2.1. Study design and procedure
A two-armed, parallel-design randomized controlled superiority trial will be conducted to evaluate the effectiveness of the GGOC-AD module intervention for adolescents compared to a non-active module (GGN-AD). The randomization procedure is described in detail in the subsection 2.5. The present study protocol complies with the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) checklist.
Figure 1 depicts the trial design and times of assessment. Students will be recruited from public schools in the metropolitan area of Valencia (Spain). Researchers will invite different schools to participate in the study. After the education center agrees to participate, members of the research group will meet students in the designated classroom, provide information about the study, and invite them to participate. Students will then receive a sheet with information about the study and the informed consent for parents or guardians and the student. Finally, participants that explicitly agree to participate, provide the informed written consent, and meet the inclusion criteria will be randomized in two groups: the intervention group that will use GGOC-AD and the control group that will use the GGN-AD module. On weekdays, students will complete the app during school hours at a pre-arranged time with the education center in order to interfere as little as possible. At least two members of the research group will be available to assist in the assessment sessions (T1-T3) and monitor the use of the app in order to promote participants’ retention. Students will be asked to complete the app at home during the weekend. As shown in Fig. 1, on the first day of the study, participants will be requested to complete the baseline assessment measures (T1, pre-treatment) and download the GGtude platform app using their personal mobile device. It will be freely available from the App Store or Google Play. The intervention group will access the GGOC-AD module of said app and the control group will use the GGN-AD module. Users will carry out cognitive training for 14 consecutive days, completing approximately three levels per day. On the day that they finish the app (day 14), participants will be instructed to complete the assessment instruments again (T2, post-treatment). Finally, a one-month follow-up evaluation will be conducted (T3). Both groups will complete the outcome measures and the app at about the same time, using their personal mobile device. At all-time assessments, questionnaires will be completed by Lime Survey software in a designated classroom.
Note
14 days of GGN-AD: use of the GG Neutral – Adolescents module during 14 days; 14 days of GGOC-AD: use of the GG Obsessive-Compulsive Disorder – Adolescents module during 14 days.
2.2. Participants
This study will be conducted in a formal educational setting with students from schools in the metropolitan area of Valencia (Spain). The expected age of participants will be between 15 and 17 years old. Inclusion criteria for participants will be: a) providing written, informed consent (both the adolescent and their parents or legal guardian), b) being enrolled in 4th level of Obligatory Secondary Education of the Spanish Education System (ISCED levels 2 and 3 in the International Standard Classification of Education), and c) availability to use a mobile phone (indistinctly Android or iOS mobile system) with Internet access.
2.3. Intervention
Both intervention and control modules will be administered in the school center. Participants will use the module on their own mobile device individually. During weekdays, they will use it in their classroom with the assistance of researchers, whereas on weekends they will be required to complete it at home on their own.
GG Obsessive-Compulsive Disorder – Adolescents (GGOC-AD) module
The intervention module will be GGOC-AD. It consists of game-like interactions that are intended to help users increase their accessibility to self-statements that facilitate more adaptive appraisals of thoughts, emotions, and coping strategies. Users of GGOC-AD will be presented with “blocks” featuring statements related to OCD relevant beliefs, such as “I can’t trust my memory” or “No one is perfect”. Participants will be instructed to respond to these statements by either accepting them (i.e., pulling the blocks downward toward themselves) or rejecting them (i.e., throwing the blocks upward away from themselves). If a participant makes a wrong choice, that is, accepts a statement representing an OCD-related maladaptive belief, a notification will appear encouraging them to try again. GGOC-AD is composed of a total of 57 levels that are divided in 14 blocks. Users are instructed to complete one block, that is, three to four levels per day. Each level has approximately eight to ten statements, which take approximately one minute to complete.
The GGOC-AD module was designed based on the Spanish version of the OC module previously used in the non-clinical adult population and OCD patients (18, 21, 24). The adolescence adaptation was carried out as follows: first, the 968 statements that compose the adult version were reviewed, one by one, by independent members of the research group. Each reviewer selected the ones that could be problematic for adolescents, considering: a) the lack of compressibility of the vocabulary and b) the contextual inappropriateness for teenagers. For instance, sentences that referred to a working environment or to a romantic relationship were considered problematic due to being difficult to generalize to the adolescent population. A total of 370 statements were proposed to be reviewed. To that end, three separated blocks of statements were administered to a group of nine adolescents from different schools. The adolescents were instructed to carefully read each statement and evaluate each one according to a 10-point Likert scale (from 1 = “I don’t understand it at all” to 10 = “I understand it completely”). After that, they had to propose a better alternative for those that were difficult to understand (scores < 8). The research group discussed each of the sentences and the alternatives proposed by the adolescents in different sessions and agreed to change a total of 75 statements by consensus. With this version of the OC module, five adolescents completed the module over 14 days. After using the module, the research group members met the students individually to ask for their opinion on the statements. Considering the feedback given by the participants, the research group reviewed all the statements in the module again. Finally, a total of 335 statements (34.6 %) were aapted (e.g., “My partner accepts me” was changed to “My friends accept me”).
GG Neutral – Adolescents (GGN-AD) module
The control module will be GGN-AD. It is a non-active intervention that follows a similar structure and functioning to the intervention module. It includes the same number of levels as GGOC-AD and participants will also complete three levels per day. The main difference from the intervention module is the statement content. In this case, the statements are true/false options related to general knowledge, geography, math, or language issues. Participants will then have to accept the statements that are true, dragging them to the bottom of the screen (e.g., "Madrid is the capital of Spain"); and reject false sentences, dragging them to the top of the screen (e.g., "China is the capital of Spain"). The sentences of GGN-AD were reviewed and adapted to the Spanish context by the research team, because they were originally written to suit the English-speaking population. For example, phrases such as “Arizona is a state of the USA” were changed to “Valencia is a Spanish city”.
2.4. Primary and secondary outcome measures
All measures will be completed at pre-, post- and follow-up assessment. Participants will also complete a sociodemographic data sheet (see additional measures). The primary outcomes measures will be:
Obsessive Compulsive Inventory - Child Version (OCI-R-CV) (27). The OCI-R-CV is a self-report instrument that assesses the discomfort associated with obsessive-compulsive symptoms in children and adolescents between 8 and 18 years old. It consists of 21 items with a 3-point Likert scale (from 0 = “never” to 2 = “always”). The total score and the six subscales (Doubting, Obsession, Hoarding, Washing/Checking, Ordering, and Neutralizing) will be used in this study. The Spanish validation of the OCI-R-CV shows good psychometric properties (28).
Obsessive Beliefs Questionnaire - Child Version (OBQ-CV) (29). The OBQ-CV is a self-report questionnaire that assesses obsessive beliefs through 44 items with a 5-point Likert scale (from 0 = “strongly disagree” to 4 = “strongly agree”). In this study, we will consider the three subscales: a) Responsibility and threat estimation, b) Perfectionism and intolerance of uncertainty, and c) Importance and need to control thoughts, and the OBQ-CV total score. The Spanish adaptation presents good psychometric properties (30).
The secondary outcome measures will be as follows:
The Patient Health Questionnaire for Depression and Anxiety (PHQ-4) (31, 32). This is a brief-screening to assess anxiety and depression symptoms. It is composed of 4 items evaluated with a 4-point Likert scale (from 0 = "never" to 3 = "almost every day"). The total score and the two subscales (anxiety and depression) will be used in this study. The Spanish validation shows good psychometric properties (33).
Single-Item Self-Esteem Scale (SISE) (34). The SISE is a self-report measure that determines the extent to which the sentence “I have a high self-esteem” describes participants on a 9-point scale, ranging from 1 (“not very true for me”) to 9 (“very true for me”). In this study, the SISE will be used as a self-esteem score, in accordance with the reliability and criterion validity reported by Robins et al. (34).
Additional measures:
Sociodemographic data sheet: The data required will be the following: age, gender, data about their school center (academic year, class group, and school center).
2.5. Randomization
All the participants will be randomly assigned to GGOC-AD (intervention) or GGN-AD (control) groups (allocation 1:1) with the randomizer.org platform (no block randomization). It will take place prior to the T1 assessment and will be conducted by a researcher not involved in the assessment and intervention procedure.
2.6. Data management
Data will be collected and managed in line with the Ethics Committee of Research in Humans of Universitat de Valènica. Participants will access the questionnaires using a unique personal identification number (ID), created beforehand. It is a code that ensures anonymity and does not allow participants to be identified. Their ID will be used to match the collected data. No other personal data is requested. Data from the questionnaires will be stored on a secure server of the Universitat de Valènica. In addition, persons who do not adequately answer the two control questions (e.g., “In this question, please check "always") included in the questionnaires at each evaluation time will be eliminated from the data base.
2.7. Statistical analyses and sample size
Analyses will be performed using the free software R. The effect of the intervention on primary and secondary measures over time will be appraised using a generalised linear mixed model (GLMM), and, specifically, a linear multiple regression model where time (T1, T2 and T3) and group (experimental and control) are the independent variables whereas the primary and secondary outcome measures are the dependant variables. To avoid overly optimistic estimates of the effectiveness of the experiment, an ITT (intention to treat) approach will be used, using the method of multivariate imputation by chained equations algorithm (MICE). Descriptive statistics such as means, standard deviations, and frequencies will be also calculated.
The sample size of the study has been estimated a priori using G-power 3 (35) based on the formula provided by Cohen (36). It was revealed that a sample of at least 55 participants will need to be recruited to accomplish a power study of .805 at alpha = .05 and a medium size effect (.15). Considering the dropout rates in previous studies in our context (21), we will invite approximately 200 individuals to participate in the current study.