Abstract
This study produced a collection of feedback statements that might be useful in occupation-based interventions to create a sense of agency and responsibility among patients with Parkinson’s disease when performing daily or social occupations.
Participating in daily and social occupations is crucial for health, quality of life, and well-being (Steger et al., 2008). People with neurological conditions encounter limitations in their occupational participation (Ekstam et al., 2011; Goverover et al., 2015; Isaksson et al., 2007; Kennedy-Behr & Hatchett, 2017; Nyman et al., 2021; Sandstedt et al., 2018). The ability to perform occupations and a sense of agency (SoA) are inextricably linked (Lindström et al., 2013). SoA, defined as the experience of controlling an external event through one’s own activities, is disrupted among people living with a neurological condition (Haggard, 2017; Haggard & Eitam, 2015; Khalighinejad & Haggard, 2015; Zito et al., 2020). For example, people living with Parkinson’s disease (PD) have a variety of deficits in emotional and cognitive domains, as well as with sensorimotor integration (Lewis & Byblow, 2002; Saito et al., 2017). Considering the impairment of sensorimotor integration among people with PD, it has been argued that they also experience alterations in action awareness (i.e., in feeling responsible for voluntary and goal-directed activities and their consequences; Haggard, 2017; Saito et al., 2017; Yano et al., 2020). As PD progresses, people with the disease experience difficulty performing daily occupations, reduced quality of life, and restricted social participation (Eggers et al., 2018; Thordardottir et al., 2014). Responsibility is heavily reliant on the neural circuits that underpin SoA (Haggard, 2017). Evidence suggests that SoA is associated with a sense of responsibility, which is essential when performing goal-directed occupations (Caspar et al., 2016; El Zein et al., 2019; Khalighinejad & Haggard, 2015).
According to Kielhofner’s Model of Human Occupation (MOHO), the client’s sense of responsibility should be taken into account during the intervention process (Kielhofner, 2002; Taylor, 2017). When people engage in an occupation based on their motivation, values, interests, and habits, they begin to develop occupational identity and competence (Kielhofner, 2002). When habits are lost as a result of an illness, occupational identity and SoA may be disrupted, leading to problems meeting role-related responsibilities (Maersk et al., 2021). People living with PD struggle to maintain their roles and experience role changes after their diagnosis (Benharoch & Wiseman, 2004; Thordardottir et al., 2014). Disruptions in role-related responsibilities are associated with SoA and contribute to decreased quality of life (Taylor, 2017). Greater SoA and sense of responsibility change health-related behaviors and promote autonomy, self-encouragement, and well-being among individuals (Brown et al., 2019; Normann et al., 2013).
Diverse professions can help individuals with a neurological disorder acquire a stronger SoA (Aoyagi et al., 2019; Kranick et al., 2013; Lindström et al., 2013; Miyawaki et al., 2020; Normann et al., 2013) and, as a result, a greater sense of responsibility when doing daily or social activities (El Zein et al., 2019). Occupational therapy interventions for people living with PD have been recognized (Dixon et al., 2007; Foster et al., 2021; Radder et al., 2017; Sturkenboom et al., 2013, 2014) as maintaining and improving daily routines and roles. However, the concept of responsibility in occupational therapy intervention has yet to be investigated.
Using feedback in the context of rehabilitation is recognized in the literature (Doig et al., 2009; Houle-Johnson et al., 2019; Rendos et al., 2021; Stanton et al., 2015). Extrinsic feedback, which is categorized as knowledge of performance (i.e., information about the quality of activity performance during or after task completion) or knowledge of result (i.e., information delivered at the end of task performance with respect to the goal), may lead to enhanced motor learning outcomes and participation (Subramanian et al., 2010). The provision of meaningful feedback has the potential to cause long-term neurological and behavioral changes (Levin & Demers, 2021). However, to the best of our knowledge, extrinsic feedback statements that incorporate the client’s sense of responsibility have not yet been considered in rehabilitation. Sense of responsibility has a key link to motivation (i.e., a conscious or unconscious internal state that drives one to act) and volition (i.e., the process of initiating and carrying out an activity that is under one’s control) and leads to increased engagement in activities and occupations (Beck et al., 2017; Haggard, 2008; Li et al., 2011). Hence, using extrinsic feedback statements may help people with PD to achieve improved participation, health, and well-being.
The aim of the current study was to reach consensus on a set of extrinsic feedback statements that have the potential to create a sense of responsibility among people with neurological disorders in the course of (before, during, and after) performing daily or social occupations. The evidence generated from this study is essential to applying future occupation-based interventions that are based on the concept of responsibility. For this purpose, we used an international Delphi technique with a heterogeneous group of experts. Additionally, the feedback statements derived from experts were presented to people living with PD to ensure that these statements resonated with them. Delphi methodology is widely used in health care to establish consensus on clinical issues (de Meyrick, 2003). This article outlines the outcomes of this Delphi procedure.
Method
Study Design
The Delphi study was performed in two rounds with experts and one round with people living with PD in compliance with Guidance on Conducting and REporting Delphi Studies (Jünger et al., 2017). The Delphi technique is a structured and iterative process whereby participants reveal their opinions anonymously, get insight into group opinions, and reconsider their answers until an a priori determined level of consensus is reached (Vernon, 2009).
The study was approved (IR.IUMS.REC.1400.753) by the ethical committee of the Iran University of Medical Sciences. All participants consented to take part in the study by choosing “yes” from a multiple-choice menu in response to a question regarding their consent.
Participants
Experts
We sought to identify experts in neurological disorders who practiced in academic or clinical settings. We defined an expert as either a clinician with a special interest in the management of neurological conditions or an academic who holds a tenured or honorary university position who has published at least one peer-reviewed journal article in the field of neurological conditions in the past 5 yr. To identify potential participants, we looked at (1) academic university profiles; (2) mailing lists of five organizations (Association of Occupational Therapists of Ireland, Canadian Association of Occupational Therapists, WA Occupational Therapy Association, Australian Physiotherapy Association, and Occupational Therapy Australia) to identify clinicians with an interest in neurological disorders; and (3) the corresponding or first author of systematic reviews with MeSH headings of neurological rehabilitation, neurologic disorder, Parkinson’s disease, stroke, and multiple sclerosis.
The experts were from a variety of geographical locations: Europe, North America, South America, Australia, Africa, South and East Asia, and the Middle East. They had various professions and included occupational therapists, physical therapists, speech- language pathologists, general practitioners, neuroscientists, neurologists, health researchers, and nurses. As illustrated in Figure 1, 853 potential participants were sent invitation letters. There may have been some overlap in invitations to individuals who were both academics and clinicians. We recruited 100 experts in the first round to reach saturation, based on the recommendations of the Consensus-based Standards for the Selection of Health Measurement Instruments (Terwee et al., 2018). A large sample size is suggested for Delphi studies to improve their quality by reducing group error and enhancing external validity (Stennett et al., 2018).
Flowchart of Delphi experts’ recruitment process.
Patients
From lists of participants in previous research projects, we identified people diagnosed with idiopathic PD according to U.K. Brain Bank criteria (Gibb & Lees, 1988) and Hoehn and Yahr Scale Stages 1 to 5 (Hoehn & Yahr, 1967) who had previously visited Iran University of Medical Sciences movement disorder centers. Ninety-four patients were invited to participate; 73 consented to participate, and 21 declined. Acceptable results can be obtained via a relatively small group of homogeneous participants who are selected on the basis of strict criteria (Akins et al., 2005).
Delphi Methodology
Experts
Feedback statements were developed with a focus on responsibility to oneself, family members, friends, and people in the community (Eliasa, 2014). The terms activity and occupation are used interchangeably depending on country, culture, or author. Therefore, each feedback statement included the terms activity and occupation but did not specify an activity or occupation. (Salles & Matsukura, 2016; White et al., 2014). Feedback statements were written as general statements with no activity or occupation specified. Participants were instructed that statements would be delivered to the client before, during, or after the activity or occupation and that one or more feedback statements might be used for each activity or occupation.
After the generation of a set of 32 feedback statements based on our clinical experience and semistructured interviews with rehabilitation specialists (i.e., occupational therapists, speech-language pathologists, and physical therapists), two professionals conducted a pilot study before the first Delphi round occurred. After minor grammatical changes to the statements in the pilot study, the survey was distributed via Google Forms to all potential experts for the first round.
First, demographic information was collected from all experts, including age, gender, profession, degrees held, years of experience, and region of practice. Next, participants anonymously rated their level of agreement with each survey statement and its effectiveness in creating a sense of responsibility in the course of performing daily or social activities or occupations on a 5-point Likert scale (5 = strongly agree/very effective, 4 = agree/effective, 3 = neither agree nor disagree/uncertain, 2 = disagree/ineffective, 1 = strongly disagree/very ineffective). The experts were then asked to provide additional feedback and comments in response to an open-ended question at the end of the survey. Experts were allowed to contact us if they required further information or clarification. Items on which experts reached consensus in the first round were eliminated from the next round, consistent with the Delphi methodology (Jünger et al., 2012; van Balen et al., 2019). Free-text comments provided by experts were reviewed to revise and rephrase statements for clarity. Furthermore, the comments were synthesized, and 13 additional feedback statements were added to the list of statements in the second round (Che Daud et al., 2015; van Balen et al., 2019).
In addition, to avoid double- or triple-barreled statements (i.e., statements that included two or three outcomes while offering room for only one answer), three statements (21, 24, and 30) from the first round were turned into seven statements on the basis of experts’ opinions. The survey was then sent to the experts who had participated in the previous round. They were allowed to rerate the statements by viewing anonymous response data from the previous round to establish consensus with other experts. To increase the completion rate, reminders were sent after 1 and 2 wk in each round. We used the respondents’ email addresses to determine who had not responded and hence required a reminder. Each round was completed in 4 wk.
Patients
The feedback statements that were prepared and revised in these two rounds with experts were presented to patients via Google Forms to see whether they resonated with them. First, the patients were contacted, the study aims were explained in detail, and a convenient time was scheduled for completion of the survey questionnaire. Next, the concept of responsibility was explained and demographic and PD-related data were collected. Then, patients were asked to rate on a 5-point Likert scale (5 = strongly agree/very effective, 4 = agree/effective, 3 = neither agree nor disagree/uncertain, 2 = disagree/ineffective, 1 = strongly disagree/very ineffective) their level of agreement with and the effectiveness of each statement in creating a sense of responsibility in the course of performing daily or social activities or occupations. Patients were also asked to provide additional feedback in response to an open-ended question given at the end of the survey. They proposed no additional feedback statements. This round was completed in 10 wk.
Eventually, an expert panel of occupational therapists was formed to classify feedback statements on which consensus had been reached, based on patients’ opinions, so that they could be used in the intervention process.
Data Analysis and Delphi Consensus Criteria
Descriptive statistics (Ms, SDs, ranges, frequencies) were used to analyze the demographic data. The median and interquartile range (IQR) for each statement’s ratings on the 5-point Likert scale were calculated. Consensus was set as an IQR of ≤1 and ≥70% agreement (i.e., 70% of participants’ votes) in two adjacent Likert scale categories fall into two Che Daud et al., 2015; Kales et al., 2019; van Balen et al., 2019). There are no firm rules regarding consensus levels for Delphi studies (Powell, 2003; Steele et al., 2008). Hence, we established this level after a literature search that revealed that consensus is typically set between 60% and 90%, depending on the authors (Antonini et al., 2018; Deane et al., 2003). A 70% level of consensus has been successfully used in previous Delphi studies (Antonini et al., 2018; Hasson et al., 2000). A subgroup analysis of experts looked at respondents’ profession, gender, degrees held, region of practice, and years of experience in neurological management. Kruskal–Wallis and Mann–Whitney tests were used to examine the differences in experts’ median ratings. Statistical significance was set at p < .05.
Results
Delphi Participants’ Characteristics
A total of 100 experts from 25 countries contributed to this Delphi consensus study. Of the 100 experts who completed Round 1, 84 completed Round 2. The majority of the experts were female (65%), age >31 yr (M = 42.15 yr, SD = 11.93, range = 24–69), with an average of 17.14 yr (SD = 11.40, range = 2–50) of clinical experience. The mean age of the 73 patients was 66.03 yr (SD = 10.49), and time since diagnosis averaged 9.97 yr (SD = 5.82). Detailed data are presented in Table 1.
Demographic Characteristics of the Participants
Delphi Process
Experts
Round 1
Consensus was reached on the level of agreement with and effectiveness of 18 of 32 statements (56%), with the median ranging from 3 to 5. Double- or triple-barreled statements (Statements 21, 24, and 30) were removed, and Statements 22, 23, 25, 26, 31, 32, and 33 were substituted for the next round. Consensus was not achieved on level of agreement and effectiveness for six statements (Statements 3, 13, 18, 20, 29, and 38). Moreover, consensus was not reached on level of agreement with two statements (4 and 34) or on effectiveness for three statements (5, 6, and 17).
Round 2
Participants were asked to rate the discordant statements from the previous round (i.e., the 11 statements for which consensus was not reached), 13 new statements generated from the free-text comments, and the seven statements substituted for double- or triple-barreled statements. As shown in Table 2, most ratings were very similar to the first round, and medians did not change. Consensus was achieved for all statements presented in this round.
Feedback Statements That Create a Sense of Responsibility in the Course of Doing Daily or Social Activities and Occupations
Note. Statements are rated on a 5-point Likert scale, ranging from 1 (very ineffective/strongly disagree) to 5 (strongly agree/very effective). Agreement and effectiveness are defined as level of agreement with each statement and the effectiveness of each statement in creating a sense of responsibility along performing daily or social activities/occupations, respectively. Dashes indicate statements that were removed; hence, Mdn and IQR were not calculated. IQR = interquartile range; Mdn = median; NA = not available (i.e., either eliminated in Round 2 owing to the reached consensus in Round 1 or being proposed in Round 1 by experts).
Percentage of experts who selected two adjacent categories.
Double- or triple-barreled statements removed in Round 2.
New statements added in Round 1.
In general, 40 and 38 statements were identified as, respectively, having a high level of agreement and effectiveness (i.e., respondents rated them as strongly agree/very effective or agree/effective). Nine and 11 statements, respectively, were rated as neither agree nor disagree/uncertain for agreement and effectiveness, and 38 statements were rated as agree/effective or strongly agree/very effective for both agreement and effectiveness. The number of statements generated from these two rounds totaled 49 (40 statements rated as strongly agree/very effective and 9 statements rated as agree/effective).
Subgroup Analysis
In the subgroup analysis, consensus was reached among experts who were occupational therapists on the ratings of level of agreement and effectiveness for each extrinsic feedback statement. We found a few insignificant differences between respondents by gender (female, on agreement level for Statement 14, p = .47, and effectiveness for Statement 36, p = .20; IQR = 4 [1.5]) and degrees held (BSc, on agreement level for Statements 5, p = .21, and 11, p = .18, and on effectiveness for Statements 14, p = .49, IQR = 4 [2], and 52 effectiveness, p = .73, IQR = 4 [2]; PhD, on effectiveness for Statement 16, p = .84, IQR = 4 [1.25]). Experts who work in the Middle East reached consensus on all statements except agreement level for Statement 14, p = .44, IQR = 4 [2], and effectiveness for Statement 35, p = .89, IQR = 3 [2]). The median ratings of level of agreement with and effectiveness of each statement were similar, regardless of years of experience (p > .05).
Patients
As shown in Table 3, consensus was reached on the level of agreement with and effectiveness of all 49 statements presented. Twelve statements were rated as neither agree nor disagree/uncertain. Patients agreed (i.e., rated as agree or strongly agree) with 37 statements and rated the effectiveness of 3 of these 37 statements as uncertain. Ultimately, patients rated 34 feedback statements as strongly agree/very effective or agree/effective on level of agreement and effectiveness.
Patients’ Perspectives on Feedback Statements and Their Classification
Note. Statements were rated on a 5-point Likert scale ranging from 1 (very ineffective/strongly disagree) to 5 (strongly agree/very effective). Agreement and effectiveness are defined as level of agreement with each statement and the effectiveness of each statement in creating a sense of responsibility in the course of performing daily or social activities and occupations, respectively. IQR = interquartile range; Mdn = median; NA = not applicable; − = negative; P = prevention; + = positive; SF = symptom management and function improvement.
Percentage of patients who selected two adjacent categories.
Finally, the expert panel classified 34 statements that were agreed on by patients and deemed as being effective by type (negative or positive), target (self or others), and purpose (prevention or symptom management and function improvement).
Discussion
This Delphi study produced a collection of statements that might be useful in future occupation-based interventions. These statements were produced and agreed upon by internationally renowned experts and patients living with PD. Such consensus has not previously been reached.
Reaching consensus among experts on rating the statements “Doing this activity/occupation will affect your health and quality of life in the future” and “Doing this activity/occupation will help prevent the disease from getting worse in the future” as strongly agree is in line with previous literature. Strong SoA has been identified as a key health determinant. Furthermore, disruption of SoA among patients with neurological disorders has serious consequences for quality of life (Frith et al., 2000; Haggard, 2017). Therefore, these two statements may be helpful in creating a greater SoA and, hence, a sense of responsibility. Moreover, patients rated these two statements as effective.
Statements regarding health-related quality of life, including both motor and nonmotor symptoms, were rated agree/effective by both experts and patients. Statements such as “Doing this activity/occupation will improve your anxiety and depression,” “Doing this activity/occupation will decrease your pain,” and “Doing this activity/occupation improves your hand function” stimulate patients to change their attitude toward their condition and gain an increased SoA. These occupations are performed intentionally on the basis of each individual’s condition (Taylor, 2017). SoA and sense of responsibility are higher when individuals are performing intentional and voluntary motor activities (Nierula et al., 2021).
Feedback statements such as “By doing this activity/occupation, you have a greater role in your improvement and progress,” “Doing this activity/occupation gives you a sense of empowerment,” and “By doing this activity/occupation, you will achieve your goals” were rated as agree/effective by both experts and patients. SoA has many facets, including awareness of goal, purposiveness, sense of control, and authorship (Pacherie, 2007). These statements can be attributed to these aspects of SoA. Empowerment and goal achievement are basic principles of occupational therapy and are associated with health and quality of life (Custer et al., 2013; Moghimi, 2007).
Negative statements, such as Statements 7, 8, 35, and 39, were rated as neither agree nor disagree/uncertain. Negative outcomes reduce the SoA in performed activities as a result of self-serving bias (i.e., reporting agency over positive actions but denying it over negative consequences; Christensen et al., 2019; Haggard, 2019; Yoshie & Haggard, 2017). The literature has shown that intentional actions that have unpleasant effects on others depend on the individual’s SoA with respect to these actions (Cushman, 2008; Harris et al., 2012; Kreft & Jetz, 2007; Lepron et al., 2015). Therefore, as indicated by experts and patients, these feedback statements may not provoke clients’ sense of responsibility. Additionally, it appears that negative feedback is equivocal or has a negligible effect on learning (Djamshidian et al., 2010; Martini et al., 2018). “Not doing this activity/occupation will be disadvantageous to your family members” was identified as ineffective by experts, but patients regarded this statement as being very effective. This discrepancy may be explained by propositions in previous literature that when people have responsibility for others’ negative experiences, empathetic responses are increased (Cushman, 2008; Harris et al., 2012; Kreft & Jetz, 2007; Lepron et al., 2015).
Feedback statements regarding family members or friends, such as “Not doing this activity/occupation will be disadvantageous to your friends” and “By doing this activity/occupation, you reduce the burden on your friends,” were also rated as neither agree nor disagree/uncertain. People with neurological conditions are susceptible to social cognitive disruptions (Henry et al., 2016). Social cognition is critical for interpersonal relationships. People regulate their own behavior and actions according to others’ emotions and points of view (Coundouris et al., 2020). However, individuals are held more responsible for the direct effects of their actions than they are for unspecified and unpredictable effects (Haggard & Tsakiris, 2009). The effects of these statements are not direct and may occur after some delay. Accordingly, these statements may not be effective in creating a sense of responsibility in patients with neurological disorders.
The feedback statement regarding work (i.e., “By doing this activity/occupation, your co-workers will feel more satisfied with you at work”) was rated as agree/effective by experts, but patients rated this statement as neither agree nor disagree/uncertain. This discrepancy may be explained by the fact that patients with PD experience early retirement or unemployment (Schrag & Banks, 2006), and most of the patients we recruited (95.9%) were not employed. Therefore, feedback regarding work may not be suitable for patients living with PD. This statement may be effective for people with other neurological disorders who can remain employed despite their disease. Hence, further research is needed to determine the efficacy of work-related feedback statements.
Classification of derived feedback statements can help clinicians to use proper statements in regard to patients’ characteristics, personality, and attitudes. Incorporating these statements with the proposed classification may be helpful for clinicians while delivering interventions. First, patients become familiar (familiarization) with statements and try to understand them. Then, they will absorb (absorption) the concept of responsibility and observe the effects of their actions. Eventually, they will become aware (self-awareness) of their actions and its effects. Hence, this process would improve SoA and sense of responsibility.
An important strength of the current study was the tailored iterative study design. External validity was enhanced by including a large panel of key international experts in both research and patient care. Moreover, statements were strengthened by rating both level of agreement with them and their effectiveness. Most of the experts were active in both clinical settings and research; thus, they may have rated statements using both their clinical and research experience. We did not conduct face-to-face meetings with participants between Delphi rounds to discuss ratings and disagreements. This enabled panelists to react without bias and pressure from others. Moreover, the effectiveness of the derived statements was confirmed by the patients’ opinions.
Study Limitations
Although it was carefully conducted, this study is not without limitations. Even though we aimed to include experts from all relevant fields, the majority of participants were occupational therapists. Thus, the opinions of other professions (e.g., physiotherapy, neurology) were underrepresented internationally. Moreover, different professions might interpret the statements differently. However, the proportion of occupational therapists in our sample appears to be consistent with the use of occupation and activity as an intervention in clinical settings by occupational therapists (American Occupational Therapy Association, 1997). Nonetheless, a random sampling method could reduce this sample bias. Regarding the balance of regions of practice, Europe and the Middle East were overrepresented. Communication with experts was established via email; thus, opportunities for interaction were limited. The feedback statements were presented only to patients living with PD. Therefore, the presented results may not generalize to people with other neurological conditions. Feedback statements were composed with no specifications. Further research is needed to investigate each feedback statement for specified activities and occupations.
Despite these limitations, these findings are the result of the first Delphi study to explore and develop feedback statements regarding the concept of responsibility. Further research could explore to what extent rehabilitation specialists use these statements in practice. Including caregivers among the participants could have made the current study more complete. Future qualitative research could examine other stakeholders’ (i.e., caregivers) insight of the effectiveness of each statement.
Implications for Occupational Therapy Practice
The findings of this study have the following implications for occupational therapy practice: ▪ Feedback regarding quality of life stimulates individuals to gain increased SoA over their occupations. ▪ Negative feedback statements may not provoke clients’ sense of responsibility in clinical practice. ▪ Feedback regarding other people may not be effective in creating sense of responsibility while performing occupations.
Conclusion
This study identified a set of 34 extrinsic feedback statements with high levels of agreement and effectiveness that can create a sense of responsibility among individuals with PD in the course of performing daily or social occupations/activities. We do not suggest the strict use of these consensus statements in clinical practice; patient characteristics, personality, and attitudes differ, so some modifications to the statements may be required. These statements are, rather, an attempt to develop a shared understanding of the concept of responsibility in rehabilitation.
Footnotes
Acknowledgments
We appreciate all experts and patients who participated in this Delphi study. This work was supported by Iran University of Medical Sciences, Tehran, Iran.