In the first published RCT of a social communication intervention for persons with ABI, Helffenstein and Wechsler [57] compared the impact of an interpersonal process recall (IPR) treatment to a nontherapeutic attention control. IPR is a training method that was originally developed for education of counselors and focused on insights into the counselor–client and teacher–learner relationship [58–60]. This approach utilizes videotape playback of various situations to stimulate recall of the various dynamics that are involved in interpersonal communication. In a typical format, the client–counselor interaction is videotaped, and then a second counselor reviews the videotape with the client to discuss client-recalled feelings or to elaborate on the meanings of various aspects of the communication [60]. The proposed advantages of the IPR method for addressing social communication abilities for persons with TBI include: immediate and direct feedback, which is particularly important when memory difficulties are present; the opportunity for the person with TBI to provide self-feedback and receive the perspectives of others; flexibility of the approach to cover various content and communication goals; opportunities for generation of alternative behaviors, modeling, rehearsal, and role-play; individualized attention; and the ability to utilize videotape of conversational exchanges in naturalistic environments as well as clinic-based conversation samples. In addition, this individual approach can be utilized in a variety of treatment settings (e.g., rural communities, general rehabilitation service settings, etc.) where a group treatment format may be less feasible.

The IPR therapeutic approach also allows for the individualized targeting of specific social communication goals, including the flexibility to focus on all three aspects of social communication: receiving, processing, and sending skills. For example, social cue perception, can be addressed within the IPR framework through a focus on observing the listener during videotape playback, followed by overt monitoring, modeling, and role-play within the treatment session. Processing skills, such as difficulty in generating communication alternatives, can be enhanced through the mutual recall and generation of alternatives within the therapy setting, which is a key aspect of the IPR approach. Finally, the IPR approach provides great opportunities for addressing expressive social communication skills, with immediate feedback available within the treatment structure and multiple opportunities for modeling and role-play.

In Helffenstein and Wechsler’s study, 16 individuals with “nonprogressive” brain injury were randomized to receive either 20 h of IPR treatment or 20 h of nontherapeutic attention [57]. Treatment for the IPR group consisted of 20 sessions involving: (1) participation in a videotaped interaction, (2) structured review of the taped interaction with feedback provided by self, conversational partner, and therapist, (3) development of alternative skills, (4) modeling, and (5) rehearsal. At post-treatment assessment, those receiving IPR treatment reported significantly reduced anxiety and improved self-concept. More importantly, the IPR group participants were rated to have significantly greater improvement in specific interpersonal skills by both treating professional staff members and independent observer raters who had been masked to treatment condition. Additionally, communication improvements were maintained at a 1-month follow-up period for a small subset of the study sample for whom data were available. Strengths of this early study included its randomized controlled design, use of independent outcome ratings, reliance on multiple measures to assess effectiveness of the intervention, and multiple methods to assess generalization of skills to outside-of-treatment settings. However, characterization of the small study sample was limited as to definition of injury etiology and severity, a clear description of the sample selection (e.g., consecutive series, convenience sample), and details of treatment procedures for each condition. Despite these limitations, results of this study were encouraging, particularly given the positive results for interpersonal skill improvements noted for the treated group by masked raters in nontreatment settings.

While this is the only published study exploring the use of IPR in addressing social communication after ABI, it is noted that this type of approach, or variations of this approach, are widely utilized in clinical settings. An ongoing clinical trial designed to replicate the Helffenstein & Wechsler [57] study with a larger, more clearly defined sample using a manualized approach to treatment is currently underway and may provide additional information about the utility of the IPR approach for persons with TBI [61].

There are a few case studies presented in the literature with fairly strong methodological rigor which show the impact of individually delivered interventions for specific social communication abilities in patients with more chronic TBI (all participants were greater than 18 months post-injury). In an early study, either feedback or self-monitoring procedures were provided to two individuals in a group setting using other-administered or self-administered colored light cues (“red” for negative and “green” for positive) in response to communication behaviors in a multiple baseline across treatment design, with nontreatment baseline sessions conducted prior to beginning the intervention [62]. Conversational behaviors, as rated by independent observers, improved to within the range of a comparison group of noninjured individuals for both patients during implementation of both feedback and self-monitoring conditions showing an impact from the intervention; however, there was a failure to generalize to nontreatment conditions. This may have been due to the limited number of treatment sessions provided. Unfortunately, there was lack of further follow-up to assess maintenance of treatment effects after the second set of treatment sessions.

Self-monitoring was used to learn conversational skills with female peers for two adult men with severe TBI [63]. The men were trained to count the numbers of specific target behaviors (compliments, asking others, and self-disclosure) that they performed when interacting with female peers; however, no specific instructions to increase or decrease these behaviors was provided. The study was conducted in a multiple baseline across behaviors design with each conversational behavior addressed in a different stage to training. Both participants in the program showed an increase in the number of compliments and “asking other” communication behaviors that fell within the range of communication behaviors exhibited by a noninjured comparison group. Decreases in self-disclosure were also noted for these participants, however there was greater variability in performance and self-disclosing statements were still observed with greater frequency than performed for a social comparison group. Results were maintained over a 1-month follow-up period for the participant that had such data available.

Giles and colleagues showed the impact of a focused feedback intervention where the goals of treatment were to reduce verbosity and circumstantial speech in a 27-year-old man [64]. The intervention consisted of verbal instruction regarding the rationale for behavior change with an emphasis on the phrase “short answers” to cue concise responses. Half-hour sessions were provided 5 days per week for a 1-month period, with cues for “short answers” and “permission to think before responding” given at least twice per treatment session. During each session, the following tasks were practiced: (1) responses to questions for which the patient was to respond with 1-word answers, (2) responses to questions with specific content that would require brief answers, and (3) unstructured conversation. Verbal praise was uses as an immediate social reinforcement for successfully completed tasks, with “time out on the spot” used for failure to complete tasks successfully. Significant main effects were revealed using ANOVA for single-subject design for question type and time period (baseline, treatment, post-treatment), but no significant interactions were observed. Use of one-word responses to structured questions showed the greatest improvement and was significantly better than attempts to provide brief responses to semi-structured questions. Interestingly, performance at follow-up was significantly better than at baseline assessment, suggesting some maintenance of gains.

In a case series of four individuals with chronic TBI, Brotherton and colleagues conducted a skills training program that was individualized to target communication behaviors identified for each participant during baseline assessment, with utilization of a multiple baseline across behaviors methodology [65]. The skills training program was conducted in 1-h sessions provided twice weekly and contained the following components: role-play, increasing understanding of the rationale for changing the target behaviors, modeling the correct behaviors, behavioral rehearsal, videotape feedback on performance, and social reinforcement of correct behaviors. Results showed that two of the four participants demonstrated clear improvements and maintenance of improvements over a 1-year follow-up for motoric communication behaviors (e.g., posture, self-manipulation), and some improvements during training for verbal behaviors, although maintenance of such improvements was limited. The other two participants had variable findings, with no evidence of improvements in performance on the target behaviors at 1-year follow-up.

Group interventions for treating social communication impairments following TBI are a common component of many post-acute rehabilitation programs [66]. These group interventions typically present a set of important social skills (e.g., greeting another person), and train clients to perform these skills correctly, and are likely best suited for therapy clients with similar rehabilitation goals. The inherent advantages of group interventions are that there are enhanced opportunities for feedback and observation of others’ communication styles [50].

In a more recent RCT study of a group intervention for social communication functioning following TBI, Dahlberg and colleagues randomized 52 adults, ranging in age from 18 to 65 who were at least 1 year post-injury, to either a group social skills intervention or to a waitlist control [50]. The intervention included twelve 90-min group sessions offered weekly during which a structured curriculum was followed, with group size limited to eight participants per group. The intervention was designed to utilize co-group leaders from different clinical backgrounds to facilitate collaboration and varying perspectives; to emphasize self-awareness and self-assessment to enhance goal-setting; to use group process to support interactions, feedback, problem-solving and social support; and to focus on generalization of skills through involvement of a friend or family member. Initial sessions addressed self-assessment and goal-setting; intermediate sessions targeted instruction of strategies for communication goals, feedback, and practice of skills; and the latter sessions emphasized generalization of skills and problem-solving. Results of this study revealed a significantly more improved performance at 12-week follow-up evaluation on seven of ten independently rated communication rating scales for those in the treatment group as compared to the no-treatment group. In addition, self-ratings of social communication skill were significantly improved as compared to controls at this time point. However, ratings of these abilities by significant others did not show significant group differences. In addition, group differences on secondary outcomes (self- or other-report measures of social integration, productivity, satisfaction with life) were not statistically significant. Inspection of change over time for the combined treatment and deferred treatment groups revealed significant improvements above baseline performance on nine of ten independently rated communication subscales, self-reported social communication skills, and on self-, other-, and group leader-rated individualized goals as measured by goal attainment scaling (GAS) methodology. At 6-month follow-up evaluation, maintenance of gains were demonstrated with six of ten rating scales, and self-reported social communication abilities, individualized goal performance, and satisfaction with life were significantly better than baseline performance. Strengths of this study included randomized design, multiple social communication outcome measures, independence of both communication partners and communication raters for key outcome measures, analyses using both per-protocol data and intent-to-treat models, careful delineation of the participant flow and study design, and use of multiple follow-up time points. Limitations of this work included somewhat vague criteria for study eligibility regarding level of social communication impairment required for inclusion, missing data for the primary outcome measure for a large percentage of cases (25 % missing for repeated measures analyses), and lack of a nontherapeutic attention control, which makes it unclear to what extent the treatment impacted function as compared to nonspecific treatment effects. However, given the strength of the initial findings, this approach is described as a model within the American Congress of Rehabilitation Medicine Brain Injury-Interdisciplinary Special Interest Group’s (ACRM BI-ISIG) manual presenting evidenced-based cognitive rehabilitation interventions [67].

McDonald and colleagues published findings of a RCT for social communication intervention involving 39 persons with severe, chronic ABI [51]. These participants were randomized to receive a social skills training program, a social activity program (nontherapeutic attention control), or to a waitlist control condition. The social skills training program included once weekly sessions for 12 weeks, which included 2 h of manualized group component to train social behaviors (such as greetings, introducing self and other, topic selection, etc.) and 1 h of training on social perception (including emotional perception and understanding of social inferences). In addition to the 3 h of group therapies, participants each received 1 h of individualized therapy with a clinical psychologist to address personally identified issues such as self-esteem, anxiety, or depression. The social activity program included 12 weekly 4-h sessions focused on group social activities, such as cooking, crafts, and board games. Results revealed that social activity alone did not lead to improved performance on any outcome variables as compared to waitlist controls. Participants in the skills training group were noted to have improved functioning on partner-directed behaviors, specifically with relation to self-centered behavior and partner-involvement behavior as compared to the other groups. However, no significant treatment effects were observed for social perception or emotional adjustment outcomes, nor were there effects observed for self- or other-report measures of social functioning. The authors concluded that social skills interventions produced circumscribed improvements, particularly with direct measure of social behavior. The lack of significant improvement in social perception and emotional adjustment were not entirely unexpected and may be due to the fact that participants selected for study participation were those with social communication behavioral deficits, and may or may not have had significant difficulties with social perception or emotional adjustment. Since baseline performance for some would fall within the normal range, the ability to demonstrate improvements would have been restricted.

Future studies would benefit from employing selection of participants with specific skill difficulties (i.e., social perception impairments, poor emotional adjustment) to assess more directly the ability to improve as a result of treatments designed to impact these abilities. Despite having a fairly rigorous design, several limitations are noted for this study. Attrition rates for the study led to reduced power, which was further impacted as effect sizes were smaller than anticipated. Another issue was that authors utilized reassignment to treatment condition as some subjects were unable to attend for scheduling reasons the group to which they were initially randomly assigned. Given the small numbers, the ability to use a more conservative intent-to-treat methodology would have eliminated any potential treatment effects, so this was not done.

In addition to a more rigorous randomized clinical trial methodology, several cohort studies examining the impact of social communication interventions have been published. Johnson and Newton conducted a prospective study of a group of ten individuals that participated in a group that met for 90 min each week over a 1-year period [68]. Group sessions were divided into two parts: the first half involved the entire group meeting as a whole to consider a specific issue and the latter half would consist of smaller breakout groups to allow for more detailed individual work. Sessions consisted of a review of the previous meeting, introduction of a specific topic, discussion of the main issues, practice on specific issues, role-play, and feedback from peers and therapists. Finally, generalization was encouraged by developing social opportunities that would allow for group members to work on the selected skills in real-world social settings. Following treatment, there were no significant group changes on measures of social adjustment, social performance, social anxiety, or self-esteem. However, categorical analysis revealed that while only one participant performed within the range of a normal social comparison group at pre-treatment, six individuals performed within this range at post-treatment assessment. This study had several methodological problems, including multiple statistical tests with small sample size, limited generalization attempts, and an intervention that would likely be impractical for clinical use given its year-long involvement.

Wiseman-Hakes et al. conducted a group intervention for six adolescents with ABI, four of whom were less than 8 months post-TBI, one of whom was 8 years post-TBI, and one of whom had ABI of unspecified etiology [69]. These six participants participated in an intervention of Sohlberg and colleagues “Improving Pragmatic Skills in Persons with Head Injury” [70] modified for use with a group. Four modules were taught: initiation, topic maintenance, turn-taking, and active listening, and each module consisted of an awareness phase, a practice phase, and a generalization phase. The intervention emphasized repetition, consistency, and feedback, with peers providing feedback and cueing. Significant improvements were found for ratings of pragmatic communication skills made by independent observers in nontreatment contexts for participants following treatment, and these improvements were maintained at 6-month follow-up. Given that the majority of participants were less than 8 months post-injury and no control group was utilized for the study, it is unclear to what extent changes reflect the effects of intervention versus spontaneous recovery.

Braden and colleagues conducted a cohort study with pre–post intervention and follow-up assessments as a feasibility study to explore whether a group social communication intervention that had demonstrated effectiveness in a TBI cohort without complications would be effective for a sample of persons with TBI with co-morbid neurological or psychiatric conditions [50, 52]. Participants were 30 individuals with TBI who were at least 1 year post-injury and had identified social communication impairments. In this cohort using paired t-tests, participants showed significant improvements in self- and other-ratings of social communication behaviors, in goal attainment scale achievement, and in self-reported satisfaction with life. Gains on behavioral rating measures of social communication did not reach statistical significance, nor did measures of awareness or social participation. This study suggests the potential for social communication interventions to be used in a broader population of persons with TBI and warrants further investigation.