Overall study design
The Sleeping Sound with ADHD project is an RCT of a behavioural sleep intervention versus usual care (see Figure ). The project runs from mid 2010 to the end of 2012. This period encompasses participant recruitment (including determining eligibility for the study), baseline data collection, intervention delivery and participant follow-up (3-12 months). This project has been funded by the National Health and Medical Research Council of Australia (project grant number: 607362) and has been granted ethics approval by the Royal Children's Hospital (#30033) and the Victorian Department of Education and Early Childhood Development (#2010_000573) Human Research Ethics Committees.
Participants include families of children aged 5 to 12 years with paediatrician diagnosed ADHD (any subtype) and at least one of the following sleep problems: sleep onset association disorder, limit setting disorder, delayed sleep phase or insomnia (idiopathic or psychophysiological), as defined by the American Academy of Sleep Medicine (see below). Children are also eligible if they have parent-reported night time anxiety (defined as having both difficulties falling asleep and significant worry at bedtime).
Recruitment - Stage 1
Victorian paediatricians (public and private) pre-identify their patients with ADHD either through their medical software (eg Genie, Medical Director) or through case notes. Paediatricians send a study-designed letter to the child's primary caregiver inviting them to take part in the study. The letter advises parents that the research team will phone them to ask about their child's sleep and ADHD symptoms. An 'opt out' approach is used, whereby parents are asked to contact the study team if they do not
wish to learn more about the study. If parents do not opt out within a two week period, the paediatrician provides the research team with the contact details of the families by fax or post. The team then contacts the parent to explain the study further. In our 2006 survey, this approach resulted in a 74% uptake rate and was well accepted by parents [5
Recruitment - Stage 2
The research team telephones all parents who do not opt out to assess inclusion/exclusion criteria (see section below) and whether or not the parent is interested in participating in the study (10 minute phone call). Eligible families are mailed an information sheet, consent form, and baseline survey. To determine the extent of any participant bias, data are collected and compared with non-participants; this includes child gender, child age, and socioeconomic status of the family's immediate neighbourhood.
Eligible families include children who meet all of the following criteria:
a) Full diagnostic criteria for ADHD, using the 18-item DSM ADHD Rating Scale IV, which is a validated scale measuring the core symptoms of ADHD [28
]. We also use study designed questions to assess ADHD symptom duration ('Did your child have these symptoms for six months or longer before he/she was diagnosed with ADHD?), onset ('Did your child have these symptoms before he/she turned seven?) and impairment (Are these symptoms present at home, school or when out socially e.g., in the park, visiting friends)?
b) Moderate/severe sleep problems by parent report [5
c) American Academy of Sleep Medicine diagnostic criteria for at least one sleep disorder including sleep onset association disorder, limit setting disorder, delayed sleep phase and/or insomnia (idiopathic or psychophysiological) [29
]. Children experiencing significant night time anxiety and difficulty falling asleep at night are also eligible.
Children are excluded from the study if they meet any of the following criteria:
a) Receiving specialised help for their child's sleep from a psychologist or a specialised sleep clinic (apart from their paediatrician).
b) Have a serious medical condition (e.g., severe cerebral palsy) or an intellectual disability (IQ < 70). Children with other developmental or mental health comorbidities are not excluded.
c) Have suspected obstructive sleep apnoea (OSA). OSA is assessed using the three OSA items from the Child Sleep Habit's Questionnaire (CSHQ) [30
]. This scale can help identify children who may suffer from OSA. Parents who report that their child sometimes or usually snores, stops breathing and/or snorts/gasps during their sleep are contacted by the lead investigator (HH), a paediatrician, for further assessment. If OSA is suspected, these children are excluded and referred to appropriate clinical services. After approximately six months, we will follow-up with children excluded due to suspected OSA to see whether they are eligible to participate in the study following clinical assessment and possible treatment i.e. if they have ongoing behavioural sleep problems despite treatment of their OSA.
The primary outcome is ADHD symptom scores. In order to detect a 0.4 standard deviation (clinically meaningful) shift in the mean ADHD score between the treatment and control group at the three-month follow-up, with 80% power and at a significance level of 0.05, we would require 99 children in each arm (198 total children). Allowing for up to 20% reduction in the proportion of children with sleep problems amongst the control arm, this sample size would also ensure adequate power to detect a further 20% reduction in the proportion of children with sleep problems between the intervention and control arms. It would also enable us to detect a shift in other secondary outcomes, such as the mean quality of life scores between the two treatment arms.
In order to have follow-up data on 198 children at the 3 month follow-up, we need paediatricians to send study information to approximately 1124 children with ADHD. This assumes an initial 70% response rate (i.e. consenting to hearing more about the study, n = 787), and that of these, 45% will have a moderate/severe sleep problems (based on 2006 survey [5
], n = 354). This also takes into account 70% of those with moderate/severe sleep problems consenting to taking part in the study (n = 248) and 20% participant drop out over the follow-up period.
Upon receiving the completed consent form and baseline survey, an independent research assistant randomises families to either the behavioural sleep intervention group or the control group of 'usual care'. Families are randomised using a pre-generated random number sequence developed by a statistician, which is contained in sealed opaque envelopes stored in the independent research assistant's office. We used varying block sizes of 2, 4 and 6 in the randomisation sequence to maintain balance between the trial arms over the course of the trial and so that allocations could not be predicted from the previous sequence. It is expected that the patient group will predominantly be boys, so that randomisation is stratified by gender in order for distribution proportions of males and females to be comparable across the randomisation groups. All families are mailed a letter to inform them of their group allocation. Intervention families are then telephoned to book a consultation time at their paediatrician's office. Control families can access usual care for ADHD from their child's paediatrician - our previous survey suggests this does not routinely involve sleep management [5
]. If the parent consented to teacher participation, we mail the child's teacher a baseline survey to complete at the point of randomisation.
The behavioural intervention is evidence based [18
] and consists of two face-to-face, one-on-one sleep consultations with a trained clinician (trainee consultant paediatrician or psychologist) held two weeks apart and one follow-up telephone call another two weeks later. The first session focuses on an assessment of the child's sleep problem, providing information about normal sleep and sleep cycles, advice about sleep hygiene, and a plan specifically tailored to the child's particular sleep disorder. For example, sleep onset association disorder, typically associated with the need for parental presence at sleep time, is managed with adult fading (i.e., graduated extinction). This technique requires gradual withdrawal of parental presence from the child's bedroom over 7-10 days. Limit setting disorder is managed by ignoring child protests and rewarding compliance with bedtime routines. Delayed sleep phase is managed by temporarily setting the child's bedtime later, gradually bringing it forward, and waking the child at a pre-set time in the morning and encouraging early morning light exposure [18
]. Parents are offered a range of management strategies and are free to choose the strategies they would like to try. All parents are also asked to complete a sleep diary. The second session is held two weeks later to review the sleep diary, reinforce strategies, trouble shoot and monitor progress.
A standardised consultation record is kept for all children, as per our pilot. This includes the presenting sleep problem/s, possible contributors to the problem (eg TV in bedroom), ADHD medication use, comorbidities, and usual bedtime routines. The clinician records the duration of each consultation, sleep problem diagnoses, family sleep management goals, handouts given to parents, and management strategies chosen by the family.
A follow-up telephone call is made at a time convenient for the family, approximately two weeks after the second consultation, to provide an opportunity for the child's parents to ask further questions, and to reinforce strategies, trouble shoot, and monitor progress.
Our pilot study demonstrated that this program is feasible to deliver and acceptable to families. However, a barrier to participation was travel to the Royal Children's Hospital Melbourne and we have addressed this by conducting consultations at the treating paediatrician's consulting rooms.
Data are collected using parent, teacher and self-report questionnaires, as well as objective measures including a face-to-face assessment of working memory and actigraphy - all measures are outlined in Table . Outcomes are measured at 3, 6 and 12 months post randomisation with the exception of actigraphy (3 months only), the working memory test (6 months only) and teacher report (3 and 6 months only).
Study measures and time-points
Child ADHD symptoms, using the ADHD Rating Scale IV
- parent and teacher versions. An 18-item validated scale measuring the core symptoms of ADHD (inattention, impulsivity, hyperactivity), which has been demonstrated to be sensitive to change [28
- primary caregiver report of child sleep problem (none, mild, moderate or severe) [5
Children's Sleep Habits Questionnaire (CSHQ)
- 33-item, validated measure of disorders of initiating and maintaining sleep, which can distinguish clinical from community samples [30
- The Actiwatch 2 (Philips Resprionics) is a small, motion sensor that is attached to the non dominant wrist to measure body movements and is used to provide an objective measure of sleep. Movement patterns are analysed and used to differentiate between sleep and wake times, thus providing an objective measure of sleep onset and total sleep duration. Use of actigraphy data in children has been shown to be reliable and valid and it also correlates well with data obtained using polysomnography (overnight observation of sleep) [32
]. The Actiwatch is accompanied by an instruction sheet and sleep log which covers napping, medication use, time in bed, night awakenings, morning awakening time etc. The sleep log assists in the interpretation of Actiwatch data. Children wear the Actiwatch for seven days during school term to assess both weekday and weekend sleep behaviour.
Strengths and Difficulties Questionnaire (SDQ)
- parent and teacher versions. A 25-item validated measure of behavioural and emotional problems for children aged 4 to 16 years. It provides standard scores on 5 subscales (hyperactivity/inattention, conduct problems, emotional symptoms, peer relationship problems, and prosocial behavior); a total problems score is derived from the first 4 subscales [33
Pediatric Quality of Life Inventory (PedsQL - child and parent versions)
- a 23-item validated measure for children aged 2 to 18 years. Provides total, physical, and psychosocial health summary scores, with higher scores indicating better health-related quality of life [34
- number of days missed or late for school over the preceding three months [5
Daily Parent Rating of Evening and Morning Behaviour (DREMB)
scale - an 11-item rating of core ADHD symptoms and behavioural problems typically experienced over the past month [35
Working Memory Test Battery for Children
- a face-to-face assessment which provides an objective measure of the impact of the sleep intervention on the child's working memory, a critical executive function. Backwards Digit Recall, Counting Recall, and Listening Recall subtests are administered [36
Other sleep help - parent report of other professional help sought for their child's sleep eg GP, school nurse.
Sleep program evaluation (intervention group only) - parent report of the usefulness of program strategies and ability to put strategies into practice.
Depression Anxiety Stress Scale (DASS)
- a validated 21-item measure of adult mental health with clinical cut points for each of the three subscales of depression, anxiety and stress [37
- number of days missed or late for work over the preceding three months [5
Socio-demographic questions are also included in the baseline questionnaire - these cover family composition, parental education and age, language spoken at home, annual household income, and child medication use and diagnosed comorbidities.
A comprehensive assessment of comorbid diagnoses is completed using the Anxiety Disorders Interview Schedule for Children (ADIS-C-IV).38
This assessment is conducted over the telephone shortly after randomisation (the ADIS-C-IV is validated for administration over the telephone) [38
]; assessing comorbidity is important in understanding any differential effects of the intervention as a result of the child's comorbidity. We will also assess inter-rater reliability to ensure consistency in the way that interviewers code parental responses on the ADIS-C. We will audio record the first 10 interviews of parents who give verbal consent; interviewers will code parental responses from these ten interviews and inter-rater reliability coefficients will be calculated.
Analyses will be by 'intention to treat' at the level of the individual child. At the 3, 6 and 12 month follow-up, we will compare change in mean primary caregiver and teacher scores (3 and 6 month follow-up only) on the ADHD Rating Scale IV between the two trial arms using t tests. We will also compare proportions of children with moderate/severe vs no/mild sleep problems between the two arms using chi squared analysis at each of the follow-up periods. We will compare mean scores at follow up for sleep latency and duration (actigraphy), child working memory, behaviour, health-related quality of life, daily functioning and school attendance, as well as compare mean number of days late for work and parent mental health (depression, anxiety and stress) between the two arms, using either Mann-Whitney or t tests, as dictated by data distribution.
At each follow-up, we will conduct a set of regression analyses (linear regression for continuous data, logistic regression for categorical data) for each outcome, adjusted for potential confounders identified a priori. Confounders will include child age, medication use, comorbidities, family socio-demographic factors, and family socioeconomic status, which will be assigned according to postal code of residence using the Index of Relative Socioeconomic Disadvantage (mean 1000, s.d. 100) from the Australian Bureau of Statistics census-based Socio-Economic Indexes for Areas (SEIFA) [39