An online intervention for carers to manage behavioral symptoms in motor neuron disease (MiNDToolkit): a randomized parallel multi-center feasibility trial

Abstract Background Evidence on management of behavioral symptoms in motor neuron disease (MND) is lacking. The MiNDToolkit, an online psychoeducational platform, supports carers dealing with behavioral symptoms (BehSymp). The study objectives were to ascertain recruitment and retention rates, carer and healthcare professional (HCP) use of the platform, and completion of online assessments, to inform a full-scale trial. Design: Randomized, parallel, multi-center, feasibility trial. Setting England and Wales, across diverse MND services; recruitment from July/21 to November/22; last participant follow-up in March/23. Participants Carers of people with motor neuron disease (PwMND) with BehSymp, recruited through MND services. After confirming eligibility, participants completed screening and baseline assessments online via the MiNDToolkit platform and were randomized centrally in a 1:1 ratio to MiNDToolkit or control. Intervention MiNDToolkit offered tailored modules to carers for the 3-month study period. Carers in the intervention group could receive additional support from MiNDToolkit trained HCPs. The control group was offered access to the intervention at the end of the study. Data were collected on platform usage and psychosocial variables. Main outcomes One hundred and fifty-one carers from 11 sites were invited to join the study (letter, face-to-face); 30 were screened; 29 were randomized. Fifteen people were allocated to the control arm; 14 to intervention. Carers were mostly female; median age for was 62.5 (IQR: 58, 68; intervention) and 57 (IQR: 56, 70; controls). Study retention was high (24/29 = 82.76%); carers engaged with the platform on average 14 times (median (IQR):14.0 (10.0, 18.5)) during the study period. Conclusion The MiNDToolkit study was feasible and well accepted by carers and trained HCPs. A definitive trial is warranted.


Introduction
Around 50-75% of people with motor neuron disease (PwMND) present with progressive (1) behavioral symptoms (BehSymp) commonly seen in frontotemporal dementia (FTD) such as apathy, disinhibition, rigidity (2), and deficits in social cognition (3).International consensus criteria on the diagnosis of motor neuron disease (MND) with FTD-like symptoms have been published (4) to support clinical identification, management, and research studies.
Behavioral symptoms can negatively affect clinical decision making, which are particularly pressing in MND given the rapid progressive changes in mobility, swallowing, and breathing (5).Additionally, PwMND with severe apathy and disinhibition (6,7), or those with MNDFTD (8) have worse prognosis.Finally, cognitive and BehSymp have been associated with greater carer burden and distress in spouses (9,10), children (11), and healthcare professionals (HCPs) (12).
The National Institute for Health and Care Excellence (NICE, London, UK) guidelines for the management of MND highlight the importance of professionals' recognition and assessment of these symptoms (13), but, despite the negative impact of such symptoms in the prognosis of MND, and the wellbeing of their carers, there is marked paucity of evidence on the clinical management of behavioral and cognitive symptoms in MND.
No research study testing an intervention to support carers in managing BehSymp in MND has been identified.This study aimed to assess the feasibility of conducting a multi-center, randomized controlled trial of MiNDToolkit, a novel online intervention to support carers in the management of BehSymp.

Design
A multi-site, two-arm, parallel, randomized, controlled feasibility trial was conducted, allocating carers of people with MND, in a 1:1 ratio to a "usual care" arm or an intervention arm who were given access to the MiNDToolkit online platform for three months.After three months, all participants completed follow-up questionnaires and the intervention was made available to the control participants, with all participants offered continued access to the platform until the end of the trial.A nested qualitative process evaluation, including HCP acceptability of MiNDToolkit, is described separately (14).Ethical approval: London Queen Square Research Ethics Committee (19/LO/0692, IRAS260290).

Participants
Potential participants were identified and referred by clinical teams.Inclusion criteria were: family or live-in professional carers of PwMND with BehSymp, or MNDFTD (4); aged 18 years; having 7 hours of contact with the PwMND/week; able to communicate in English without support; not a carer of a PwMND who already has a carer recruited into the study.Eligibility was checked in two stages: sites identified carers fulfilling the above demographic characteristics, and the online platform screened for symptoms via online questionnaires.

Sample size
A sample size of 20-30 participants was chosen following published recommendations for feasibility trials (15,16).It was considered that this would be a sufficient sample to report on the practicalities of delivering the novel online intervention, recruitment, uptake, and attrition.

Procedures
Clinical teams were asked to consider all families affected by MND in their caseload to identify potentially eligible carers.Carers attending routine appointments were informed about the study (in-person, telephone, and post).Carers consented to share contact details with the research team, with staff completing a consent to contact form on behalf of the potential participant.All study information was co-designed with carers from the carer-research involvement group.
Upon receipt of a referral, a researcher telephoned potential participants, and sent an information sheet via email.If the participant confirmed interest in the study, they were issued a login for the MiNDToolkit platform, allowing them to provide an electronic informed consent and clinical screening to confirm eligibility.The platform algorithm behind the user-friendly interface calculates assessments automatically, so a participant was then taken to the trial activities or diverted to a "thank you" message if they did not pass screening.Additional information is given in the Appendix.

Randomization
A randomization list created by the Norwich CTU, using random blocks of 2 and 4, with equal allocation, i.e. a ratio of 1:1 for intervention to control was used.Allocation was concealed prior to randomization; the list was held only by members of the data management team.Due to the nature of the psychoeducational intervention, blinding of allocation was not possible.

Intervention: MiNDToolkit
MiNDToolkit is a psychoeducational intervention delivered via a bespoke online platform.HCP reinforcement of the intervention during routine clinical contacts is recommended, with MiNDToolkit training (14) provided to professionals involved at trial sites.
MiNDToolkit contains 16 modules, comprising BehSymp in MND and strategies to deal with symptoms (Table 1).Following advice from carers, modules are short and time to complete is shown before modules are selected.Animations are simple and audio content is prioritized, e.g.video Online intervention for carers to manage behavioural symptoms in MND 507 clips, simple animations.The online platform is the result of an extensive adaptation process of the original material in 2020 (17), transforming the paper-based to an online-based intervention, ensuring that MiNDToolkit could be used even during lockdowns.
The MiNDToolkit platform is adaptive.Modules offered, defined by an algorithm, vary per participant at each assessment point.Modules match BehSymp reported, e.g. if the PwMND has few symptoms, few modules are shown.Carers are asked to complete two modules/week and are informed that the access to the platform will cease after three months.Automatic reminders for module completion are emailed weekly, and carers can opt-in to receive a SMS reminder.Carers can, however, choose to complete all modules at once if they wish.There is no restriction as to how long they can access the modules per week, and they can replay modules during the 3-month period, based on carer consultation a priori.
A minimum of two HCPs from each site were trained to reinforce the content of the MiNDToolkit at every opportunity, e.g.clinical appointment, phone call, home visit.HCPs were asked to record their points of contact on MiNDToolkit, if they were able to reinforce the content, and what the result was of that discussion, including if they did not have time to reinforce MiNDToolkit content.
The decision to move from primary HCP delivery (paper, pre-pandemic) to HCP reinforcement (online MiNDToolkit) was a deliberated response.The goal was to facilitate clinical teams' engagement with MiNDToolkit, while ensuring carers could have access to research studies regardless of COVID-19 lockdowns-also potentially facilitating implementation and national scaling-up of MiNDToolkit in future.

Control: treatment as usual
Current MND specialist care does not entail standardized provision of care for BehSymp management.Some HCPs may advise if they have clinical experience, but no clear approaches are recognized.This lack of standardized provision and training has been further explored in the process evaluation (14).As such, carers in the control group were effectively a waiting list control group.The screening assessment was automatically scored by MiNDToolkit, and carers meeting the eligibility criteria were seamlessly taken to baseline data collection.If the criteria were not met, the carer was invited three months later for a repeat screening assessment, when symptoms of the PwMND may have changed.Participants were followed-up at three months after randomization.

Primary outcome measures
Feasibility outcomes were collected to enable an estimation of key parameters to inform a future trial, and to provide preliminary information about the impact of the intervention:

Analysis
Feasibility outcomes were reported descriptively and narratively.The analysis of efficacy outcome measures was descriptive, reported as medians Online intervention for carers to manage behavioural symptoms in MND 509 and interquartile ranges (IQRs) and numbers and percentages for binary and categorical variables.Descriptive statistics on MiNDToolkit engagement were reported.Formal analyses were conducted in terms of efficacy for questionnaire outcomes using generalized linear models (GLMs).However, these are for information purposes only, as efficacy is not our focus for this study.HCP and carer acceptability are reported separately (14,19).

Trial monitoring and safety
A Trial Management Group assisted with developing the design, co-ordination, and strategic management of the trial, overseeing safety.Carers could report adverse events at follow-up assessment.Researchers and linked HCPs were automatically notified if a carer indicated that they were experiencing thoughts of self-harm (PHQ-9), and a list of contact details for support and crisis lines would appear on the platform for the carer.

Recruitment and retention
Two hundred and eighty-four MND families were assessed for eligibility, with 151 carers invited to the study via letter or face-to-face.Of these, 30 were screened, and 29 passed the clinical screening in their first attempt and were randomized, 14 to the intervention and 15 to the treatment as usual group.
Reasons for exclusion and number of people eligible but not invited are shown in Figure 1.Followup occurred between March/22 to March/23, with the trial ending when the last participant completed their 3-month follow-up assessments two participants were lost to follow-up (one per arm), two participants did not start the intervention or complete follow-up assessment (Figure 2).Carer attrition rate due to PwMND decline/death was much lower than anticipated, with only one carer discontinuing the study before the follow-up assessment.Overall, 24 participants were assessed for the feasibility and acceptability objectives (Table 2).

Participant characteristics
Carers were mostly female in both groups.Median age was 62.5 for carers in the intervention group, and 57 for carers in the control arm.Ethnicity was mixed, with 71.4% Caucasian, and remaining categorized as Other or Black.Approximately, 50% of carers were engaged in paid work, part-or fulltime; around a quarter of all carers reduced work hours to provide care for the person with MND (Table 3).Demographic characteristics of PwMND are shown in eTable 1.

MiNDToolkit-carer use
From those allocated to intervention, 12/14 carers interacted with the intervention modules, on average, 14 times (median (IQR): 14.0 (10.0, 18.5)) during the 3-month intervention period.Length of engagement per interaction on MiNDToolkit was measurable, since MiNDToolkit was pre-set to log out after 60 minutes, and people usually do not log out when moving to another browser or activity, leading to inflated values of engagement.For this reason, we are not reporting the length of engagement per interaction.

MiNDToolkit-completion of outcome measures
Outcome measures of participants were completed in full, without accidental missed items.Questionnaire items cannot be missed because the platform was set up to ensure participants would only move forward after entering a response, with pre-set parameters, e.g.selecting a response for multiple choice questions, or for example, entering a digit if the question was open and referring to months.The median time taken to complete baseline questionnaires was 36 minutes (median ¼ 36, IQR ¼ (21.0, 153.0), range ¼ (6.0, 38965.0)),compared to 28 minutes at follow-up (median ¼ 28.0, IQR ¼ (19.0, 52.0), range ¼ (10.6, 45010.0)),with some participants completing the full set of questionnaires in one session, while others completed this over several days.Note that the platform calculated the time between the sessions, over days, not per session.
Regarding the PwMND, MNDFRS-R scores were similar in both groups and reflected moderate-severe disability (overall 27/48 at baseline).Behavioral symptoms were prominent in both groups (Table 2).
We performed GLMs to formally test for a between group difference but found no evidence of a difference between groups for any of the questionnaire outcomes (eTable 3).The study was powered for feasibility and not efficacy; hence, caution is required to interpret effect estimates.

Site recruitment
All sites referred recruited participants.Ten were National Health Service (NHS) UK sites and one charitable service was funded by the NHS to provide MND care.In the UK, NHS services are funded through central government via taxes, translating in free care at the point of access; as such, all PwMND for whom the participant were caring for, were receiving free care.Sites varied: five were recognized specialist MND Care Centers, one offered palliative care services, two were community-based and three services were from secondary hospitals.This diversity Online intervention for carers to manage behavioural symptoms in MND 511 meant different frequencies of contact, types of care provided, and team composition.For example, specialist centers usually review families 3-monthly and include a neurologist or palliative care consultant, while community-based teams may have weekly contact, via a team of allied health professionals and nurses.Site principal investigators had varied professional backgrounds, with most being new to research.
Securing 11/24 sites took longer than intended.Three sites fell through, one for lack of communication after initial discussions, and two at the late stages of set up: team ill health, IT policy incompatibility.Another 10 sites expressed interest but were unable to deliver the study for various reasons, e.g.capacity issues despite team's interest.Administrative delays in study set up also contributed to slow recruitment progress, indicating that early identification of or reconnection with previous sites from the feasibility study would be helpful in a future trial.

Carer recruitment
151/284 families of PwMND identified as potential participants were invited to be screened for MiNDToolkit.Site initial screening was inconsistently reported, but the proportion of recruited carers from those invited (19.87%) indicates that recruitment to a larger trial would be feasible.Unfortunately, many reasons for non-response are unknown.
Protocol and procedural changes were implemented to facilitate recruitment and encourage HCP/site engagement: Eligibility: Reduction on number of contact hours with PwMND (from 14 to 7 hours/ week), after carer feedback, recognizing that individuals could be the primary carer but not have many hours of contact.Introduction of mail invitations to enable participants to self-refer to the study, on site feedback.
HCPs were sent thank you cards and care packs (e.g.chocolate) to celebrate achievement of recruitment target or over-recruitment.
Monthly newsletters highlighted sites that performed well, acknowledging their performance and engagement.Letters of invitation sent directly to potential participants not only facilitated study recruitment; it also demonstrated to HCPs that PwMND with BehSymp could be missed if not assessed.The gap between HCPs' knowledge of, and experience in managing BehSymp, was also identified in the MiNDToolkit process evaluation (14).Uncertainty of how to deal with families' response to the study may have initially created a barrier to study referral, which was likely overcome by the carers' positive response to the feasibility trial mail out.
Carer retention rate was excellent compared to the few other studies focusing on MND carers, where almost 50% of carers dropped out (24), while in our study, 82.76% completed follow-up measures.
Potential outcome measures for future studies were explored in this study: carer quality of life, depressive symptoms, and anxiety.These are psychological wellbeing domains often evaluated in studies involving carers of PwMND (9,25), but unlikely candidates for future trial as median scores were mild for anxiety and non-clinical range for depression.There is evidence in other contexts that psychoeducational interventions can reduce anxiety symptoms in carers (26,27), but this is not a universal finding (28).Quality of life may not respond well to the intervention, as the potential benefits of MiNDToolkit may be tempered by the uncertain trajectory yet rapid progression of MND.Carers face numerous losses in rapid succession in MND, and their wellbeing will be affected by multiple factors.Outcomes relating to carer competence or skills, are more likely to respond to a psychoeducational intervention as shown in other carer trials in stroke (29) and dementia (30).Indeed, feedback from carers through the process evaluation (14) revealed that learning about BehSymp, and perceiving an ability to make changes in daily management was empowering and supported acceptance of these nonmotor symptoms.

Strengths and limitations
Strengths include high engagement of participants in MiNDToolkit usage and study completion, compared to another trial involving carers of PwMND (31), and echoes findings of carer demand for individual support (32).Our carers were diverse regarding ethnicity and employment status.Statistics showed that carers accessed MiNDToolkit early or late in the day; the online nature of MiNDToolkit made it accessible for carers who spend many hours providing care.Additionally, carer feedback in interviews (14) confirmed that MiNDToolkit was easy to use and engaging.
We identified simple modifications that can be made to the platform and training, e.g.additional video to demonstrate HCP role play, and paper documents as memory aides for the HCPs (14).
Limitations include small sample size and lack of blinding.One site was randomly allocated only carers in the control arm, thus limiting their ability to reinforce the intervention, which future per site randomization approaches would rectify.Reinforcement from the HCPs was variable and intended to be higher (14).This could be addressed via changes in the training and materials provided to increase clarity of the HCP role in the intervention, or increasing hours dedicated to the trial.As it was not possible to report the length of engagement per platform interaction, we were unable to examine variables that may have influenced this.

Conclusion
To the best of our knowledge, the MiNDToolkit is the first multi-center RCT to investigate the feasibility of testing an online intervention for carers dealing with BehSymp in MND.Currently, no guidelines for the management of BehSymp in MND exist.Carers' need of specialist support cannot be underestimated.MiNDToolkit is an acceptable, feasible psychoeducational intervention that addresses the needs of carers-and HCPs, to learn, understand and successfully manage these symptoms.A definitive RCT is warranted.potential sites for a future fully powered clinical trial.Start site set-up contracting a year before the sites are due to receive green light.Approach several sites concomitantly.
Create mini video clips to support HCPs in further understanding which actions are research processes, and which actions are parts of the MiNDToolkit intervention per se.Provide specific training and associated visual materials for completion of the screening log.

Figure 2 .
Figure 2. MiNDToolkit feasibility trial recruitment progression from first site opening for recruitment in July/2021, to last (11th) site starting site recruitment in November/2022.

Table 1 .
MiNDToolkit modules offered to carers.Compulsory modules were presented at the beginning and end of the menu to all.Modules on symptoms and strategies were tailored to carers' questionnaires responses or were made available by the HCP.
a FU: follow-up after 3 months.b Exact 95% confidence interval used (Clopper-Pearson method) due to small sample size and proportions often close to 0 or 1.

Table 3 .
MiNDToolkit feasibility trial: carer descriptive statistics at baseline and follow-up.