Design
ssSIP is a multicentre, prospective, assessor blinded, three arm randomised controlled feasibility trial with embedded process and health economic evaluation.
Participants
Adults with a new moderate to severe dysphagia will be recruited from inpatient acute stroke and rehabilitation hospitals, from 7 days to 2 months after a new stroke. Moderate to severe dysphagia is defined as a Dysphagia Severity Rating Scale score of ≥4 OR safety concerns with thin fluids identified by the clinical speech-and-language therapist (SLT). Participants will have sufficient visual, cognitive and communication skills to participate in the intervention which will be verified by a screening test. This will involve being able to perform a volutional swallow, being able to demonstrate ability to point to their swallow signal on the screen and follow the instructions of the task. Clinical teams will obtain written consent from participants or their proxy if they do not have the capacity to consent.
Participant inclusion criteria
- Adults over 18
- Clinical diagnosis of stroke
- >1 week and < 8 weeks post stroke
- New moderate to severe dysphagia (DSRS ≥4) OR patients where the SLT has identified safety risks with full amounts thin Level 0 fluids
- Not rapidly improving dysphagia (2 clinical SLT assessments over a week showing minimal change in DSRS)
- Pass an eligibility screen – ability to perform a volitional swallow and have sufficient visual, cognitive and communication skills to participate in the intervention
Participant exclusion criteria
- Medically unwell, GCS <10, on >4L oxygen, poor prognosis, end of life care
- Previous dysphagia
- Degenerative neurological condition
- Patient likely to be repatriated to or rehabilitated at another organisation with 10 days
- Participation in another trial aimed at improving dysphagia
- Unwilling to remove beard/hair from under chin
Approvals & amendments
ssSIP has been approved by the Bromley Research Ethics Committee (Ref: 23/LO/0131), and by local research and innovation teams and is registered on ClinicalTrials.gov (NCT05744245).
A substantial amendment to include participant interviews in the protocol and a non-substantial amendment clarifying inclusion criteria were approved on 18 August 2023 and 29 January 2024 respectively.
Randomisation, allocation concealment and blinding
Once a participant has consented and has passed the eligibility screening assessment they will be randomized to one of three groups; ST-0 which is no treatment sessions per day, ST-1 is one treatment session per day and ST-2 is 2 treatment sessions per day. All three groups will continue to receive usual care (UC). Participants will be randomized by minimization on age, stroke severity and dysphagia severity through the secure server of the database. This will ensure allocation concealment. Outcome assessors will be blinded to group, but participants, therapists and clinical teams will not be blinded given the nature of the intervention.
Trial procedures
All trial procedures are summarised in Table 1.
Table 1 Schedule of enrolment, assessments and interventions.
|
|
Screen
|
Baseline
|
Day 1-14
|
Day 15
|
Discharge/ inpatient death
|
Day 90
|
|
Location
|
Hospital
|
Hospital
|
Hospital
|
Hospital or outside
|
Hospital
|
Hospital or outside
(Phone)
|
|
Eligibility
|
+
|
|
|
|
|
|
|
Consent/assent
|
+
|
|
|
|
|
|
|
Eligibility screening assessment
|
|
+
|
|
|
|
|
|
DSRS, FSS, NIHSS
Swallow strength & Swallow skill
|
|
+
|
|
+
|
|
+
|
|
VFS (Sub-group) ASPEKT
|
|
+
|
|
+
|
|
|
|
Participant interviews (Sub-group)
|
|
|
|
+
|
|
|
|
Randomisation
|
|
+
|
|
|
|
|
|
ST0 vs ST1 vs ST2
|
|
|
<>
|
|
|
|
|
Targeted outcomes: pneumonia & number of abx for LRTI
|
|
|
<
|
>
|
+
|
|
|
All SAEs
|
|
|
<>
|
|
|
|
|
Intervention-related (S)AEs
|
|
|
<
|
>
|
|
|
|
Fatal SAEs
|
|
|
<
|
=
|
|
>
|
|
Usual care data
|
|
|
|
+
|
+
|
|
|
All-cause mortality
|
|
|
|
|
|
+
|
|
QoL: EQ-5D, EQ-VAS
|
|
+
|
|
+
|
|
+
|
|
mRS, BI, discharge date & location, pneumonia & number of abx for chest infections, DHI
|
|
|
|
|
|
+
|
|
Resource use
|
|
|
|
|
+
|
+
|
BI: Barthel index; mRS: Modified Rankin Scale; DSRS: dysphagia severity rating scale; DHI Dysphagia Handicap Index; EQ-5D: EuroQoL-5-dimension; EQ-VAS: EuroQoL visual analogue scale; FSS; feeding status scale; NIHSS: National Institutes of Health stroke scale; QoL: quality of life; VFS: Videofluoroscopy; SAE: serious adverse events; LRTI: lower respiratory tract infection; SSSIP: Swallow Strength and Skill Training In Post stroke dysphagia; ASPEKT: Analysis of Swallowing Physiology - Events, Kinematics & Timing; Abx: antibiotics.
Intervention
Intervention refinement
The intervention protocol has been further refined since the protocol described by Benfield et al 2023. (10) The refinement process encompassed reviewing the findings from the previous trial, testing the latest technology in collaboration with a public and patient involvement group (PPI) with lived experience of dysphagia and appraising new research literature in neuro-rehabilitation and post stroke dysphagia. The detailed process will be published elsewhere; here we describe the current treatment protocol.
Treatment protocol
Participants receiving the intervention will be set up with electrodes placed on the submental muscles, connected through a surface electromyography device to a laptop. Each therapy session will be 35 minutes and will involve 40 repetitions of a combination of swallow strength and swallow skill exercises with visual biofeedback about their performance on the laptop screen in front of them. The starting point of the exercises is calibrated to the individual and is based on the mean sEMG amplitude of the three effortful swallows which we refer to as the repetition maximum (RM). In each block of strength exercises participants will perform five effortful swallows aiming for a visual target at 100% of their calibrated RM. The exercises become more or less challenging depending on performance. Each block of skill exercises involves performing five swallows with different timing and amplitude targets between 20-50% of their RM (Figure 1). Participants will be asked to aim for the peak timing and amplitude of their swallow to fall within a box on the screen. Visual as well as verbal feedback will be given by those facilitating the intervention. Each session will aim for 20 strength and 20 skill repetitions (swallows) with 60 second breaks after each five-trial block but can be tailored to the individual if the tasks are challenging. Those in the ST-1 group will receive one session per day for 10 days and those in the ST-2 group will receive two sessions per day for 10 days. The therapy will be delivered by trained speech and language therapists (SLT) or SLT assistants (SLTA) who are trained to deliver dysphagia rehabilitation.
Usual care
All participants will receive usual standard care by the stroke team and the speech and language therapy team which usually includes assessment and management recommendations, patient and family education and rehabilitation.
Training and delivery co-design
The training package was designed with a group of five SLT and SLTA stakeholders over a series of online workshops. This also included approaches for implementing high intensity therapy in the acute inpatient stroke setting. The process will be described elsewhere, here we describe the content of the package.
Training package
An initial cohort of SLTs and SLTAs will receive training comprising of a 15-minute theory video and a 90-minute face to face practical training session including practice setting up and using the equipment. Teams will be given a manual detailing how to deliver the training. They will have access to video clips demonstrating set up and delivery of the intervention and crib sheets for use in the therapy sessions to remind them of key steps. A poster about the intervention will also be provided. This aims to spark the interest and support of the multidisciplinary team to help ensure participants are available and ready for therapy sessions. A treatment planning log will also provided so that sessions can be planned in advance. The resources were designed to enable those that have received the training and have some experience delivering the intervention to train up new colleagues if needed. During the trial the SLTs and SLTAs will be able to contact the researchers via email or telephone for any further support. They can also request joint sessions with the researchers if there are any uncertainties about delivery.
Outcomes (including safety & process evaluation)
Primary Outcome – Feasibility
The primary outcome is trial delivery, determined by a) the trained clinicians’ ability to deliver the intervention in real life inpatient acute stroke settings and b) suitability of trial procedures for use in a future larger scale trial. An embedded process evaluation will explore treatment fidelity, effectiveness of training and facilitators and barriers to implementation. The following methods will be used to collect data on different aspects of feasibility.
Recruitment and retention
The number of participants recruited, number eligible for the intervention and number remaining in the trial at the day 90 data collection will all be gathered.
Treatment fidelity
Observation sessions with clinicians delivering intervention at the three sites (6 in total) will be conducted by the research team. A fidelity checklist (Additional file 1) will be used to record fidelity based on the prespecified core components of the intervention. Interviews will be conducted with clinicians across different sites who have been trained and have delivered the intervention (description below). The number of sessions, length of sessions, repetitions of the exercise per session and deviations from the protocol will be collected after each treatment session by clinicians on case record forms. Usual care data will be collected at day 15 and discharge.
Feasibility of implementation
Interviews will be conducted with trial participants and clinicians who delivered the intervention around feasibility of implementation. Interviews will be semi-structured and iterative, the topic guides (Additional files 2 and 3) will be informed by the Conceptual Framework for Intervention Fidelity (CFIF) (11), Conceptual Framework for Implementation Research (CFIR) (12). The latter also includes the key domains of the COM-B model (13) which is relevant for the participant interviews. Clinical teams will be able to contact the research team for support with intervention delivery during the trial and the type of problems encountered will be recorded in a request for support log.
Feasibility of trial processes
To ensure trial processes are feasible and can be carried forward to a future trial, completeness of case record forms at each follow up stage and the use of a participant diary will be captured.
Progression criteria for taking the trial forward into the next stage are proposed to be:
- At least one participant recruited per month per site using any necessary strategies required to achieve this
- Over 80% of the dose is delivered or facilitators identified to achieve this.
- Over 80% of the core components of the intervention are delivered or facilitators are identified to achieve this.
Additional file 4 further outlines the progression criteria including criteria for determining feasibility of trial processes.
Secondary outcomes
Swallowing, clinical, quality of life and health economic secondary outcomes will be collected to explore treatment effects and cost effectiveness of different doses of the intervention:
Dysphagia severity rating scale (DSRS)
The DSRS is a validated tool to grade the severity of swallowing impairment in post stroke dysphagia (14). It is a 13-point scale, capturing diet and fluid intake and level of supervision required and higher scores indicate more severe dysphagia. It will be used to determine treatment effect of the different doses of ST on dysphagia severity at days 15 and 90 and speed of recovery.
Swallow strength and skill
Swallow strength is estimated by mean normalized effortful swallow (15) captured by calculating the mean difference between five effortful swallows and five regular swallows using sEMG and the BiSSkiT software. Swallow skill is estimated by percent of accurate attempts at swallowing at a target time and amplitude window on the BiSSkiT software. Similar paradigms have previously been described (16, 17) to capture ability to modulate swallowing based on peripheral stimuli which is necessary in safe swallowing of food and liquids. Swallow strength and skill assessment will be carried out at baseline and at day 15 to explore treatment effect on swallow strength and skill.
Analysis of Swallowing Physiology: Events, Kinematics and Timing (ASPEKT)
Participants at one site will be invited to have a Videofluoroscopy (VFS) at baseline and day 15. The assessment protocol will be embedded in the local clinical protocol and will include four comfortable sips of thin fluids (International Dysphagia Diet Standardisation Initiative, IDDSI, Level 0), one comfortable sip of Level 2 mildly-thick fluids, and one 5ml teaspoon of Level 4 puree diet. All trial stimuli are prepared to 40% weight to volume EZ HD 98% Bracco UK LTD barium and drinks thickened with Nutilis Clear Nutricia. Continuous images were recorded onto DVD with a Phillips system at a frame rate of 25fps. Each recording will be spliced into separate boluses and anonymised. Each bolus will be analysed with the ASPEKT method (18) to explore any treatment effects on swallow physiology.
Clinical outcomes
Mortality, number of chest infections and length of hospital stay will be captured at day 90.
Quality of life; EuroQol 5 Dimension 5 Level Scale (EQ‑5D‑5L) and Dysphagia Handicap Index (DHI)
The EQ-5D (19), a valid and reliable tool for measuring health related quality of life will be collected at baseline, day 15 and day 90. The DHI has undergone a degree of validation with mild to moderate dysphagic patients including an unspecified number of stroke patients. It was found to differentiate between controls and those with dysphagia, has high internal validity and test–retest reliability, and is sensitive to significant differences in severity of dysphagia (20)19]. PPI members felt that participants early post stroke may not have sufficient insight into their dysphagia so the DHI will be collected at day 90 only.
Resource use
Data on resource use will be collected from participant’s notes at day 15 and discharge. On discharge, participants will be given a diary to record any contact with health professionals related to their swallowing up to day 90. On the day 90 follow up participants will be asked questions about their health and the services they have accessed since discharge referring to the diary.
Safety
The intervention has an excellent safety record (10) but it is important to collect any related adverse events. Therefore, adverse and serious adverse events (SAEs) will be captured from 0-15 days, procedure-related SAEs over days 0-15 days, and discontinuations due to SAEs and fatal SAEs over days 0-90.
Data collection & management
A trial database (REDCap) will be used by local site investigators to enter data in electronic case record forms. The investigators will use the database to collate and organize anonymised data, this ensures secure storage and maximises data quality. Local trial sites will collect and store personal data for participants for one year after the trial ends in order to follow up and share findings of the research, after which it will be archived.
Trial management (including data monitoring)
The trial coordination centre will be responsible for set up and day to day management of the trial. They will also carry out central and remote data monitoring to ensure timely and complete data entry. Remote monitoring will consist of reviewing the site file and checking that source data matches the collected data and trial protocol and procedures have been followed. A proportionate steering/trial management meeting will be held quarterly with co-investigators and two independent members. Progress will be monitored by the committee and decisions made to continue with the trial. A data monitoring appointee will review the data (unblinded to treatment allocation) mid-way through the study to review the scientific merit of the trial and verify there are no related adverse or serious adverse events.
Sample size
This is a feasibility study and therefore it is not powered to detect a statistically significant intervention effect. There is no definitive recommendation for participant numbers in feasibility studies [42]. Sample size target will be 120, this will allow for a dropout rate of 10% and an estimated 4% who will fail the screening assessment to achieve 34 participants per group for final analyses. Similarly the VFS is exploratory and not powered, but a pragmatic sample size of 30 (10 per group) will be used. Given the incidence of dysphagia and data from previous trials (10) this recruitment rate of 1-2 per month per site over 24 months should be feasibly.
For the clinician & participant interviews, the maximum sample size for each will be 12 due to what is practical for this trial. Fewer participants may be needed to achieve sufficient information power (21) as the aim of the study is narrow, the sample is highly specific to the study, established theoretical frameworks have been used to plan the topic guide and will structure thematic analysis, the quality of dialogue should be strong as interviewers are experienced clinicians. Clinicians will be recruited across all sites, and participants from one site.
Statistical methods
Feasibility & Process evaluation data
Quantitative data e.g. intervention session length, number & type of usual care sessions etc. will be analysed with descriptive statistics using SPSS. Normally distributed data will be presented as means ±SDs; non-normally distributed data as medians (interquartile ranges).
Qualitative data e.g. interviews and observations, will be analysed using a qualitative descriptive approach (22) which involves largely documenting what the participants say and summarizing without extensive interpretation. Transcripts will be coded by two separate individuals on the ssSIP trial using N-VIVO. The data collected will be analysed using thematic analysis (23) as a way of identifying, analysing and reporting any meaningful patterns and themes. Data familiarisation, line-by-line coding and development of broad themes will be used. The themes will then be mapped on to the abovementioned CFIR, CFIF and COM-B model framework domains.
Exploratory analysis of secondary outcome measures
The trial is not sufficiently powered to formally assess efficacy. However, exploratory analysis will be conducted on an intention to treat basis to compare treatment groups using ANOVA for parametric data, Kruskal-Wallis for ordinal data and Chi-squared for dichotomous outcomes. Further planned analyses including are detailed in the Statistical Analysis Plan (Additional file 5).
Health economic evaluation
A pilot health economic evaluation will be conducted to test the health economic methods for use in future definitive economic evaluations and feasibility of data collection and sensitivity of measures of health-related quality-of-life measures.
The primary health economic analysis will adopt an NHS and personal social services cost perspective in alignment with NICE guidance (2022) (24). Secondary analysis will incorporate a broader societal perspective to account for wider impacts of the stroke dysphagia, such as time lost from paid employment, and effects on families and friends.
Data will be gathered using a purposively designed patient resource proforma, which patients will complete themselves. This tool will collect comprehensive data on all aspects of patient treatment and follow-up, including medication, equipment and supplies, inpatient and outpatient hospital visits, social care, primary and community care, time lost from work, and informal care.
Key collection points for resource use data are set for day 15, discharge, and day 90. Patient/carer diaries, distributed at hospital discharge, will be trialled to collect more accurate data on resource use, such as healthcare professional input, during data collection on day 90. Intervention-specific costs, including treatment (equipment), staff time, and staff training to deliver the intervention, will be recorded using intervention logs completed by the study team.
The outcome measure for the economic evaluation will be the number of QALYs (Quality Adjusted Life Years) based on a 90-day time horizon. The EQ-5D-5L questionnaire will be administered at baseline, day 15, and day 90.
An incremental approach will be applied across the three arms, calculating an incremental cost-effectiveness ratio (ICER) and incremental net monetary benefit. Cost Effectiveness Acceptability Curves (CEACs) will illustrate the probability of effectiveness relative to willingness to pay at the NICE threshold of £20-30k per QALY gained. Sensitivity analyses will examine key cost drivers.
Dissemination
Trial results will be written up and shared in academic journals, through conference presentations and through webinars for clinicians and patients, further dissemination will be informed by discussions with the PPI group.