Feasibility trial design
The current study was a two-arm randomised controlled trial (RCT) which investigated the feasibility of carrying out a definitive RCT to evaluate the effectiveness of an early social communication intervention in addition to standard NHS speech and language therapy (SaLT) (compared with standard NHS SaLT alone) for young children with Down syndrome. The protocol was developed in line with the Standard Protocol Items: Recommendations for Interventional Trials [22] and published [23]. The results are reported in line with the CONSORT extension to pilot and feasibility trials [24].
Setting
The study was conducted in three NHS sites in England, providing SaLT services across three geographical regions: Berkshire Health NHS Foundation Trust (BHFT), Oxford Health NHS Foundation Trust (OHFT) and North-East London NHS Foundation Trust (NELFT), and each site had a principal investigator. All assessments were conducted remotely using either online or paper questionnaires, with support by telephone. This was in response to the COVID-19 pandemic when most face-to-face services stopped. The protocol was amended so that there was no face-to-face contact between the research team and the participants and their families.
Public and patient involvement (PPI)
The study protocol was developed with the help of our PPI representatives, which consisted of two parents of children with Down syndrome. They contributed to the finalising of the procedure, improving the readability of the parent manual and actively contributing to all decision-making regarding the feasibility trial as members of the Trial Steering Group which also included a highly specialist paediatric speech and language therapist, all three principle investigators, an independent statistician, an independent clinical research consultant, a developmental and heath psychologist with an interest in child development, a professor of developmental clinical psychology with a track record of clinical trials, the clinical trial manager and a representative of the sponsor (Berkshire Health Foundation Trust).
Participants
a) Children with Down syndrome and their parents/carers
Children with Down syndrome and their families were recruited through SaLT services in BHFT, OHFT and NELFT who distributed information sheets and consent forms about the study to the parents/carers of every child with Down syndrome between the ages of 11 and 36 months on their caseloads. It was up to the parents/carers to decide whether one parent or both parents/carers participated and delivered the intervention. Parents who either declined to participate or did not engage with the research project were invited by their child’s SaLTs to give their reasons for not participating.
Eligibility criteria
Inclusion criteria
-
1.
Parent or guardian willing and able to provide informed consent on behalf of participant
-
2.
Confirmed diagnosis of trisomy 21 (Down syndrome)
-
3.
Male or female child, 11 to 36 months old at study entry
-
4.
Parent/guardian has the literary and language skills needed to use the parent intervention manual.
-
5.
The participant is not currently taking part or due to take part in a language-based intervention study.
Exclusion criteria
-
1.
Children with comorbid conditions (for example autism spectrum disorder) as determined by the principal investigator for each NHS site
-
2.
Any reason that may hinder participation, such as complex health issues requiring repeated hospital admissions
-
3.
Prior knowledge of the intervention as specified in the parent manual
b) Speech and language therapists
SaLTs were recruited through clinical and professional networks and current NHS sites to take part in an interview on parent-delivered interventions, the acceptability of the intervention from a SaLT service delivery point of view and their views on clinical trials.
Inclusion criteria
-
1.
Currently practising SaLT in the UK with a paediatric caseload (duration of practising with a paediatric caseload was not considered)
-
2.
Currently working within the NHS or having recently worked within the NHS (not more than 2 years have passed since last NHS post)
Procedure and intervention
The SaLTs supporting the children who took part in the intervention attended a 1-h training session delivered by the research team on the main goals of the intervention and the different stages so that they could support the parents if needed. The SaLTs had the opportunity to ask any questions and to feed into the intervention materials prior to the participant recruitment. The SaLTs acted as a point of contact for the families, to support the parents/carers with delivering the intervention when needed.
The intervention is focused on promoting and supporting the development of early social communication skills and, in particular, the child’s ability to respond to shared attention. During the sessions, the parent used the toys provided to encourage their child to engage in shared attention. A shared focus of attention can be achieved through seven levels, depending on the child’s developmental stage. During the sessions and over the course of the 10 weeks, parents/carers progressed through different levels of responding to joint attention from level 1 (the adult gently puts the child’s hand on object to signal that the specific object is the focus of attention) to level 7 (the adult placed a toy outside the child’s visual field, and the child then followed the adult’s gaze to establish shared attention with the object) (see Appendix 6 for a full description of the different levels based on Whalen and Schreibman [25]). Once shared attention was established (through any of the levels), the adult used this as an opportunity to provide rich language input to the child by the following: (a) labelling the object/labelling activities around the object (e.g. it is a bus; it drives around); (b) describing the toys/objects in terms of colour, shape, size and noise they make and how we can play with the toy; and (c) inviting the child to play/interact with the object. The approach is based on the social-pragmatic account of language acquisition, which assumes that shared attention and understanding the intentions of another person are the prerequisite skills for language development [26]. The parents recorded every session they had with their child on the diary form provided (see Appendix 1) and sent weekly to the trial manager.
Recruitment of children with Down syndrome and their families
SaLTs from participating NHS trusts identified potential participants (children with Down syndrome) and their parents by reviewing their caseloads against the inclusion criteria. They then introduced the study to the parents of all potentially eligible children at a routine appointment or via email/telephone call and provided the parent/caregiver with a participant information sheet giving details of what study participation would involve. They invited them to contact the research team if interested in participating or if they had questions. Parents/carers took the time they needed to consider participation and were able to delay study entry by a few weeks to fit in with other family commitments. Where possible, reasons for nonparticipation were gathered by the child’s SaLT.
Sample size
The target sample size was 25 children with Down syndrome based on literature recommendation of a minimum of 24 participants for feasibility studies [27,28,29] in order to estimate a standard deviation (SD) for the purposes of informing a subsequent sample size calculation.
Randomisation and allocation
After written consent was given by the children’s parents/carers and baseline assessments completed, the children with Down syndrome were randomised by the clinical trial manager or other designated team member via Sortition® (a secure web-based clinical trial randomisation software developed by the University of Oxford) using block randomisation to receive standard care (Control) or standard care plus the intervention (Intervention) in a 1:1 ratio, stratified by site, to account for regional differences in standard care. Following randomisation, the parents/carers were contacted by the research team, who explained their child’s study allocation, what that meant and what would happen next.
Blinding
Due to the nature of the intervention (parent delivered), it was impossible to blind the parents and the children’s SALTs to group allocation. The research assistant, who administered the pre- and post-intervention measures and who entered all the data, was blind to group allocation and also did not have access to the intervention materials until the end of the project.
Intervention group
Once randomisation was complete, the intervention manual (paper based and printed using non-tear paper), blank diary forms (Appendix 1), a bag of age-appropriate toys and links to short video demonstrations of the intervention were sent to the parents/carers in the intervention group. The video materials showed each step of the intervention via short clips (2–3 min long). Each video clip was clearly labelled as to which level of shared attention was being shown. Each video showed a parent–child interaction, and there was a commentary (as a subtitle on the screen) explaining what the parent is doing and how the child is responding. The intervention was designed to be delivered by parents/carers over a period of 10 weeks. The intervention was delivered in the child’s home by one or both of their parent(s)/carer(s) over 10 weeks. Parents/carers were advised to deliver it for 1 h in total each week, over three to six individual sessions. The sessions could last between 10 and 20 min, and parents/carers chose how to allocate the time.
Support to deliver the intervention from the child’s SaLT was available at the request of parents by telephone/email. Parents also had access to the principal investigator of each site and to the chief investigator. The SaLTs and investigators recorded all contacts from parents including duration and content of each contact.
Families continued to access standard NHS SaLT for the duration of the project. All contacts related to standard care were recorded, including the duration, number of contact points and activity type (assessment, advice, intervention/review).
Control (comparator) group: standard care
The control group received standard NHS SaLT care for this patient group. Standard care varied depending on each individual child’s needs and on the pathway specific to each NHS site, ranging from two contacts per year to monthly contacts. This typically included assessment, advice being given to parents on how to support their child’s general communication skills, review and intervention on feeding and use of baby sign Makaton. The SaLTs recorded all contacts with the family for the duration of the study.
At the end of the 6-month follow-up period, all families in the control group were provided with the intervention manual, accompanying materials and video links and had access to their child’s SaLT and/or members of the research team for support with delivering the intervention.
It is important to note that standard care did not include receiving a structured manual on how to support their children’s early social communication skills including shared attention nor videos showing how a child’s responding to shared attention could be supported. Standard care was more holistic covering different aspects of child development including communication, whereas our intervention specifically focused on building the children’s early social communication skills and language through shared attention skills, and it was very structured.
Measures
Assessments were administered at baseline, immediately post intervention (10–14 weeks after baseline) and follow up (6 months later) and scored by a research assistant blind to group allocation.
Primary outcome measure
I) Reading Communicative Development Inventory (R-CDI) [21] is a widely used parental checklist which assesses receptive and expressive language based on the MacArtur-Bates Communicative Development Inventories [30,31,32]. Parents are asked to tick the words their child understands, understands and says or understands and signs. We computed children’s understanding of words and expressive language. Expressive language was measured by adding together all spoken words and signs the child was reported to use. For the bilingual children, a total vocabulary was computed which included a sum of all the words (spoken, signed and understood) in both/all of their languages. The CDI was chosen as the primary outcome measure because the aim of the intervention is to increase children’s language by increasing their vocabulary and this measure directly assesses expressive and receptive vocabulary. Parental reports of language (such as the CDI) are widely used because parents have extensive experience with their children in a variety of naturalistic settings [33]. The MacArthur-Bates Communicative Development Inventories have been widely used in theoretical studies and in studies of importance for public health [31, 32, 33). This measure has a reasonable predictive and concurrent validity. For example, children’s scores on the CDI at ages 2 and 3 correlate significantly positively with standardised receptive language measures [33]. Importantly, the concurrent validity of the CDI has been established for children with Down syndrome [34]. The measures of children’s language obtained on the CDI and standardised measures of language correlated strongly between 0.70 and 0.82.
Secondary outcome measure
Ii) Communication and Symbolic Behaviour Scale (CSBS) [35]: This is a norm-referenced standardised tool available as an online or paper questionnaire, completed by parents/carers. The Infant–Toddler Checklist (which is part of the CSBS) has sensitivity of 78% and specificity of 84% [35]. It assesses communicative functions, gestural communicative means, vocal communicative means, verbal communicative means, reciprocity, social-affective signalling, and symbolic behaviour. This measure was chosen as a secondary outcome measure because the intervention focuses on increasing children’s early social communication behaviours and these can be assessed using this scale. The measure has reasonable reliability and validity [35].
-
b) Quality of life
-
The Infant Toddler Quality of Life (ITQOL-SF47) is completed by parents/carers and is a measure of infant quality of life. It is a reliable, valid and precise measure, and it was found to exceed item-level scaling criteria [36]
-
The Adult Quality of Life Questionnaire [37] is completed by parents/carers and is a measure of parent/carer quality of life. It is a simple non-standardised instrument suitable for use with adult carers that measures quality of life in eight separate domains: support for caring, caring choice, caring stress, money matters, personal growth, sense of value, ability to care and carer satisfaction. The questionnaire was developed by initially putting together information from a range of sources including review of the literature on carers, scales used in previous carer research, an expert informed panel, and with the involvement of carers.
-
The Hospital Anxiety and Depression Scale [38] is completed by parents/carers and is a self-assessment measure of symptoms of anxiety and depression. The scale is fully described in Zigmond and Snaith [38]. The internal consistency was established by item-subscale correlations, and significant associations of between 0.76 and 0.41 for the anxiety scale and between 0.60 and 0.30 for the depression scale were reported. Subsequent studies established further the psychometric properties with Cronbach’s alpha of 0.93 for the anxiety scale and 0.90 for the depression scale [39]. The validity of the scale was found to be satisfactory by Clark and Fallowfield (1988) [40].
In addition, all parents/carers of participating children completed a demographic questionnaire devised by the research team at baseline which asked questions about parental age, employment status and education and also contained the Vineland Adaptive Behaviour Scale [41], which is a standardised measure used to assess the children’s general cognitive and adaptive abilities and has acceptable validity and reliability and is the assessment of choice for educational, clinical and research purposed. For full details and a critical evaluation of its psychometric properties, see Pepperdine and McCrimmon [42].
All questionnaires were completed by parents/carers using online links, or using paper copies posted to the participants, with support from a member of the research team if required. The Communication and Symbolic Behaviour Scale (CSBS) was administered over the phone.
Adherence to the intervention and contamination
Adherence to the intervention was monitored by asking the parents/carers to complete a weekly diary (see Appendix 1). The diary provided information on how many sessions the parent/carer carried out with their child, their duration, the number of different toys used at each session, the level at which they were working and additional comments. The principal investigators contacted the parents/carers by telephone in weeks 4 and 8 to check adherence, with a window of + / − 7 days. They asked a standard set of questions specifically designed to obtain information on how closely the parents were following the manual, how often they carried out the intervention sessions and their duration, the range of toys they used and what they usually did in order to engage their child (see Appendix 2 for a standard set of questions asked).
Contamination was assessed at the following: (1) study entry for those who were randomised to the intervention group and (2) before the final follow-up for those randomised to the control group. Parents/carers were sent a short questionnaire (Appendix 3) asking them whether they were familiar with the social communication intervention, whether they had seen the materials or whether they had seen the intervention being carried out. There was no evidence of contamination during the trial based on the data provided by the questionnaires. Given that the trial was ongoing during the height of the COVID-19 pandemic, the opportunities for parents/carers who live in different regions, to share the hard copies of the manual and toys provided, were minimal. Importantly, none of the participating families reported having seen the intervention manual and materials prior to being provided with their own set of the study materials.
Parent/carer satisfaction with the intervention
Parent/carer satisfaction with the intervention was evaluated via a brief questionnaire comprising six questions (see Appendix 4), the first five of which were closed questions and the sixth, asking for comments on the intervention, was open-ended. Only parents/carers who completed the intervention were sent the questionnaire.
Acceptability of the intervention to SaLTs
We invited all the SaLTs who had been involved with the current study to take part in in an interview with a member of the research team. We also opened the call to a wider group of SaLTs who had not been involved in the current study across the three NHS sites and SaLTs in other areas of the UK working in paediatric services for children with complex needs to take part in an interview as we wanted to have a broad range of views on this type of intervention and in SaLTs being involved in clinical trials, not only those who had been involved in the project. We distributed the information sheet through the Down Syndrome Research Forum, professional networks, through social media and SaLT managers and personal contacts. We aimed to recruit a diversity of SaLTs specifically in relation to gender, ethnicity and geographical area. In the last 3 months of the study, SaLTs were interviewed to explore their views on a parent-delivered intervention for young children with Down syndrome and their views and willingness to participate in future RCTs.
A topic guide was developed by the researchers addressing the study aims and used flexibly following the lead of the interviewees (Appendix 5). Data were collected via one-to-one interviews conducted via MS Teams. These lasted between 25 and 30 min each. Researchers reviewed and edited the interview transcripts auto produced by MS Teams.
Analysis
Descriptive statistics (socio-demographic, language and cognitive abilities, and health status) were collated and summarised. Parent/carers’ satisfaction with the intervention was also summarised.
Quantitative analyses
All statistical analyses were prespecified in a statistical analysis plan (SAP) which was agreed and signed off by the trial statistician and chief investigator prior to commencement of any analyses. As this study only aimed to address feasibility objectives, no formal hypothesis testing was undertaken to make between-group comparisons, but rather summary statistics were calculated by allocated group and overall at each time point. For each outcome, point estimates of standard deviations (SDs) as well as associated 60%, 80% and 95% confidence intervals (CIs) are presented in line with Browne’s recommendation to use the limit of the one-sided 80% CI for an SD obtained from a pilot study to inform subsequent sample size calculation [43].
Qualitative analyses
The data from the parental responses (n = 11) regarding their reasons for not taking part in the study were summarised.
The data from the interviews with the SaLTs on their views of a parent-delivered social communication intervention were analysed using reflexive thematic analysis [44], an approach to data analysis that enables patterns of meanings across a dataset to be developed. Data were coded inductively using an essentialist approach to report the experiences, meanings and reality of participants [45]. The third author familiarised herself with the data and generated codes using NVivo (Version 20.6.1.1137). Codes were then discussed with the first author and refined. Where disagreements arose, agreement was reached by consensus. The third author then grouped the codes into themes. These themes were discussed and refined by the first author.
Add Comment