Protocol and registration
This study has received ethical approval from the Queensland Children’s Health Services (Lady Cilento Children’s Hospital) Human Research Ethics Committee (approval number: HREC/15/QRCH/32) and the University of Queensland Ethics Committee (approval number: 2015000456). The study methodology was documented in a protocol and registered prior to starting recruitment (Australian New Zealand Clinical Trials Registry, ID: ACTRN12615000419561). This is version 1 of the study protocol completed on 10 September 2015. Methods have been documented in accordance with the Consolidated Standards of Reporting Trials (CONSORT 2010) [43] and Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT 2013) [44] statements. The intervention has been described using the Template for Intervention Description and Replication (TIDieR 2014) guidelines [45].
Design and setting
This study is a single-center, superiority, parallel-group, prospective randomized controlled trial (see Fig. 1). Eligible participants are randomized to receive either (1) medical hypnosis (intervention group) or (2) standard care (control group).
Participants are recruited from the Pegg Leditschke Paediatric Burns Centre (PLPBC) at Lady Cilento Children’s Hospital (LCCH), Brisbane, Australia (AUS). The PLPBC is the major specialist tertiary burns center for Queensland and Northern New South Wales, Australia. The center’s clinical multidisciplinary team treats approximately 800 new burn patients per year.
Eligibility criteria for participants
Eligible participants are children aged between 4 and 16 years (inclusive) who meet the inclusion criteria of (1) an acute burn of any depth (excluding erythema only) and (2) presentation to the PLPBC for treatment (inpatient or outpatient). Children are excluded from the study if they are non-English speaking; cognitively impaired; under the care or investigation of the Department of Communities, Child Safety, and Disability Services; on ventilator support; or if they have initial burn wound care procedures carried out in the operating room under general anesthesia.
Patients of the PLPBC are identified by the center’s research manager, and participants’ eligibility is assessed by the nursing staff. All eligible patients presenting to the PLPBC are approached and invited to participate in this RCT by a clinical research team member not involved in their primary care. Parent (the term parent includes legal guardian) informed consent is obtained and recorded. Child consent is obtained for all children able to read and write. Verbal assent is obtained for all other eligible children.
Interventions
Recruited participants are randomized into either (1) the intervention group: medical hypnosis or (2) the control group: standard care immediately prior to their first wound care procedure (i.e., dressing change or application, which may involve cleaning the wound) at the PLPBC:
-
1.
Intervention group
-
2.
Control group
-
Standard care
-
Standard procedural distraction is available to this group, including music, toys, the Ditto™ (Diversionary Therapy Technologies, QLD, Australia) [6] and other electronic devices (e.g., TV, hand-held games, portable DVD players), books, and parental presence. The consulting hypnotherapist is positioned in the procedure room with the participant but does not provide hypnosis to ensure that the only difference is the use of medical hypnosis. The hypnotherapist does not interact with participants in the control group to avoid potentially using hypnotic techniques and is only present to help record outcome measures associated with wound care procedures.
Outcomes
Previously validated scientific measures are used for all outcomes. The duration of each wound care procedure is timed. Prior to the application of new dressings at the first dressing change, the burn depth is calculated by measuring blood perfusion at the burn site using a Moor LDI2-BI2 Laser Doppler Imager (Moor Instruments Limited, Devon, UK). Data is collected at every dressing change until ≥95 % re-epithelialization occurs, and the total number of dressing changes is recorded.
Primary outcome measures
The two primary outcome measures are pain intensity and wound healing (i.e., re-epithelialization time). Pain measurements are taken several times for each participant. The first time point (baseline) is immediately before premedication prior to removal of the wound dressing in clinic. The second time point is immediately after the new dressing is in place; the worst pain is assessed reflecting the maximal pain intensity experienced by the participant during the procedure (retrospective). For the third time point, a final pain assessment occurs to gauge pain intensity immediately after the new dressings are in place and hypnotherapy has ceased. No pain measurements are taken during hypnosis. Assessing pain intensity in this manner effectively gives three time points relative to the procedure: before, during, and after. These measurements take place at each subsequent dressing change until ≥95 % wound re-epithelialization.
Pain
A range of pain scales are utilized to measure pain intensity (assessed by child, nurse, or parent). The Faces Pain Scale-Revised (FPS-R) [49] is used to assess the child’s self-report of pain and will be the primary outcome measure. Convergent validity of the FPS-R is supported by a strong positive correlation (r = 0.93, p < 0.001, N = 76) with a visual analog scale (VAS) pain intensity measure in children aged 5–12 years and by strong positive correlations with the VAS (r = 0.92, p < 0.001, N = 45) and the color analog scale (r = 0.84, p < 0.001, N = 45) in a clinical sample of pediatric inpatients aged 4–12 years [49]. Even among the youngest patients sampled (four-year-olds), there is evidence of usage of the FPS-R and analog scales in a consistent and reliable manner [49]. Nurses report a behavioral/observational rating on the Face, Legs, Arms, Cry, Consolability (FLACC) scale [50, 51]. The FLACC scale has been validated for use in settings such as postoperative pain [50]. Despite a recent systematic review rescinding recommendation of the FLACC scale for procedural pain assessment [51], we have chosen this scale in the absence of an acceptable alternative due to its extensive use in prior clinical trials examining procedural pain. Parents also rate their child’s pain using an 11-point (0 to 10) numeric rating scale (NRS) [52]. Pain scores reported verbally by the parent (NRS) and child (FPS-R) are documented by the primary investigator. Nurses document the FLACC pain scores.
Wound healing
Re-epithelialization is defined as ≥95 % wound healing and no further wound dressings required. Scabs or crusts are defined as unhealed areas [53]. For the purpose of the main analysis, this outcome measure will be assessed using the percentage re-epithelialization assessed from 3D digital photography by an independent surgeon and nurse blinded to study treatment group. The percentage re-epithelialization reported by the independent surgeon will be used in the main analysis if appropriate. The outcome will also be measured from 3D digital photography by the investigator (SJC), and the time to wound healing recorded in the medical records will also be used for comparative purposes. Wound photographs are taken at each dressing change using 3D LifeViz System™ (Quantificare, Sophia Antipolis, France) [54]. The independent blinded surgeon and nurse will mark out the wound edges after the photographs are taken, along with any unhealed areas. Surface area computer mapping will then be used to determine percentage re-epithelialization [53, 54].
Secondary outcome measures
Secondary outcome measures collected at baseline include a self-reported procedural anxiety measure, a saliva sample (for measurement of stress biomarkers: salivary cortisol and salivary α-amylase), and heart rate (HR). Baseline measurements are taken before nurses administer pharmacological analgesia according to PLPBC standard practice.
Procedural anxiety
The visual analog scale for anxiety (VAS-A) [55, 56] will be used to measure procedural anxiety in children. Self-reported anxiety measures will only be administered to children 8 years old and above. For participants younger than 8 years, the parent will be asked to assess their child’s anxiety using the same scale. The anxiety measurement is obtained prior to premedication and immediately after new dressing application.
PTSD
PTSD severity three months following injury will be assessed using the Child PTSD Symptom Scale (CPSS) [57] for children aged 7 years or older. The CPSS is designed to assess PTSD diagnosis and symptom severity in children ages 8 to 18 who have experienced a single-incident traumatic event [57]. Total symptom score and the three symptom clusters of the CPSS demonstrate high internal consistency (α = 0.89 for total score) [57]. The percentage agreement between PTSD diagnoses at two separate time points was 84 %, indicative of moderately high reliability [57]. Test-retest reliability of the total CPSS score is acceptable (κ = 0.84) [57]. Convergent validity of the CPSS has been supported (Pearson’s r = 0.80, p < 0.001) when measured against the Child PTSD Reaction Index (CPTSD-RI) [57].
The Young Child PTSD Checklist (YCPC) [58] will be used for children younger than seven years old. Face validity of the YCPC items is excellent based on a series of studies by Scheeringa et al. that used these items in an interview format and formed the basis for the new DSM-5 disorder titled “Posttraumatic stress disorder for children six years and younger” [59–62]. The test-retest reliability (intraclass correlation coefficient = 0.87) [63] was acceptable, and the predictive validity [64] of PTSD symptoms using an interview format has been supported.
Parent satisfaction
Ease of the child’s wound care procedure as assessed by the parent is a secondary outcome measure. At the conclusion of every dressing change, a parent rates how easy they believe the wound care procedure was for their child on a 5-point Likert scale (from “not at all easy” to “extremely easy”) with higher values indicating greater satisfaction.
Biochemical stress markers
Salivary cortisol, representing hypothalamic-pituitary-adrenal axis activity, and salivary α-amylase (a proxy for norepinephrine indicative of sympathetic adrenomedullary system activity) are measured as biomarkers of stress associated with wound care procedures [65]. The participant places a Salivette™ (Sarstedt Australia Pty., Ltd. Mawson Lakes, SA, Australia) under their tongue for 2 minutes for saliva collection at these times: (1) immediately before premedication prior to removal of wound dressing in clinic, (2) immediately after the new dressings are in place, and (3) three months post-injury to obtain a baseline. The parent completes a saliva collection survey which records variables pertinent to salivary analysis: collection time, time participant last woke up, time participant last brushed teeth, any medication given, any food/drink/gum during the previous hour, time participant last had any caffeine, and pertinent smoking or tobacco history.
The date, time, and volume of saliva collection are recorded in the laboratory and samples are refrigerated at 4 °C and processed within 7 days. Samples are spun in a centrifuge at 1400 × g at room temperature for 10 minutes and the saliva frozen at −80 °C until analysis. Salivary cortisol and α-amylase will be quantified using ELISA kits (Stratech Scientific, Avalon, NSW, Australia) with saliva samples analyzed in triplicate. Heart rate will also be recorded as a physiologic measure of pain and distress at time points (1) and (2) for each dressing change.
Hypnotic responsivity
The Stanford Hypnotic Clinical Scale for Children (SHCS-C) [66] will be used to assess and record hypnotic responsivity for participants in the intervention group as recommended by prior methodological reviewers of relevant literature [32]. This assessment will only be conducted within the intervention group to ensure that control participants remain naïve to hypnotherapy. Normative data for hypnotic responsivity are available for children aged 3–16 years. The SHCS-C correlated 0.67 with a slightly modified version of the Stanford Hypnotic Susceptibility Scale, Form A for pediatric use [66], supporting concurrent validity.
Demographic and clinical information
Participant demographics and medical history are recorded from the caregiver and hospital chart: mechanism and site of injury, estimated percentage total body surface area (TBSA) of burn, burn depth, any first aid treatment applied, and medication given. TBSA is determined by a consultant surgeon using the Lund and Browder method [67].
Participant timeline
Participants are enrolled after presenting to the PLPBC and after their study eligibility has been assessed (Fig. 1). Regardless of which trial arm they are randomized into, all participants receive treatment during their scheduled appointment times for burn wound care at the PLPBC. Primary outcome data is collected concurrently with participants’ scheduled wound care procedures until the burn wound is ≥95 % re-epithelialized. The endpoint for secondary outcome data collection is 3 months post-burn. No extra participant visits to the PLPBC are required for the sole purpose of data collection.
Sample size
A sample size estimate was derived from the primary outcomes: days to re-epithelialization and pain. Based on re-epithelialization within 15 (SD = 4) days and a minimum clinically important difference (MCID) of 3 days [26], the sample size required was estimated as 29 per group, using 80 % power and an α of 0.05. Allowing for 10 % loss to follow-up, a total of 66 participants will be required. Additionally, this sample size is adequate to show an MCID of 2 (SD 2.5) in the pain outcome measures [26]. Recruitment will continue until at least 33 participants in each arm have been obtained with complete data for the primary outcomes.
Randomization
A computerized random number generator is used to randomize study participants. Simple randomization is overseen by staff not involved with the study. Third-party concealment of group allocation occurs by using a numbered series of opaque, sealed envelopes prepared in advance. The primary researcher is then told to which group the participant is allocated.
Blinding
Medical hypnosis provided throughout a procedure cannot be masked. This study’s nature prevents full blinding, but certain outcome measures are blinded. The re-epithelialization assessors are blinded, as these measurements take place using 3D digital photographs. Trial group allocation remains unknown to this assessor. If discrepancy arises between the two re-epithelialization assessors, the assessment of the blinded assessor is taken as definitive to reduce potential performance bias. Burn depth and salivary analysis are also blinded measures, as data for these variables are provided to the investigators in a non-identifiable format and as the assessor of these variables is blinded to trial group allocation.
Discontinuation
Study participants can withdraw from the trial at any time. The number of adverse events will be documented, and any adverse event will be described in detail for both treatment groups. Current relevant literature has not reported any serious harmful effects associated with indicated pediatric medical hypnosis or hypnotherapy.
Data analysis
Data will be analyzed using SPSS 23 (IBM Corporation, Armonk, NY, USA). Descriptive statistics such as the mean and standard deviation, median and interquartile range, and confidence intervals will be used to report the sample demographics (i.e., age, gender, and mechanism of injury) and to summarize outcome measures, as appropriate. Between-group comparisons will be conducted for potential confounding variables for the primary outcomes. Potential confounding variables affecting wound healing that will be examined include burn depth, days taken to present to the PLPBC, ethnicity, mechanism of injury, percent TBSA, age, and gender [25]. Potential confounding variables affecting pain intensity that will be examined include age, the presence/absence of skin grafting, state anxiety (determined by VAS-A), pharmacologic analgesia given immediately before or during wound care procedures, percent TBSA, and days to re-epithelialization [26, 68]. If significant between-group differences are present for potential confounding variables, those variables will be controlled for in the primary analyses. Between-group differences will be investigated using univariate parametric or non-parametric analyses as applicable (e.g., linear regression, Student’s t test, or Mann-Whitney U test for continuous data and the chi-squared test or Fisher’s exact test for categorical data). All data will be analyzed on the intention-to-treat (ITT) principle as the primary approach. However, a sensitivity analysis will be conducted for data collected as per protocol. A repeated measures analysis will be undertaken using generalized estimating equations (GEEs) [69] including the main effect of treatment group and time on the primary pain and healing outcomes, as well as on the secondary outcomes of procedural anxiety, PTSD, parent satisfaction, and biochemical stress markers.
Analyses will be conducted with data stratified for burn depth (superficial partial-thickness/deep partial-thickness/full-thickness) and participant age (e.g., <8 years/≥8 years, with age strata based on age-group validity of the VAS-A) [69]. Differences in hypnotic responsivity between the intervention group and a normative comparison group will be analyzed using z-scores or using equivalent non-parametric tests such as the Mann-Whitney U test where applicable. Differences in re-epithelialization by the independent blinded raters (surgeon and nurse) and the investigator (SJC) will be examined using reliability coefficients and measures of agreement (e.g., percentages of exact agreement). Differences between measures of pain (e.g., observer report by parents versus child self-report) will be examined using correlational analyses where appropriate.
The influence of demographic and clinical factors, and primary and secondary outcomes not included as dependent variables, on primary and secondary outcomes will be examined using regression models and GEE models. Post hoc adjustment for multiple comparisons will be conducted using the Šidák correction [70] where appropriate. Statistical significance will be set at p < 0.05.
Data storage
Data are protected in locked filing cabinets within the secure area of the Centre for Children’s Health Research, University of Queensland. Data are entered into a spreadsheet using Excel. Any incomplete data are coded as unknown, missing, or not applicable. The data set will be cleaned, checked, and then locked for analysis. Upon trial completion, data will be stored for 15 years as stipulated by the Queensland Children’s Health Services (LCCH) Human Research Ethics Committee.
Dissemination
Outcomes will be published in a peer-reviewed medical journal (publication target Burns) and will also be reported at relevant conferences.