Effectiveness of Communication-specific Coping Intervention for adults with traumatic brain injury: preliminary results

ABSTRACT

People with traumatic brain injury (TBI) describe everyday interactions as a long-term challenge frequently associated with ongoing stress. Communication-specific Coping Intervention (CommCope-I) is a new treatment developed to target coping in the context of communication breakdown. The intervention incorporates principles of cognitive behavioural therapy, self-coaching and context-sensitive social communication therapy. The purpose of this study was to examine the effectiveness of CommCope-I in a group of adults with severe TBI and ongoing functional communication difficulties. Participants were 13 adults with severe TBI (GCS = 3–8; mean age = 35.2 years; mean time post-injury = 7.6 years). The project involved three phases: (1) Control/pre-intervention wait phase (multiple assessments), (2) Treatment (6 weeks), and (3) Follow-up (12 weeks). Repeated measures ANOVA with planned pairwise comparisons were used to test the significance of change. Intervention elicited statistically significant improvements in communication-specific coping, functional communication and stress that were maintained for three months. Improved use of communication-specific coping strategies was evident in clinician blind ratings. Clients reported significant reduction in stress at the end of treatment and one and three months later. This intervention provides a promising means of improving communication-specific coping and reducing communication dysfunction and its negative consequences for people with TBI.

KEYWORDS:

• Brain injury
• communication
• coping
• intervention

Introduction

Traumatic brain injury (TBI) is typically acquired through motor vehicle accidents, assaults or falls and is the most common cause of disability in young people (Australian Institute of Health and Welfare, 2007). The complex neuro-behavioural effects of TBI result in a range of cognitive, communication, personality, behavioural and social consequences disrupting the quality of life of survivors and their families over many years (Access Economics, 2008).

In the domain of communication, as many as 70% of those with TBI report word retrieval problems (Ponsford, Olver, & Curran, 1995), while 35% are reported to have comprehension difficulties (Olver, Ponsford, & Curran, 1996) and 26% exhibit impaired social communication (Ponsford et al., 1995). Although few longitudinal studies of communication outcome have been conducted, evidence suggests that communication deficits after TBI persist into the long term (Snow, Douglas, & Ponsford, 1998, 1999). Further, impaired communication skills correlate significantly and substantially with long-term well-being, employment and social participation (Dahlberg et al., 2006; Knox & Douglas, 2009; Snow et al., 1998; Struchen et al., 2008; Struchen, Pappadis, Sander, Burrows, & Myszka, 2011). A significant relationship between communication and social participation is to be expected, given that communication is the means by which we negotiate daily activities and maintain relationships.

As communication deficits are not only common but also persistent, people with TBI frequently continue to experience communication breakdown. They describe everyday communicative interactions as stressful (Bracy & Douglas, 2005; Douglas & Spellacy, 2000) and relatives, teachers, employers and friends who interact with them regularly, frequently view social communication problems as one of the most problematic consequences of the injury (Bootes & Chapparo, 2010; Douglas & Spellacy, 2000; Galski, Tompkins, & Johnston, 1998; O’Flaherty & Douglas, 1997; Rietdjik, Simpson, Togher, Power, & Gillett, 2013; Shorland & Douglas, 2010). Stress and anxiety are associated with difficulty expressing and comprehending a message. Typically when faced with communication breakdown, communication-specific coping strategies are used. Productive coping strategies can be seen to assist message transfer, reduce stress and support participation, while non-productive strategies do little to resolve message transfer problems, frequently exacerbate stress, and reduce social activity. In a series of studies, we found previously that people with TBI demonstrate a pattern of communication-specific coping that differs from the norm (Friedman & Douglas, 2005; Mitchell & Douglas, 2011; Muir & Douglas, 2007). Participants with TBI used significantly more non-productive strategies and significantly fewer productive coping strategies than age, sex and education matched control participants. In addition, communication-specific coping was significantly correlated with social outcome (r = .51, p < .05; 25% shared variance; Friedman & Douglas, 2005) accounting for more variance than communication ability alone (13% shared variance; Snow, Douglas, & Ponsford, 1999).

Given support for the functional importance of communication-specific coping, in combination with the well-established negative impact of communication problems on outcome following TBI and a relative lack of evidence-based interventions to address these problems (Cicerone et al., 2011; Togher et al., 2014), we developed a treatment programme specifically to target coping in the context of communication breakdown. This structured treatment, Communication-specific Coping Intervention (CommCope-I), was designed to increase the use of productive and reduce the use of non-productive communication-specific coping strategies. We have completed preliminary evaluation of CommCope-I using single-case experimental design (SCED) applying an A-B-A design with follow-up using multiple probes and one replication (Douglas, Knox, De Maio, & Bridge, 2014). The two adults (1 male, 1 female) with chronic social communication problems who participated in this study had both sustained a very severe TBI (loss of consciousness > 4 weeks) more than 5 years prior to participating in the study. Percentage of non-overlapping corrected data (PNCD) (Manolov & Solanas, 2009) was used for analysis. PNCD involves a data-correction procedure to remove baseline trend from the data series prior to calculating the change produced as a result of intervention. A large treatment effect was demonstrated in both participants (PNCD: end of treatment = 100%, 3-month follow-up = 100%). These results are consistent with highly effective treatment (Scruggs & Mastropieri, 1998).

Having demonstrated single case evidence for the effectiveness of CommCope-I, we undertook the current study to establish further proof of concept of the intervention approach and to investigate its feasibility in a group study. In this study, we applied a pre- vs. post-intervention repeated measure design with an extended baseline pre-intervention phase to evaluate the stability of the dependent measures and act as a control phase. Our aim was to evaluate intervention and maintenance effects on proximal measures of communication coping and more distal measures of functional communication and stress at post-treatment and 1-month and 3-month follow-up assessments. The hypotheses were as follows: When compared to pre-intervention performance, participants would show significantly improved performance in (1) communication coping, (2) functional communication, and (3) stress, at the end of treatment, and at 1 month and 3 months after the end of treatment.

Method

Participants

Potential participants were contacted by mail through a brain injury rehabilitation unit or through private allied health practitioners working in the community setting. Participants were required to have sustained severe TBI as indexed by either a Glasgow Coma Scale (GCS) score ≤ 8 or duration of post-traumatic amnesia (PTA) ≥ 7 days, to be at least 2 years post-injury and to have ongoing functional communication difficulties as evidenced by self and/or close other report (more than 1.5 standard deviations above the normative mean for frequency of deficits) on the La Trobe Communication Questionnaire (LCQ, Douglas, O’Flaherty, & Snow, 2000). People with a medically documented pre-injury history of neurological or psychiatric disability and those with clinical evidence of receptive or expressive aphasia were excluded from the study.

Thirteen individuals (8 men, 5 women) participated in the study. All were able to provide consent to participate and no individual had a guardian or administrator appointed to make decisions on their behalf. Severity of injury as indexed by GCS scores ranged from 3 to 8 and length of PTA ranged from 14 to 150 days. Participants’ demographic, injury-related, functional communication and activity limitation characteristics are provided in Table 1.

Table 1. Demographic, injury-related, functional communication and handicap characteristics (n = 13).

Variable	Mean	Standard deviation	Range
Age at injury (years)	27.54	10.51	16–54
Education at injury (years)	13.62	2.60	10–17
Time post-injury (years)	7.62	5.16	2–18.5
Age at assessment (years)	35.15	9.30	25–58.2
Functional Communication: La Trobe Communication Questionnaire (LCQ)
Self-report (total score)	70.15	18.20	43–89
Close other report (total score)	77.73	15.67	53–97
Activity Limitations: Craig Handicap Assessment and Reporting Technique Short Form (CHART-SF)
Physical independence	91.08	21.18	36–100
Cognitive independence	61.31	28.68	15–100
Mobility	92.77	12.26	64–100
Occupation	46.58	21.53	18–100
Social integration	77.19	27.19	23–100

Two participants lived alone with attendant care support, one lived in shared accommodation with a single flat-mate, two lived with their partners and the remaining eight participants lived with family. At the time of injury, two participants were unemployed and looking for employment, six were in full-time employment, two were pursuing study and in part-time employment and one was a full-time student. At the time of participating in the research, three participants demonstrated activity limitation across all five domains (physical, cognitive, mobility, occupation and social integration) of the Craig Handicap Assessment Reporting Technique–Short Form (CHART-SF) (Whiteneck, Charlifue, Gerhart, Overhosler, & Richardson, 1992), three across four domains, four across three domains, one each on two and one domain and one participant received the maximum score on all five domains. Two participants were employed and the remainder unemployed. One participant lived in a regional city and the others lived in a large metropolitan city.

Each participant was invited to nominate a relative or close other to participate in collection of assessment data through the course of the study. Seven close others (six women, one man) participated and completed close other assessments across all time points of the study. These close others were either parents (five mothers) or spouses (one husband, one wife) of the participants with TBI.

Measures

Four tools were used to measure change across three areas of function in which change was hypothesised (communication-specific coping, functional communication, and stress) during the course of the study. Each of these measures had been trialled in the SCED study (Douglas et al., 2014) that preceded this project and had been found to be responsive to change in the domains evaluated.

Communication-specific coping

Communication-specific coping was measured from three perspectives: self, close other, and clinician ratings.

The Communication-Specific Coping Scale—Research version (CommSpeCS) (Douglas & Mitchell, 2012) was developed to measure communication-specific coping strategies. It consists of two subscales (comprehension and expression) and has two versions (self-report and close other). Each subscale has 35 items loading onto two domains: productive communication coping strategies that support message transfer, and non-productive strategies that act as a hindrance to the interaction. To date, the CommSpeCS has been psychometrically evaluated in two studies. Internal consistency and temporal stability have been evaluated on a normative control sample of 182 (101 males, 81 females; mean age 38.2 ± 13.5 years) healthy adults and close others recruited from the community across several states in Australia (Victoria, Tasmania, South Australia and Queensland). Respondents with self-reported history of hearing impairment, psychiatric illness, neurological conditions, head injury or severe concussion were excluded. Internal consistency and test-retest reliability coefficients equalled or exceeded .80 (.80–.84) in the normative sample for self and close other versions of the scale measuring message expression and message comprehension. On the comprehension scale, scores for self-report ratings in the normative group were normally distributed around a total mean score of 71.40 ± 8.40 (Productive 52.08 ± 7.68; Non-productive 19.36 ± 4.73) and for close-other ratings there was a total mean score of 67.2 ± 9.45 (Productive 48.24 ± 8.16; Non-productive 18.81 ± 5.28). On the expression scale, scores for self-report ratings in the normative group were normally distributed around a total score mean of 75.25 ± 9.45 (Productive 57.0 ± 8.75; Non-productive 18.0 ± 4.40) and for close-other ratings scores were a mean of 73.85 ± 10.15 (Productive 55.0 ± 9.25; Non-productive 18.85 ± 5.50). In the second study, the ability of the CommSpeCS to distinguish between a group of 26 adults with TBI and a control group of 26 neurologically normal adults matched on sex, age and education was evaluated. Scores on the CommSpeCS successfully distinguished between the TBI group and the control group with significant differences between the groups (p < .01) evident on both the self-report and close other versions of the scale.

The Discourse Coping Scale—Clinician Rating (DCS–CR) was developed as a clinician rating of communication coping behaviour during interaction for the previous study (Douglas et al., 2014). The scale uses a 10 cm visual analog scale with anchor points 0 and 10 defined (0 = person shows little or no appropriate use of communication-specific coping strategies and 10 = person shows full and appropriate use of communication-specific coping strategies). The assessor is required to view a video recording of a participant interaction and rate the extent to which the specified individual demonstrates full and appropriate use of communication-specific coping strategies within the interaction. To assist with the rating, the assessor is provided with a list of 35 communication-specific coping strategies used in situations where people are struggling to understand the message being conveyed by a communication partner (i.e., comprehension strategies from CommSpeCS­–Research version) and 35 communication-specific coping strategies used in situations where people are struggling to get their message across to a communication partner (i.e., expression strategies from the CommSpeCS–Research version).

The DCS-CR procedure demonstrated excellent inter-rater reliability (ICC = .996; 95% confidence interval .990–.999) in the SCED study (Douglas et al., 2014). In the current study, inter-rater reliability on the DCS-CR was again evaluated through calculation of intraclass correlation coefficients (ICC) using the two-way random effects model. Two clinicians independently rated 27 discourse samples (38%) using the DCS procedure. The average of the scores of the two raters was highly reliable (ICC = .995; 95% confidence interval .988–.998) and the single measure ICC showed excellent single rater reliability (ICC = .989; 95% confidence interval .976–.995).

Functional communication

The La Trobe Communication Questionnaire (LCQ) (Douglas et al., 2000; Douglas, Bracy, & Snow, 2007a, b) was used to measure the frequency of functional communication problems from the perspective of the person with TBI and their nominated close other. The LCQ has two forms, a self-report form and a close other/relative form each of 30 items. Item content reflects the four domains of Grice’s (1975) Cooperative Principle of Conversation: Quantity, Quality, Relation and Manner supplemented by items reflecting social communication deficits associated with TBI (see Douglas et al., 2000; Douglas et al., 2007a, 2007b for detailed description of item content).

Each item has four possible levels of response: (1) never or rarely, (2) sometimes, (3) often, and (4) usually or always, yielding a total score range of 30–120. High scores are consistent with a perception of frequent difficulties (i.e., more disability). Douglas et al. (2000) provided LCQ normative data for the perceptions of 147 healthy adults and 109 close others. Total LCQ scores for self- and close-other reports in the normative group were normally distributed around a total score mean of 52.47 (SD = 9.62) for self-report ratings and a mean of 47.17 (SD= 9.93) for close-other ratings.

The LCQ has been psychometrically evaluated on young healthy adults (Douglas et al., 2000) and adults with TBI (Douglas et al., 2007a, 2007b) with excellent results. In the case of TBI participants, internal consistency is high for self (Cronbach’s α = .91) and close other (Cronbach’s α = .92) report forms. Test-retest coefficients across a two-week interval for TBI participants (r = .81) and their close others (r = .87) are also acceptable. Principal component factor analysis supports the construct validity of the LCQ (Douglas et al., 2000; Douglas, Bracy, & Snow, 2007b; Struchen et al., 2008) and its sensitivity and clinical utility in the context of TBI have been supported by the results of several published studies (Bracy & Douglas, 2005; Douglas, 2010; Douglas, Bracy, & Snow, 2016; Struchen et al., 2008; Struchen et al., 2011; Tu, Togher, & Power, 2011; Watts & Douglas, 2006).

Stress

The Stress subscale of the Depression Anxiety Stress Scales-21 (DASS-21) was used to index stress in participants with TBI. The DASS-21 is a short form of Lovibond and Lovibond's (1995) 42-item self-report measure of depression, anxiety, and stress (DASS). It consists of three 7-item subscales each with a possible score range of 0–21. Construct validity and normative data for the general adult population are available (Henry & Crawford, 2005). Each subscale has acceptable to high internal consistency (Cronbach’s α: depression = .88, anxiety = .82, stress = .90). Confirmatory factor analysis demonstrates the three subscales index a common factor (general psychological distress) but also account for significant variance specific to each scale.

Activity limitations

The Craig Assessment and Reporting Technique—Short Form (CHART-SF) (Whiteneck et al., 1992) was used to describe the participants’ level of activity limitation at the time they were recruited to the study. The CHART-SF has six sub-scales (physical independence, cognitive independence, mobility, occupation, social integration and economic self-sufficiency) that quantify an individual’s functioning in a particular domain. Five sub-scales (physical independence, cognitive independence, mobility, occupation, and social integration) were used in this study. Each sub-scale of the CHART-SF has a maximum score of 100 points. Higher scores for each of the CHART sub-scales reflect higher levels of functioning and participation, with a score of 100 considered the level of performance typical of the average non-disabled person. Achieving the maximum score indicates that the roles within that domain are fulfilled at a level equivalent to that of the norm (Mellick, 2000). Scores have been normed on a non-disabled population and Rasch analysis has verified scaling and scoring procedures (Whiteneck et al., 1992). Test–retest and participant-proxy reliability have been shown to be high (Zhang et al., 2002).

Intervention

CommCope-I has been described previously by Douglas et al. (2014). It is a structured intervention programme developed to incorporate the procedures and principles of cognitive behavioural therapy and context-sensitive social communication therapy. The aim of CommCope-I is to improve the client’s use of productive communication coping strategies by shaping and increasing the use of productive strategies that are already evident in the client’s behavioural repertoire. The client’s productive strategies are then used to replace his/her non-productive strategies. The programme has three components through which the client progresses: (1) facilitation of self-awareness of coping strategies, (2) skill development/strategy practice in personally relevant scenarios, and (3) evaluation of performance through video review.

The core procedural methods employed in CommCope-I reflect the principal concepts of self-coaching, a treatment method developed for people with TBI by Ylvisaker (2006). These methods include development of personally relevant scripts, use of personally compelling imagery and concrete reminders, repeated rehearsal, use of video learning trials, use of reminder scripts by communication partners, supported and self-managed use in the real world environment and self-evaluation of performance. These procedural methods also reflect the common elements and individualised nature of communication treatments that have to date yielded positive results (Behn, Togher, Power, & Heard, 2012; Dahlberg et al., 2007; Togher, McDonald, Tate, Power, & Rietdijk, 2013) and recent practice recommendations based on evidence reviews (Cicerone et al., 2011; Togher et al., 2014).

The CommCope-I programme delivered in this project required 22 hours: pre-treatment assessment (4 hours: 2 × 2-hour individual sessions), intervention phase (12 hours: 6 × 1-hour individual sessions, 6 × 1-hour sessions with communication partner/s in the community), post-treatment assessment (2 hours), and two follow-up assessments (4 hours: 2 × 2-hour individual sessions). The intervention phase is delivered at a rate of two sessions per week over a 6-week period. The intervention phase is followed by a re-assessment at 1 month and then at 3 months. The beginning of treatment to the 3-month follow-up assessment spans a period of 4.5 months.

Clinician training

Four clinicians were trained to deliver the intervention. Training involved a 3-hour group interactive workshop where the treatment with associated written support materials was described and demonstrated. The clinicians were qualified practising speech pathologists. They all had experience working with clients with neurogenic communication disorders. One had less than 5 years experience, while the remaining three had more than 5 years experience.

Procedure

Approval to conduct the study was obtained from university and health institution ethics committees. Data collection took place in the participants’ homes and in their local community. The first visit commenced with completion of informed consent procedures and collection of demographic information. Participants with TBI then completed the DASS-21, the self-report versions of the LCQ and the CommSpeCS. Close others completed close other versions of the LCQ and the CommSpeCS. These assessments were repeated 6 weeks later, then 1 week after the end of treatment and two follow-up intervals (1 month and 3 months). Thus assessment was completed at five time points: Times 1 and 2—pre-intervention, Time 3—1-week post-intervention, Time 4—1-month follow-up, and Time 5—3-month follow-up.

An independent assessor (speech pathologist not involved in the individual’s treatment) evaluated performance across assessment sessions using the DCS–CR. Session recordings were presented in random order for independent rating.

During the final follow-up session, participants were asked to provide feedback about the programme and to suggest any potential modifications that may have enhanced their experience of participation. The audio-recordings of these feedback sessions were transcribed and analysed using thematic analysis.

Design and data analysis

Repeated measures analysis of variance (ANOVA) with planned pairwise comparisons was applied to evaluate statistically the significance of intervention-related change over time in the dependent variables. Where the assumption of sphericity was violated as indicated by Mauchly’s test, Greenhouse-Geisser correction was applied. An alpha level of .05 was applied to all tests and effect size was indexed by partial eta squared (). Given the exploratory nature of the study and small sample size, no alpha level adjustment for multiple comparisons was made for statistical testing. However, exact effect size indices are reported for ANOVA results and planned comparisons that reached significance (p < .05). Effect size interpretations were based on Ferguson’s (2009) recommendations for : .04 = recommended minimum effect size representing a practically significant effect (RMPE); .25 = moderate effect; .64 = strong effect and Cohen’s (1992) recommendations for d: .2 = small effect; .5 = medium effect; .8 = large effect.

The feedback component of the final sessions was audio-recorded and transcribed. The transcribed data were analysed using thematic analysis following the principles described by Braun and Clarke (2006).

Results

Intervention-related changes were hypthesised to occur in the following three areas: (1) communication-specific coping measured by the CommSpeCS (self- and close-other report) and the DCS (clinician rating); (2) functional communication measured by the LCQ (self- and close-other report); and (3) stress measured by the stress subscale of the DASS-21.

Communication-specific coping

CommSpeCS

Descriptive statistics and results of repeated measures ANOVA for the self-report and close other versions of the CommSpeCS are shown in Table 2.

Table 2. Descriptive statistics and results of repeated measure ANOVA for measures of communication-specific coping.

Measure	Time 1 Mean (SD)	Time 2 Mean (SD)	Time 3 Mean (SD)	Time 4 Mean (SD)	Time 5 Mean (SD)	F	p
Communication-specific Coping Scale (CommSpeCS)
Self-report (n = 13)
Comprehension: Productive	50.23 (14.28)	52.50 (14.48)	47.69 (14.11)	47.85 (12.80)	46.23 (14.77)	1.142	.348	.087
Comprehension: Non-productive	19.62 (6.87)	19.42 (7.45)	16.46 (6.85)	15.47 (4.00)	15.08 (5.69)	3.987^GG	.028*	.249^b
Expression: Productive	59.31 (16.05)	60.15 (10.49)	60.84 (14.52)	58.76 (15.83)	58.54 (15.71)	0.188	.943	.015
Expression: Non-productive	17.15 (6.95)	17.29 (5.02)	15.77 (4.71)	14.47 (5.70)	14.23 (3.85)	2.09^GG	.129	.148^a
Close other (n = 7)
Comprehension: Productive	46.21 (6.59)	46.47 (6.17)	48.17 (8.52)	51.67 (10.17)	46.45 (6.05)	.930	.467	.157^a
Comprehension: Non-productive	22.17 (5.38)	21.21 (6.14)	19.83 (3.92)	20.33 (6.25)	20.00 (6.51)	.512	.727	.093^a
Expression: Productive	45.98 (3.72)	45.76 (2.73)	55.85 (12.21)	62.17 (13.70)	53.33 (8.43)	5.691	.003**	.532^b
Expression: Non-productive	20.33 (3.98)	21.17 (4.79)	18.33 (7.06)	20.17 (6.21)	18.17 (5.00)	.696^GG	.518	.122^a
Discourse Coping Scale (DCS)
Clinician (blind rated)
Comprehension Task	3.42 (1.09)	3.72 (1.21)	5.13 (1.44)	5.25 (2.21)	6.10 (1.91)	8.92^GG	.001**	.403^b
Expression Task	4.28 (1.68)	4.72 (1.36)	5.55 (1.74)	6.39 (2.02)	6.65 (2.13)	12.72^GG	< .001**	.515^b

^GGGreenhouse-Geisser correction applied to degrees of freedom; ^aexceeds RMPE (recommended minimum effect size representing a practically significant effect); ^bmoderate effect.

*p < .05; **p < .01

Self-report: Participants self-reported reduced use of non-productive coping strategies in the context of both comprehension and expression situations. Repeated measures ANOVA with a Greenhouse-Geisser correction determined that mean frequency of use of non-productive communication-specific coping strategies in situations characterised by comprehension breakdown differed significantly between time points and this difference was commensurate with a moderate effect (  = .249). Post hoc tests of planned comparisons revealed that the two baseline measures (Time 1 and Time 2) were similar (p = .897), there was a significant reduction in the use of non-productive strategies at the end of CommCope-I (Time 3) when compared to baseline (Time 2) (p = .013; d = .41) and the performance at the end of treatment (Time 3) was maintained at 1-month (Time 4) (p = .490) and 3-month follow-up (Time 5) (p = .139). Comparison between baseline (Time 1 and Time 2) and follow-up measures (Time 4 and Time 5) revealed that significantly reduced use of non-productive strategies was maintained (Time 1:Time 4, p = .045, d = 0.76; Time 2:Time 4, p = .025, d = 0.69; Time 2:Time 5, p = .013, d = 0.66). Although change over time did not reach statistical significance on the measure of non-productive coping strategies in situations characterised by breakdown in message transfer (expression), the measured effect (  = .148) exceeded the recommended minimum effect size representing a practically significant effect (RMPE) (i.e., .04).

Close other report: Close others reported increased use of productive strategies and decreased use of non-productive strategies by participants with TBI in comprehension and expression situations. ANOVA with a Greenhouse-Geisser correction determined that mean frequency of use of productive communication-specific coping strategies in situations characterised by breakdown in message transfer (expression) differed significantly between time points and was commensurate with a moderate to large effect (  = .532). Post hoc tests of planned comparisons revealed that the two baseline measures (Time 1 and Time 2) were similar (p = .905). There was an increase in the use of productive strategies at the end of CommCope-I (Time 3) when compared to baseline (Time 1) but this did not reach significance (p = .071) until comparisons between baseline scores and follow-up scores (Time 1:Time 4, p = .037, d = 1.86; Time 2:Time 4, p = .025, d = 2.00; Time 2:Time 5, p = .05, d = 1.09) were made. Improved performance was maintained and there was no significant change between Time 3 and Time 4 (p = .199) and Time 3 and Time 5 (.571). The positive effect observed on the use of productive communication-specific coping strategies in situations characterised by comprehension breakdown represented a practically significant effect (  = .157), but it did not reach statistical significance. Similarly, practically significant positive effects were evident with a reduced use of non-productive strategies (comprehension  = .093; expression  = .122).

DCS-CR

Descriptive statistics and results of repeated measures ANOVA for the DCS-CR can be found in Table 2. ANOVA with a Greenhouse-Geisser correction demonstrated that mean discourse coping clinician ratings on the expression and comprehension tasks differed significantly between time points and this difference was commensurate with a moderate effect (DCS-CR: Expression Task,  = .515; DCS-CR: Comprehension Task,  = .403). On the Expression Task, the two baseline ratings (Time 1 and Time 2) were similar (p = .258). The rating was significantly higher at the end of CommCope-I (Time 3) when compared to baseline (Time 1:Time 3, p = .008, d = 0.74; Time 2:Time 3, p = .030, d = 0.54). Ratings at Time 4 and Time 5 follow-ups were significantly higher than Time 3 (p = .005, d = 0.45; p = .001, d = 0.57). Although ratings continued to increase after treatment, there was no significant change between Time 4 and Time 5 (p = .132). Comparison between baseline (Time 1 and Time 2) and follow-up measures (Time 4 and Time 5) revealed that significantly improved discourse ratings were maintained (Time 1:Time 4, p = .003, d = 1.14; Time 1:Time 5, p = .001, d = 1.24; Time 2:Time 4, p = .003, d = 0.99; Time 2:Time 5, p = .002, d = 1.10). The results on the comprehension task repeated this pattern. The two baseline ratings (Time 1 and Time 2) were similar (p = .08). The rating was significantly higher at the end of CommCope-I (Time 3) when compared to baseline (Time 1:Time 3, p = .001, d = 1.35; Time 2:Time 3, p = .001, d = 1.06). Improved ratings at Time 3 were maintained at Time 4 and Time 5 follow-ups (p = .866; p = .076). Maintenance of the improved ratings was supported by comparison between baseline (Time 1 and Time 2) and follow-up measures (Time 4 and Time 5) which revealed significantly improved discourse ratings (Time 1:Time 4, p = .03, d = 1.11; Time 1:Time 5, p = .001, d = 1.79; Time 2:Time 4, p = .049, d = 1.57; Time 2:Time 5, p = .002, d = 1.53).

Functional communication

LCQ

Descriptive statistics and results of repeated measures ANOVA for the self-report and close other versions of the LCQ are shown in Table 3.

Table 3. Descriptive statistics and results of repeated measure ANOVA for measures of functional communication and stress.

Measure	Time 1 Mean (SD)	Time 2 Mean (SD)	Time 3 Mean (SD)	Time 4 Mean (SD)	Time 5 Mean (SD)	F	p
La Trobe Communication Questionnaire (LCQ)
Self-report (n = 13)
Total score	70.15 (18.20)	69.54 (23.29)	58.38 (18.33)	62.58 (18.96)	58.35 (17.27)	4.78^GG	.017*	.285^b
Close other (n = 7)
Total score	77.73 (15.67)	79.93 (17.89)	70.21 (9.82)	71.70 (8.05)	70.57 (9.90)	1.05^GG	.360	.149^a
Depression Anxiety and Stress Scales (DASS-21)
Depression (n = 13)	12.83 (13.09)	11.83 (13.00)	8.17 (8.84)	7.31 (7.70)	6.33 (9.30)	1.80^GG	.183	.141^a
Anxiety (n = 13)	7.99 (11.25)	9.83 (12.04)	3.67 (5.58)	5.39 (7.04)	6.50 (8.79)	3.50^GG	.060	.241^a
Stress scale (n = 13)	12.33 (14.37)	14.00 (14.30)	6.50 (6.50)	4.62 (4.30)	3.50 (4.50)	5.62^GG	.026*	.338^b

^GGGreenhouse-Geisser correction applied to degrees of freedom; ^aexceeds RMPE (recommended minimum effect size representing a practically significant effect); ^bmoderate effect.

*p < .05.

Self-report: Improvement with a moderate effect (  = .285) was self-reported in the domain of functional communication. There was no significant change in frequency of social communication problems from Time 1 to Time 2 (p = .869), there was a significant reduction in problems at the end of treatment (Time 1:Time 3, p =.004, d = 0.64; Time 2:Time 3, p = .031, d = 0.54) and the gains at the end of treatment were maintained at 1 month (Time 4) (p = .085) and 3-month follow-up (Time 5) (p = .989). The maintenance of significantly reduced social communication problems at follow-up was also evident on comparison with baseline measures (Time 1:Time 5, p = .012, d = 0.67; Time 2:Time 5, p = .021, d = 0.55).

Close other report: A practically significant positive effect was evident with decreased functional communication problems reported by close others (LCQ  = .149) following intervention, however this change did not reach statistical significance.

Stress

DASS-21

Descriptive statistics and results of repeated measure ANOVA for the DASS-21 are presented in Table 3. Although all three scales of the DASS-21 were administered, a reduction in scores on the stress scale was the only change around which a specific hypothesis was made and thus subjected to planned comparisons. Repeated measures ANOVA with a Greenhouse-Geisser correction revealed stress differed significantly between time points (  = .338) with a similar pattern as evidenced in the communication-specific coping measure and the measure of functional communication problems. Post hoc tests revealed that scores on the stress scale at the two baseline time points (Time 1 and Time 2) were similar (p = .147), there was a reduction in stress scores at the end of CommCope-I (Time 3) when compared to baseline (Time 2) but this did not reach significance (p = .062) until comparisons were made between baseline scores and follow-up scores (Time 1:Time 4, p = .047, d = 0.83; Time 1:Time 5, p = .024, d = 0.94; Time 2:Time 4, p = .024, d = 1.01; Time 2:Time 5, p = .011, d = 1.12). Although stress scores continued to reduce after treatment, there was no significant change between Time 3 and Time 4 (p = .216). At the 3-month follow-up point (Time 5) there was further reduction in stress scores compared to the end of treatment (p = .05).

Participant feedback about the programme

Several key themes emerged from analysis of the feedback provided by participants when they were interviewed about their experience of the programme and suggested modifications. Overall, participants were overwhelmingly positive in their responses.

Benefits of participation in the programme

Participants discussed a range of benefits that they had identified as a result of the programme. The themes evident within the interviews ranged from a more developed sense of awareness about their communication (particularly their communication strengths and weaknesses) to a sense of achievement and increased independence as a result of their experiences during the programme.

Being aware of own communication strengths and weaknesses: The first key theme to emerge reflects participants’ increased awareness of their own communication strengths and weaknesses, and an improved sense of self-efficacy in being able to address communication issues that they may experience.

I thought the programme was good because it has highlighted where you have problems or weaknesses, and then you can work on them. (BJO)

I’ve learnt that most of the time, they (communication partners) don’t understand because of the way that I am letting them know, so I need to change the way that I am letting them know. (DAN)

I got to see things from other points of view. I’ve got my point of view but to see it (through) somebody else’s eyes made me see it in a broader range … opened my eyes up. (RJA)

Putting strategies into action immediately: Participants reported that a positive aspect of the programme was the ability to put the strategies that they had identified and were practising with their treating clinician into real world action. This was particularly reflected in participants’ comments about the nature of the programme, which targeted the increased use of specific strategies that already exist in the individual’s repertoire within situations that are most relevant and important to them.

(We) have done a variety of things, haven’t just been sitting at home doing things, we’ve been out and about doing role plays in real situations, as well as doing role plays at home. (BJO)

The things we’ve been learning, I’ve been able to put them into practice and they do make a difference. (MAN)

Breaking the strategies down into steps, numbered points (made it easier to remember). (MGR)

Having a sense of accomplishment: Participants’ feedback reflected a strong sense of accomplishment and achievement when reflecting on what they had done over the course of the programme. This sense of accomplishment was reflected in feelings of achievement and increased confidence.

I think I probably still don’t speak perfectly but I feel more confident. (BAN)

By the third session, I was trying really hard … trying to get 10/10. In my head, I was trying really hard to get 10 for that one, and I did because I was using all the strategies in everyday conversation … practice makes perfect.(ABR)

In the beginning, I hated you and the things that you made me do, because I wasn’t used to seeing myself (on video) but now I see what strategies I’ve done, and I like watching it. I feel like, “That’s me and I did it” … I definitely know there’s improvements, seeing myself at the start and finishing, I’m a lot clearer. I think all around it has been a good seven weeks.

(WAM)

Achieving greater independence: Several participants described increased independence as a significant outcome for the programme. For some participants, this reflected independence in communication tasks, such as being able to order their own meal. For others, this reflected increased confidence both in their own skills and the perceptions of others to be capable to undertake a range of tasks on their own.

I am going to order at the restaurant myself from now on.

(BAN)

Areas for improvement

A number of participants proposed some minor modifications to improve the programme.

Having more opportunities to practise in the community: Some participants, particularly those who reported an initial reluctance to participate in community-based sessions, reported that they would have preferred a higher number of sessions to take place in their “real life” settings. In some cases, the availability and priorities of support workers in supporting clients to access the community proved to be a barrier.

I enjoyed (this) a lot, probably because I’m always home. (DAN)

… probably should have got out more. (BAL)

Maybe more practice in society, rather than practice at home … (it helps) put theory into practice.

(MAN)

Extending the length of the programme: Several participants reported that they still had a number of communication goals that they wished to achieve and would prefer if the programme were longer.

I thought it was too long in the beginning, but now I wish it was longer … there is much more improvement to be made. (BAL)

It was okay, but a few more (sessions would be) even better, I think.

(VA)

Discussion

CommCope-I is a new approach to functional communication intervention following acquired brain injury. This novel intervention focuses on communication-specific coping strategies that can be readily applied across all situations a person encounters on a daily basis. It takes the emphasis off deficits and focuses primarily on increasing the use of productive strategies already evident in the day-to-day interactions of the individual with brain injury. The intervention allows not only people with brain injury to participate but also their everyday communication partners.

In the current study, the CommCope-I programme elicited statistically significant improvements on expression and comprehension discourse tasks rated independently by a clinician blind to the phase during which the task was completed. These improvements were maintained up to 3 months post-treatment and reflected a moderately large treatment effect. The clients also self-reported positive changes in their interpersonal communication including significant reductions in their use of non-productive communication-specific coping strategies and functional communication problems. These changes were maintained at 1 and 3 months following treatment. Participants also reported a significant reduction in stress at the end of treatment and this reduction too was maintained 1 and 3 months later. Self-reported reductions in non-productive coping strategies, functional communication problems and stress were all consistent with a moderate clinical effect.

Close others who interacted regularly with the clients also perceived positive intervention-related changes. They reported a statistically significant increase in observed use of productive communication-specific coping strategies which was maintained 1 and 3 months following treatment. This change represented a moderate clinical effect. Close others also observed positive changes in functional communication that was consistent with a practically but not statistically significant effect.

The pattern of improvement reported by participants themselves and by their close others provides an interesting comparison of these perspectives. Participants themselves seemed to be more aware of reducing their use of non-productive strategies, while their close others were more aware of the increased use of productive strategies. It is noteworthy that the quality of the interaction will improve through both these channels and the frequency of functional communication problems is expected to decrease through either or both a decrease in use of non-productive strategies and an increase in use of productive strategies. Indeed, both the self-report and close other results on the LCQ demonstrated this outcome.

The qualitative feedback about the programme from clients and close others was particularly encouraging and in line with the positive changes in communication ability and confidence associated with CommCope-I. The personal feedback effectively highlighted important strengths of the intervention including its focus on strategy development, use of video feedback, inclusion of everyday communication partners and real world practice in the community. It is notable that a number of these treatment variables are also contained within other contemporary treatment programmes that have been found to be effective (Behn et al., 2012; Dahlberg et al., 2007; Togher et al., 2013).

Although the programme was found to be effective in its 6 week delivery mode, some participants commented that they would have liked it to be longer. Feasible delivery over 6 weeks with positive outcomes has advantages from an efficiency (cost and time) standpoint. However, the potential benefits of providing top-up, extended or staggered doses of intervention could also be investigated. Participants were able to complete practice and performance of four to six personally relevant scenarios during the intervention period. An alternative approach to length of treatment could involve identification of a number of scenarios on which to complete the CommCope-I process.

Overall, the results of this study indicate that this novel intervention focused on communication-specific coping can produce significant improvement in the functional communication skills of adults who have substantial communication problems as a consequence of severe TBI. Not only was significant improvement evident, but it was elicited in adults who sustained their injuries on average 7.6 years previously and the improvement was maintained at least 3 months post-intervention. Further, positive change in well-being reflected by significantly reduced stress levels was present at the end of the intervention, and this reduced stress persisted three months after the intervention had concluded.

These findings are particularly significant when considered in the context of the scope and consequences of communication problems after TBI. A large proportion of adults with TBI experience persistent communication problems and these problems significantly and negatively affect social and emotional outcome in the short and long term. Clearly, CommCope-I can provide an efficient (2 hours assessment + 12 hours of treatment) means of reducing communication dysfunction and its negative consequences for people with TBI.

Limitations of the study

This study brings with it several limitations that require consideration. First, the design employed in this proof of concept project did not include a control group. Instead, we applied a pre- vs. post-intervention repeated measures design with an extended baseline pre-intervention phase included to evaluate the stability of the dependent measures in a non-intervention phase. While not ideal, this 6-week baseline phase afforded some degree of control by enabling comparison of performance across the 6-week treatment phase with that of the 6-week pre-intervention control phase. Second, the sample size was small, particularly in the case of close other participants, and no adjustments were applied to correct for multiple statistical comparisons. Nevertheless, statistical significance and the magnitude of effect size for these comparisons were demonstrated sufficiently to inform the design of future evaluation utilising randomised controlled trial designs. Finally, the participants were drawn exclusively from the severe end of the injury continuum in the chronic phase. While the chronicity and severity of participants could be seen to limit the generalisability of the findings, they also minimised the likelihood that results would be compromised by ongoing spontaneous recovery. Further, although all participants had sustained severe TBI determined by GCS or PTA duration, their current levels of activity limitation showed substantial variation across physical, cognitive, mobility, occupation and social integration domains of activity. In addition despite considerable therapy input previously, these participants came to the programme with substantial ongoing communication problems that impacted negatively on their everyday activities. Thus although the participants had longstanding problems that were apparently resistant to traditional therapy approaches, we saw them as being well placed to demonstrate the effectiveness of the programme and benefit from it.

Conclusion

The results of this project provide support for the effectiveness of CommCope-I for people who have sustained severe TBI and are continuing to experience communication problems more than 2 years post-injury. Given the strength of the results, it is likely that the intervention will prove to be at least equally as effective with adults with communication problems and less severe TBI and also be suitable for application earlier post-injury. These findings represent sufficient proof of concept and feasibility data to support further large-scale research using controlled trials and including direct comparison with alternative treatments as well as systematic investigation of participant characteristics and treatment delivery modes.

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Jacinta M. Douglas http://orcid.org/0000-0003-0940-6624

Lucy Knox http://orcid.org/0000-0001-8331-6032

Additional information

Funding

This work was supported by the Transport Accident Commission, Victorian Neurotrauma Initiative under [grant number DO71].