All study procedures will take place at or through the VA Connecticut Healthcare System (VACHS) in West Haven, Connecticut. All telehealth delivery will be administered using VACHS-approved platforms.

Participants will be recruited from VACHS through the Polytrauma Support Clinic Team, Speech Pathology Service, and flyers. The study team will access Polytrauma/TBI and Speech Pathology consult lists, as well as patient lists generated for biweekly Polytrauma/TBI team meetings, and will conduct chart reviews of patients on these lists. The study team will ask the aforementioned recruitment sources to describe the study to potentially eligible participants identified by the study team or the recruitment sources and to ask whether they would be interested in learning more about the study. If so, their names and contact information will be forwarded to the study team.

Following informed consent and screening, phase 1 participants will undergo a preliminary version of the TALE treatment. No baseline, posttreatment, or follow-up assessments will be administered during this initial treatment and manual development phase, as the primary focus at this stage is to refine the intervention in preparation for the subsequent treatment feasibility trial. Administration of the treatment will occur serially to allow for iterative revision and refinement. Thus, in the phase 1 treatment development trial, feedback from each participant regarding their experience with TALE will be used to revise the treatment for the next participant.

Individuals who express interest in the study will be prescreened by telephone to confirm that they meet major eligibility criteria. Those deemed eligible will be invited for an in-person initial screening.

In addition to reviewing the medical record, screening measures will be used to verify eligibility for participation during the in-person screening (Table 1). A psychosocial intake will be administered to collect demographic and psychosocial data. Screening measures will include the Ohio State University Traumatic Brain Injury Identification [37], the 4-item Mayo-Portland Adaptability Inventory (MPAI-4) [61], the LCFS [60], and the Bedside WAB-R (cutoff score: ≥ 93 for inclusion) [40]. The in-person screening is expected to take approximately 1 hour.

Baseline and posttreatment assessments will take approximately 2.5 hours for each assessment period. The first posttreatment assessment will occur after treatment ends, approximately 3 months after the first treatment session. The follow-up assessment will occur about 1 month thereafter. The posttreatment and follow-up assessments will be the same as those administered at baseline. As the MPAI-4 is already being administered at screening, it will be repeated only after treatment and at the 1-month follow-up. Given that baseline testing will occur within about 1 month of screening and that all participants will be beyond the acute stage of TBI, when most spontaneous recovery is expected to occur, MPAI-4 performance is expected to be stable during the period between screening and baseline. This approach reduces participant burden associated with readministration of the same assessment within a short interval. Informants will complete the TBI Quality of Life Communication Item Bank (TBI-QOL COM) short form [54] to provide collateral information about functional communication. Participants will be offered breaks and, if requested, testing will be split into multiple sessions. To accommodate participants and allow for testing flexibility, self-report surveys and several assessments may also be administered by phone or virtually to shorten the in-person testing session. The modality of assessments will be tracked.

In both phases 1 and 2, participants will receive treatment from a trained study therapist. Individual, hour-long sessions will occur 2 times a week over approximately 2-3 months. In the phase 1 treatment development trial, there will be about 16-20 sessions in total. Feedback obtained in phase 1 will be used to reduce the number of sessions to about 16. Thus, following the iterative revision process, we anticipate that the number of sessions will range from 14 to 18 for the phase 2 feasibility trial.

Participants will be offered the choice of telehealth or in-person sessions and will be encouraged to adhere to their chosen treatment delivery format for the duration of treatment. Some flexibility will be allowed to accommodate participants who wish or need to change formats during treatment. Although the treatment structure aims to deliver 2 sessions each week, some flexibility will be allowed to accommodate issues such as missed appointments, holidays, leave, and “catch-up” appointments, which may result in 3-4 sessions in a week. Some participants may not be able to complete “catch-up” sessions. As such, to allow flexibility, study participation may extend up to 6 months.

Treatment questionnaires will be administered to all phase 1 participants and to phase 2 participants randomized to the TALE intervention arm at the end of treatment to ascertain information regarding training satisfaction and training content. Assessment questionnaires will be administered to all phase 2 participants at the posttreatment and follow-up assessments. The questionnaires will take approximately 15 minutes. A 30- to 45-minute semistructured exit interview will also be administered posttreatment to obtain qualitative feedback regarding the treatment. The questionnaires and exit interview are further described in the “Primary Outcome Measures: Treatment Feasibility” section.

The individuals (n=5-10) recruited in phase 1 will not be randomized, as phase 1 focuses on the initial treatment and manual development phase. For phase 2, individuals will be randomized to either the discourse treatment condition, TALE, or a TAU condition using permuted block randomization (Figure 2). Randomization will be stratified by TBI severity (mild vs nonmild) and education (high school vs college degree). Grading of injury severity will follow VA/Department of Defense TBI severity guidelines [62]. The study statistician (RF) will generate the sequentially numbered randomization scheme. The PI will assign participants according to the randomization scheme. The randomization sequence will be concealed from the study assessors and study therapist, except for the PI, who will also conduct assessments. Personnel who enroll and assign participants will be separate staff. Personnel who enroll participants will not have access to the randomization scheme.

This is an unblinded feasibility study. Participants and outcome assessors will not be blinded to intervention assignments. Blinding was considered but is not feasible for practical reasons, namely, personnel constraints. The research team does not currently have sufficient staff to carry out the study while maintaining blinding. Given that the primary outcomes focus on feasibility rather than treatment effects, the risk of bias due to lack of blinding was deemed to be low. Blinding will be implemented in future fully powered trials of the intervention, in which hypothesis testing (eg, examining treatment efficacy) will be possible.

TALE aims to improve narrative discourse ability by targeting story organization and story content in the context of social communication. Story organization will be addressed through story grammar training by teaching specific episode elements: setting, initiating event, attempt, and direct consequence. Lay language will be used to convey these elements (ie, setting, trigger, action, and result), and the acronym “STAR” will be implemented as a central concept and mnemonic device to support learning of these elements. Story content will be addressed by targeting aspects foundational to story completeness and global coherence. Thus, treatment of story content will encompass lessons on main ideas and details, intentions, and word retrieval.

The manual will feature individual session content along with clinical examples, scripts, and vignettes, as well as handouts and homework assignments. Treatment materials will include publicly available photos, videos, and other media to elicit narratives. Microsoft PowerPoint will be used to deliver lessons, allowing for the presentation of mixed media.

All handouts and homework assignments for each week will be provided in paper form, electronically, or both, depending on the participant’s preference. Although participants will be given about 1 week to complete homework assignments, homework will be assigned at each session to address aspects not yet completed or to provide continued practice for completed components. Homework will include learning and practicing strategies and activities to promote functional carryover of skills learned in session to interactions with the informant and other communication partners in the participant’s daily environments.

The initial discourse training stimuli will be developed by the first author (KL) and distributed to the second (CC) and last authors (JMF) for feedback on the following training elements: organization, content, appropriateness of training techniques, and appropriateness of training stimuli. The treatment development process will be iterative and will occur on a monthly basis during the initial development phase. The expected outcome of this phase will be a preliminary version of the treatment manual, specifying session length, structure, the specific training exercises to be used in each session, and how training should progress through the activities.

A panel of 4 experts and stakeholders has been recruited to provide an external review of the treatment protocol and materials during treatment development. We will also recruit a Veteran with TBI, who will not be a participant in the study, to serve as a key stakeholder reviewer.

A preliminary version of the treatment will be administered serially to 5-10 participants (phase 1: treatment development). Feedback from these participants will be obtained at the end of treatment using the measures (training satisfaction and training content questionnaires) described in the “Primary Outcome Measures: Treatment Feasibility” section. Brief, informal feedback measures will also be obtained after each session.

Using information obtained during evaluation of the preliminary version of the discourse treatment protocol, we will continue to revise training content and refine manual development. Treatment refinement, based on information obtained from the serial administrations, will continue until the PI and study team concur that information from administered sessions, including but not limited to participant feedback, has been appropriately incorporated into the manual and no further modifications are warranted. This final stage of treatment development will culminate in a near-final version of the treatment manual and training materials in preparation for use in the subsequent treatment feasibility trial (phase 2).

The proposed discourse treatment approach uses systematic instruction as a foundation to guide learning [63]. Systematic instruction is an evidence-based approach to optimize learning in individuals with learning impairments through structured training that incorporates (1) explicit models, (2) minimization of errors during the initial learning phase, (3) strategies to promote learner engagement, and (4) carefully guided practice to promote mastery, maintenance, and generalization. Treatment techniques that have the potential to improve discourse ability will be implemented, including feedback; use of simulated and actual social contexts; functional practice of learned skills; development of metacognitive and metalinguistic strategies; and hierarchical training [21].

General therapeutic techniques will be used throughout the phases of treatment, including brief and simple instructions, visual cues/reminders, and “teach-back” repetitions/summaries of information by the participant to check understanding. Scaffolding and feedback will be provided to correct mistakes and guide participants toward target productions. Training will be incremental and will build upon each earlier section of treatment. The discourse treatment will also employ general learning principles, including errorless learning and spaced retrieval. The treatment will build upon prior studies by integrating the aforementioned techniques and components into a systematic instruction approach to maximize the potential for change.

There will be 3 treatment sections (discussed later). Treatment components specific to each section are described in the following sections.

The SBF will be used to guide remediation of both story content and story organization abilities, as it provides an organizational framework for mapping story information. Participants will be taught to lay the foundation for a story through instruction in the story grammar elements that form episodes. They will then learn how to build upon those elements as they produce the story episode (eg, mapping relevant information), which will involve judgments of the relevance and salience of story content. When the story topic changes, participants will be taught to recognize the transition and shift to form new episodes to support the novel story content. Throughout the treatment process, participants will actively develop and hone the ability to enhance or suppress story information depending on its relevance and salience (see Textboxes 1-3).

Veterans randomized to TAU will continue to receive standard rehabilitative care if they choose to engage in these services. This may include other cognitive-communication treatment, speech-language therapy, psychosocial treatment, physical therapy, occupational therapy, pain management, and pharmacological interventions. No TAU services specifically provide discourse treatment.

This section describes the 5 screening measures mentioned in the study procedures. The intake will collect psychosocial and demographic data. Major intake sections include housing, employment, current medications, rehabilitation history, MH and neurologic history, and substance use. The Ohio State University Traumatic Brain Injury Identification is a standardized screening tool for assessing the lifetime history of TBI [37]. The MPAI-4 is frequently used in the VA to assess impairment and functioning following TBI and will be used as a key screening tool for the proposed study [61]. The MPAI-4 items related to verbal communication, alcohol use, and drug use will be used to ensure participants meet the inclusion criteria on those items. The LCFS will provide a measure of the severity of current functional cognitive impairment in carrying out activities of daily living, as well as an index of orientation and ability to follow directions [60]. The Bedside WAB-R will be used to rule out the presence of aphasia and auditory processing problems (cutoff score: ≥93 for inclusion) [40]. To pass the Bedside WAB-R, individuals must pass the 3 subsections that assess auditory-based skills. See Table 1 for a summary of screening measures and corresponding psychometrics.

As this is a treatment development and feasibility study, the primary hypotheses concern the tolerability and acceptability of the treatment. To this end, we will calculate rates of treatment attendance and study retention. Analysis of tolerability will be performed by calculating the number of participants who (1) attend treatment sessions, (2) withdraw from treatment, (3) complete each of the assessment batteries, and (4) withdraw from assessment. The number of training and assessment sessions attended will also be calculated. Treatment dropouts, who only participate in pretraining assessments, will be compared with those who engage in the intervention on demographic, neurocognitive, and psychosocial variables. Treatment acceptability will be assessed by averaging participant satisfaction ratings for the treatment and assessment methods and participant feedback ratings regarding training content. Preliminary recruitment and participant flow data will be collected by calculating the number of participants who (1) call to inquire about the study, (2) qualify based on the brief telephone screening and sign informed consent, (3) qualify following intake screening, and (4) are randomized. Exit interviews will be analyzed qualitatively using a thematic analysis approach to provide an understanding of the impact of communication difficulties on the individual and participant feedback on various aspects of the treatment.

Descriptive statistics, including means and SDs for quantitative variables (including discourse, cognitive, MH, functional communication, and daily functioning variables in the assessment battery) and frequencies and percentages for qualitative variables, will be calculated for participant demographic variables, including age, education, and race, as well as ratings related to participant characteristics. We will also examine the data for outliers.

Although hypothesis testing is not the primary focus because this is a treatment feasibility study, we will perform several statistical analyses to provide preliminary estimates of effect sizes, score ranges, and variability (eg, means, SDs, measures of dispersion). For measures in the assessment battery, scatterplots will be constructed to allow visualization of performance patterns at each assessment time point. Measures of dispersion and distribution will also be obtained. Correlations will be calculated between the 2 discourse measures (story content and story organization) cross-sectionally, as well as between the discourse measures and other measures in the assessment battery. Associations between demographic and outcome variables (eg, discourse, cognition, MH, and functional measures) will be assessed via correlation and cross-tabulations. To compare treatment groups and estimate effect sizes, linear mixed models analyses will be performed on changes in outcomes from T1 (baseline) to T2 (after treatment) and from T2 to T3 (1-month follow-up). Linear mixed models are robust statistical analyses that account for missing data. In summary, these analyses will provide information regarding the feasibility of the discourse treatment, as well as inform our conceptual model of the relationships among the outcome variables.

This project received initial approval from the Human Subjects Subcommittee (HSS), the institutional review board of VACHS, on July 2, 2021 (IRBNet ID 1628797; HSS # KL0001). All protocol amendments will be submitted to the VACHS HSS for approval before implementation. Where applicable, revisions will be made to consent forms, and participants will be reconsented as necessary.

Informed consent will be obtained from all participants by study personnel in a private office. Potential participants will read the consent document (or have it read to them) and will be asked whether they have any questions about the study or the study procedures. Those who wish to be recontacted for a future study may check a box on the consent form indicating this.

Study participants will be paid as follows:

Privacy will be protected by conducting all study procedures in private offices. The PI will supervise and monitor all assessments and interventions. Identifying information will be protected by storing patient identifiers separately from other study data, coding patient data, and storing all information in locked offices or on a secure, access-restricted server.

Data will be collected during assessments and the study intervention. Data with no protected health information or personally identifiable information will subsequently be entered into electronic VA REDCap (Research Electronic Data Capture; Vanderbilt University), with access limited to designated study personnel.

Data will be analyzed by group, and identifying information will be removed from all forms and will not be directly linked to the data, as participants will be identified only by an assigned research number to protect their confidentiality. Compliance with the Health Insurance Portability and Accountability Act and confidentiality guidelines will be maintained, and only qualified research staff will have access to the database or coding key.

Research data will be maintained in accordance with current VA guidelines. Deidentified data will be used by research personnel to address study hypotheses. Any publications resulting from the study will use only deidentified data. No protected health information or individually identifiable health information will be disclosed to a non-VA entity.

This is a minimal risk study. The risks associated with this study are related to privacy, confidentiality, and psychological factors. Identifying information and sensitive medical information (ie, psychiatric history, substance use history) will be collected in this study. There is a potential risk to privacy and confidentiality if someone else becomes aware of the Veteran’s study participation; this will be communicated to participants in the consent form. Participants may experience psychological changes as they become more aware of their communication difficulties. Participants may become tired or frustrated while participating in the intervention or assessment battery. Participants will be offered breaks as needed and will always have the option of withdrawing from study participation.

In the unlikely event that a participant experiences a significant increase in psychiatric symptomatology (eg, exacerbation of PTSD symptoms), the consent form will specify that the participant’s MH provider will be contacted and that the participant may be hospitalized, even if this is not the participant’s wish.

Attribution of risk categories will be defined by the PI, with input from the coauthors, and will include grading by severity and relation to the study (definite, probable, possible, unlikely, or unrelated).

A data monitoring committee was not indicated for this study, as this is a minimal-risk, nonpharmacological study. The PI will oversee all study procedures and monitor the trial at least weekly during laboratory meetings.

Final data sets underlying all publications resulting from the research may be shared approximately 6 months after publication. Deidentified data sets may be shared with other investigators, upon written request, under a data use agreement prohibiting the recipient from identifying or reidentifying any individual whose data are included in the data set. The data sharing agreement will also restrict redistribution to third parties and stipulate proper acknowledgment of the data source. These data sets will be shared in accordance with VA guidelines.