Daily Symptom Dynamics and App Usage Patterns in a Digital Aftercare Intervention for Depression: Insights from the iCAN Project's Pilot Study SMARDY

Context Depression is a leading cause of disability worldwide, with an increasing prevalence among the youth. A major gap in the treatment of severe courses of illness is the timely follow-up treatment after discharge from the hospital due to barriers such as a lack of outpatient treatment capacities. Digital mental health interventions, such as “Intelligent, chatbot-assisted outpatient aftercare of depression in adolescents and young adults” (iCAN), aim to address this gap. The iCAN app integrates depression-specific skills training, a chatbot for motivation, and navigation to outpatient follow-up care. A previous study (Culmsee et al., 2024) demonstrated the potential to improve symptoms over a 14-day period. However, little is known about the daily symptom dynamics and their relationship with smartphone usage during intervention. Methods This analysis focuses on longitudinal data collected during the iCAN pilot study called “Piloting a Self-help Mobile Algorithm-based intervention for Depression in Youth” (SMARDY). The data was collected between October 2022 and April 2023. 25 participants aged 13–25 years (PHQ-9 ≥ 8) used the prototype of the iCAN app for 14 days. Depressive symptoms were assessed using ecological momentary assessments (EMAs) and standardized pre- and post-intervention measures (e.g., PHQ-9). Automatically recorded smartphone usage metrics, such as screen time and apps used, are analyzed to explore depressive symptom changes in the context of smartphone usage patterns. Intervention The participants accessed the iCAN app three times a week to complete tasks in skill-training modules on behavioral activation, self-esteem, interpersonal relationships, dealing with negative thoughts, day structure, and sleep and additionally completed daily EMAs. Results The Results showed an overall decrease in PHQ-9 scores from baseline to post-intervention, with 80% of participants improving. However, considerable individual variability in symptom trajectories was observed. App usage patterns differed strongly between users, with no clear "one-size-fits-all" effect. Higher creativity app use was positively associated with symptom severity, while more intervention app use was associated with slightly higher symptoms. Conclusions These findings suggest that digital behavior may reflect mental state and inform tailored interventions. Limitations of the study include the lack of a control condition, small sample size, and the inclusion of only Android users. Future research should investigate how apps are used exactly and their impact on treatment outcomes.