Objectives
The objectives were to gauge consensus between HPs and parents of children with neurodisability on the interventions for EDSD, and their outcomes, that should be evaluated in future research, and to agree on frameworks for such research in the NHS.
Methods
Participant recruitment and selection
We sent invitations for the workshops to HPs and parents who took part in the national survey and/or the Delphi survey who had expressed an interest in taking part in subsequent stages of the research, and to members of HP professional networks (via e-mail). Parents who were interested in taking part in the workshops provided their location, the age of their child (preschool, primary school or secondary school) and the nature of their child’s EDSD (physical, non-physical or mixed). HPs stated their profession, the types of EDSD that they worked with (physical, non-physical or mixed) and the geographical location and types of services that they worked in (community, secondary care, tertiary service and education). We purposively selected participants to maximise variation in their experience of EDSD and service provision.
Procedure
We held two half-day workshops: one in Newcastle and one in London. Members of the research team designed and facilitated the workshops. Jeremy Parr, Lindsay Pennington and Helen Taylor led both workshops, with other members of the research team facilitating. The Newcastle workshop was facilitated by Charlotte Buswell, Allan Colver, Deborah Garland, Christopher Morris, Helen McConachie, Johanna Smith and Julian Thomas, and the London workshop was facilitated by Jill Cadwgan, Deborah Garland, Diane Sellers and Johanna Smith. We aimed for detailed discussion about the frameworks of future research, including the PICOTS and the elements of study design that related to feasibility of research in the NHS. At the start of the workshops, workshop leaders presented an overview of the study stages and the findings from the completed stages. Attendees then discussed individual topics in small groups. Most groups contained a mix of parents and HPs. One member of the research team facilitated each small group; notes from the discussions were written on flip charts.
The two workshops were iterative, with the results of the first workshop in Newcastle presented to participants at the second workshop in London. To do this, Helen Taylor summarised the notes from the Newcastle workshop. Jeremy Parr, Lindsay Pennington and Helen Taylor formed bullet points from the summaries to be presented in the small groups at the London workshop.
Topics discussed in the small groups included:
Interventions for EDSD. The concept of an intervention ‘toolkit’, which was first proposed and discussed at the second round of focus groups and formalised during subsequent co-applicant meetings, was presented (see ). The following aspects of a toolkit intervention were discussed –
How could the essential interventions identified in the Delphi survey be presented to parents as a list of treatment options?
What level of detail would parents need on each intervention?
How would a menu of treatment options be individualised?
What level of support would families need from HPs to use the toolkit?
Ways of measuring outcomes.
We presented the list of essential outcomes from the Delphi survey. Participants suggested the tools that they knew from professional and personal experience that could be used to measure each outcome, and discussed the pros and cons of each tool. In both workshops, one group of parents considered the three published parent-report measures of Children’s mealtime behaviour, Parent behaviour and Food intake that had been identified as the highest quality in the measurement properties review – BPFAS, PediEAT and a food frequency questionnaire.
Designing future research – defining treatment as usual; the concept of the toolkit and its implementation, including feasibility and barriers; unit of randomisation and acceptability of RCTs in EDSD; methods of recruitment and selection.
Parents received a £100 shopping voucher as a thank you for their time and to cover travel costs. Professionals received a £25 shopping voucher to cover their travel costs.
Data processing
Flip chart feedback compiled during the workshop discussions and the notes made at each group formed the data for analysis. Morag Andrew and Lindsay Pennington reviewed all data, identified key themes and presented findings to the full research team for discussion.
Results
A total of 15 parents and 19 HPs took part in the workshops. Nine of the parents had children with physical EDSD, two had children with non-physical EDSD, two had children with mixed EDSD and two had one child with physical EDSD and one child with non-physical EDSD. HP participants comprised speech and language therapists (n = 5), dietitians (n = 4), paediatricians (n = 4), occupational therapists (n = 3), clinical psychologists (n = 2) and a physiotherapist. One member of the NIHR-appointed external steering committee (speech and language therapist) participated in the Newcastle workshop and one (nurse) in the London workshop.
Use of eating, drinking and swallowing difficulty interventions in current clinical practice
Participants reported that multiple EDSD interventions are used in current clinical practice, but in an unstructured and unco-ordinated manner.
Toolkit of interventions
Participants agreed that no single intervention would be suitable for all children with EDSD; for many children several interventions may be delivered concurrently or sequentially. Participants liked the concept of an intervention ‘toolkit’ that parents and HPs could use together to identify the most appropriate interventions for individual children and their families.
Participants agreed that the toolkit should be represented visually and available in digital and hard copy, with interactive properties to aid communication with HPs. They thought that it should be flexible to allow families and teams to individualise intervention selection. Parents thought that some parents would want to have ownership of the toolkit and to be integral to decisions on which interventions were selected for their child. In that context, parents and HPs agreed that thorough information on each available intervention would allow families to share decision-making with HPs.
Support for families using the toolkit
Health professionals thought that toolkit use should be supported by a lead HP (e.g. a speech and language therapist) and MDT. Professional support would be required throughout toolkit use and may include psychological input; however, the nature of the support that was required would vary between families.
Participants raised several issues about the delivery of an intervention toolkit:
how to deliver the toolkit to meet the needs of a heterogeneous population with diverse EDSD
how to deliver the toolkit in geographical areas in which multidisciplinary EDSD teams are unavailable or under-resourced
how to avoid/overcome problems caused by delays in obtaining appropriate equipment (e.g. optimising positioning for feeding) to ensure efficient use of the toolkit
how best to deliver the toolkit for children with non-physical EDSD who are not currently linked to HP teams.
Ways to measure outcomes
The focus group and Delphi survey stages identified the important outcomes of any EDSD intervention (see Chapters 9 and 10). These outcomes were organised into three categories for presentation at the workshops:
category 1 – improvement in physical aspects of a child’s EDSD
Safety
General health
Nutrition
Growth
Oral motor control
category 2 – changes in behaviour of child or parent
category 3 – changes in child or family’s well-being
The groups discussed potential measures for each EDSD outcome and their strengths and weaknesses in clinical practice. For implementation in research, measures must be acceptable, valid and reliable. Not all of the measures that were suggested by participants fit all three of these criteria. A full list of measures generated in the stakeholder consultation workshop is provided in Appendix 20.
No acceptable measures that have evidence of reliability and validity were identified for Safety (measures discussed were videofluoroscopy, frequency of chest infections, frequency of choking episodes and observation of meals) or General health (use of a structured questionnaire capturing general health information, school attendance, bowel function including medications needed, skin pallor, hair thickness, dental health, hospital admissions or outpatient attendances, energy levels, concentration and sleep). For Nutrition, participants suggested 3-day food diaries; measuring blood micronutrient levels; food frequency questionnaires; body fat percentage using smart scales and the The Eatwell Guide [Public Health England, UK, URL: www.gov.uk/government/publications/the-eatwell-guide (accessed 23 December 2020)] (see Appendix 20); photographs of food at the beginning and end of a meal, including photographs of the floor to assess spillage; and the need for dietary supplements. Participants identified several measures of Growth including skinfold thickness, bioelectrical impedance, weight and height, and upper arm circumference (see Appendix 20 for further information). Several of the measures suggested for Growth and Nutrition have evidence of reliability and validity, but each would require further investigation in terms of their acceptability and validity in children with neurodisability.
As these outcomes had not been included in the measurement properties review, we undertook a brief review of their properties (Box 6).
Brief literature review of Nutrition and Growth outcome measures
For Oral motor control, participants identified the SOMA, which was included in the measurement properties review.
The measurement properties review showed that the PediEAT and the BPFAS had evidence of validity and reliability as measures of the Child’s enjoyment of meals and Parents’ understanding of child’s eating and drinking difficulties. In the consultation workshop, groups of parents considered these two measures in detail. Parents liked that the PediEAT has clear, specific questions, which feel quick to answer, and that the wording is factual and does not feel judgemental. The tool was deemed likely to identify the main issues and provide sufficient detail to help inform joint goal setting and identify comorbidities. The front page was considered off-putting owing to its length, as it included an outline of intended use, disclaimers and references relevant to the tool. However, the questionnaire was judged as lengthy for use during clinic appointments and better completed during a separate session to enable collection of the most detailed information. Although it is not a measure of participation, parents thought that some items were relevant to participation.
Parents liked the specificity of the BPFAS, but disliked its length. They felt that parents would need an explanation of why it was being used and why it was so long, so that they felt it would be worth the time taken to complete. However, they thought that it would be useful to consider results taken at different time points to review progress using this measure.
Participants did not identify measures of Participation, Quality of life of children or family members or Mental Health; however, generic measures have been validated, such as the CAPE117 and KIDSCREEN,118 that measure children’s Participation and Quality of life, respectively.
Designing future research (including randomisation and recruitment selection)
Usual care
Participants agreed that multiple interventions were recommended by HPs and delivered by parents. However, the way of introducing interventions to families, in terms of the methods of information sharing, the level of detail and the personnel providing the information, was not consistent within or across teams. It was acknowledged that interventions were not introduced in a systematic manner, resulting in inconsistency in treatment as usual.
Intervention delivery
Some participants questioned whether or not a trial of a toolkit of interventions would be necessary, as the toolkit provides a framework for existing practice and may feel onerous if tested within a trial. If a trial was conducted, there was uncertainty about which trial design would be most appropriate; some thought that a RCT may not be appropriate for evaluation of a toolkit containing multiple interventions, selected based on individual need. Case series studies were discussed as a potential alternative study design.
Randomisation
If a RCT design was chosen, participants agreed that randomisation would be acceptable to the majority of families. To avoid contamination between intervention and control arms, it was proposed that cluster randomisation by geographical area/region/service would be more appropriate than randomisation on an individual patient basis.
Population
Participants discussed that a future trial could include all children with EDSD (physical, mixed and non-physical EDSD), or focus on a specific condition or type of feeding difficulty (physical and mixed EDSD or non-physical EDSD). If the trial population included all children with EDSD, stratification by EDSD type or by diagnostic category was suggested. There was some support for focusing on a particular age group (e.g. preschool children), as this may reduce variability in service provision, for example access to therapy services for those children attending special school compared with those attending mainstream school.
Role of tertiary services in trial delivery
It was acknowledged that tertiary services cover large geographical areas with potentially important variability in resource. As tertiary services do not typically lead local delivery of therapy interventions, participants thought that tertiary teams should not lead the use of toolkit interventions directly.
Participant selection
A number of participant selection processes were discussed, including clear inclusion/exclusion criteria. Interviewing was discussed as a possible way of assessing family readiness for trial participation. Participants advised that careful consideration should be given to avoid selection bias.
Recruitment strategy
Participants thought that multiple recruitment sources would need to be considered, and may include:
health – health visitors, GPs, nursery nurses, allied health professionals, community nurses, paediatricians, feeding clinics/teams and dysphagia clinics
research registers (e.g. ASD-UK)
education (including home education communities) – special schools, additionally resourced schools, mainstream schools and independent schools
charities [e.g. National Autistic Society, Bliss (London, UK), SCOPE and Mencap (London, UK)]
local parent/carer groups – raising awareness and identification.
Potential barriers to recruitment
Identification of participants
Participants highlighted that a number of eligible children will not have been formally identified to have EDSD, and may not have recent input from HP teams (e.g. children with ASD prior to diagnosis or children with an ASD diagnosis whose EDSD have not been formally identified and are not under follow-up). In these circumstances, participants thought that staff in education settings may be well placed to identify eligible children.
Agreement of clinical services to participate
Participants thought that HPs may perceive trial participation as a burden on limited time and that the trial could generate an increase in referrals to feeding teams, particularly for children with non-physical EDSD. Current service capacity and treatment costs for any trial would need to be carefully considered.
Parent/carer and child factors
Parental stress, language and literacy, information technology skills and access, parents with additional needs and comorbidities affecting engagement in research were identified as additional potential recruitment barriers.
Challenges to intervention delivery
Participants thought that any trial must be deliverable within the current NHS infrastructure. The challenge of managing local resource variability within future intervention trials was highlighted. Equipment waiting times may also affect delivery of planned interventions. Further challenges may be faced where children attend school out of borough and have a lack of funded local therapy support for home-delivered interventions.
How would services manage the potential increase in referrals?
Participants thought that the actual increase in referrals to feeding teams would probably be small; it would be necessary to stress the potential benefits associated with trial participation to trusts to ensure engagement. It was suggested that trial-related (but clinically relevant) training and resources could be put in place to provide lasting benefit throughout participation.
Summary
The stakeholder consultation workshops showed consistent support from parents and HPs for the concept of a toolkit of interventions that could be worked through by HPs and parents. There was debate about how a toolkit might be best presented. The parents and HPs in the consultation workshops agreed with most of the outcomes that were deemed essential by the Delphi survey participants: Safety, General health, Nutrition, Oral motor control, Growth, Children’s enjoyment of mealtimes (incorporating reduced frustration and distress and associated behaviours), Parental understanding of child’s eating and drinking difficulties, Quality of life of child, Quality of life of family, Parental mental health and Children’s social participation. Ways of measuring most of these outcomes were proposed. Parents also endorsed the Pedi-EAT tool, which was found to have the strongest measurement properties (see Chapter 6). Parents and HPs felt that a trial of a toolkit could be achieved and supported a cluster design in which services were allocated to implement the toolkit or treatment as usual. However, treatment as usual would need defining in comparator group services, as this varies considerably. Participants also identified challenges to the identification and recruitment of participants for future NHS research, including increased referrals to services if the trial was advertised within trusts.
Patient and public involvement in the stakeholder consultation workshops
The parent co-investigators were involved in designing and delivering the stakeholder consultation workshops. The PAG reviewed the materials to be shown to parents and professionals, and commented on the structure of the workshops and the timings of the individual tasks. Changes to the layout and wording were subsequently made to the slide presentation summarising the study and key findings, and a document was created to send out to all attendees prior to the workshops to help provide a background to the study and clarity on what would happen on the day.
Strengths and limitations of the stakeholder consultation workshops
Consultation workshop strengths include participation of both parents and HPs across two diverse geographical areas. Parent/carers of young people with physical, non-physical and mixed EDSD participated, representing a broad range of EDSD experiences. Parent/carers had accessed secondary or secondary and tertiary-level EDSD services. Multidisciplinary professional representation was achieved during both workshops, with participation from HPs working in secondary and tertiary feeding services. The iterative nature of the workshops facilitated collection of detailed information on the topics discussed. We decided not to include young people with EDSD in the consultation workshops because of the abstract nature for older respondents of interventions for young children. This could be considered a limitation of the workshops; however, young people gave their opinion about outcomes during separate young people’s focus groups.
How did the stakeholder consultation workshops inform the next study stage?
The opinions of young people with EDSD on the outcomes identified as important to parents and HPs during the two stakeholder consultation workshops were sought during two young people’s focus groups (see Chapter 12).
Output from the stakeholder consultation workshops was used to generate recommendations, specifically those concerning the further development of an intervention toolkit, outcome measures and the design of future research to establish the effectiveness of an EDSD intervention toolkit.
