Addressing Participation in Adults With Postconcussive Symptoms Using Cognitive Strategy Training: A Feasibility Trial

An estimated 2.5 million traumatic brain injuries (TBIs) are evaluated each year in the emergency department; roughly 75% to 90% are classified as mild TBIs (mTBIs; Taylor et al., 2017). Despite an increased awareness of mTBIs (i.e., concussions), this number likely remains an underrepresentation of the actual number of mTBIs because of a lack of appropriate evaluation and follow-up with health care professionals (Prince & Bruhns, 2017). As awareness and identification of mTBI grows, the number of people requiring rehabilitation intervention will rise (Choe & Giza, 2015).

People with mTBI may acutely experience— separately or in any combination—somatic (e.g., dizziness), cognitive (e.g., decreased attention), and affective symptoms (e.g., irritability; Bergersen et al., 2017). Notably, the interactive effects of multiple symptoms can amplify the total impact on everyday life (e.g., lack of sleep may lead to reduced attention; Prince & Bruhns, 2017). These symptoms lead to decreased academic performance, difficulty returning to work, and an overall struggle in life participation (Chu et al., 2017; Cooksley et al., 2018). These limitations are of relatively minor consequence to the majority of people with mTBI, whose symptoms resolve within 10 days postinjury (McCrory et al., 2013). However, approximately 10% to 15% of people with mTBI have persisting symptoms and associated functional decrements for many months or years; in other words, persistent concussive symptoms (PCSs; Zemek et al., 2016).

Of the limited existing literature on PCS treatment, the majority of interventions have a primary goal of improving specific impairments. However, a focus on impairment remediation may be short-sighted, because improvements in impairments often are not sustained and largely do not translate to improvements in daily life functioning (Cicerone et al., 2011; Poppelreuter et al., 2009). A 2019 review of PCS clinical trials identified a small amount of studies that used cognitive–behavioral therapy to improve symptom and psychological outcomes, with positive results (Jaber et al., 2019; Kjeldgaard et al., 2014). The review found that interventions with the goal of improving symptoms or psychological outcomes (i.e., bottom-up interventions) did not explicitly target occupational performance outcomes.

In contrast, top-down interventions are used in an attempt to improve performance at the level of everyday activity. For example, metacognitive strategy training (MCST) interventions help clients enhance control over higher level cognitive functions (i.e., executive function) to overcome cognitive, emotional, and behavioral obstacles for improvement in occupational performance. The Cognitive Orientation to daily Occupational Performance (CO-OP) intervention, a type of MCST, is an evidence-based intervention for improving occupational performance poststroke. CO-OP uses a global problem-solving strategy: Goal–Plan–Do–Check. Use of a broadly applicable strategy in the context of meaningful occupations ensures that the intervention is salient to the client and increases the likelihood of skill transfer (Dawson et al., 2017). Skill transfer reflects optimal learning and refers to the ability to use a previously learned skill within a new context or task (Geusgens et al., 2007). CO-OP therapists use guided discovery methods, such as asking a series of probing questions, to support participants in analyzing their own task performance and generating potential solutions for improving performance. Strategy training interventions, such as CO-OP, use metacognition and executive functioning to teach clients how to approach everyday life difficulties, recognize performance barriers, and generate potential solutions.

Current evidence supports the use of CO-OP for addressing occupational performance in acquired brain injury populations (Dawson et al., 2009, 2013; McEwen et al., 2015; Polatajko et al., 2012). Previous literature has demonstrated skill transfer with improvement in activities that were and were not addressed in intervention (Dawson et al., 2009, 2013; McEwen et al., 2015). CO-OP has also been applied, to a limited extent, to people with PCS in two prior studies. Hunt and colleagues (2019) found CO-OP to be feasible for use in adolescents with PCS. An additional study with a single-case design (n = 3) piloted CO-OP with participants with TBI; one of the participants presented with PCS and demonstrated occupational performance improvements (Dawson et al., 2009). Evaluation of the feasibility of CO-OP in adults is warranted because of the differing occupations of adulthood (Wrzus et al., 2013) and the likely differential impact of PCS symptoms on occupational performance in adults as compared with that in adolescents. Literature is lacking on how best to enable return to participation in adults with PCS; however, existing literature suggests that MCST may be a viable option. In the present study, we sought to evaluate the feasibility of CO-OP in a sample of adults with PCS and to evaluate the preliminary effect on outcomes of occupational performance, cognition, and PCS symptoms.

A single-group, prospective design was used, with outcomes gathered pre- and postintervention. A small-scale design for feasibility evaluation is recommended early in intervention evaluation to be fiscally conservative and to inform intervention refinement should a large-scale study be warranted (Bowen et al., 2009). All study procedures were approved by the University of Missouri Institutional Review Board. All participants provided written informed consent.

Participants (n = 15) were recruited through colleague referral and through a university-based e-mail communication that reaches the campus and members of the community. Interested people contacted researchers by phone or e-mail and were asked to complete a prescreening survey. Those who met the initial eligibility criteria completed screening procedures (i.e., Patient Health Questionnaire–9; PHQ–9) and baseline testing in a university research laboratory. Participants were included if they had

• ▪ a concussion diagnosis,

• ▪ one or more postconcussive symptoms for 4 wk or more that inhibit performance of daily activities per self-report, and

• ▪ ages 18 to 60 yr.

Participants were excluded if they had

• ▪ severe neurological or psychiatric conditions;

• ▪ severe depressive symptoms (a score over 21 on the PHQ–9; Kroenke et al., 2001);

• ▪ an inability to read, write, or speak English fluently; or

• ▪ a lack of transportation.

We evaluated intervention acceptability through administration of the Client Satisfaction Questionnaire– 8 after intervention completion (Table 1). Feasibility outcomes also included recruitment, retention, and intervention adherence rates. Each feasibility outcome was chosen on the basis of recommendations of Bowen et al. (2009) and Tickle-Degnen (2013). Acceptability feasibility ensures that a proposed intervention is perceived as satisfactory by the intended population before large-scale employment (Bowen et al., 2009).

Assessment data were gathered by a trained rater who was blind to the study purpose. We evaluated occupational performance using the Canadian Occupational Performance Measure (COPM; Law et al., 2014). We evaluated the functional impact of PCS symptoms by using the Neurobehavioral Symptom Inventory (NSI; King et al., 2012). The impact of sleep and visual symptoms was evaluated further with the Pittsburgh Sleep Quality Index (PSQI; Carpenter & Andrykowski, 1998) and the College of Optometrists in Vision Development–Quality of Life Outcomes Assessment (COVD–QoL; Mozlin, 1995). Finally, we assessed cognition using the National Institutes of Health (NIH) Toolbox Cognition Battery (Carlozzi et al., 2017,; Weintraub et al., 2013), the Weekly Calendar Planning Assessment (WCPA; Weiner et al., 2012), and the Dysexecutive Questionnaire (DEX; Burgess et al., 1998; Wilson et al., 1996). Each of these measures has demonstrated adequate reliability and validity; supporting citations are provided in the following text.

The COPM is a semistructured interview guide for evaluating occupational performance within self-care, leisure, and productivity domains (Law et al., 2014). COPM results may be used to set goals and measure change over time. Within this study, the COPM was used to identify a minimum of five occupational performance goals. For each goal, perceived performance and satisfaction levels are separately rated on a scale of 1 to 10, with higher scores indicating greater levels of performance and satisfaction.

The NSI is a 22-item self-report measure of neurobehavioral symptoms intended to track symptoms over time in the TBI population (Cicerone & Kalmar, 1995). The degree to which each symptom disturbs functioning over the prior 2 weeks is rated using a 5-point Likert scale ranging from 0 (none) to 4 (very severe). Meterko et al. (2012) identified a four-factor model with the following domains: affective, somatosensory, cognitive, and vestibular.

The PSQI is a 24-item self-report measure of sleep dimensions, including quality, latency, duration, efficiency, medication use, and daytime dysfunction (Buysse et al., 1989; Carpenter & Andrykowski, 1998). A global score is derived from summing each of the sleep dimension scores, with higher scores indicating worse sleep and a score of more than 5 representing poor sleep quality.

The COVD–QoL is a 30-item self-report measure of the impact of visual symptoms on quality of life (Maples, 2000; Mozlin, 1995). Each item is rated according to symptom frequency with a 5-point Likert scale ranging from 0 (never) to 4 (always). There are five subscales, including accommodation, binocularity, orientation, oculomotor, and perception, with total scores ranging from 0 to 120.

The DEX is a 20-item self-report measure of executive function in daily life (Burgess et al., 1998; Wilson et al., 1996). Items cover four domains: emotional, motivational, behavioral, and cognitive. Each item is rated for frequency using a 5-point Likert scale ranging from 0 (never) to 4 (very often), with total scores ranging from 0 to 80.

The NIH Toolbox Cognition Battery is a series of cognitive tests that evaluates language, episodic memory, executive function, working memory, and processing speed (Weintraub et al., 2013). In this study, we used the iPad application for administration. We report the fully corrected T-score values with a normative mean of 50 (SD = 10). Specifically, we report the fluid cognition composite score, which reflects abilities related to problem-solving, abstraction, and adaptation to novel situations (Horn & Cattell, 1966).

The WCPA is a performance-based test of executive function in which a series of appointments are sorted into a blank weekly calendar (Toglia, 2015; Toglia et al., 2017; Weiner et al., 2012). Some appointments are required to be entered into fixed times, and others are not. During the assessment, five rules must be followed (e.g., do not answer questions from the examiner). This study used the Level 3 version for adults and for older adults in which the appointments and errands to be entered are provided in paragraph form. Performance on the WCPA can be evaluated using appointment accuracy, efficiency, planning time, total time, number of rules followed, and number and types of strategies used.

Participants engaged in 10 weekly, 45-min CO-OP sessions with a trained interventionist. Five goals were identified using the COPM. One goal was not discussed within the intervention (i.e., the untrained goal), to serve as a measure of skill transfer. CO-OP is a metacognitive strategy training intervention with a primary purpose of improving occupational performance (Dawson et al., 2017). The client and therapist collaborate to identify areas of performance breakdown within client-chosen goals. The foundation of the CO-OP process is the learning of the Goal–Plan– Do–Check metacognitive strategy. Client use of this broad strategy can increase the likelihood of skill transfer, because it can be applied in a range of occupations and environments. Clients identify a specific goal to address and then develop a plan to address areas of performance breakdown. The plan is then trialed in the do phase within the intervention session as appropriate and in everyday life. Finally, the client checks whether the plan was successful or whether modifications are needed for improved performance (Figure 1). The Goal–Plan–Do–Check global strategy is used to develop more tailored, domain-specific strategies that are used within the Plan. During this process, the therapist facilitates gradual client independence in application of the Goal–Plan–Do–Check strategy, including opportunities for generalization of learned cognitive strategies through a process called guided discovery. Guided discovery takes the form of methods such as asking a series of probing questions or modeling for the client. For a more thorough description of the CO-OP Approach^TM, please see prior CO-OP literature (Dawson et al., 2017; Polatajko et al., 2001).

We calculated descriptive statistics for sample demographics, intervention acceptability, and performance at pre- and postintervention for each preliminary effect outcome measure. Additionally, recruitment rate, retention rate, and adherence rate were calculated as outlined in Table 1. To determine change from pre- to postintervention, we calculated effect sizes (Hedge’s g) with associated 95% confidence intervals. Hedge’s g was calculated on the basis of a paired sample calculation in which the mean change from pre- to postintervention is divided by the standard deviation of the difference. Hedge’s g is a more conservative estimate of effect for small sample sizes (Borenstein et al., 2021). Effect sizes were interpreted as 0.2 indicating small; 0.5, medium; and 0.8, large (Cohen, 1988). Null hypothesis significance testing was not conducted, because it is not recommended for feasibility designs (Bowen et al., 2009; Lancaster et al., 2004).

A total of 14 participants with PCS completed the study (Table 2). The sample was predominately female (64.28%) and White (71.42%). The majority of the sample had experienced multiple concussions (M = 3.0, SD=1.66), and reported mild depressive symptoms.

There was a recruitment rate of 32% and a retention rate of 93% (Figure 2). There was a 100% adherence rate to intervention sessions. Finally, CO-OP was seen as highly acceptable as indicated by an average score of 3.85 (SD = 0.25) on the Client Satisfaction Questionnaire–8. All feasibility criteria set a priori were met, with the exception of recruitment rate, which fell short of the intended benchmark by 17% (Figure 2).

From pre- to postintervention, large improvements were demonstrated for performance (Hedge’s g = 1.97) and satisfaction (Hedge’s g = 2.59) scores for trained COPM goals (Table 3). Large improvement was also noted for satisfaction (Hedge’s g = 0.81), and medium improvement was noted in performance levels (Hedge’s g = 0.58) for untrained goals. Overall, larger gains were observed for goals that were directly trained compared with untrained goals. Large effects were noted for the affective and cognitive domains of the NSI (Hedge’s gs = 1.14 and 0.85, respectively), sleep outcomes per the PSQI (Hedge’s g = 0.87), and perceived executive function in daily life per the DEX (Hedge’s g = 0.80).

Medium-sized effects were observed for visual symptoms in daily life (Hedge’s g = 0.54). The sample demonstrated large gains in fluid cognitive abilities on the NIH Toolbox (Hedge’s g = 1.08). Changes were negligible for functional cognition, as measured by the WCPA accuracy scores, time to completion, or efficiency; however, medium effects were present for time spent planning before the task (Hedge’s g = 0.72) and for the number of rules followed during task performance (Hedge’s g = 0.52).

The primary goal of this research was to evaluate the feasibility of the CO-OP intervention in adults with PCS; secondarily, we sought to determine whether CO-OP elicited positive changes for concussive symptoms, functional cognition, and occupational performance outcomes. Much of the existing literature attempts to improve specific impairments experienced postconcussion and does not adequately account for how clients will engage in daily life activities in the presence of the persistent symptoms or for how participation in daily activities affects symptoms. In contrast, CO-OP is likely to be fitting for this population, because it teaches clients the problem-solving skills needed to grade activity participation and optimize quality of life despite symptoms.

Evaluation of feasibility is indicated for a completely novel intervention or when there is limited literature within a population (Bowen et al., 2009). Study findings supported the high acceptability of CO-OP for meeting the needs of adults with PCS, which is arguably one of the most key feasibility aspects. Participant satisfaction is also reflected in the high retention rate (93%) and adherence rate (100%), which further indicates that participants assigned value to continued study participation. Although the recruitment rate in this study did not meet the established benchmark threshold, compared with other rehabilitation trials, the present study had an adequate recruitment rate (32%; Geed et al., 2021). Collectively, these feasibility findings support an efficacy trial as a prudent next step for addressing occupational performance needs in adults with PCS.

The study sample demonstrated improvements in occupational performance, fluid cognitive abilities, executive functioning in daily life tasks, and affective and cognitive symptoms. Because this study was designed to test feasibility (e.g., small sample size, single group), effect sizes are not generalizable at the population level, but they do suggest that further research is warranted. Similar to samples in prior CO-OP literature (Dawson et al., 2009, 2013; McEwen et al., 2015), this sample experienced improvements in the performance of occupations that were directly trained and untrained; however, improvements on untrained goals were smaller in nature. It is expected that clients will demonstrate greater learning and skill improvement on goals that are directly addressed within rehabilitation, because this allows for actual skill repetition and problem-solving related to potential difficulties that could arise (Geusgens et al., 2007). Improvement on untrained goals requires more independence on behalf of the client; not only must a skill be retained, but the client also needs to recognize when and how to apply the skill within unlearned tasks and contexts. Therefore, although skill transfer is notoriously difficult, intentional structuring of learning experiences is needed to optimize skill transfer as much as possible.

Large effects were also present for fluid cognitive abilities and executive functioning abilities within daily life. Changes in fluid cognitive abilities suggest a potential downstream effect of the behavioral CO-OP intervention on cognitive processes. Prior studies have also found improvements on neuropsychological cognitive outcomes in the chronic stroke population after the CO-OP intervention (Wolf et al., 2016). Although cognitive processes are not the main target of strategy training interventions, gains in fluid cognition is not wholly unexpected, because it overlaps with many of the same cognitive processes engaged by CO-OP (e.g., problem-solving, awareness; Akshoomoff et al., 2013). Observed decreases in affective and cognitive symptoms in everyday life may be the result of participants’ increased ability to problem-solve and self-modify participation in activities in a way that minimizes symptom elicitation.

Study findings should be interpreted with reasonable caution because of several factors. The majority of our sample were undergraduate and graduate college students and may not be representative of the population. Additionally, this study was conducted in a university research setting and may not fully reflect constraints that are present in a clinical environment.

The findings of this study have the following implications for occupational therapy practice:

• ▪ People with ongoing concussive symptoms perceive CO-OP as appropriate for addressing their daily life needs.

• ▪ The use of the CO-OP to address postconcussive symptoms appears to be feasible for occupational therapy; however, future efficacy trials are required.

Results suggest that, in the PCS population, CO-OP is feasible to administer and acceptable to the client. The occupation-based and client-driven nature of CO-OP aligns well with functional limitations and needs of those with PCS. CO-OP may improve occupational performance outcomes and PCS symptoms. Future work will verify the efficacy of CO-OP within PCS, including whether gains are retained over time.

We thank the University of Missouri Research Council for funding of this work and the participants for their engagement in this study.