"Clicking without understanding": A mixed-methods analysis of user agreements in digital mental health services

This study evaluates privacy policies and terms of service agreements from digital mental health platforms, focusing on accessibility, comprehensibility, and alignment with informed consent principles in healthcare informatics.

We applied mixed methods combining content analysis and computational linguistic assessment to 139 user agreements from international mental health applications and Singaporean providers, including commercial platforms and social service agencies serving vulnerable populations. We evaluated readability, communicative practices, regulatory compliance, and power asymmetries. Only 1.67% of services implemented comprehension verification for informed consent. User agreements required approximately 16 years of education for comprehension and exhibited significant linguistic power asymmetries favoring providers. Privacy policies comprehensively addressed data collection but systematically neglected post-service communication regarding data retention and deletion. Among local services, only 8.33% adequately communicated data breach notification procedures as required by Singapore's Personal Data Protection Act. Terms of service failed to establish bidirectional communicative exchange necessary for meaningful healthcare informed consent. Findings reveal fundamental misalignment between digital mental health agreements and collaborative communication principles essential to therapeutic relationships and healthcare informatics best practices. Communication barriers pose particular risks for individuals with serious mental illness requiring accessible health information for decision-making. Results have implications for health informatics policy, consumer health technology design, and digital health regulatory frameworks. Digital mental health platforms demonstrate significant user communication deficiencies. Our findings point to the need for user agreements that are written in plain language, that incorporate essential informed consent components, that balance linguistic power between providers and users, and that accommodate the cognitive needs of vulnerable populations seeking mental health support.