Background
Attention deficit hyperactivity disorder (ADHD) is a leading cause of child referrals to men-tal health services, and major risk factor for early criminality, school drop-out and exclusion. It imposes significant burdens on children, their families and a wide range of public services.
Long-term evidence for the effectiveness of main and non-mainstream treatments is weak. Current approaches to generating evidence are expensive and slow. New approaches are required to identify treatments which can improve outcomes. This thesis rethinks the identi-fication and testing of interventions to improve outcomes for ADHD.
Methods.
Using the Trials within Cohorts (TwiCs) approach, the Sheffield Treatments for ADHD Re-search (STAR) project recruited a cohort of children with ADHD and collected outcomes at 0, 6 & 12 months from carers and teachers. For the first randomised, controlled trial (RCT) embedded in the cohort an eligible proportion were randomly selected and offered treat-ment by homeopaths, or nutritional therapists, additional to usual care. At 6 months, their outcomes (Conners Global ADHD Index) were compared with those not offered interven-tions, and the feasibility of the methods and interventions assessed.
Results
Between September 2015/2016, the target number of 144 participants were recruited to the cohort. 124 were eligible for the 1st trial and randomised. 83 were offered a treatment of which 72 accepted and 50 attended 1+ appointments. 89/124 paired (baseline & 6-months) carer and 31/100 paired teacher questionnaires were available for analysis. Teach-ers’ responses were too few, and unstable, but there were preliminary indications of treat-ments’ effectiveness according to carers: t = 1.08, p= .28 (-1.48, 4.81) SMD .425 for treat-ment by homeopaths; t = 1.71, p = .09 (-.347, 5.89), SMD = .388 for treatment by nutritional therapists. No serious adverse events attributable to treatments were reported.
The TwiCs approach was feasible but required some modifications. Return of carer ques-tionnaires was improved by addition of an incentive. Delivery of treatments was feasible, but attendance rates affected by therapists’ contacting strategies, and one therapist dropped out.
Discussion
A representative cohort was quickly recruited, and the first pilot RCT efficiently conducted. Attrition and uptake were comparable with other pragmatic studies in ADHD. Although therapists reported that delivery was challenging, 70% of participants accepting treatment received at least three consultations and reported they were helpful. No serious adverse events attributable to treatment occurred.
Modifications are required to improve poor return of teacher outcomes, therapist’s contact-ing strategies, crossover from treatment to usual care (40%), and cohort representativeness.
Conclusion
The TwiCs design can make an important contribution in the search to improve outcomes for those with ADHD. The STAR project demonstrated the feasibility of the TwiCs approach to pragmatic RCT design for testing interventions for children with ADHD. In the pilot RCT the novel interventions showed preliminary indications of effectiveness according to carers.
