Objectives

The main objectives of this study are: (1) to design and implement a serious game called emoTICare to promote physical and emotional health, as well as the development of socio-emotional skills among adolescents with Type 1 Diabetes Mellitus (T1DM) through a psychological intervention conducted on a technological platform that incorporates artificial intelligence and Ecological Momentary Assessment (EMA); (2) to analyze the effectiveness of this serious game on indicators of disease adjustment, quality of life, and socio-emotional well-being, as well as improvements in socioemotional competencies and skills.

The study hypothesizes that adolescents will maintain or improve their physical and emotional health indicators following the emoTICare program. Specifically, statistically significant improvements (p ≤ .05) are expected in the following variables: health-related quality of life, social skills and communication, emotional awareness, resilience, self-esteem, as well as coping and problem-solving abilities. Additionally, a statistically significant reduction (p ≤ .05) is anticipated in variables related to the perceived threat of the disease, as well as emotional and behavioral problems. The experimental group is expected to show improvements in all these areas when comparing pre- and post-treatment evaluations, and also in comparison to the control group on the waiting list.

Study design

A clinical longitudinal experimental design with inter-subject comparisons was employed, with the aim of comparing an experimental group (i.e., individuals receiving psychological treatment) and a control group on the waiting list (i.e., those without treatment) at three time points: before treatment (T1), and two post-treatment time points (T2, T3). The control group hasn’t any alternative activities or exposure to placebo-like interactions. Furthermore, an intra-subject analysis was conducted by comparing repeated measures to observe changes within the same individuals before (with pre-treatment measurements, T1), after 6 weeks (T2, in the control group and post-treatment in the experimental group), 6 months after T2 (T3, in the control group and post-treatment in the experimental group).

Initially, immediately after completing the informed consent online, a first evaluation will be conducted for all participants (T1), covering the main areas that constitute the psycho-emotional state of the participants. Following this first evaluation, participants in the experimental group will engage in six weeks of gameplay, with each week dedicated to completing one mission of emoTICare. At the six-week mark after the initial evaluation, and immediately following the completion of the emoTICare application by participants in the experimental group, a second evaluation (T2) of all participating adolescents (control and experimental groups) will take place. Following a period of six months from T2, a subsequent measurement, designated T3, will be conducted in all participants.

T1 will serve as the baseline measure of the psycho-emotional state of participating adolescents before the commencement of the emoTICare intervention. The homogeneity of the control group will be evaluated by comparing the T1 scores of both the control and experimental groups to ensure no significant differences exist, thereby allowing for comparable study groups. T2 will also function as a control measure for all participants concerning the evaluated variables. However, in the control group, only the changes resulting from the natural progression of time without intervention will be examined, while in the experimental group, the benefits achieved in these variables following the emoTICare intervention will be analyzed. T3, the follow-up measure, will also function as a control measure after six months for all participants with regard to the evaluated variables.

This experimental study includes a randomized clinical trial aimed at assessing the effects of completing the emoTICare program concerning the socio-emotional competencies and psychological well-being of these adolescents. The method employed for generating allocation sequences was a computerized random number generator. The sequence generation was based on a process of simple randomization, founded upon a single sequence of random allocations. The R programme was employed to allocate the participants at random to distinct groups (control and experimental) with equal probability. Consequently, it was guaranteed that every participant had an equal probability of being placed in any group. In order to guarantee allocation concealment and minimise selection bias, the allocation process was conducted in a controlled and secure environment. The sequence generated by the R program was stored in such a way that the researchers responsible for allocation did not have access to allocation information prior to the inclusion of participants in the study. Furthermore, the allocation was executed by an independent researcher external to the team, who interacted directly with the participants, thereby facilitating the maintenance of blinding. The integrity of the procedure was ensured through regular monitoring, and any possibility of manipulation of the allocations was avoided. This ensured that both the participants and the researchers responsible for the intervention were blinded to the assigned group, thus minimising any possible selection bias and ensuring the validity of the results.

The principal investigator will be responsible for obtaining informed consent from the participants (adolescents and parents/legal guardians) by informing them about the study. Adolescents aged between 12 and 16 inclusive will be required to sign the consent form in conjunction with their guardian, while those aged 16 and over will be permitted to sign it directly.

The collection of data will be facilitated through the utilisation of the Limesurvey platform, wherein the participants will be prompted to complete questionnaires in an anonymous manner. The principal investigator will supervise access and data security, and the research team will only access anonymised data for analysis.

The research team will measure baseline data at the commencement of the study through online questionnaires, and also the two post-test measures. The randomisation process will be overseen by the principal investigator and team, and will be conducted using a computerised random number generator.

The research team will be responsible for configuring and training participants and staff on the use of the Limesurvey platform, ensuring proper usage of the system.

According to the Standard Protocol: Recommendations for Interventional Trials statement [49], the current study protocol outlines the details of the study rationale, objectives, interventions, methods and statistical analyses, organization, and ethical considerations (See Figs 1 and 2).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. SPIRIT schedule of enrolment, interventions and assesments.

https://doi.org/10.1371/journal.pone.0325763.g001

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Study design of the serious game intervention.

https://doi.org/10.1371/journal.pone.0325763.g002

The literature review process, the design of the psychological content of emoTICare, and the technological design of the serious game were conducted from December 2022 to June 2024. In July and August 2024, a beta study (pilot study) was carried out with a smaller sample of adolescents without Type 1 Diabetes Mellitus, during which the emoTICare program was tested to correct any potential errors or issues that arose during gameplay, as well as to gather feedback from the adolescents. Since September 2024, recruitment has commenced for the study sample of adolescents with Type 1 Diabetes Mellitus.

An accidental or convenience sampling method was employed. Participants were recruited as follows: all diabetes-related associations in Spain (over 100) were contacted via email and/or phone to request their voluntary participation in the study. These associations were asked to disseminate the study information to their members so that interested individuals could contact the researchers via email to participate. Regarding potential confounding variables, we acknowledge that differences between groups may arise due to the non-randomized recruitment method. To address this, we collected demographic and clinical variables (e.g., age, gender, diabetes type, disease duration) to assess and statistically control for potential confounders in the analyses.

In terms of selection bias, the recruitment process relied on voluntary participation through patient associations, which may lead to an overrepresentation of individuals who are more engaged in diabetes-related activities and support networks. This potential bias was mitigated by reaching out to a broad range of associations across Spain, ensuring geographic and organizational diversity in participant recruitment.

The phase encompassing the evaluation of the variables of interest, T1-T2-T3, will take place through the survey platform of the Universitat de València, Limesurvey, from September 2024 to the end of February of 2027. The response platform for the questionnaire battery includes a consent form for anonymous participation in the study. In this platform, informed consent will be obtained from the parents or legal guardians of the participants, as well as from the minors themselves, ensuring that all are informed about the anonymity and confidentiality of their data, confirming that their participation in the study is voluntary.

To access and play the emoTICare serious game online for six weeks (one week per mission of the game), participants will be provided with a link at the end of the evaluation survey conducted at T1. This serious game contains six distinct areas, each designed to be completed within a maximum of one week, thus allowing sufficient time to progress at different paces. During the period in which the adolescents are completing the serious game, an email address and a mobile phone number will be made available for participants to contact in case they need assistance (see Table 1).

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Intervention timeline.

https://doi.org/10.1371/journal.pone.0325763.t001

Sample size

In Spain, the report on “Prevalence of Diabetes Mellitus” [50] highlights that more than three million people in the country have diabetes, of which 3.4% have Type 1 Diabetes Mellitus (T1DM). The Spanish Diabetes Society [51] estimates that T1DM accounts for 1 in every 10 cases in Spain, totaling approximately 90,000 individuals with this condition. Specifically, the incidence of diabetes in individuals under 15 years old varies between 11 and 24 cases per 100,000 inhabitants per year [51].

To determine the sample size, we used a 95% confidence level and a 5% margin of error, which is a standard approach in medical and epidemiological research. This calculation was based on the estimated population of individuals with T1DM under 15 years of age (approximately 90,000 individuals). Considering the anticipated dropout rate and the expected effect size, the total sample size was calculated to be 372 participants. This estimate assumes a normal distribution of the population and that the outcomes of interest are representative of the general T1DM population in Spain. Assumptions made during the calculation include: A 95% confidence level and 5% margin of error; A normal distribution of the studied variables, consistent with similar research in the field.; A moderate effect size based on previous studies investigating similar interventions and a 20% dropout rate, which is typical for this type of intervention study.

Limitations of the sample size calculation include potential variability in the effect size, which could differ from our assumptions. Additionally, since participants are recruited through specific diabetes associations, this sample may not fully represent the broader T1DM population, potentially introducing selection bias. Given the potential variability in healthcare access, lifestyle factors, and diabetes management across different regions, participants’ region of origin was recorded to assess potential regional differences. This information allows us to explore whether geographical factors may influence the study outcomes and, if necessary, include region as a covariate in statistical analyses to control for potential biases due to regional heterogeneity.

Eligibility and recruitment criteria

To be included in this study, participants must be aged between 12 and 16 years, have a diagnosis of Type 1 Diabetes Mellitus (T1DM) for at least six months, not have a diagnosis of any other chronic physical illness, and not have any prior diagnosis of psychological disorder. Additionally, they must have completed the informed consent form prior to participation.

Exclusion Criteria: Physical diagnosis other than type 1 diabetes mellitus; No access to internet or new technologies; Not speaking Spanish.

The opportunity to participate in the study, along with relevant information will be communicated through various channels. In order to recruit participants, the study will be publicized throughout Spain through the dissemination channels of diabetes associations. To participate, parents or legal guardians of the adolescents must provide prior written informed consent.

Regarding the status and timeline of the study (we offer an estimate): A) participant recruitment will be completed by April of 2026; B) data collection will be completed by February of 2027 and C) final results are expected by April of 2027. None of these stages have already been completed.

Interventions

The emoTICare technological platform intervenes with participants through a serious game, which is defined as a video game primarily designed to teach or provide the player with certain skills that serve to achieve a purpose beyond the game itself [43,44]. Accordingly, using the theoretical model of chronic illness adjustment developed, a series of playable areas were created that address the main factors promoting adaptive adjustment to chronic illness, specifically Type 1 Diabetes Mellitus in the adolescent population.

Firstly, when selecting the narrative context of the game, the aim was to generate a certain degree of immersion, as this has been shown to increase the likelihood of changing beliefs and attitudes in the real world [52].

Development of the platform

Health-related quality of life (HrQoL) is a multidimensional construct that encompasses the physical, emotional, and social aspects of well-being. Unlike general quality of life, HrQoL primarily refers to the effect that illness and its subsequent treatment have on the various functional areas of an individual [53–55]. As previously explained, adolescence is a stage of life marked by critical changes that require personal skills to manage. Given the importance of coping with a chronic illness such as Type 1 Diabetes Mellitus (T1DM) during this period, and the need to promote socioemotional resources during this developmental transition to enhance adolescents’ quality of life, an initial review of the existing scientific literature on conceptual models of illness adjustment and health prevention was conducted. This review sought to identify common indicators of optimal positive development during adolescence, establishing a framework to promote adolescent quality of life.

As a result, a model was developed—the Disease Adjustment Model from an Integrative Perspective (DAMIP) [56]—which integrates a solid theoretical framework to facilitate the design, justification, and development of psychological interventions aimed at preventing psychological difficulties and enhancing adjustment to chronic illness during adolescence, thus improving quality of life and emotional well-being in this population.

The theoretical model was developed through the integration of different concepts and indicators present in pre-existing theoretical models, leveraging their complementarity. DAMIP draws upon Livneh’s Integrative Model [57,58], supplemented by Antonovsky’s Salutogenic Model [59], Leventhal’s Self-Regulation Model of Illness [60], and Hochbaum’s Health Belief Model [61], as well as elements from the proposal for promoting quality of life in adolescents from the National Academies of Sciences, Engineering, and Medicine (Health and Medicine Division; Division of Behavioral and Social Sciences and Education; Board on Children, Youth, and Families; Committee on Applying Lessons of Optimal Adolescent Health to Improve Behavioral Outcomes for Youth), led by Kahn and Graham [62], and the Kidscreen Project by Ravens-Sieberer et al. [63,64]. The DAMIP model provides the scientific community with a theoretical framework for optimizing socioemotional diagnosis and intervention in this population.

Based on the theoretical configuration of the DAMIP model, its core elements were incorporated into the development of the emoTICare serious game. The game consists of six missions aimed at fostering the most frequently cited concepts related to the structure underlying optimal development and well-being, which are crucial when implementing interventions to improve the biopsychosocial health of adolescents.

emoTICare

The emoTICare program engages users through a serious game aimed at developing socio-emotional skills and promoting physical and emotional health in adolescents with T1DM. This graphic adventure game consists of a time-travel journey across different parts of the world and various historical periods. Using the DAMIP model, a series of playable areas were developed to address the key factors that promote adaptive adjustment to chronic illness, specifically tailored to T1DM in the adolescent population.

According to the principles of serious games, players need to experience a certain level of immersion, as this increases the likelihood of changing beliefs and attitudes in real life [52]. To achieve this, a narrative context was established to enhance immersion, supported by a fictional storyline and well-defined goals that guide the player through the challenges necessary to complete the intervention.

After reviewing possible genres and settings and consulting the target population about their preferences, it was concluded that the storyline and setting should feature a graphic adventure incorporating time travel mechanics.

Initially, the player awakens disoriented in an unfamiliar place, which turns out to be the Temporal Control Center. Here, they meet an Artificial Intelligence called Alura, who oversees the center and will serve as their guide to help them understand the game mechanics.

Alura explains that a temporal rupture has occurred, causing the clock of time to stop, and its cores have been scattered in unknown places, altering the world as it was known. From this point on, the character embarks on a space-time adventure to recover all the cores and repair the clock (for more detailed information on emoTICare, access the S1 File).

Throughout the game, the player will travel to six distinct areas: Egypt (Psychoeducation Area), Venice (Emotional Awareness Area), the Amazon (Emotional Regulation Area), NewGlobal (Cognitive Coping Area), Petra (Identity Area), and India (Social and Communication Skills Area).

Each of these areas will focus on one of the indicators of quality of life in adolescents specified in our theoretical model, as follows:

Minigames

In addition to the areas that users will progressively access throughout the intervention, specific tasks have been developed for each area. These tasks aim to ensure users continue practicing the acquired techniques and reinforce their learning. The first five areas will feature a repeatable “minigame” that users can play as many times as they like while waiting for the next area to become available. The minigames are: Knowledge of Diabetes, Emotion Identification, Relaxation Techniques, Key Point Analysis, and Changing Irrational Beliefs about Identity.

Assessment tools

Measurements

Missing data will not be an issue, as the survey is conducted through LimeSurvey, where all questions are mandatory and cannot be skipped by participants.

To be included in the data analysis, participants must have completed all three measurement points and answered all questionnaires.

1. T1: First measurement: After the participant had signed the informed consent, the initial score for this outcome was assessed.
2. T2: Second measurement: POST Up to 6 weeks after the second measurement, the second measurement of the same variable was carried out.
3. T3: Third measurement: POST Up to 6 months after the second measurement, the third measurement of the same variable was carried out.

[Time Frame: Baseline up to 30 weeks]

Primary Outcome Measure:

1. Change Quality of Life (measured by KIDSCREEN-27).
2. Change Emotional and Behavioral Problems (measured by SDQ).
3. Change Emotional Competences (measured by EAQ-30).
4. Change Self-concept (measured by CAG)
5. Change Coping (measured by SPSI-R).
6. Change Social Skills (measured by CHASO).

Secondary Outcome Measure:

1. Change Threat of disease (measured by BIP-Q)
2. Change Psychoeducation Diabetes

Psychoeducation of the disease: the ECODI Diabetes knowledge scale will be used to assess this variable [79]. This scale consists of 25 items and four dimensions: basic knowledge and self-care, laboratory values, diet and physical exercise. It is a scale that is corrected by the number of hits and misses.

[Time Frame: Start of the game and during the game]

1. Change Resilience (measured by CD-RISC).

Other Pre-specified Outcome Measures:

1. Glycaemic control

HbA1c: It will be used to evaluate the glycemic control of the diabetic patient in endocrinology diseases.

Data analysis

In the present study, the per-protocol analysis (PP) approach was utilized to evaluate the effectiveness of the treatment and the management of missing data. This method includes only participants who fully adhered to the study protocol, ensuring that the results reflect the true efficacy of the intervention under ideal conditions. Participants will be included in the main analyses only if they have completed the treatment as per the protocol, that is to say, if they have adhered to all instructions without significant interruption. Participants who withdraw or fail to complete the treatment will be excluded from the analyses. This methodological decision is underpinned by the principle of minimizing potential bias from missing data, thereby ensuring that the results of the study accurately reflect the outcomes of participants who adhered to the treatment protocol. To address the issue of missing data during follow-up or treatment, predefined procedures will be implemented. The last observation carried forward (LOCF) imputation method will be employed to address missing data, with the last available measurement being utilised to complete the missing values. Furthermore, sensitivity analyses will be conducted to assess the impact of missing data on the study outcomes. These procedures are designed to minimise data loss, maintain the internal validity of the study, and ensure the robustness of the results.

Several statistical analyses will be performed to determine the effectiveness of emoTICare. As a first step, chi-square tests and t-tests were performed to examine potential differences between the control and experimental groups in key demographic and clinical variables (e.g., age, gender, disease duration). Although randomization is expected to ensure comparability, these tests were conducted to verify baseline equivalence. Propensity Score Matching (PSM) was not used to adjust for baseline differences, as the study followed a randomized controlled design.

Additionally, chi-square tests will compare the number of participants who withdraw from the experimental group with those who withdraw from the control group, ensuring no significant differences exist.

The primary focus of this study is on the evaluation of the programme’s overall effectiveness, as opposed to the presentation of definitive assertions concerning the outcomes of individual participants. It is imperative to emphasise that the tests are embedded in pre-specified models, with the primary focus being on assessing the overall effectiveness of the programme.

Before testing the change models, descriptive analyses, Pearson correlations, and multivariate and univariate analysis of variance (MANOVA and ANOVA) will be conducted using pre-intervention scores for participants in experimental and control groups to identify potential baseline differences. Furthermore, multivariate and univariate analysis of covariance (MANCOVA and ANCOVA) will be performed to detect changes in post-intervention scores (short-term effect), with pre-intervention scores as a covariate. These analyses will focus on latent variables derived from the composite scores of multiple Likert items, which represent underlying constructs such as emotional symptoms or health-related quality of life. Effect sizes (Cohen’s d) for each variable will also be calculated to estimate the magnitude of differences between the experimental and control groups. These analyses will be conducted using SPSS V.28.

To ensure the reliability of the results, we will begin with a series of diagnostic tests and validation procedures. This will include an analysis of residuals to assess the goodness of fit of the model, using residual plots and tests for normality (e.g., Q-Q plots and Shapiro-Wilk test). Homoscedasticity will be evaluated through the Breusch-Pagan or White’s test, and the Durbin-Watson statistic will be used to check for autocorrelation in the residuals. These tests will allow us to ensure that the assumptions of the parametric models are met. Additionally, we will conduct cross-validation (k-fold) to ensure that the model generalizes well to unseen data and is not overfitted. We will also perform outlier detection using Cook’s distance and leverage plots to examine the influence of individual data points on the model.

Multiple hierarchical regression analyses will be conducted to examine the added predictive value of the experimental condition in explaining the intervention program’s efficacy. Hierarchical regression is a stepwise method where predictors are added to the model in blocks. The rationale behind the hierarchical approach is to assess the incremental contribution of each new block of predictors while controlling for the variables entered in the previous blocks. This allows for a clearer understanding of the unique variance explained by the experimental condition, after accounting for initial factors.

Step 1: The first block typically includes variables measured at time 1 that are believed to influence the dependent variable (i.e., the change between T2-T1 and T3-T1 for each variable). These could include baseline measures of emotional symptoms, quality of life, or disease-related factors.

Step 2: In subsequent blocks, new independent variables (predictors) are added. For example, in the second block, we would add the experimental condition (treatment vs. control) to assess its impact on the change in the dependent variable, controlling for the baseline scores. The dependent variable will be the change between pre- and post-intervention scores, as well as between pre-intervention and follow-up scores. A statistically significant change in the coefficient of determination in Model 2 (predictor: experimental condition, controlling for pre-intervention score) will be interpreted as added predictive value. A significant prediction from the experimental condition, controlling for pre-intervention score, will allow attribution of significant change to the intervention.

The Reliable Change Index (RCI) will be calculated for the scores of the evaluated variables. Reliable change will be determined by the extent to which an individual’s progress exceeds the natural fluctuation of measurements [80]. The RCI will be calculated using pre- and post-emoTICare means, along with the pre-intervention standard deviation and Cronbach’s alpha index.

Following completion of the emoTICare program, statistically significant improvements (p ≤ .05) are expected in scores related to health-related quality of life, social and communication skills, emotional awareness, resilience, self-esteem, as well as coping and problem-solving skills. Conversely, a statistically significant reduction (p ≤ .05) is anticipated in variables associated with perceived disease threat, as well as emotional and behavioral problems.

An independent data safety monitoring board (DSMB) is to be convened at six-month intervals in order to review serious adverse event (SAE) reports. Should the DSMB determine that there is no association between the serious adverse events and the intervention under evaluation, the study may proceed. Conversely, if the events are found to be associated with the intervention, the DSMB is able to implement measures to ensure the safety of participants. In the event of the identification of serious adverse events, the DSMB may take the following actions: to halt the study, modify the protocol, adjust the treatment, enhance monitoring, or recommend early termination, with the overarching aim of ensuring participant safety.

Ethics and dissemination

This study adheres to the ethical guidelines of the 2013 World Medical Association Declaration of Helsinki and has received approval from the Human Research Ethics Committee of the Universitat de València (Reference 2023-PSILOG-3178945, UV-INV_ETICA-3178945). Appropriate measures have been taken to ensure the complete confidentiality of participants’ data, under the Personal Data Protection Law (LOPD) 3/2018 of December 5.

This study and its protocol have been approved and registered as a clinical trial in the ClinicalTrials.gov PRS (Protocol Registration and Results System) (Reference – ID: EmoTICare NCT06331429).

This protocol follows the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) 2013 guidelines.”

The results obtained will be disseminated to the main educational, social-health, and scientific stakeholders.