A randomised controlled trial of the Neuro Emotional Technique (NET) for childhood Attention Deficit Hyperactivity Disorder (ADHD): a protocol

The objectives of this research are to:

Ethics approved, registered, interdisciplinary, placebo controlled, double blind, randomized clinical trial.

Four private clinics in Sydney, Australia.

Four chiropractors, three male and one female volunteered their time and services to administer interventions to the participants of the study at no cost. The chiropractors had varied tertiary qualifications: one had Bachelor of Medical Sciences and a Master of Chiropractic; two had a Bachelor of Science and a Master of Chiropractic and one had Bachelor of Science and a Graduate Diploma of Chiropractic. All of the chiropractors had undergone the four NET training workshops over a two year period, and had passed the certification exams in both theory and practice, and a thus were qualified as NET certified practitioners. These practitioners were chosen on the basis that they were experts in the field, were certified and agreed to abide by the study protocols.

Parents were initially screened on the phone, and their children were included in the study if inclusion criteria were met. Inclusion criteria comprised: children aged between 5–12 years; children with a medical diagnosis of ADHD for minimum of 2 months, diagnosed by Pediatrician, Psychiatrist or Clinical Psychologist; informed verbal and written consent from parents and participants; children currently undergoing any interventions (pharmacological, psychosocial therapy, educational, occupational therapies etc) or children with the diagnosis of ADHD who have not undergone any treatment; and children who presented with the following comorbid disorders – Conduct Disorder, Oppositional Defiant Disorder, Learning Disabilities, and Anxiety Disorders. Any children who did not meet the inclusion criteria, or who met the exclusion criteria were excluded from the study. Exclusion criteria comprised: A diagnosis of Language difficulties, Mood Disorders (eg. Depressive Disorder or Bipolar Disorder), Communication Disorder, mutism, Pervasive Developmental Disorder (eg. Autism or Asperger's Disorder), Psychosis, profound deafness and other physical or mental disabilities that would prevent participants interacting physically and verbally with the care provider during the consultation; any changes in medications during the course of the study, and any child or parent who had previously undergone any NET treatment.

The Diagnostic and Statistical Manual of Mental Disorders ADHD diagnostic criteria [53, 54] were used by the participants' Pediatricians and Clinical Psychologists to establish a diagnosis. The reliable and validated Conners' Parent Rating Scales-Revised Long Version (CPRS-R:L) and Conners' Teacher Rating Scales-Revised Long Version (CTRS-R:L) [55‐59] were used pre and post intervention.

The ratings scales were administered pre-intervention to establish baseline data, and were used as outcome assessments to measure the intervention effects at the end of the first month and at the conclusion of the study i.e. in seven months. Of the 14 subscales described in the CPRS-R:L and the 13 subscales described in the CTRS-R:L, two were chosen as primary outcome measures (POMs) and three were chosen as secondary outcome measures (SOMs). Primary outcome measures included Conners' ADHD Index and Conners' Global Index Total subscales H and K respectively. The secondary outcome measures included DSM-IV:Inattentive, DSM-IV:Hyperactive-Impulsive and DSM-IV:Total, subscales L, M, and N respectively.

Eligible participants were randomly allocated to one of three groups A, B, & C (i.e. group A-sham, group B-treatment, and group C-control). The "QuickCalcs" program from the GraphPad Software [60] was used to randomly sequence 150 numbers (repeated 1×) into the three groups.

Each number and its randomised letter (A, B, or C) were placed in an opaque envelope. Each envelope was numbered consecutively from 1–150.

As each parent called to have their child screened they were allocated the next consecutive number once they were eligible, so that eligible participant no. 1 was allocated to envelope no. 1, etc. The screener was unaware of the sequence of group allocation at the time of screening, hence allocation was concealed at time of recruiting.

Participants were blinded as to which intervention group they were allocated to, in an attempt to control for the Hawthorne Effect. Parents and teachers were blinded to which intervention group the children were allocated to, in an attempt to minimize rater bias when completing rating scales i.e. CPRS-R:L and CTRS-R:L. The registered Psychologist and the Clinical Psychologist were blinded as to which group participants were allocated to, in order to minimize bias when scoring and interpreting the data from the CPRS-R:L and CTRS-R:L.

The NET practitioners were not blinded in this study as they needed to know which participants were receiving which intervention, i.e. treatment versus sham protocol. The parents of the control group were also aware that their children were in the control group awaiting treatment at the completion of the study.

Calculations for the sample size were set with a significance level of 0.05 and the power of 80%.

As there are no prior studies conducted in the area of chiropractic or NET for an ADHD cohort, for power analysis purposes a SD of 6.7 was adopted based on data from Shortt [61], yielding a sample size of 63. The value of 6.7 for power testing is supported by findings from a paper on the meta-analysis of RCTs in ADHD responses to atomoxetine using the CRS [62]. This study revealed an average SD score of 6.7, which is consistent with the SD found in the Shortt data [61]. In a meta-analysis of psychotherapy dropouts the mean attrition rate was 46.86% [63, 64]. Based on this drop-out rate it was prudent to increase the sample size by this figure, increasing it from 63 to n = 93.

Recruitment for the study commenced in July 2006 and the first participant was enrolled in August 2006 and the last participant completed the study in March 2008. Each participant that qualified and was enrolled into the study and had their own start date, hence the staggered starts and as a result the study took 17 months to complete.

As part of the phone screening and the informed consent forms the parents of the participants were advised that they were free to withdraw their child/children from the study at any stage without any consequences.

Group A, the sham group, continued with their existing program and the sham protocol was added to their regimen. Group B, the treatment group, continued with their existing program and the NET protocol was added to their regimen. The participants in groups A and B attended a clinical facility for the first month and received eight treatments (two treatments per week) for approximately 15–20 minute sessions. Then, they were scheduled for another six treatments over the next six months (i.e. one treatment per month). Group C, the control group, consisted of children who continued with their existing ADHD treatment regimen as prescribed by their pediatrician, clinical psychologist and/or medical doctor.

The sham protocol was designed to mimic the real NET treatment protocol, however without the therapeutic benefits. The sham protocol incorporated the MTP throughout the procedure, with different sets of statements and tapping on the back. The set of statements used were not related to the treatment protocol, but instead used for testing of congruent and non-congruent reactions to statements like, " my name is John" and then "my name is Robert" i.e. the child's name first then another name. See Table 2 for a complete list of the first Sham protocol. In total eight different sham protocols were designed with different lists of referential statements.

As there are no known side effects and there have not been any reports of adverse reactions to the NET protocol, we do not anticipate that any participants will be exposed to any unnecessary risks. However, in the event that a child becomes emotionally distressed during the protocol, the procedure will be terminated. Support counselling will be offered by qualified psychotherapists and counsellors for the child at no expense to the parents. These professionals have all been approved for the study by the university Ethics Committee and have also undergone background screening and have been approved to work with children.

The trial protocol was reviewed by the Macquarie University Human Ethics Committee (Sydney, Australia) and was granted Human Ethics Approval (HE26AUG2005-M04261). As a prerequisite to conducting research with children, the Ethics committee required all those involved with the study undergo background screening as part of the "Working With Children Check". All those involved in the study were approved and ethics was granted.

Furthermore, as part of this study, an application was made to the Department of Education and Training and a State Education Research Approval was granted (SERAP #06.350). This enabled the researchers to send questionnaires to the teachers of the participants of the study.

All of the parents of the participants were issued with two copies of the informed consent forms (one to keep and one to return to the university). The informed consent form conveyed the following information: the names of the chief investigator, the supervisors and the four practitioners administering the interventions; the name of the university and the department conducting the research; a brief description of the control, treatment and sham groups; the duration of the study; the questionnaires used; the Ethics approval and contact details of the Ethics officer; the right to withdraw their child at any time; in the event a child becomes distressed the back-up counselling services available; the involvement of teachers; and the fact that the results of the study will be disseminated via conferences and publications whilst maintaining the children's privacy.

Demographic and baseline characteristics of participant variables will be compared between the three randomised groups using the Restricted Maximum Likelihood (REML) in GenStat 10^th Edition [69].

The primary analysis will compare outcomes for each primary and secondary outcome measure between the three groups (treatment, sham and control) using the Restricted Maximum Likelihood (REML) in GenStat 10^th Edition [69], to estimate treatment effect with 95% CI. REML is a method for fitting mixed linear models (the models for ANOVA being special cases) which can produce unbiased (or less biased) estimates of variance and covariance parameters. This is a relatively new and powerful technique that allows for data sets that are unbalanced, have missing values, have treatments with unequal variances and/or have correlated error terms [70].

An intention-to-treat (ITT) analysis will be undertaken of the one month and seven month data from the CPRS-R:L and CTRS-R:L for the participants who commenced the study but did not complete.