered RCT will include (1) >70% of participants successfully contacted agree to participate; (2) 70% of participants with a Credibility and Expectancy score and Client Satisfaction score over each scale’s midpoint; (3) >70% of participants participated in at least three of four sessions; (4) >70% therapist adherence (measured via checklists and independent rater agreement on audio recordings); (5) >70% of participants practice at least one skill on 3 days per week (6) >70% of participants have no questionnaires missing; and (7) minimal adverse events. If these criteria are not met, revisions will be necessary. Benchmarks will be reported separately for the TOR-C and HE-C. These benchmarks were used previously in federally funded studies by our research team [34,52,53] and are consistent with guidelines for intervention development [35,67].

For quantitative measures, we will use descriptive statistics to characterize the sample and paired sample two-tailed t tests to assess within-group changes between baseline and postintervention and between postintervention and the 3-month follow-up. Cohen d, including 95% CI, will be used to determine effect sizes using conventional standards (small effect sizes of 0.2 SD units, medium effect sizes of 0.5 SD units, and large effect sizes of 0.8 SD units) [68]. In line with recommendations for analyses in pilot studies [69,70], we will refrain from conducting between-group analyses of efficacy.

Analyses of the semistructured qualitative exit interviews in phase 2 will broadly follow the methods described in phase 1 above and will gather information on (1) perception of the skills taught in the program, (2) barriers and facilitators to completion of program and home practice, (3) perception of program structure and experience in the sessions, and (4) perception of the assessments.

Power analysis is not appropriate for qualitative analyses. For the individual interviews, a sample size of n=20 is typically enough to achieve saturation of themes [71]. With a sample size of N=50 in phase 3 and conservatively assuming that the feasibility criteria in Table 2 are independent, the study will have 80% power to confirm feasibility of all criteria if the expected rate of each criterion is at least 83%. Of note, we achieved greater retention in some of our other studies, including the TOR study [34] as well as previous mind-body studies using live video [41]. Importantly, as mentioned above, this power analysis is not calculated to detect group differences or significant changes in outcome measures but rather to establish feasibility.