Study design and data analysis

We will conduct an open trial, using a pretest post-test within-groups design to evaluate change on outcome measures. This research design is recommended when conducting preliminary research on the effectiveness and feasibility of novel treatments.32 Following screening, eligible participants will complete the outcome measures at time 1 and then again following completion of the programme at time 2 (post-test). The flow of participants through the study is shown in figure 1.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Flow of participants through study.

Sample size

Sample size calculations were based on a moderate effect size at the conventional α level of 0.05. A moderate effect size was estimated on the basis of results from a review of self-help treatments for anxiety disorders in young people.33 Based on this, it was calculated that a minimum of 45 participants is required to adequately power the study. The literature on which to base estimated attrition rates was limited, however, a randomised controlled study of internet-based CBT for child anxiety disorders indicated a screening exclusion/dropout rate of 38% and a further 25% dropout rate from the treatment condition.34 It should be noted that in this study, screening and pretesting assessments were conducted with trained interviewers via telephone. Based on this estimate, and taking into account the fully automated nature of the screening and pretesting process in the current trial, we aimed to recruit 150 participants to the screening process to account for participants lost to screening, pretesting and later attrition during treatment.

Participants

We aim to evaluate the effectiveness of the programme in a sample of young people (aged 12–18) residing in Australia. Eligible participants are Australian residents aged 12–18 years who are currently exhibiting symptoms of OCD, as determined by their score on the Short OCD Screener (SOCS),35 who speak English and who have internet access and an email address. In addition, participants are not eligible if they report current moderate-high self-harm or suicide risk as measured by the suicide risk module of the Mini International Neuropsychiatric Interview for Children and Adolescents (MINI-KID).36 This module contains 14 yes/no questions that produce a dichotomous suicide risk rating (present or not present) as well as a numeric suicide risk score with identified anchors for ‘low’, ‘moderate’ and ‘high’ suicide risk. This module has shown good to excellent inter-rater and retest reliabilities for current and lifetime suicidality (area under the curve=0.89–0.99, κ=0.81–0.96).36 Further exclusion criteria include elevated symptoms of eating disorder or psychosis, as measured by the SCOFF questionnaire37 and the Adolescent Psychotic-Like Symptom Screener (APSS).38

Recruitment

Recruitment of participants to the OCD? Not Me! programme is via existing referral networks of general practitioners mental health professionals, as well as through advertising in print media and on social networking sites. As part of the referral process, information about the programme is provided via conference/workshop presentations and direct communication with specialist healthcare providers.

Procedure

Young people and their parents and caregivers are invited to register for the study by accessing the website (http://www.ocdnotme.com.au), and are able to provide informed consent to participate in the programme by following an online consent procedure. Once informed consent has been provided, participants receive secure login details that enable them to log into the screening assessments and, if eligible, to access the programme. To determine whether the OCD? Not Me! programme is appropriate for the young person's current symptoms, they are invited to complete the screening measures described above. These measures are administered online via the website and are automatically scored to immediately determine programme suitability. If the programme is deemed suitable, the young person is then invited to complete the young person pre-programme assessment battery described below. In addition, parents and caregivers receive an email notification informing them that the OCD? Not Me! programme has been deemed suitable for their young person's current needs. Parents and caregivers are also required to complete the parent/caregiver pre-programme assessment battery described below, by clicking on a link provided in the email. As with the screening measures, the pre-programme assessments are administered and scored entirely online, via the website. Once eligible participants and their parents/caregivers have completed the pre-programme assessment batteries, young people then receive cumulative access to the eight-stage online programme. As participants begin each stage, an automated email is sent to parents/caregivers with a link to the supporting resources for that stage. As the programme is fully automated, no clinician contact is utilised throughout the assessment or treatment process, or in the follow-up of idle accounts. Participants who do not log into their account for a week receive an automated email reminding them to login. Should the participant's account remain idle, this email is resent once a week for 3 weeks. At the end of this period, participants who have not logged into the programme in the preceding month are locked out of their account. They are able to access the programme if they complete the registration and pretesting period again. A participant is considered a dropout if their account is idle for 1 month.

If the programme is not deemed suitable for the young person, they are immediately notified, and their parent/caregiver will receive an email informing them of the outcome of the screening assessment. In addition, if young people indicate during the screening assessment that they are currently experiencing moderate-high risk of self-harm or suicide, their parents/caregivers will receive a separate email informing them of this risk, with information and resources on how to manage it. Parents and caregivers are invited to seek out more intensive and specialised services using the national referral database.

Diagnostic assessment

In order to establish a diagnosis of OCD and to determine the presence of other comorbid diagnoses, we developed the Youth Online Diagnostic Assessment (YODA). The YODA is an online self-report measure designed to assess psychiatric symptoms in young people according to the diagnostic criteria outlined in the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-V).39 Development of the YODA for the purpose of this study was necessary because there were no comprehensive psychiatric assessments available for young people that were (A) self-report format; (B) designed for online administration and (C) available free or at minimal cost. The YODA covers the following disorders: OCD, separation anxiety disorder, social phobia, specific phobia, panic disorder, generalised anxiety disorder, major depressive episode, dysthymia, post-traumatic stress disorder, bulimia, anorexia, tic disorders, Tourette's disorder, attention deficit/hyperactivity disorder, conduct disorder, oppositional defiant disorder and pervasive developmental disorder.

Outcome measures—young people

To capture detailed data on young people's OCD symptoms and their well-being before and after the programme, we constructed a comprehensive assessment battery from a number of extant measures. These measures are described in detail below.

The Children's Florida Obsessive Compulsive Inventory

The Children's Florida Obsessive Compulsive Inventory (C-FOCI)40 was selected to assess OCD symptoms. This measure is a brief self-report measure of OCD symptoms in children and adolescents, comprising a 17-item symptom checklist and a 5-item severity scale. The C-FOCI was selected because it has been validated for internet administration and is suitable to compare with the Children's Yale-Brown Obsessive Compulsive Scale (CY-BOCS).35 While the clinician-administered version of the CY-BOCS is often cited as the gold standard measure of OCD symptoms in young people, the self-report version performs less well—as a result, it has been argued that this measure is not feasible for use outside clinical settings.40 The C-FOCI has demonstrated adequate internal consistency, and convergent and discriminant validity.40 In addition to pretest and post-test administration, participants are asked to complete the C-FOCI prior to starting each stage of the programme.

Children's Obsessive Compulsive Inventory—Revised

The self-reported Children's Obsessive Compulsive Inventory—Revised (ChOCI-R)41 was selected as a secondary measure of OCD symptoms. This 32-item measure comprises 20 questions that evaluate the presence of specific obsessions and compulsions in children and adolescents, and 12 questions that assess severity of OCD symptoms and associated impairment. The ChOCI-R has demonstrated excellent internal consistency, and good convergent and discriminant validity.41

The Youth Quality of Life Instrument—Short Form

The Youth Quality of Life Instrument—Short Form (YQoL-SF)42 was chosen to measure quality of life. The YQoL-SF is a 16-item self-report scale that provides a multidimensional assessment of quality of life among children and adolescents. Psychometric data for the longer version of the scale (the Youth Quality of Life Instrument—Revised) support the reliability and validity of this instrument.42

Rosenberg Self-Esteem Scale

The Rosenberg Self-Esteem Scale (RSES)43 was chosen to measure self-esteem. This 10-item measure was designed to assess self-esteem among children and adolescents, and is the most widely used measure of self-esteem available. The RSES has adequate internal reliability and evidence supports the construct validity of this measure.43–45

Feedback questionnaire

A 10-item online feedback questionnaire was developed to assess participants’ experience of the OCD? Not Me! programme. This self-report questionnaire uses a combination of Likert-type response scales and open-ended questions to assess various dimensions of participant experience including enjoyment, relevance, satisfaction and degree of completion. This survey is automatically issued via email link when participants exit the programme (subsequent to withdrawal, dropout or completion).

Outcome measures—parents and caregivers

The aim of including outcome measures for parents and caregivers was threefold: (A) to collect parent/caregiver reports of their young person's OCD symptoms; (B) to assess parent/caregiver involvement in their young person's OCD rituals and (C) to assess parent/caregiver distress. The measures included in the parent/caregiver assessment battery are described below.

ChOCI-R (parent report)

The parent-report version of the ChOCI-R was selected to corroborate the self-report version of the ChOCI-R administered to the young person. Similar to the self-report version of this scale, the parent report comprises 20 questions that evaluate the presence of specific obsessions and compulsions in children and adolescents, and 12 questions that assess severity of OCD symptoms and associated impairment. Psychometric evaluation of this scale has indicated that it displays good internal consistency, and convergent and divergent validity, and that item and scale scores share strong correlations with scores on the self-report scale.41

Child Obsessional Impact Scale—Revised

The parent-report version of the Child Obsessional-Compulsive Impact Scale—Revised (COIS-R)46 was selected to assess the degree to which OCD symptoms were causing functional impairment in the young person. This 33-item measure assesses OCD-specific functional impairment across four domains (daily living skills, school, social, family/activities) and has good internal consistency, concurrent validity and test-retest reliability.46

Family Accommodation Scale for Obsessive Compulsive Disorder

The Family Accommodation Scale for Obsessive Compulsive Disorder—Self-Rated Version (FAS-SR)47 was chosen to measure the degree to which family members facilitate or participate in a young person's OCD-related rituals and/or avoidance. This 19-item measure has demonstrated excellent internal consistency and strong convergence with criterion measures.47

21-Item Depression Anxiety Stress Scales

The 21-item version of the Depression Anxiety Stress Scales (DASS-21)48 was selected to measure parental distress. This scale is designed to measure psychophysiological symptoms of depression, anxiety and stress, in clinical and non-clinical populations. Psychometric evaluation of this scale has indicated that it demonstrates adequate internal consistency and concurrent validity.

Data analyses

Multilevel mixed effects linear regression will be used to evaluate change on the outcome measures. A generalised linear mixed model (GLMM) procedure will be used to test for the effect of time in the context of a hierarchical design, with time treated as a fixed effect, and participants treated as a random effect. The primary benefit associated with using the GLMM maximum likelihood procedure is that it maximises statistical power and reduces the impact of sampling bias due to participant attrition, as it does not rely on all participants providing data at each time point.49 In addition, GLMM can account for unequally spaced data collection points and correlations between repeated measurements.50 Because this procedure uses all available data at each time point, it is particularly useful for conducting intent-to-treat analyses.50

Significant main effects for time will be followed up with pairwise contrasts to compare pretreatment to post-treatment scores on each of the outcomes measures. Effect sizes will be calculated within groups, using the formula provided by Cohen.51 In addition, we will calculate the proportion of participants who report reliable and clinically significant change on the C-FOCI and ChOCI, using the methodology proposed by Jacobson and Truax.52

Data collection and monitoring

Data will be collected and managed in line with the Australian Code for the Responsible Conduct of Research.53 As part of the intervention development, software was developed to automatically and securely monitor, score and download data collected online. Members of the research team are responsible for overseeing and monitoring data collection, and for providing progress reports to the Australian Government Department of Health.

To motivate participants to engage with the programme and to promote participant retention, participants are offered one entry into a prize draw for a gift card valued at $A30 when they reach the halfway point of the programme. Identifying data collected from participants for the purpose of notifying them about the results of the prize draw is automatically separated from other data collected.

Ethics

The Curtin University Human Research Ethics Committee approved the study protocol, and the trial was registered with the Australian New Zealand Clinical Trials Registry (ACTRN12613000152729).

Methodological considerations

The current study has several strengths, including the repeated measurement of OCD symptoms, and suicide and self-harm risk at the beginning of each stage of the OCD? Not Me! programme. The use of repeated measurements allows us to monitor and respond to risk, as well as to analyse trajectories of change in OCD symptoms over the course of the programme. In addition, the inclusion of a broad range of outcome measures, such as self-esteem, family accommodation and family distress, allows us to evaluate the impact of the programme on clinically meaningful outcomes relating to quality of life and family functioning. Some limitations must also be acknowledged. First, the validity of the study may be affected by the use of a diagnostic measure (YODA) that has not yet been validated, and the reliance on self-report measures.

Second, the absence of a control condition due to the open trial design limits conclusions about the efficacy of the programme.