About this paper symposium

---

Panel information
Panel 6. Developmental Psychopathology

Paper #1
Developing Sidekick, a Digital Just-in-Time Adaptive Intervention (JITAI) for Adolescent Depression Treatment in Collaborative Care
Author information	Role
Jessica Jenness, University of Washington, United States	Presenting author
Edward Preble, RTI International, United States	Non-presenting author
Meghan Hegarty-Craver, RTI International, United States	Non-presenting author
Ty Ridenour, RTI International, United States	Non-presenting author
Jonathan Holt, RTI International, United States	Non-presenting author
Michael Wenger, RTI International, United States	Non-presenting author
Sean Munson, University of Washington, United States	Non-presenting author
Kirsty Weitzel, RTI International, United States	Non-presenting author
Christina James, University of North Carolina - Chapel Hill, United States	Non-presenting author
Sofia Olivares, University of North Carolina - Chapel Hill, United States	Non-presenting author
Warren Szewczyk, University of Washington, United States	Non-presenting author
Kate Benjamin, University of Washington, United States	Non-presenting author
Tricia Aung, University of Washington, United States	Non-presenting author
Adam Bryant Miller , University of North Carolina - Chapel Hill, United States	Non-presenting author
Abstract
Background: Over 3 million adolescents are diagnosed with major depressive disorder (MDD) each year, leading to significant risks for lifelong medical, psychosocial disability, and increased suicide risk (Mojtabai et al., 2016; Clayborne et al., 2019). Evidence-based practices (EBPs) like standard cognitive behavioral therapy (CBT) have limited effects in treating MDD (Weisz et al., 2006), leave sleep and fatigue disturbances unresolved (Kennard et al., 2006), and do not provide adequate support for skill generalization outside of therapy sessions (Bru et al., 2013). Just-in-Time Adaptive Interventions (JITAIs) are a promising alternative to enhance standard CBT practices by processing real-time self-report and wearable device data, enabling timely and appropriate support when needed most (Klasnja et al., 2015; Cerrada et al., 2017; Haug et al., 2020). JITAIs may improve treatment effectiveness by identifying vulnerable states through passively sensed wearables and self-reported measures, focusing on often unresolved issues like sleep and physical activity. They can then provide immediate interventions that enhance both proximal outcomes (e.g., sleep, activity) and distal outcomes (e.g., mood, functioning). Despite their promise, a JITAI approach targeting adolescent MDD has not been empirically studied. Method: Our ongoing mixed-methods work includes development of an algorithm for an adaptive JITAI intervention, Sidekick (Figure 1), that monitors sleep and physical activity treatment targets measured via wearable smart watches and daily self-report, selects tailored interventions, and delivers these interventions through a smartphone app. Once developed, we will test the adaptive Sidekick algorithm in a small, human-centered usability trial with adolescents aged 13-17 who have undergone MDD treatment (n = 10) and mental health clinicians (n = 5 pilot users) within a collaborative care program. The pilot usability trial will test whether the Sidekick algorithm (1) monitors treatment targets such as sleep duration, bed/wake times, step count, and activity intensity via passively (wearables) and actively (self-reported ecological momentary assessment) collected data and (2) applies decision rules to select appropriate interventions to match treatment targets, such as individualized activity suggestions on sedentary days. Pilot users will also provide feedback on intervention messages drawn from existing digitally based EPBs for improving sleep (Dolsen et al., 2021) and physical activity (Phillips et al., 2018). Using a human-centered design framework, we will assess Sidekick’s usability, feasibility, acceptability, appropriateness, and engagement through interviews and surveys after four weeks of use. Results: User testing data collection will conclude by February 2025. We hypothesize that the Sidekick algorithm will successfully detect vulnerable proximal states and deliver tailored, just-in-time interventions targeting sleep and physical activity. We also predict users will rate Sidekick as feasible, acceptable, appropriate, usable, and engaging. Conclusion: Once our user-testing pilot study is complete, we will conduct a standard micro-randomized trial (MRT; Klasnja et al., 2015) that randomizes, within-person, whether an intervention is delivered to determine if Sidekick interventions are effective. During this phase, Sidekick will partner with an existing, manualized brief CBT intervention adapted for collaborative care (Blossom et al., 2024). Ultimately, we aim to create a scalable, digital intervention tool that can be applied in multiple settings to improve adolescent depression care.

Paper #2
Adapting a Dyadic Application for Suicidal Adolescents and their Caregivers seen in Primary Care
Author information	Role
Sarah Danzo, Ph.D., University of Washington, United States	Presenting author
Adia Abler, University of Washington, United States	Non-presenting author
Laura Richardson, University of Washington, United States	Non-presenting author
Abstract
Background: Youth with suicidal thoughts and behaviors (STB) often present to primary care (PC; Kemper et al., 2021; Luoma et al., 2002) and nearly half of youth who die by suicide contact a PC clinician within one month prior to suicide (Luoma et al., 2002). Unfortunately, management of STB in PC remains a challenge due to barriers such as lack of time and space to address concerns in the clinic (Byrne et al., 2021; Cutler et al., 2019; Torio et al., 2015). Thus, digital interventions that can be integrated within PC settings and used by patients and families at home may be an important avenue to assisting PC clinicians in managing suicide risk within this setting. We aimed to adapt a digital application, iKinnect, for use with youth expressing STB in PC. The tool was initially designed to improve parent and adolescent communication and parental self-efficacy to prevent risk behaviors among youth involved in the juvenile justice system. However, we hypothesized that app components were relevant and could be adapted for suicide prevention in PC. Methods: We engaged 2 members of the app development team, 2 PC staff, 2 young adults, and 2 caregivers of adolescents with lived experience in 3 focus groups to identify initial adaptations and design requirements for the intervention. Following this, our research team partnered with the app development team to develop a preliminary intervention prototype for iterative user testing and refinement using cognitive walkthroughs and think-aloud-approaches. Acceptability, feasibility, appropriateness, and usability were assessed using the Acceptability of Intervention Measure, Feasibility of Intervention Measure, Intervention Appropriateness Measure (Weiner et al., 2017), and the Intervention Usability Scale (Lyon et al., 2021). Results: The co-design group highlighted key needs for the intervention content and digital platform including: finding a balance of what caregiver input and oversight should be embedded into app functioning and what aspects should remain confidential to the teen user, a need for balance of giving providers insight into patient risk with ability for providers to respond in real time, need to integrate the intervention platform into provider electronic health systems for easy access, need for updating language, building in additional coping tools activities, and needs for unique alert systems for caregivers and PC staff. At the end of the co-design phase, the program’s average acceptability rating was 16.6/20 (83% acceptability), average appropriateness was 15.6/20 (78% appropriateness), and average feasibility was rated as 16.2/20 (81% feasibility). Initial intervention adaptations and preliminary results from user evaluation with adolescents, young adults, caregivers, and PC professionals will be discussed. Discussion: Findings from this study highlight important barriers, facilitators, and needs to adapting a digital just-in-time intervention for use within PC settings, while taking into account the diverse needs of youth, caregivers, and PC stakeholders. Ultimately, adaptation of a digital intervention that can be feasibly and effectively delivered within PC settings and engages both youth, caregivers, and PC staff has the potential to increase access and engagement with lifesaving suicide care across widely accessed settings.

Paper #3
Leveraging AI and Just-In-Time Adaptive Interventions to Improve Family Mental Health Using the Colliga App
Author information	Role
Jacqueline Duong, University of Texas at Austin, United States	Presenting author
Abstract
Introduction: Early childhood mental health problems are a significant public health concern with long-term impacts on children's development and well-being. Maladaptive parent-child interactions, such as conflict and coercion, are associated with increased risks of both externalizing and internalizing symptoms while nurturing relationships can buffer against these risks. Traditional interventions, such as in-person therapy and parent training, face barriers related to accessibility and flexibility, especially for underrepresented populations. In response, our AI-assisted parenting intervention, using the Colliga app, incorporates Just-In-Time Adaptive Interventions (JITAIs) to provide real-time, personalized support based on dynamic psychological and contextual factors in everyday life. Methods: This study involves a micro-randomized clinical trial, with 170 families randomly assigned to either the intervention group (n = 85), receiving personalized JITAI feedback, or the control group (n = 85), receiving standard parenting and child development resources without real-time adaptive feedback. The sample includes families from diverse backgrounds, with eligibility criteria such as having a child aged 6-9, proficiency in English or Spanish, belonging to an underrepresented racial-ethnic group, or having a child with elevated mental health symptoms (defined as scoring at or above the 70th percentile on at least one difficulty subscale of the Strengths and Difficulties Questionnaire (SDQ)) or low household income (≤ 33rd percentile of their residing county, adjusted for family size). Participants in the intervention group will engage in daily educational modules and activities designed to promote positive family interactions and improve emotional regulation. Activities range from mindfulness exercises to family play, and feedback is delivered through the Colliga app, tailored in real time based on data collected from mobile devices. Micro-randomization will be used to test and optimize the timing and type of feedback delivered at each decision point. Both groups will be monitored for two months, with biweekly check-ins to assess changes in family dynamics and child behavior. Expected Results: We hypothesize that families in the intervention group who receive JITAI-based, real-time feedback will show greater improvements in family functioning and reductions in child externalizing symptoms compared to the control group. Specifically, we expect tailored interventions targeting positive parent-child interactions and emotional regulation will lead to stronger attachment bonds and fewer maladaptive interaction patterns. We also anticipate that micro-randomization will help determine the most effective types and feedback timing, further enhancing the intervention’s impact. Discussion: This study has the potential to transform the way mental health support is delivered to families, particularly those from underrepresented backgrounds. By utilizing AI and JITAIs, the Colliga app offers personalized, scalable, and accessible interventions that adapt to families' real-time needs. The findings will contribute to the growing research on digital mental health tools, illustrating how real-time, data-driven interventions can improve child and family outcomes. If successful, this approach could offer a cost-effective, flexible model for delivering mental health support, addressing critical gaps in traditional service delivery methods.

⇦ Back to session