Traumatic brain injury (TBI) may cause deficits in language and social communication, sometimes experienced by delayed word recall or a diminished ability to detect emotion while communicating with others. Such impairments may lead to frustrating or embarrassing experiences and affect an individual’s family dynamic, social life, and employment status. Cognitive rehabilitation therapy (CRT) interventions for language and social communication impairments may target social or emotion perception, social skills, or communication skills. Aphasia is another possible language impairment following acquired brain injury, although more common after stroke than TBI. The committee did not identify literature describing CRT interventions for aphasia after TBI. The following chapter describes controlled studies in language and social communication, followed by the committee’s conclusions.
The committee identified and reviewed four randomized controlled trials (RCTs) of language and social communication cognitive rehabilitation (Bornhofen and McDonald 2008a, 2008b; Dahlberg et al. 2007; McDonald et al. 2008). The committee found no studies of CRT for the domain of language and social communication for mild TBI, or for moderate-severe TBI in the subacute phase. All four trials were in the outpatient setting and enrolled moderate-severe TBI patients in the chronic phase of recovery. Two of the four RCTs focused solely on CRT for emotion perception deficits, one RCT focused on social communication skills training, and one RCT incorporated a combination of both social skills training and social/
emotion perception training. To be included, participants generally had to have sufficient language and cognitive capability to participate in a group, and have impairment in social communication skills either based on a questionnaire or a referring clinician’s assessment. One of the four RCTs had some form of CRT in both trial arms but also included comparison to a waitlist arm. The committee also identified one nonrandomized, parallel group controlled design (Hashimoto et al. 2006). This study was in the chronic phase of recovery for patients with moderate-severe TBI. Subjects were instructed on social skills training; no treatment was provided to the comparator arm (Hashimoto et al. 2006). Table 9-1 presents a summary of all included studies in this review.
Randomized Controlled Trials
Two trials focusing on treatment of emotion perception deficits were reported by Bornhofen and McDonald (2008a, 2008b). Emotion perception was defined as “accurate decoding and interpretation of visual and aural stimuli that signal 1 of 6 emotional states.” The CRT program reported by Bornhofen and McDonald (2008a) included group activities, and a notebook and home practice to teach increasingly complex skills on emotion perception. Sessions were held twice weekly, for 1.5 hours each over 8 weeks; 25 hours total. One therapist (background not described) was assigned to every two or three participants. The 12 participants were receiving outpatient services for TBI and were recruited and allocated at random to treatment or to a waitlist group; there was one dropout. Study outcomes were measures of facial expression (naming and matching), The Awareness of Social Inference Test (TASIT), and psychosocial reintegration. Immediately posttreatment, the intervention yielded significantly better TASIT scores relative to the waitlist group. While the intervention group scored better posttreatment on one form of the facial expression measure (matching), the groups scored the same on the alternate form of the facial expression measure (naming), and psychosocial reintegration. One month follow-up scores in the treatment arm were significantly higher than scores prior to treatment on all measures.
The other trial reported by Bornhofen and McDonald (2008b) had the goal of teasing apart the effective components of the intervention in the trial described above, by separating and comparing an errorless learning strategy with self-instruction training (which were combined in the 2008a study intervention), with a waitlist control group; both interventions also aimed to remediate emotion perception deficits. The interventions comprised a total of 25 hours of treatment across 10 weeks, divided into weekly, 2.5-hour
|Study||N||TBI Severity Level||Brief Narrative||Comparator||Outcome Measures||Findings|
Bornhofen and McDonald 2008a
|12||Severe||This study investigated whether social perception deficits could be remediated through cognitive rehabilitation, using a treatment program that incorporated techniques previously known to be effective with the TBI population.||Y
No Content: Waitlist control group
• Facial Expression Matching Task
• Facial Expression Naming Task
• The Awareness of Social Inference Test (TASIT), Parts 1, 2, and 3
• The Sydney Psychosocial Reintegration Scale (SPRS), Current Status - Self Ratings
|At immediate post-treatment, the intervention yielded significantly better social inference (TASIT) scores relative to the waitlist group. While the intervention group performed better posttreatment on scores of one form of the matching measure, there was no difference between groups on the alternate form of the matching measure, naming facial expression, or psychosocial reintegration. One month follow-up scores in the treatment arm were significantly higher than prior to treatment on all measures.|
|Bornhofen and McDonald 2008b||18||Severe||The objective of this study was to compare the efficacy of two strategies, errorless learning (EL) and self-instruction training (SIT), for improving deficits in emotion perception.||Y
No Content: Waitlist control group
• Primary outcome measures:
■ Audiovisual emotional displays: TASIT, Part 1 (Forms A and B) and social inferences based on emotional demeanor
■ Higher order social inference making: TASIT, Parts 2 and 3 (Forms A and B)
■ Identification of static emotion: Facial Expression Same/Different, Naming, and Matching Tasks
• Generalization measures:
■ Current Status.Relative Ratings (SPRS.Relative)
■ Depression Anxiety Stress Scales (DASS)
■ Katz Adjustment Scale.Relative Report Form (KAS.R)
■ Relative Ratings (SPSS-Positive and SPSS-Negative)
■ Social Performance Survey Schedule
■ SPRS and SPRS-Self
|Both treatment groups improved modestly in emotion perception; there is limited evidence to suggest that SIT may be a favorable approach for this type of remediation.|
|Dahlberg et al. 2007||52||Moderate- Severe||The study evaluates the efficacy of a group treatment program that targets the broader definition of social skills, uses a group process approach, emphasizes self-assessment and individual goal setting, and encourages generalization through homework and family or friend involvement.||Y
No Content: Waitlist control group
• Community Integration Questionnaire social integration and productivity subscales (CIQ)
• Craig Handicap Assessment and Reporting Technique Short Form social integration and occupation subscales (CHART-SF)
• Goal Attainment Scaling (GAS)
• Profile of Functional Impairment in Communication (PFIC)
• Satisfaction with Life Scale (SWLS)
• Social Communication Skills Questionnaire-Adapted (SCSQ-A)
|PFIC subscales showed more improvement for treatment versus control; SCSQ-A selfreport ratings showed more improvement for treatment versus control. Scales showed immediate improvement, with some preserved improvement at 3- and 6-month follow-up.|
|McDonald et al. 2008||39||Severe||The aim of this study was to determine whether remediation would be effective in improving social skills deficits, such as unskilled, inappropriate behavior; social perception; and mood disturbances (e.g., depression and anxiety).||Y
No Content and Non-CRT Content: Waitlist control group; Social activity group
• Primary outcomes:
■ Emotional adjustment: DASS
■ Social behavior: BRISS-R, PDBS, and PCSS
■ Social perception: TASIT
• Secondary outcomes:
■ Katz Adjustment Scale.RI
■ La Trobe Communication Questionnaire
■ Social Performance Survey Schedule
■ Sydney Psychosocial Reintegration Scale
|Relative to the waitlist control, social activity alone did not lead to improved performance on any outcome variable. The skills training group did improve differentially on the PDBS of the BRISS-R, while no treatment effects were found for the other primary outcomes or any of the secondary outcomes.|
|Nonrandomized, Parallel Controlled Group
Hashimoto et al. 2006
|37||Moderate-Severe||This study assessed the efficacy of a comprehensive day treatment program.||Y
No Content: Patients who did not join the day treatment program
• Activities of daily living:
■ FIM version 3.0
• Societal participation:
■ Community Integration Questionnaire (CIQ)
|The enrolled subjects displayed significant improvements in speech intelligibility, problem solving, memory, attention, and social integration scores in the FIM/FAM and scores in social integration and productive activity in the CIQ.|
sessions; in each session, a therapist worked with a group of two or three patients. The 18 participants were randomized to one of the three study arms; of these, there were five dropouts. Outcome measures included facial expression recognition, facial expression naming and matching, psycho-social reintegration, and depression and anxiety, as well as relative ratings of adjustment, social performance, and psychosocial reintegration. There were few statistically significant differences across these very small (four or five patients per arm) arms on study outcome measures.
Dahlberg et al. (2007) used a randomized trial to evaluate an outpatient group treatment program aimed at improving social communication skills after TBI. They employed a treatment workbook (Social Skills and TBI: A Workbook for Group Treatment) and limited each group’s size to eight participants. Each group met weekly for 1.5 hours for 12 weeks (18 hours) and was co-led by professionals from social work and speech pathology. Early sessions focused on self-assessment and goal setting, middle sessions focused on learning strategies for those goals, and later sessions focused on generalization; homework was assigned between sessions. Family members were involved outside the group setting. The 60 adults with TBI were randomized to either immediate participation in the social communication program or delayed treatment 3 months later; 52 people completed the study. The early treatment arm was followed for 36 weeks following completion of the program, and the delayed treatment arm was followed for 24 weeks. Primary outcomes were an objective measure of social communication skills (based on blinded raters’ assessments of videotaped interactions of the participant with research assistants, who were blinded to group assignment); a subjective assessment of social communication; and a Goal Attainment Scaling measure. Secondary outcomes were two assessments of community integration and one measure of life satisfaction. The researchers found that 12 weeks after the treatment sessions had ended, the intervention versus the control group had better scores on 7 of 10 scales of the primary outcome measure, which was the objective measure of social communication skills, as well as on the subjective assessment of social communication. There were no differences on the secondary outcome measures. Score improvements were maintained in both groups through 6-month follow-up.
McDonald et al. (2008) conducted a randomized trial of social behavior and social/emotional perception training compared to one control group receiving the same amount of time in grouped social activities; a second control group was waitlisted. The CRT intervention was 12 weeks at 4 hours per week, or 48 hours total, at an outpatient or community facility. It included group sessions each week focusing on social behavior training (2 hours) and social perception training to help decode expressions of emotion and social inferences (1 hour). The fourth hour each week was an individual session with a clinical psychologist who employed cognitive
behavioral therapy (CBT) techniques to address emotional adjustment. Across the three trial arms, 51 subjects were enrolled and randomized. Due to scheduling conflicts, nine subjects were reassigned to other arms after randomization and to balance numbers across arms. Outcomes measured included social behavior (based on blinded raters’ assessments of videotaped encounters of participants with an actor), measured by the Partner Directed Behavior Scale and the Personal Conversational Style Scale; both scales are part of the Behaviorally Referenced Rating System of Intermediary Social Skills (Revised). Other primary outcomes were the TASIT to assess social perception, and self-reported depression and anxiety. Secondary outcomes included a relative’s rating of social behavior on the Katz Adjustment Scale, a social performance survey, a communication questionnaire, and both self-and relative ratings on a psychosocial reintegration scale. Findings showed that the social skills treatment arm performed significantly better on the Partner Directed Behavior Scale compared to the social activity or waitlist trial arms (p = 0.004; effect size 0.70). There were no other differences across arms on any other primary or secondary outcome measures. Study limitations included insufficient power due to both attrition and to smaller effect sizes than anticipated, as well as the reassignment of participants from their initial randomization arms.
Nonrandomized, Parallel Group Studies
Hashimoto et al. (2006) evaluated an outpatient, day treatment program in Japan targeting social skills training. The treatment ranged from of a minimum of therapy for 2 hours per day, twice each week over 3 months (52 hours), to 4 hours per day, twice per week for 6 months (208 hours). The rationale for the variation in volume of day treatment program sessions was not provided. CRT content included social skills training by a clinical psychologist/speech therapist based on an approach of teaching improved behaviors by “redesigning the subjects’ environment.” CRT interventions also included occupational therapy, family conferences, sports, vocational rehab, and cooking. Services were delivered by a rehabilitation team, including the following: doctor/nurse, social worker, clinical psychologist/speech therapist, vocational rehabilitation counselor, physical therapist, rehabilitation gymnastic trainer, occupational therapist, and others. The sample was 25 adults (22 with TBI) ages 19 to 56. A control group consisted of 12 outpatients with TBI from the same medical center who met eligibility criteria but did not participate in the program. The study does not explain how participants were selected or why some selected participants did not participate in the program. Functional Independence Measure (FIM) and Functional Assessment Measure (FAM) scores and the Community Integration Questionnaire (CIQ) were collected before and after
participants completed the program (although it is not clear when the data were obtained for controls). CRT recipients were compared with controls on mean improvement in scores on these measures. While the groups did not differ on total social cognition, communication, or FIM motor score improvement, the participants improved more than controls on 5 of 12 FIM/FAM scales including social integration, attention, memory, problem solving, and speech intelligibility. On the CIQ, program participants improved significantly more on the total score and on subscale scores of social integration and productive activity than did controls; there was no difference in improvement on home integration.
The committee found the evidence of language and social communication CRT not informative about impact (efficacy) on patient-centered outcomes (quality of life, functional status). The evidence does not rule out a potentially meaningful effect of social communication skills or emotional perception skills training on psychosocial outcomes of community reintegration in adults with chronic, moderate-severe TBI (Hashimoto et al. 2006).
The committee found limited evidence for sustained effect of language and social communication CRT among chronic, moderate-severe TBI patients from the two RCTs that assessed sustained treatment effects. These studies found that beneficial effects on social communication skills or emotion perception were maintained through 1 month (Dahlberg et al. 2007) and 6 months (Bornhofen and McDonald 2008a).
The committee found modest evidence from a synthesis of findings across four RCTs and one nonrandomized trial for benefit of CRT on social communication skills among chronic, moderate-severe TBI patients. Efficacious interventions were small group, outpatient programs, meeting once to twice weekly for approximately 3 months. These interventions also employ a standardized protocol for social communication skills training, with or without emotion/social perception deficit training or CBT. In general, appropriate candidates for these programs were individuals with demonstrated language and social communication deficits, and who had sufficient language and cognitive capacity to participate in a group program (Bornhofen and McDonald 2008a, 2008b; Dahlberg et al. 2007; Hashimoto et al. 2006; McDonald et al. 2008).
In summary, the committee identified and reviewed four RCTs of language and social communication cognitive rehabilitation (Bornhofen and
McDonald 2008a, 2008b; Dahlberg et al. 2007; McDonald et al. 2008), all with chronic phase, moderate-severe TBI patients. Two studies focused solely on CRT for emotion perception deficits, one focused on social communication skills training, and one incorporated a combination of both social skills training and social/emotion perception training. Participant eligibility included having sufficient language and cognitive capability to participate in a group, and impairment in social communication skills either based on a questionnaire or a referring clinician’s assessment. The committee also identified a nonrandomized, parallel group controlled design study of social skills training versus a “no treatment” comparator arm (Hashimoto et al. 2006), for a total of five studies reviewed. There were no studies on CRT for language and social communication deficits among patients in the subacute phase of TBI or patients with chronic, mild TBI. One noteworthy aspect of these five CRT interventions was their relative feasibility in terms of service delivery. These CRT interventions ranged in time from 18 to 52 hours of services over 3 months; they all included delivery with small groups of patients; one employed an available workbook/manual; and most involved no more than two therapists (either social work, clinical psychology, or speech pathology, where specified). The types of intervention in these trials were either social communication skills training, emotion perception deficit training, or both; one trial also included 12 sessions with a clinical psychologist to deliver CBT.
Despite the fact that none of the five trials had more than 30 subjects in a given treatment arm, four of the trials yielded positive findings of the CRT intervention relative to controls on primary study outcomes of either improved social inference, where emotion perception deficits was a target (Bornhofen and McDonald 2008a), or social communication skills (Dahlberg et al. 2007; Hashimoto et al. 2006; McDonald et al. 2008); the exception to these findings was one very small trial (Bornhofen and McDonald 2008b). Only two studies examined outcomes after the immediate follow-up after the CRT program ended. One RCT (Dahlberg et al. 2007) found persistence of improvements in social communication skills through 6 months after the program ended, and another (Bornhofen and McDonald 2008a) found persistence of improvements in awareness of social inference through 1 month after the program ended. Only the nonrandomized, parallel group study (Hashimoto et al. 2006) showed improvements on more “distal” outcomes of social integration and productive activity. While not powered to detect smaller but potentially meaningful effects, Dahlberg et al. (2007) and McDonald et al. (2008) found that scores across treatment and waitlist groups on psychosocial outcome measures did not trend toward a difference in magnitude.
There is evidence to support benefit of small group outpatient programs, meeting once to twice weekly for approximately 3 months, and
employing a standardized protocol for social communication skills training. Applied in the community setting, such a program may or may not include concurrent emotion/social perception deficit training and CBT. Evidence shows these programs have beneficial impact on social communication skills among adults with moderate-severe TBI in the chronic phase of recovery. Patients with demonstrated language and social communication deficits should have sufficient language and cognitive capacity to participate in a group program. Evidence does not show if any subgroups are more likely to benefit than others.
Bornhofen, C., and S. McDonald. 2008a. Treating deficits in emotion perception following traumatic brain injury. Neuropsychological Rehabilitation 18(1):22–44.
———. 2008b. Comparing strategies for treating emotion perception deficits in traumatic brain injury. Journal of Head Trauma Rehabilitation 23(2):103–115.
Dahlberg, C. A., C. P. Cusick, L. A. Hawley, J. K. Newman, C. E. Morey, C. L. Harrison-Felix, and G. G. Whiteneck. 2007. Treatment efficacy of social communication skills training after traumatic brain injury: A randomized treatment and deferred treatment controlled trial. Archives of Physical Medicine and Rehabilitation 88(12):1561–1573.
Hashimoto, K., T. Okamoto, S. Watanabe, and M. Ohashi. 2006. Effectiveness of a comprehensive day treatment program for rehabilitation of patients with acquired brain injury in Japan. Journal of Rehabilitation Medicine 38(1):20–25.
McDonald, S., R. Tate, L. Togher, C. Bornhofen, E. Long, P. Gertler, and R. Bowen. 2008. Social skills treatment for people with severe, chronic acquired brain injuries. A multicenter trial. Archives of Physical Medicine 89(9):1648–1659.