Using the TIDieR checklist to describe the intervention of the Sedation and Weaning in Children (SANDWICH) trial

Background: Published reports of complex interventions in randomized controlled trials often lack sufficient detail to allow trial replication and adoption into practice. Aim: The aim of this

clinicians to understand the theory behind the intervention, how it should be delivered, and the resources required.
Relevance to Clinical Practice: The SANDWICH intervention had a significant beneficial effect on reducing time on ventilation for children.The detailed description of the team-based intervention will aid replication, implementation and monitoring of fidelity in other paediatric intensive care units.

| INTRODUCTION
The Sedation and Weaning in Children (SANDWICH) trial was the largest recruiting trial in critically ill children ever undertaken, with over 10 000 children enrolled. 1The stepped wedge cluster trial and process evaluation (PE) evaluated a team-based approach, with greater involvement of nurses in assessing and weaning children from sedation and mechanical ventilation, with the aim of reducing the duration of invasive ventilation. 2The trial, conducted between 2017 and 2019 in 18 UK paediatric intensive care units (PICUs), reported a significantly reduced time to successful extubation for all children by around 7 h (6 h in a prolonged ventilation cohort) with no significant increase in adverse events. 1 This paper describes the SANDWICH intervention using a recognized framework (Template for Intervention Description and Replication [TIDieR]) for detailing complex health care interventions, 3,4 allowing critical care nurses to gain a greater understanding of the intervention applied in this trial.The TIDieR provides a 12-point checklist and guide and was developed as an extension to the Consolidated Standards of Reporting Trials (CONSORT) 2010 reporting guidance to improve intervention reporting in clinical trials. 3,5e SANDWICH trial was complex in that it incorporated multiple components and involved a change in practice for health care professionals in paediatric intensive care.The key elements of this intervention were: patient-relevant sedation plans linked to regular assessment using the validated COMFORT sedation assessment tool (a tool that quantifies clinical parameters indicating a critically ill child's level of distress 6 ); regular assessment of ventilation parameters with a higher than usual trigger for undertaking an extubation readiness test; and progression to a spontaneous breathing trial (SBT) on low levels of respiratory support to test extubation readiness.The comparison was 'usual' PICU care in the United Kingdom: this was primarily medically led by PICU intensivists with little involvement of junior medical or nursing staff. 7ring the 20-month stepped wedge cluster randomized trial, PICUs were sequentially randomized at 4-week intervals to cross over from usual care to the SANDWICH intervention (Figure 1).Prior to cross over, there was an 8-week training period on the trial intervention to facilitate practice change for all PICU clinical staff.This paper reports the detail around the trial intervention using the TIDieR in a way that will allow both replication of the study and effective implementation of the intervention into PICUs.

| Item 2: rationale, theory or goal of the elements essential to the intervention
Weaning from invasive mechanical ventilation (IMV) is a complex process involving several stages: recognition that the child is ready to begin the weaning process; steps to reduce ventilation while optimizing sedation in order not to induce distress; and removing the endotracheal tube. 8Delay at any stage can prolong the duration of IMV, therefore an intervention targeted at assisting clinicians to safely expedite this process will minimize the risks associated with IMV. 8 What is Known About the Topic The judgement and experience of clinicians are critical in guiding weaning from ventilation; however, as data from a feasibility study on paediatric usual practice showed, there was wide variation both in sedation and ventilator weaning practices and junior staff were rarely involved in the process. 7spite strong evidence that coordinated care improves quality and saves money in health care, it depends on the approach used, how well it is implemented and in a particular environment. 9Within PICU, the dynamic, complex and time-pressured environment necessitates a team approach to care delivery that requires effective communication and collaboration. 10Various intensive care unit (ICU) studies have reported associations between rates of high inter-professional collaboration and patient mortality [10][11][12] ; and improved clinician-to-clinician communication with reductions in ICU length of stay. 12,13A team-led approach that maximizes engagement of all staff in early recognition of readiness and preparation for weaning ventilation could potentially reduce the duration of IMV and PICU length of stay.This in turn would allow maximum utilization of PICU beds, frequently a scarce resource in the United Kingdom.As 67% of nurses employed in UK PICUs are staff nurses (many junior) (in the United Kingdom, a Band 5), this would greatly maximize nursing contribution to the weaning process. 14Qualitative research indicates that inter-professional collaboration and communication are major factors that influence weaning and adoption of weaning protocols. 15 ventilator weaning, there is strong evidence that mechanically ventilated patients should have their readiness to wean assessed daily and weaning should be initiated on the basis of objective clinical criteria, rather than the clinician's subjective impression. 16Weaning generally involves either a period of spontaneous breathing (SBT), or a gradual reduction in the amount of ventilator support.The SBT was developed to identify patients who are ready to discontinue ventilation. 16The test aims at monitoring signs of respiratory muscle fatigue while the patient is still intubated.Adult studies have shown that most patients do not need gradual weaning; when assessed with a daily evaluation and SBT, approximately 75% of patients are ready to be extubated. 17Early paediatric studies have shown similar results. 18,19wever, although the introduction of weaning protocols has resulted in decreased ventilation times in adult patients, 20 at the time of designing the SANDWICH trial, only one study (n = 260) had shown that a protocol of daily screening with SBT benefited the paediatric population. 21 randomized trials of sedation weaning, a Cochrane systematic review in adults (2 single-centre studies, 633 participants) 20 and a subsequent multi-centre paediatric cluster randomized trial in 31 PICUs (2449 children) showed no clear evidence that protocoldirected sedation is more effective than non-protocolised care. 22wever, a systematic review of observational studies (6 studies, 2011 children) reported a beneficial association between the use of sedation guidelines and reduced PICU length of stay, frequency of unplanned extubation, prevalence of patients experiencing drug withdrawal, total doses delivered and duration of sedation. 23e weaning of sedative drugs and the ability to wean from mechanical ventilation are interconnected concepts: the child cannot wean from ventilation if they are too sedated and fail to 'breathe' spontaneously.Nurses are the only constant presence at the child's bedside and are in control of sedation level assessment and the titration of these drugs, thus they are in an ideal position to assess the child's readiness to wean from the ventilator.In the SANDWICH intervention, these components were bundled together, which was unique in comparison to other study interventions.The model outlined in Figure 2 shows each component of this complex intervention and the proposed mechanisms of action to achieve the intervention goal.(above PEEP).During the SBT, the child was monitored for signs of respiratory distress indicated by a 20% increase in heart or respiratory rate (above pre-SBT rates), signs of increased work of breathing, or a SpO 2 <92% or increase in FiO 2 requirement.The SBT duration was a maximum of 2 h, but if the child was breathing spontaneously with no distress at any time during this period, the medical team could decide to stop and extubate.

| Item 7: Type(s) of location(s) where the intervention occurred, including any necessary infrastructure
The intervention was delivered in 18 UK PICUs, and these comprised general, mixed general and cardiac surgical, and standalone cardiac surgical PICUs.In UK PICUs, there are no respiratory therapists, so the three professional groups involved in weaning ventilation are nurses, doctors and sometimes respiratory physiotherapists.On average, around half of nurses working in PICUs had a specialist critical care qualification, and many units employed newly qualified nurses directly into intensive care.Medical staff generally comprise consultant intensivists, PICU trainees and various trainees from other specialty areas (paediatricians, anaesthetists, emergency department and surgical trainees).In UK PICUs, the nurse-to-patient ratio is typically 1:1 for invasively ventilated children.Table 1 provides more detail around the context in which this trial took place.This facilitated bedside nurses to be integral to the co-ordination of all aspects of the SANDWICH bundle in the clinical environment.

| Item 8: Number of times the intervention was delivered and over what period of time including the number of sessions, their schedule, and their duration, intensity or dose
The intervention was delivered daily for all invasively mechanically ventilated children over the trial intervention period.

| Item 9: Tailoring of the intervention
This was a pragmatic trial and therefore tailoring and flexibility of certain intervention components were permitted.There was flexibility in intervention delivery related to: the location of the ward round; the number and proportion of multidisciplinary staff involved in the ward round; and assessment times for COMFORT scoring and the daily screen for SBT readiness.The intervention slotted into the usual organization of care, requiring no additional or specialist staff.Figure 3 shows the highly pragmatic nature of the trial as assessed using the PRagmatic-Explanatory Continuum Indicator Summary (PRECIS-2). 26e nine domains of PRECIS-2 are scored from 5 (very pragmatic) to 1 (very restricted) and show that the SANDWICH intervention's scores ranged from 4 to 5.This indicates that the SANDWICH intervention can be delivered in the 'real world' and alongside usual care.

| Item 10: Modifications of the intervention during the study
No amendments were made to the intervention during the study.

| DISCUSSION
The TIDieR checklist provided a structured method for reporting all specific details of the SANDWICH intervention sufficient to allow replication.This is important because to accumulate evidence of Cochrane systematic review of factors that impact on the use of ventilator weaning protocols has detailed this. 15r us, the real value of using the TIDieR checklist was as a framework to really articulate individual components of this complex trial than are often poorly described in trial reporting.In this paper, we wanted to demonstrate to readers and researchers how to use this checklist by describing it in the context of a recently published large multi-centre RCT.
A limitation of the checklist is that some items may be open to interpretation.For example, in our case, items 3 (materials) and 6 (delivery mode) overlapped as the materials included method of delivery (online educational course).On the other hand, the process of using TIDieR initiated discussions about how the intervention com- are reported impose strict limits on word count, meaning that information that is not immediately required may be lost.In addition, the size and number of permitted online-only supplements are typically restricted to additional tables and figures supporting the trial results and trial protocol and amendments.

| CONCLUSIONS
Weaning mechanical ventilation is a complex process, dependant on many patient and organizational factors; thus, implementing behavioural change in clinical practice is a major challenge.The SANDWICH intervention tackled this challenge by providing a pragmatic teambased approach to optimizing sedation and assessing a child's readiness for liberation from mechanical ventilation.Overall, we found the TIDieR tool beneficial as it enabled a detailed description of the intervention to improve understanding and usability across multiple research sites.Such full description is important in enabling the SANDWICH intervention to be adopted in other PICUs that were not involved in the trial.

•
Weaning from mechanical ventilation is a highly complex process • Complex behaviour change interventions are not well described, and this hampers reproduction and subsequent implementation of successful interventions • The TIDieR checklist enables more precise reporting of complex interventions used in trials What this Paper Adds • A detailed and explicit description of the behaviour change intervention used in the SANDWICH trial of weaning children from mechanical ventilation • An example of how to use the TIDieR checklist to provide greater clarity in describing functional components of a complex intervention

2. 3 | 25 2. 4 |
Item 3: Materials used in the intervention (those provided to participants or used in intervention delivery or training of providers) The 8-week training period involved a multi-faceted approach to education for all clinical staff involved in patient care.A purposively developed SANDWICH course (developed by the implementation manager and clinical experts within the trial team) delivered training and assessment of intervention components and the underpinning clinical evidence supporting protocolised weaning.The courses consisted of six short online learning modules (four modules focusing on the core components of the intervention and two additional sedation and pharmacology modules) hosted by F I G U R E 1 Schematic for the SANDWICH stepped wedge, cluster randomized trial.Twenty-month trial delivered in 22 4-week blocks.One hospital site sequentially randomized [R] from block 0 to 16.One hospital site had two PICUs.Following training, each PICU transitioned to delivering the SANDWICH intervention.PICU, paediatric intensive care unit; SANDWICH, Sedation and Weaning in Children LearnPro (https://learnpro.co.uk/products/lms/learning-managementsystem), 24 a learning management system used across the UK National Health Service (NHS) that incorporates compliance monitoring.This learning system records delivery, course uptake by staff member by unit and the participants' assessment and mark.A score of 80% was required by staff to pass and receive a certificate.Compliance measures for course completion were fed back to individual PICU trainers during the training period for monitoring purposes.The online SANDWICH course was supplemented by a 136-page Trainer's Manual with PowerPoint slides, video and online and written educational material providing detail around the components of the intervention.Additional resources were produced to provide reminders and facilitate adoption of the intervention, and these included flyers, posters, screensavers, lanyards, badges and banner message pens.All training materials, including the SANDWICH course, are freely available on the SANDWICH website at https://www.qub.ac.uk/sites/ sandwich/.Item 4: Procedures, activities and/or processes used in the intervention The SANDWICH intervention comprised four key components: 1. Greater inter-professional collaboration was enabled by multidisciplinary ward rounds and other ad hoc rounds in reviewing sedation management including COMFORT scores (a validated sedation scale), sedative regimen and setting targets.In addition, regular review (minimum twice daily) of the child's ventilation status and setting ventilation goals.A ward round checklist was used to capture adherence to goal setting and goals were fed back to the bedside nurse.2. Minimum 6-hourly assessment of sedation using the COMFORT score by bedside nurses.Nurses titrated sedatives according to unit policy to attain the child's COMFORT score within the agreed target range.3. Twice-daily assessment of criteria for readiness to perform an SBT by bedside nurses.The criteria were as follows: a. FiO 2 ≤0.45 b.SpO 2 ≥95% (or as appropriate to underlying condition) c.PEEP ≤8 cmH 2 O d. PIP ≤22 cmH 2 O e. Cough present 4. When readiness for SBT criteria was met, nurses stopped or reduced sedation and informed an appropriately trained nurse or doctor who initiated an SBT with a positive end expiratory pressure (PEEP) of 5 cmH 2 O and a pressure support of 5 cmH 2 O

2. 5 |
Item 5: Description of the expertise, background and specific training given to intervention providers A trial implementation manager (LM) was employed during the trial to develop the training materials and resources.The implementation manager was a senior paediatric nurse with a BSc, MSc in Nursing and 11 years experience in practice and education in PICU.LM was a trial team member and worked closely with the trial investigators in developing the intervention.Her responsibilities were to visit each PICU and train additional local PICU trainers and provide training and implementation support and guidance for each participating PICU.

F
I G U R E 2 Theoretical model of the proposed mechanism of action of the SANDWICH intervention components.SANDWICH, Sedation and Weaning in Children Each participating PICU nominated a SANDWICH nurse with twofold responsibility: liaising with the implementation manager and training clinical staff on the intervention during the 8-week training; and liaising with the research team and collecting trial data in the usual care and intervention periods.Payment for the SANDWICH nurses' time was reimbursed to the hospital from the trial grant.Each PICU also nominated multidisciplinary SANDWICH trainers who were responsible for planning, organizing and rolling out intervention training to all clinical staff during the 8-week training period.The number of trainers varied according to the size of the PICU and included nursing and medical clinical staff of varying grades including medical consultants, advanced nurse practitioners, nursing managers and clinical nurses.Additionally, the PICUs nominated SANDWICH 'champions' to promote implementation, assist with training queries and provide local support when necessary.Champions included varying grades of medical and nursing staff in addition to pharmacists and physiotherapists.When the PICU was randomized and informed of their cross over date, the implementation manager liaised with the SANDWICH nurse to organize training dates at the start of the training period.LM visited each PICU and provided between two and four face-to-face sessions, depending on the size of the unit, to all SANDWICH trainers and champions.The PICU designated responsibility for conducting specific intervention components to relevant staff.The bedside nurse was responsible for conducting COMFORT assessments, titration of sedation and assessment of criteria for readiness to perform an SBT.Staff responsibility for the decision to proceed to an SBT when criteria were met and the conduct of an SBT varied among PICUs depending upon their local scope of practice policy.In all PICUs, the decision to extubate consistently remained with the doctor.All clinical staff involved in the delivery of the intervention had to successfully complete the online assessment.The target for achieving staff training was 80% within the 8-week training period.

2. 6 |
Item 6: Mode of delivery The mode of delivery included both online and face-to-face engagement.Online education was delivered using an established NHS online education provider (LearnPro NHS: http://sdwhtraining. learnprouk.com).Face-to-face training using the SANDWICH manual was delivered by the implementation manager to each SANDWICH training team.Training for all staff included a mix of online and faceto-face training either in seminar rooms and/or at the bedside as necessary.
were collected by the SANDWICH nurses daily and entered into an electronic case report form.These data included: completion of the ward round checklist on setting targets; completing a COMFORT score at least every 6 h; daily screening for readiness for SBT; time and duration of SBT (if undertaken); and reasons for non-progression to an SBT when criteria were met and non-progression to extubation when the child passed an SBT.Adherence was measured by the proportion of intervention components performed, staff trained and intervention reach (admissions screened divided by admissions of children on IMV).We fed back the intervention adherence proportions on three occasions during the trial to the SANDWICH nurse to disseminate within the PICU and promote adherence.The feedback included the unit's own adherence values alongside an anonymized unit's high adherence values for comparison.This process was used to maintain/improve fidelity.We collected feasibility and acceptably from the PE, for example, interviews with key staff during the trial and approximately 2 months following the training period.
effectiveness and the processes of practice change, it is necessary to have accurate replication.The TIDieR checklist facilitated a clear description of the intervention, how and when it was delivered; how staff were trained and detail about the training and the expertise of who was involved.Such complex interventions like SANDWICH are multi-faceted with several components and it is not always possible to know which components provide the 'active ingredient(s)'.Moreover, both the effectiveness and the replicability of such interventions are F I G U R E 3 The pragmatic nature of the SANDWICH trial as shown by the PRagmatic-Explanatory Continuum Indicator Summary (PRECIS-2).SANDWICH, Sedation and Weaning in Children reliant not only on how they were provided but by whom.These are important points that are frequently inadequately unreported. 27TIDieR builds upon the 2008 extension to the CONSORT (Consolidated Standards of Reporting Trials) Statement for trials of non-pharmacological treatments to accommodate items that were frequently not reported in these interventions 28 : namely complexity of the intervention and expertise of the care provider.Thus, the more clearly the intervention components are detailed, the easier it will be to implement change successfully.Furthermore, knowing what the SANDWICH intervention components are, and how they were delivered, may allow more efficient and cost-effective implementation and lead to better decisions about the non-core components that can be adapted to suit other PICUs.The culture and context of PICUs are also important factors to consider when implementing any change in behaviour practice.It is beyond the scope of this paper to prescribe what will work best, for whom and in what circumstances.However, a ponents were described.This enabled the team to capture a clear picture of the reality of implementation into practice.Thus, describing an intervention using TIDieR can be useful for facilitating clinicians to reflect on what they are doing and see what part they play in delivering the intervention.Difficulties in reporting the checklist within or alongside the main trial paper concern available word count and inclusion of supplemental material.Typically, mainstream journals in which clinical trial results