Abstract
Background.
Stroke rehabilitation includes both restorative and adaptive interventions. There are few specific guidelines regarding adaptive interventions.
Purpose.
To determine the effectiveness of adaptive interventions on participation outcomes generally, and more specifically by theoretical focus of the intervention.
Method.
Systematic review. Relevant databases were searched for randomized controlled trials of adaptive interventions that included a participation measure. Data were extracted and Hedges’ g was calculated for all interventions and then by intervention classification.
Results.
Fourteen named interventions were examined in 24 studies. When all interventions were considered, the following was observed: a medium effect on person-specific participation goals that approached significance (0.60; [95% CI: −0.06; 1.26], p = .07), a negligible and non-significant effect on broad participation (0.10; [95% CI: −0.13; 0.33], p = .37), and a small to medium significant effect on instrumental activities of daily living (IADL; 0.37, [95% CI: 0.12; 0.62] p = .004). Studies evaluating person-specific participation were dominated by learning-focused interventions, while studies evaluating broad participation and IADL were dominated by coping-focused interventions.
Conclusion.
Learning- or motivation-focused approaches appear to have an important impact on person-specific participation goals. Future research should focus on clarifying the effectiveness of these interventions and improving impact on broader participation.
Introduction
There are two broad types of interventions in stroke rehabilitation: restorative and adaptive (Wolf et al., 2012). Restorative interventions target impairment and seek to normalize body function and structures (Cumberland Consensus Working Group et al., 2009). Adaptive interventions are focused on efficient return to valued activities and social roles and seek to adapt the environment or the activity to allow for participation (Cumberland Consensus Working Group et al., 2009). According to the International Classification of Functioning, Disability and Health (ICF; World Health Organization, 2002), participation relates to performance of an activity in a real-life situation. Adaptive approaches aim to improve participation rather than body structures and functions. Restorative exercise interventions, while important for decreasing impairment, have limited long-term impact on improving participation (Beauchamp et al., 2017). Adaptive interventions, therefore, may be particularly important in improving participation.
The Canadian Stroke Best Practice Recommendations: Rehabilitation, Recovery, and Community Participation following Stroke (Mountain et al., 2020; Teasell et al., 2020) include specific guidance for restorative interventions, such as exercise and constraint-induced movement therapy, but no specific recommendations for adaptive interventions. This could lead therapists to inaccurately conclude that specific adaptive approaches do not exist or have not been evaluated (Egan et al., 2020). This is not the case. Egan et al. (2024) performed a scoping review focused on adaptive interventions that included metacognitive or self-management components. They identified 29 named interventions examined in 77 studies. Analysis of explicit and implicit discussion of theory behind these interventions led to the identification of three overarching theoretical drivers of these interventions: learning, motivation, and coping.
Interventions focused on learning are guided by Learning Theory and Cognitive Learning Theory (McEwen et al., 2009). The idea underlying these interventions is that people acquire the ability to do things through a thinking and experiential process (metacognition), that begins with setting a goal, planning a strategy to achieve the goal, experimenting with the strategy, and then evaluating the results (McEwen et al., 2009). Examples of learning-driven interventions include the Cognitive Orientation to daily Occupational Performance (CO-OP; Skidmore et al., 2017), and Goal Management Training (Robertson, 1996).
Interventions based on motivation are guided by Self-determination Theory, Social Cognitive Theory, and adult learning theories (Bandura, 1998; Collins, 2004; Ryan & Deci, 2019). These interventions are based on the idea that motivation is harnessed when people work to re-engage in activities that meet needs for autonomy, relationship and competence; this motivation allows people to persist in problem-solving when challenges arise (Egan et al., 2024). Examples of motivation-driven interventions include Solution Focused Brief Therapy (Ratner et al., 2012), and Occupational Performance Coaching (OPC; Kessler et al., 2018).
Finally, interventions focused on coping are guided by Social Cognitive Theory (Bandura, 1998) and Psychological Stress Coping Theory (Lazarus & Folkman, 1984). These interventions view improvement in self-efficacy as the dominant mechanism of action, strengthening people's belief that they can cope with the effects of illness and carry out their valued activities and social roles. This strengthened belief supports persistence in problem-solving and experimentation required to improve engagement in valued activities and social roles (Bandura, 1998; Lazarus & Folkman, 1984). Some examples of coping-driven interventions include Restore4Stroke (Tielemans et al., 2014), the Bridges Stroke Self-Management Program (SSMP; Cadilhac et al., 2016), and Improving Participation After Stroke Self-management Program (IPASS; Wolf et al., 2016).
Other classification systems have appeared in systematic reviews. For example, Kossi et al. (2024) examined the effect of different types of rehabilitation on participation outcomes among people who had experienced a variety of neurological diagnoses. Both restorative and adaptive approaches were included. Interventions were classified as cognitive behavioral therapy (CBT), education, exercise, CBT plus exercise, and multidisciplinary. With its grounding in exercise and CBT, this classification is limited in its usefulness in classifying occupational therapy intervention. For example, CO-OP was classified as CBT plus exercise (Kossi et al., 2024).
The theoretical drivers identified by Egan et al. (2024) represent a new way of looking at adaptive interventions. While this system has not yet been applied in systematic reviews, there are reviews that align with these classifications. For example, two systematic reviews have considered adaptive interventions with a learning or coping theoretical driver.
Kiriakou and Psychouli (2024) carried out a systematic review on one learning-based intervention, CO-OP. While they did not perform a meta-analysis, they concluded that there was support for its effectiveness overall. Swanton et al. (2020) carried out a systematic review of cognitive strategy training. This category included interventions classified by Egan et al. (2024) as learning- or motivation-driven adaptive interventions on activities or participation. Swanton et al. (2020) demonstrated a positive effect on activity performance after stroke, with the standard mean difference being 0.75 (95% CI: 0.41; 1.10). However, these results combined measures of participation with measures of activities of daily living, making it difficult to determine the impact solely on participation.
There are several reviews of self-management interventions; these interventions are classified by Egan et al. (2024) as coping-focused. In terms of recent reviews, one considered only instrumental activities of daily living (IADL), rather than participation more broadly, and found a small effect that approached significance (Hedges’ g: 0.21; 95% CI: −0.03; 0.85; Oh et al., 2022). As well, Lau et al. (2022) considered the impact of interventions on functional independence but not participation. A new review could provide updated relevant estimates and resolve these differing conclusions, while indicating whether there is clear superiority among the learning-, motivation-, or coping-focused approaches.
The main objective of this systematic review and meta-analysis was to determine whether adaptive interventions, with metacognitive or self-management components, are effective in enhancing participation after stroke. The primary research question was: For people who have experienced stroke, are adaptive interventions with metacognitive or self-management components more effective than usual care or control interventions in improving participation in valued activities and social roles? The secondary research question was: Is one type of adaptive intervention (learning-, motivation-, or coping-focused) more effective than another?
Methods
Eligibility Criteria
Eligible studies included randomized controlled trials (RCTs) of adaptive interventions that were explicitly or implicitly guided by a learning-, motivation-, or coping focus. Participants were adults (≥18 years old) who had experienced stroke. Interventions could have been based in any setting, including hospital, home, or outpatient clinic. Finally, studies were included if a measure of participation was used. In their scoping review of adaptive interventions, Egan et al. (2024) noted that evaluation of participation was carried out using three distinct types of measures. They referred to these as measures of person-specific participation, broad participation, or IADL. Measures of person-specific participation evaluate performance in activities related to the specific, unique, occupation-based therapy goals of the individual. Measures of broad participation evaluate predefined generic groups of outcomes included in the ICF schema of activities and participation (World Health Organization, 2002). Measures of IADL evaluate generic activities related to housework and home upkeep.
Potential studies were excluded if the intervention focused only on the provision of generic stroke education, reduction of specific impairment (e.g., cognition), or improved adherence to medical management. Studies with interventions that were not explicitly or implicitly guided by a learning, motivation, or coping focus were also excluded. As well, studies that evaluated participation using an overall measure of quality of life that included some questions related to participation were excluded. Studies with samples that included people with conditions other than stroke (e.g., traumatic brain injury), studies of children (<18 years old), and studies written in a language other than English or French were also excluded, as were studies reported in abstracts only.
Information Sources and Search
The search was led by a health sciences research librarian (LS) and followed the principles of PRISMA-Search (Rethlefsen et al., 2021). Databases included Medline and Medline in Process via Ovid, Embase Classic + Embase via Ovid, APA PsycINFO via Ovid, and CINAHL via EBSCOHost. The search strategy was developed in Medline and then translated into the other databases (see Appendix A in the supplementary data). An RCT filter was used in the Medline, CINAHL, and Embase databases. The Medline search strategy was peer reviewed by Nicholas Dehler (Research Librarian, Health Sciences and Medicine) using the Peer Review of Electronic Search Strategies tool. All databases were searched from dates of inception to May 29, 2023. All references were entered into an Endnote file for processing, then uploaded into Covidence for duplicate removal and subsequent screening. Reference lists of literature, systematic, and scoping reviews and studies selected for full-text review were searched for potential additional studies.
Study Selection
Two reviewers (KD and ME) independently screened titles and abstracts, then obtained and assessed full texts for inclusion. Reasons for exclusion of papers after full-text review were recorded. In the event of disagreements between the two reviewers, a third reviewer (DK) was consulted.
Data Extraction and Critical Appraisal
The main outcome was participation, specific to participant-identified goals (person-specific participation) or considered more generically (broad participation). A secondary outcome was IADL, as these tend to be classified as participation (World Health Organization, 2002). The data were extracted by two of three reviewers working independently (KD, NGJ, and ME) using an Excel spreadsheet created a priori. Study authors were contacted for additional data where necessary.
Quality of included studies was evaluated using the JBI critical appraisal tool for the assessment of risk of bias in RCTs (Barker et al., 2023). Studies were considered higher quality if they attained a positive rating for questions 1 (true randomization), 2 (concealed allocation), 6 (blinded assessors), and 10 (the outcome was measured similarly in both the treatment and control groups). Two of the three reviewers (NGJ, KD, and ME) carried out these assessments independently. Disagreements were resolved by a fourth reviewer (DK).
Categorization of Interventions
Study interventions were classified as learning-, motivation-, or coping-focused in one of two ways. First, for studies of interventions included in the earlier scoping review (Egan et al., 2024), the classification determined in this review was maintained. Second, for newly identified interventions, the classification was determined by two authors (DK and ME) working independently who reflected on theory noted explicitly or implicitly in the introduction, methods, or discussion. In cases of disagreement, a third reviewer (EL) was consulted, and a decision was made by consensus. All categorizations were carried out prior to data synthesis.
Data Synthesis
Meta-analyses were performed for studies first by outcome and then by outcome and intervention classification. Standardized mean differences (Hedges’ g) were calculated for the difference in mean change from baseline to posttreatment scores between control and intervention groups, along with the 95% confidence intervals and heterogeneity statistics (tau-squared and I-squared). When standard deviations for mean change from baseline were not reported, these were calculated using p-values, t-values, confidence intervals, and standard errors. If this was not possible, they were imputed using a conservative correlation coefficient of 0.5 (Higgins et al., 2024; Pearson & Smart, 2018). Analyses were conducted in R (package ‘meta’; Balduzzi et al., 2019). Estimates of effect were categorized using Brydges (2019) guideline where 0.16 is considered small, 0.38 medium, and 0.76 large.
Results
Study Selection
The search resulted in 1822 documents, of which 1354 were unique. Title and abstract screening led to the exclusion of a further 1232 studies. Where an excluded document was a research abstract, attempts were made to locate the full report of the study (see Appendix B in the supplementary material). This left 122 studies for which the full text was obtained and reviewed. Ninety-eight were excluded, resulting in the 24 studies in this review (Figure 1).

PRISMA flow chart of study process.
Study Characteristics
The 24 included RCTs were published between 2011 and 2023 (Table 1).
Study Characteristics and Summary of Findings
Types of Interventions
The most commonly named interventions were coping-focused (10 interventions, 14 studies), followed by motivation-focused (three interventions, four studies), and then learning-focused (one intervention, six studies; Table 1). Popular coping-focused interventions included the Bridges Self-Management Program, Improving Participation after Stroke Self-Management Program (IPASS), and Restore4Stroke. Among motivation-focused interventions there were two studies regarding Taking Charge After Stroke (TaCAS). All the studies of learning-driven interventions evaluated CO-OP.
Outcomes
Person-Specific Participation
Seven studies included person-specific participation measures, including the Canadian Occupational Performance Measure (COPM) and Goal Attainment Scaling. All the learning-focused intervention studies and one of the motivation-focused intervention studies included a measure of person-specific participation. When changes between baseline and postintervention in all studies were considered, there was a medium effect that approached significance (0.60 [95% CI: −0.06; 1.26]; p = .07). When studies were considered by theoretical driver of the intervention, there were medium but non-significant effects for both learning-focused (0.58 [95% CI: −0.19; 1.35], p = .14) and motivation-focused interventions (0.68 [95% CI: −0.27; 1.62] p=0.16; Figure 2).

Forest plots by intervention focus and outcome.
In several studies, results included different periods of postintervention follow-up. However, the follow-up periods were so varied that summary estimates could not be calculated. There were three studies of CO-OP (a learning-focused intervention) in which specific participation was considered at 1- or 3-month follow-up. In two studies (McEwen et al., 2015; Poulin et al., 2017) the intervention group continued to have better outcomes. One study of a motivation focused (Kessler et al., 2017) included follow-up at 6 months. In this study, intervention participants did not continue to have better specific participation results at follow-up.
Broad Participation
Fourteen studies included a measure of broad participation: three of the learning-focused intervention studies, three of the motivation-focused intervention studies, and eight of the coping-focused intervention studies. Nine studies were included in the meta-analysis. Results demonstrated a negligible and non-significant effect on broad participation (0.10 [95% CI: −0.13; 0.33], p = .37). When studies were considered by theoretical driver of the intervention, there was no effect for learning-focused interventions (−0.00 [95% CI: −0.45; 0.45], p = .999), a medium negative but non-significant effect for motivation-focused interventions (−0.70 [95% CI: −1.64; 0.24], p = .15), and a small nonsignificant effect for coping-focused interventions (0.19 [95% CI: −0.09; 0.47], p = .17) (Figure 2).
Three motivation-driven interventions studies included outcomes at 6-month follow-up (Fu et al., 2020; Kessler et al., 2017) and 12-month follow-up (Harwood et al., 2012) and the results were mixed. Among the studies of coping-based interventions, three measured broad participation at 6 months and none found differences between the intervention and control groups (Cadilhac et al., 2011; Forster et al., 2023; Hill et al., 2019). One study measured broad participation at 3-month follow-up and 9-month follow-up (Tielemans et al., 2015) and found a significant difference favoring the treatment group only at 9 months.
Instrumental Activities of Daily Living
A measure of IADL was included in four studies and all evaluated coping-focused interventions. Meta-analysis revealed a small to medium significant effect on IADL (0.37 [95% CI: 0.12; 0.62], p = .004; Figure 2).
Critical Appraisal of Studies
Learning-focused interventions met between seven and 11 of the 13 JBI quality indicators, with two of the six (33.3%) studies considered higher quality. Motivation-focused interventions met between nine and 11 of the 13 JBI quality indicators, with all studies being higher quality. Finally, coping-focused interventions met between eight and 12 of the 13 JBI quality indicators, with six of the 14 (42.9%) studies considered higher quality (see Appendix C in the supplementary material).
Discussion
We identified 24 RCTs that included evaluation of changes in participation following 14 adaptive interventions with metacognitive or self-management components. When these studies were considered within the learning-, motivation- and coping-focused framework, two observations stood out. First, how participation was evaluated differed by intervention driver. Evaluations of learning- and motivation-driven interventions tended to use measures of specific participation, such as the COPM. In contrast, evaluations of coping-focused interventions tended to use measures of broad participation. Second, RCTs of coping-focused interventions were much more common than RCTs of learning- or motivation-focused interventions.
Neither of these differences is surprising. Learning- and coping-focused interventions are predominantly delivered and evaluated by occupational therapists. As personally valued occupation is the primary concern of occupational therapists (Egan & Restall, 2022), these professionals would be more likely to focus interventions on personally valued activities, that is, specific participation. Coping-focused interventions are delivered and evaluated by a range of healthcare providers, including occupational therapists. One would therefore expect these studies to be more numerous and to be evaluated using more generic measures of participation.
When considered altogether, learning- and motivation-focused interventions appear to have a medium effect on person-specific participation goals. It is impossible to determine the impact of coping-focused interventions on person-specific participation as no studies included this type of assessment.
Neither learning-, motivation-, or coping-focused interventions appeared to affect broad measures of participation. This is consistent with the results of recent, related reviews (Kiriakou & Psychouli, 2024; Kossi et al., 2024; Oh et al., 2022; Swanton et al., 2020). It may be that measures of broad participation are not responsive enough to register change (Magasi & Post, 2010). It is also possible that actions at the micro-context level, that form the basis of these interventions, are inadequate to effect changes in broad participation without accompanying actions at the meso- or macro-context level (Della Vecchia et al., 2021). This is an important area for future research.
While none of the interventions appeared to have a substantial impact on broad participation, learning- and motivation-focused interventions do appear promising for improving person-specific participation. The current study builds on previous systematic reviews, which supported the effectiveness of learning-driven interventions for person-specific participation outcomes (Kiriakou & Psychouli, 2024; Swanton et al., 2020), while providing an estimate of effect.
Several coping-focused studies included a measure of IADL, and these indicated a small to medium effect on IADL. IADL may be an aspect of participation that can be influenced by coping-focused outcomes. Future studies on coping-focused interventions could benefit from examining IADL and broad participation separately to determine if these interventions have distinct impacts on IADL and broad participation.
This systematic review demonstrates that learning- and motivation-focused interventions are more effective than usual care in helping people move towards the participation goals that they have identified. These person-specific participation goals are an important consideration for rehabilitation practice given that long-term difficulties with participation after stroke are common and returning to valued occupations is important for individuals with stroke and their families (Abrahamson & Wilson, 2019; Guo et al., 2021). Notably, even when impairment levels are controlled for, improved participation is related to future quality of life (Slenders et al., 2024) and protective against symptoms of depression (Egan et al., 2014). We therefore recommend that learning- and motivation-focused interventions be adopted by occupational therapists working in stroke care.
The current study provides important information that could assist therapists in selecting effective adaptive interventions to improve participation. Given the medium effect of CO-OP and OPC, occupational therapists are encouraged to implement these approaches with individuals who have experienced stroke. While only one study of OPC was included in the meta-analysis, the study effect was very similar to the overall effect of learning interventions. Further research should contrast these approaches, particularly in terms of patient experience.
The improvement in person-specific participation seen after learning- and motivation-focused interventions does not seem to translate more broadly to broad participation, highlighting an important avenue for future research. Questions to be addressed include why changes in person-specific participation do not translate to broader participation and whether the inclusion of aspects of coping interventions, which provided small but significant improvements in broad participation, may be helpful.
Finally, while the Canadian stroke best practice recommendations clearly state that stroke survivors’ goals be ascertained, they provide no guidance regarding specific interventions that might help patients achieve their participation-focused goals (Teasell et al., 2020) This systematic review demonstrates that there is randomized controlled trial evidence that learning- and motivation-focused interventions can be helpful in the attainment of patient-identified participation goals. However, the current structure of the guidelines does not contain a category for the inclusion of this evidence (Teasell et al., 2020). We suggest that recommendation categories be revised to allow inclusion of evidence regarding interventions that promote participation goals.
Limitations of the Study
This study had one main limitation. Due to variations in the reported data, not all studies could be included in the meta-analysis. Efforts were made to locate missing data, but these were not always successful.
Conclusion
This systematic review demonstrates that learning-focused interventions likely have a medium effect on person-specific participation goals and coping-focused interventions have a small effect on IADL. None of the interventions appeared to improve broad participation. Further research should be directed towards firmly establishing the effectiveness of learning- and motivation-focused interventions on specific participation and determining how improvements in person-specific participation can better translate to broader participation.
Key Messages
Adaptive interventions exist across three areas of focus—learning, motivation, and coping—and their impact on participation has been tested in randomized controlled trials.
Adaptive interventions are effective for improving people's performance in their unique and specific participation goals and IADL, but have little impact on broader participation outcomes.
Future research should determine the link between learning- and motivation-focused adaptive interventions and broad participation outcomes.
Supplemental Material
sj-docx-1-cjo-10.1177_00084174251352305 - Supplemental material for Adaptive Interventions for Enhancing Participation Poststroke: A Systematic Review and Meta-Analysis
Supplemental material, sj-docx-1-cjo-10.1177_00084174251352305 for Adaptive Interventions for Enhancing Participation Poststroke: A Systematic Review and Meta-Analysis by Mary Egan, Dorothy Kessler, Patrick Duong, Kaitlyn Duquette, Nalia Gurgel-Juarez, Lindsey Sikora and Elizabeth Linkewich in Canadian Journal of Occupational Therapy
Footnotes
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
ORCID iDs
Supplemental Material
Supplemental material for this article is available online.
Author Biographies
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
