The negative impacts of caring for a person with motor neuron disease (MND) include depression, anxiety, prolonged grief, increased burden, decreased social contacts, and reduced quality of life (Aoun et al. Reference Aoun, Kissane and Cafarella2020; Gluyas et al. Reference Gluyas, Mathers and Hennessy Anderson2017; Harris et al. Reference Harris, Thomas and Thomas2017). There is growing recognition of the need for interventions that improve the psychological well-being (PWB) of individuals caring for people with MND. Our recent systematic review on this topic identified 12 studies using non-pharmacological interventions (NPIs) specifically designed to improve hedonic (subjective) and/or eudaimonic (functioning) aspects of carer PWB. The hedonic approach includes strategies to improve the emotional state of MND carers by encouraging positive emotions and reducing psychological distress, while eudaimonic methods emphasize the functioning aspects of PWB, such as self-acceptance, environmental mastery, positive relationships, personal growth, purpose in life, and autonomy (Cafarella et al. Reference Cafarella, Effing and Chur-Hansen2022; Ryan and Deci Reference Ryan and Deci2001). NPIs are usually complex, involve numerous interacting components, and incorporate behavioral and psychosocial processes involving specific actions or behaviors required for prevention, care, or cure (Craig et al. Reference Craig, Dieppe and Macintyre2013; Ninot Reference Ninot2021). Commonly, NPIs are poorly reported compared with pharmacological interventions, with less detail on the intervention content (Hoffmann et al. Reference Hoffmann, Erueti and Glasziou2013; McCleary et al. Reference McCleary, Duncan and Stewart2013).
Recently, steps have been made to progress the evidence base for developing behavior change interventions (Leventhal et al. Reference Leventhal, Weinman and Leventhal2008; Michie et al. Reference Michie, van Stralen and West2011a). A taxonomy of behavior change techniques (BCTTv1), listing and describing 93 intervention components from 16 categories, has been developed to aggregate active ingredients that may be included in NPIs, providing a framework to categorize intervention content. Behavior change techniques (BCTs) refer to replicable intervention components designed to alter or redirect causal processes regulating behavior (Michie et al. Reference Michie, Richardson and Johnston2013).
In our recent systematic review, in which we evaluated methods and outcomes of NPIs targeting the PWB of MND carers, we noted that the active content of NPIs was not or insufficiently described (Cafarella et al. Reference Cafarella, Effing and Chur-Hansen2022). This can be an encumbrance to replication and understanding of the critical content factors contributing to the effectiveness of the NPIs. More transparency of the intervention components leading to positive effects is necessary. Applying the BCTTv1 (Michie et al. Reference Michie, Richardson and Johnston2013) framework to PWB interventions for MND carers may facilitate a shared understanding of the specific strategies used within and across studies of interventions targeting the PWB of MND carers and may advance research and practice in this domain. In addition, it may enable theoretical links and provide further important methodological detail to promote accurate replication and practice, potentially enhancing effectiveness. It also allows the evidence derived from NPIs to be appropriately synthesized.
In alignment with the strategy of the Iterative Protocol for Evidence Base Accumulation (IPEBA), the content of all BCTs used in each study arm should be assessed independently, enabling more accurate isolation of the active ingredients and greater insight regarding the relationship between outcomes and intervention content (Peters et al. Reference Peters, de Bruin and Crutzen2015). Without this insight, intervention replication and implementation are impeded, and the reporting of effect sizes for new interventions may remain uninterpretable (Michie et al. Reference Michie, Abraham and Eccles2011b). Our recent review also revealed that information regarding the support provided to control group participants of PWB interventions for MND carers lacked detail (Cafarella et al. Reference Cafarella, Effing and Chur-Hansen2022).
A more explicit theory and evidence-based approach to designing PWB interventions may identify the need to include BCTs that have not previously been used with MND carers. BCTs that have been proven effective in interventions with similar populations (e.g., carer of chronic neurological conditions, Fakolade et al. Reference Fakolade, Walters and Cameron2020) could possibly be used to optimize existing or future PWB interventions for MND carers.
For this study, we conducted detailed content-coding of interventions targeting the PWB of MND carers using BCTTv1. The primary aim of this research was to use the BCTTv1 to identify, map, and describe the active components of intervention and comparator groups of studies evaluating the PWB of MND carers. Secondary aims were to (a) identify which active ingredients were not being utilized and (b) to explore whether variability in the effectiveness of interventions targeting the PWB of MND carers (conducted in our previous systematic review) could be better explained through improved characterization of the active content of these interventions.
Method
This study was conducted following the Joanna Briggs Institute (JBI) methodology for mixed-methods systematic reviews (MMSRs) (Aromataris and Munn Reference Aromataris and Munn2020) and Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Moher et al. Reference Moher, Liberati and Tetzlaff2009).
Search strategy and study selection
The search strategy was constructed to capture a broad range of intervention studies designed to improve the PWB of MND carers. The search strategy involved 3 distinct steps and used the keywords listed in Text Box 1. The first step included a search of CINAHL, PubMed, ProQuest, Scopus, and PsycINFO databases, followed by an analysis of the text words in the title, abstract, and index terms describing the article. Identified keywords and index terms were then included in the second search across the databases. Stage 3 involved a search for further studies by analyzing the reference lists of the included manuscripts. Studies were limited to those published in English from 1990 to December 2022. Manuscripts published before 1990 were excluded as our previous systematic review of interventions designed to improve the PWB of MND carers did not locate any studies before 2013. Included studies could be quantitative, qualitative, or mixed-methods. Interventions needed to be carer-based and designed to improve carer PWB. Interventions could be delivered to individuals or groups, and any delivery mode, duration or dose was acceptable. Spousal, family, or other informal carers of people with MND were included, but professional or paid carers were not. Carers of individuals from all MND disease stages and care settings were included.
Search terms
Keywords (and their variations) used in the searches included those related to Condition: (“motor neuron* disease” OR MND or “amyotrophic lateral sclerosis” OR ALS), Population (famil* OR informal OR spous*) AND (carer* OR caregiv*) and Intervention (*behavio*r therapy” OR “cognitive behavio* therapy” OR CBT OR “acceptance and commitment therapy” OR ACT OR mindfulness OR relaxation OR meditation OR counsel*ing OR grief OR bereavement OR “social support” OR “family therapy” OR “art therapy” OR “dance therapy” OR “music therapy” OR “*complementary therap*” OR *Exercise OR Yoga OR “person centered*” OR “dignity therapy” OR psychotherap*OR psychosocial OR psychological OR treatment OR training OR education* OR program.
Outcomes of interest were any associated with the PWB (inclusive of hedonic and eudaimonic outcomes) of MND carers. In accordance with the JBI methodology for MMSRs, recommended screening protocols, design-appropriate standardized quantitative and qualitative critical appraisal forms to assess inclusion and methodological quality, and standardized data extraction tools were employed (Stern et al. Reference Stern, Lizarondo and Carrier2020). Further evaluation of selective reporting bias was conducted by comparing measurements and outcomes in studies with published protocols (where available). In studies without published protocols, methods and results were compared. All studies were evaluated by each of the 3 reviewers (PC, TE, ACH) to determine inclusion, and disagreement was resolved by discussion. The study selection is displayed in Figure 1.
PRISMA flow diagram of manuscript selection.
Data collection
The first author (PC) contacted the authors of the included articles to provide a detailed description of the intervention and any comparison groups. In addition, searches were performed to locate any published materials, including descriptive information about the study groups. Descriptive materials could include protocols (published or unpublished), intervention or study manuals, training resources or any other materials detailing the content of study groups. Authors were contacted via email and permitted 8 weeks (with 3 reminders) to provide the requested information.
Data extraction and analysis
Quality assessments and data extraction from the selected studies using design-appropriate JBI tools were conducted by two reviewers (PC and TE: quantitative data; PC and ACH: qualitative data) using agreed narrative evidence descriptors (Aromataris and Munn Reference Aromataris and Munn2020). Data were presented narratively and synthesized descriptively as recommended by the JBI guidelines for MMSRs (Aromataris and Munn Reference Aromataris and Munn2020). From our previous study results (Cafarella et al. Reference Cafarella, Effing and Chur-Hansen2022), we knew a priori that data pooling for a meta-analysis would not be possible due to the heterogeneity of the interventions, outcomes and measurements, and the small number of studies.
Content for all available study groups of the included studies was coded independently by two trained reviewers (PC and research assistant SW – see acknowledgments) using BCTTv1, a valid and reliable method of retrospectively specifying the active content of interventions, which lists and describes 93 intervention components from 16 categories (Michie et al. Reference Michie, Richardson and Johnston2013, Reference Michie, Wood and Johnston2015) (listed in Table 1). Coder training in using BCTTv1 may increase the validity of coding BCTs and coder competence (Michie et al. Reference Michie, Wood and Johnston2015). The coders (PC, SW) completed the interactive online training (Bct-taxonomy.com 2023). In accordance with IPEBA, the BCT content of each study group was assessed independently to increase awareness of the potential bias associated with BCTs used with the control groups, enable more accurate isolation of the active ingredients and greater insight regarding the relationship between outcomes and intervention content (Peters et al. Reference Peters, de Bruin and Crutzen2015). Both coders discussed any differences in the identification of BCTs until a consensus was reached. A third trained coder (TE) was available for resolution in cases where consensus could not be reached.
BCTTv1 categories and BCTs identified per study
Key: x = BCT identified
Results
Sixteen manuscripts were included (Figure 1), describing 14 unique studies. One study published quantitative and qualitative research components in separate papers (de Wit et al. Reference de Wit, Beelen and Drossaert2020, Reference de Wit, Vervoort and van Eerden2019), and another study was described in 2 manuscripts (Aoun et al. Reference Aoun, Chochinov and Kristjanson2015; Bentley et al. Reference Bentley, O’Connor and Kane2014). Six of the 14 studies included mixed-methods, while 4 used solely quantitative methods (Creemers et al. Reference Creemers, Veldink and Grupstra2014; Kavanaugh et al. Reference Kavanaugh, Cho and Fee2020; Pagnini et al. Reference Pagnini, Phillips and Haulman2021; van Groenestijn et al. Reference van Groenestijn, Schröder and Visser-Meily2015), and another 4 employed a qualitative approach (Cipolletta et al. Reference Cipolletta, Gammino and Francescon2018; Marconi et al. Reference Marconi, Gragnano and Lunetta2016; Olesen et al. Reference Olesen, la Cour and With2022; Thomas et al. Reference Thomas, Warrier and Arun2022). Four randomized controlled trials (RCTs) were included (Creemers et al. Reference Creemers, Veldink and Grupstra2014; de Wit et al. Reference de Wit, Beelen and Drossaert2020; Pagnini et al. Reference Pagnini, Phillips and Haulman2021; van Groenestijn et al. Reference van Groenestijn, Schröder and Visser-Meily2015). All papers were published between 2013 and 2022 (10 of the 16 in the last 5 years) (Tables 2 and 3).
Synthesized narrative findings and identified BCTs in studies with a quantitative component
Q = Critical appraisal of quality score (/21); PWB = psychological well-being; n = sample size.
Synthesized narrative findings and identified BCTs in studies with a qualitative component
Q = Critical appraisal of quality score (/13); PWB = psychological well-being; n = sample size.
Ten of the 14 studies were conducted in 3 countries. Four studies originated in Australia (Aoun et al. Reference Aoun, Chochinov and Kristjanson2015, Reference Aoun, Deas and Kristjanson2017, Reference Aoun, Hogden and Kho2018; Bentley et al. Reference Bentley, O’Connor and Kane2014; Ugalde et al. Reference Ugalde, Mathers and Hennessy Anderson2018), 3 in the U.S.A. (Creemers et al. Reference Creemers, Veldink and Grupstra2014; Kavanaugh et al. Reference Kavanaugh, Cho and Fee2020; Steinhauser et al. Reference Steinhauser, Olsen and Johnson2016), and 3 in Italy (Cipolletta et al. Reference Cipolletta, Gammino and Francescon2018; Marconi et al. Reference Marconi, Gragnano and Lunetta2016; Pagnini et al. Reference Pagnini, Phillips and Haulman2021). Two studies were conducted in the Netherlands (de Wit et al. Reference de Wit, Beelen and Drossaert2020, Reference de Wit, Vervoort and van Eerden2019; van Groenestijn et al. Reference van Groenestijn, Schröder and Visser-Meily2015), 1 in Denmark (Olesen et al. Reference Olesen, la Cour and With2022), and another in India (Thomas et al. Reference Thomas, Warrier and Arun2022). Overall, 530 MND carers were studied in the quantitative component of this review and 256 in the qualitative component (Tables 1 and 2).
Eleven of the 14 studies either had separately published protocols (n = 6) or the authors provided additional materials (including 1 separate unpublished protocol). The remaining studies (n = 7) described the protocol within the text of the original manuscript. An email response with additional information was received from the authors of 10 of the 14 included studies.
Forty-one of the possible 93 BCTs (44%), as defined by BCTTv1 (Michie et al. Reference Michie, Richardson and Johnston2013), were identified as active ingredients. We identified BCTs in all 14 intervention groups (Figure 2) and the 4 control groups. The number of identifiable BCTs ranged from 5 to 23 (mean: 12.1, SD: 5.78) in the intervention groups and 1 to 3 (mean: 1.8, SD: 0.96) in the control groups.
Number of studies in which specific BCTs were identified in the intervention groups.
Among the BCTs identified for the 14 intervention groups, the most prevalent were 3.1 Social support (unspecified) (100%), 1.2 Problem-solving, 2.5 Monitoring of outcome(s) of behavior without feedback, 9.1 Credible source (85.7%), and 11.2 Reduce negative emotions (78.6%) (Table 1 and Figure 2). BCTs were identified in 4 of the control groups: 2.5 Monitoring of outcome(s) of behavior without feedback (n = 4), 3.1 Social support (unspecified) (n = 2), and 9.1 Credible source (n = 1) (Table 1). Fifty-two of the possible 93 BCTs were not identified in either the intervention or control group descriptions.
The coding exercise identified 12 of the 16 overall BCT categories described in BCTTv1 (Michie et al. Reference Michie, Richardson and Johnston2013). The 4 BCT categories not described in the studies were Associations, Reward and threat, Scheduled consequences, and Covert learning (Table 1). The most frequently deployed BCTTv1 categories were Social support (100%), Feedback and monitoring, Comparison of outcomes (85.7%), Goals and Planning, and Regulation (78.6%). The other categories used included Identity (64.3%), Shaping knowledge, Comparisons of behavior (57.1%), Natural consequences (42.9%), Repetition and substitution (35.7%), Antecedents, and Self-belief (28.6%) (Table 1).
Seven of the 10 studies with a quantitative component and 9 of the 10 studies with a qualitative component reported PWB benefit(s) from their interventions. Overall, 11 of the 14 studies demonstrated some benefits. The identified BCTs and BCT categories associated with these studies are detailed in Tables 2 and 3, enabling characterization of the active content of the interventions associated with hedonic and eudaimonic PWB outcomes. Among the 11 studies reporting intervention benefits, the most prevalent identified BCTs were 3.1 Social support (unspecified) (100%), 1.2 Problem-solving (81.8%), 11.2 Reduce negative emotions (81.8%), 2.5 Monitoring of outcome(s) of behavior without feedback (72.7%), 4.1 Instruction on how to perform a behavior (72.7%), 3.3 Social support (emotional) (72.7%), and 9.1 Credible source (72.7%) (Tables 2 and 3). A higher number of BCT categories was identified in studies with intervention benefits (mean: 7.9, SD: 2.21) compared to those not demonstrating PWB gains (mean: 5.3, SD: 1.53) (Tables 2 and 3).
Table 2 provides a summary methodological quality score (range: 0–21) for the quantitative studies, calculated using the JBI critical appraisal tools (median: 9, IQR: 7.75). The 4 RCTs were the only studies that achieved summary quality scores above 50% (score range between 57.1 and 76.2%). Table 3 presents a summary methodological quality score (range: 0–13) for the qualitative studies, calculated using the JBI critical appraisal tools (median: 8, IQR: 3.75). Four studies achieved summary quality scores below 50% (score range: 7.7–46.2%). Six studies scored between 53.9 and 100%.
Discussion
This is the first study to comprehensively identify and map the active components of interventions (and comparator groups) targeting the PWB of MND carers using the BCTTv1 taxonomy. We identified 14 studies reporting interventions targeting MND carer PWB, of which 4 were RCTs. Fifty-two of the 93 BCTs listed in BCTTv1 were not identified in the reviewed studies, nor were 4 of the 16 BCT categories, highlighting the untested intervention content in this field. The results enabled a more transparent comprehension and characterization of the active content of the interventions demonstrating hedonic and eudaimonic PWB benefits for MND carers. This review also provides directions to improve interventions in this developing field of research.
The BCT category of “Social support” was identified as an intervention component in each study. Among the BCTs in this category, Social support (unspecified) (100%) was the most frequently deployed, with Social support (emotional) (64.3%) and Social support (practical) (57.1%) regularly used. In studies demonstrating PWB benefits for carers, Social support (emotional) (72.7%) and Social support (practical) (63.6%) were commonplace. However, BCTs associated with the social support category were not among the most frequently identified in a systematic review of BCTs used in health interventions for persons with chronic neurological conditions and their caregivers (Fakolade et al. Reference Fakolade, Walters and Cameron2020). Nevertheless, the need for social support for MND carers has been demonstrated in several studies, with practical aspects such as equipment provision, demonstration and support identified as pragmatic needs (Aoun et al. Reference Aoun, Cafarella and Hogden2021, Reference Aoun, Deas and Kristjanson2017). A lack of emotional support has also been expressed as a notable problem for MND carers (Aoun et al. Reference Aoun, Cafarella and Hogden2021, Reference Aoun, Kissane and Cafarella2020). The form in which social support is manifested appears relevant, as professional support is not commonly evaluated as helpful by MND carers (Aoun et al. Reference Aoun, Kissane and Cafarella2020). Peer support groups have demonstrated PWB benefits among carers of people with progressive neurological illnesses (Pasquini et al. Reference Pasquini, Margaritini and Gambella2022), while peer support groups for MND carers improved the PWB of one cohort but not another (Marconi et al. Reference Marconi, Gragnano and Lunetta2016). Although peer support may facilitate the development of coping strategies mediating hedonic PWB variables such as carer stress (Pearlin et al. Reference Pearlin, Mullan and Semple1990), some authors have argued a “goodness-of-fit” approach to carer support needs, incorporating the risk of mental health problems (Aoun et al. Reference Aoun, Kissane and Cafarella2020) and MND patient stage (Poppe et al. Reference Poppe, Koné and Iseli2020) may be more effective than a “one-size-fits-all” method (Aoun et al. Reference Aoun, Cafarella and Hogden2021). While the need for social support is evident, the impact of this active ingredient on MND carer PWB outcomes is difficult to determine from our study since this BCT category was identified in all the studies (including those not demonstrating PWB benefits), and 2 of the 4 control groups.
Our results elucidate the potential of exposure of the control group to active intervention ingredients, which may affect study outcomes. While the BCT categories of Social support and Feedback and monitoring were the most frequently deployed, they were also prevalent in the control groups, reinforcing the importance of specifying the active ingredients in all study groups. Further, a closer look at the data indicates that the high frequency of the Feedback and monitoring category is due to the prevalence of the specific ingredient “Monitoring of outcome(s) of behavior without feedback” (85.7%), essentially tracking outcomes for participants in all study groups. Therefore, our results support the IPEBA approach, where the BCT content of each study arm is independently assessed, enabling greater insight into the relationship between outcomes and intervention content (Peters et al. Reference Peters, de Bruin and Crutzen2015). Results may be rendered uninterpretable in the absence of this insight casting doubt over the accuracy of reported intervention effect sizes (Michie et al. Reference Michie, Abraham and Eccles2011b; Peters et al. Reference Peters, de Bruin and Crutzen2015). This imprecision can be further exacerbated by poor reporting of the control conditions, as the description may not provide all the included BCTs, potentially biasing the results and interpretation (de Bruin et al. Reference de Bruin, Viechtbauer and Eisma2016).
“Reduce negative emotions” was another prevalent (81.8%) BCT identified in the intervention studies demonstrating PWB benefits to MND carers. The rationale for including this component is evident from the high rates of mental health issues linked with the role (Aoun et al. Reference Aoun, Kissane and Cafarella2020; Gluyas et al. Reference Gluyas, Mathers and Hennessy Anderson2017; Harris et al. Reference Harris, Thomas and Thomas2017). Despite the high frequency of this BCT, the results of this study and our previous systematic review (Cafarella et al. Reference Cafarella, Effing and Chur-Hansen2022) found that quantitative studies using traditional, hedonic PWB outcome measures (e.g., psychological distress scales) rarely demonstrated intervention benefits on these questionnaires. It remains unclear if the mode of measurement or lack of sensitivity associated with specific quantitative, hedonic outcome measures contributed to the lack of positive effect. However, more carer PWB benefits were evident when studies used qualitative data and eudaimonic measures of PWB, demonstrating the value of including these approaches (e.g., sensitivity, responsiveness, and the importance of non-hedonic aspects of PWB) (Cafarella et al. Reference Cafarella, Effing and Chur-Hansen2022). Eudaimonic outcomes incorporate and value PWB components such as skill development, competency, self-efficacy, and a sense of control, which can help to reduce the guilt associated with their perceived caring ability and burden (Keating et al. Reference Keating, McGregor and Yeandle2021; Teahan et al. Reference Teahan, Lafferty and McAuliffe2020). An example is Kavanaugh et al.’s (Reference Kavanaugh, Cho and Fee2020) study, which focussed on upskilling young MND carers and demonstrated gains in eudaimonic PWB. The eudaimonic variable of growth is a significant mediator of resilience among carers of people with dementia (O’Dwyer et al. Reference O’Dwyer, Moyle and Zimmer-Gembeck2016). The BCTTv1 category of Identity was prevalent in our review (64.3%) and relates to the eudaimonic aspects of PWB, demonstrated to contribute positive aspects of caring and positive emotions (Beach et al. Reference Beach, Bélanger-Hardy and Harding2022; Teahan et al. Reference Teahan, Lafferty and McAuliffe2020). The category of Self-belief was less prevalent (28.6%), although perceived competency in caring among those looking after people with dementia has been associated with reduced psychological distress (Bennett Reference Bennett2018; Moore et al. Reference Moore, Davis and Gola2017) and aligns with capabilities perspectives of eudaimonic PWB (Keating et al. Reference Keating, McGregor and Yeandle2021). Such perspectives emphasize the importance of a carer’s perceived ability and resources to conduct valued caring tasks in their overall PWB (Keating et al. Reference Keating, McGregor and Yeandle2021), yet this remains an under-explored area in PWB research with MND carers.
Another frequently deployed BCT identified in positive intervention studies was “Problem-solving” (81.8%), the most prevalent component of the BCTTv1 Goals and Planning category. While the value of goals and planning in initiating and maintaining PWB has been demonstrated in chronic health populations such as spinal cord injuries, older adults, cancer, and arthritis, a recent systematic review found this BCTTv1 category was less prevalent and likely under-utilized in dyadic health interventions for persons with chronic neurological conditions and their carers (Fakolade et al. Reference Fakolade, Walters and Cameron2020). Problem-solving seems inherent in the role of an MND carer and appears to be logically linked with PWB and the prevalence of this BCT in positive intervention studies provides tentative support for this plausible connection.
The BCT labelled “Credible source” was one of the most prevalent identified among the studies and was coded frequently in those reporting intervention benefits. Qualified health professionals typically delivered interventions in the included studies. A recent systematic review of BCTs used in health interventions for persons with chronic neurological conditions and their caregivers, also found this BCT to be among the most frequently identified (Fakolade et al. Reference Fakolade, Walters and Cameron2020). Characteristically, dyadic psychosocial interventions for neurodegenerative diseases have been delivered by qualified health professionals (Fakolade et al. Reference Fakolade, Walters and Cameron2020). However, being a “credible source” may not guarantee that professionals delivering the BCTs were sufficiently trained, as some health professionals did not believe they were adequately trained to deliver the BCTs during the intervention effectively (Fakolade et al. Reference Fakolade, Walters and Cameron2020; Hardeman et al. Reference Hardeman, Michie and Fanshawe2008; Keogh et al. Reference Keogh, Matthews and Hurley2018).
This study also identified BCTTv1 categories and individual BCTs that were absent or infrequently utilized. Fifty-two of the possible 93 individual BCTs were not identified in either the intervention or control group descriptions. Some of the absent BCTs may not be applicable for implementation with MND carers; however, the majority (55.9%) of these techniques remain untested with this population. The knowledge void regarding the value of absent BCT categories is further exacerbated by a lack of direct comparison with similar populations because there is currently only 1 review of the BCTs in caregivers of chronic neurological conditions (Fakolade et al. Reference Fakolade, Walters and Cameron2020). That review included 27 studies and identified 10 of 16 BCTTv1 categories. Twelve of the 16 BCTTv1 categories were identified in our study, while those absent were Associations, Reward and threat, Scheduled consequences, and Covert learning. The potential value of these absent categories and individual BCTs in fostering MND carer PWB is yet to be determined. For example, the Associations category was present in a systematic review of BCTs used in health interventions for persons with chronic neurological conditions and their caregivers (Fakolade et al. Reference Fakolade, Walters and Cameron2020). Prompts, cues, and associative learning, individual BCTs from the “Associations” category, have been shown to support the development of habit formation in critical areas such as medication delivery among carers of ageing people (Stawarz Reference Stawarz2017). Similarly, in relation to the “Reward and threat” category, various forms of reward have been associated with facilitating the formation of routine-based strategies essential for perceived competence in caregiving, such as medication monitoring (Stawarz Reference Stawarz2017). Consistent with our study, the absent categories in the systematic review of BCTs used in health interventions for persons with chronic neurological conditions and their caregivers, were Reward and threat, Scheduled consequences, and Covert learning. Categories identified in our study, but absent in their research, were Regulation, Identity, and Self-belief (Fakolade et al. Reference Fakolade, Walters and Cameron2020), Additionally, the categories of Repetition and substitution, Antecedents, and Self-belief were less prevalent in our study. Without more studies enabling direct comparison with similar reference populations, it is unknown if absent BCT categories and/or absent individual BCTs represent under-utilized intervention components, potential novel interventions, or whether they are less relevant or effective. An alternative explanation may be that the BCTs reported may not accurately represent the intervention content. While 11 of the 14 studies either had separately published protocols or the authors provided additional materials, a known problem in coding exercises is that BCTs are often poorly described or under-reported in published materials (Chakraborty et al. Reference Chakraborty, Bailey and Seidler2022). BCT coding remains dependent upon and therefore limited by the detail provided in the reported content (Fakolade et al. Reference Fakolade, Walters and Cameron2020). The reporting of NPIs frequently lacks precision and detail even after authors have been contacted (Hoffmann et al. Reference Hoffmann, Erueti and Glasziou2013). Published descriptions may report less than half the BCTs evident from a protocol (Lorencatto et al. Reference Lorencatto, West and Christopherson2013). Hence, the possibility remains that some BCTs may remain uncoded.
While BCTTv1 is a standardized, systematic approach to classifying the active content of interventions and has been described as the most comprehensive and frequently used of such taxonomies (Michie et al. Reference Michie, Richardson and Johnston2013; Wood et al. Reference Wood, Richardson and Johnston2015), there is always some degree of subjectivity and variability in the application of taxonomies (Chakraborty et al. Reference Chakraborty, Bailey and Seidler2022) even when coding intervention content independently to decrease bias. Another challenge is related to the uncertain intensity of individual BCTs and the unclear value of this knowledge (JaKa et al. Reference JaKa, French and Wolfson2019) as questions remain regarding the appropriate coding and reporting of BCT intensity and dose (Chakraborty et al. Reference Chakraborty, Bailey and Seidler2022).
Strengths and limitations
Although we have identified the active content of MND carer interventions for PWB and characterized those associated with positive outcomes, these inferences should be treated tentatively. The data pooling for a meta-analysis was not possible due to the heterogeneity of the interventions, outcomes and measurements, and the small number of studies. Additionally, many of the studies only achieved low scores on the JBI critical appraisal tools evaluating methodological quality, indicating that methodological development is needed in this nascent field of research. These methodological weaknesses are discussed in-depth in our systematic review of interventions targeting PWB for MND carers (Cafarella et al. Reference Cafarella, Effing and Chur-Hansen2022).
Our study included quantitative and qualitative methods and was undertaken with an integrative approach to PWB, incorporating hedonic and eudaimonic aspects of PWB, given that both have been linked with positive outcomes (Vazquez et al. Reference Vazquez, Hervas and Rahona2009). In comprehensively mapping the active components of PWB interventions for MND carers using BCTTv1, we have identified the BCTs that are yet to be tested or potentially under-utilized and have facilitated a broader capture of potential intervention benefits. However, heterogeneity issues and the small number of studies precluded data-pooling, and methodological limitations were commonplace. There are known limitations associated with the use of taxonomies, such as subjectivity, lack of direction regarding the coding of intensity, and dependency on the detail provided in the reported content (Chakraborty et al. Reference Chakraborty, Bailey and Seidler2022; Fakolade et al. Reference Fakolade, Walters and Cameron2020; JaKa et al. Reference JaKa, French and Wolfson2019).
Future recommendations
It is imperative that future research ensures that descriptions of the active content of study arms are explicit and sufficiently detailed to enable generalizability and replicability. Precision is essential in each study group, facilitating better comprehension of the relationship between results and intervention content, increasing confidence in effect sizes, and reducing the risk of bias in the outcomes and interpretation (de Bruin et al. Reference de Bruin, Viechtbauer and Eisma2016; Peters et al. Reference Peters, de Bruin and Crutzen2015). Taxonomies such as BCTTv1 can assist in the design and description of the active content in study arms (Michie et al. Reference Michie, van Stralen and West2011a, Reference Michie, Wood and Johnston2015) and facilitate the systematization of reporting standards in this field, as in others (Johnston Reference Johnston2021).
Although this study identified and characterized BCTs associated with positive PWB outcomes for MND carers such as 3.1 Social support (unspecified), 1.2 Problem-solving, and 11.2 Reduce negative emotions, the nature of these inferred connections requires further investigation due to reporting and methodological quality concerns. Moreover, most BCTs remain untested in this population. Absent BCTs, particularly those demonstrating PWB-related benefits among carers of other populations (e.g., prompts/cues, associative learning, rewards, Stawarz Reference Stawarz2017) require closer scrutiny and potential investigation with well-designed intervention studies.
Future research would benefit from employing existing frameworks (e.g., JBI, PRISMA) to guide study design. Taxonomies of approaches to intervention development may inform future intervention design, piloting, evaluation, and implementation. This process emphasizes methodological rigor and supports theory development, which is particularly relevant for nascent fields such as interventions targeting the PWB of MND carers (O’Cathain et al. Reference O’Cathain, Croot and Sworn2019).
Our recent systematic review concerning PWB interventions for MND carers revealed a need for theory identification and development (Cafarella et al. Reference Cafarella, Effing and Chur-Hansen2022). Applying a taxonomy of intervention components such as BCTTv1 facilitates comprehension of the relationships between intervention content and reported outcomes (Michie et al. Reference Michie, Wood and Johnston2015). This allows the potential introduction of theory to interventions since the BCTs are based on theoretical constructs and can add data to elucidate the “mechanisms of action” responsible for change (Carey et al. Reference Carey, Connell and Johnston2019). BCTs may be evident in interventions, but unfortunately, theory is often absent in the development of interventions (Prestwich et al. Reference Prestwich, Sniehotta and Whittington2014).
Conclusions
This review provides a snapshot of the active content of the interventions (and comparator groups) targeting PWB for MND carers. Identified and absent BCTs and BCTTv1 categories were mapped, and those associated with positive outcomes were characterized. These inferences should be treated tentatively due to reporting and methodological quality issues. However, the results elucidate a more transparent comprehension of the active content of the interventions demonstrating hedonic and eudaimonic PWB benefits for MND carers and provide direction to improve interventions in this developing field of research.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S1478951524000877.
Acknowledgments
The authors thank Sharni Whitburn (SW) for her coding assistance.
Funding
This research received no specific grant from any funding agency, commercial or not-for-profit sectors. SW was a Research Assistant at the School of Psychology, University of Adelaide.
Competing interests
None.
Open access
