Project Scope¶
Trove provides study management for digital health research: study definitions, participant enrollment and consent, and collection of health data. The aim is to let studies share these services while retaining their own participant interfaces and specialized procedures.
Trove is in beta and is being actively tested and used at Stanford. The platform tour shows the current interfaces. The design pages describe both implemented contracts and planned extensions; GitHub milestones and issues track delivery.
Study Definitions¶
A study definition describes participant-facing content, activities, collection requirements and execution rules. Collection requirements identify the data needed and permitted sources, rather than the internals of one operating system. Applications report the capabilities they support.
Published revisions preserve the content a participant accepted. The server owns enrollment, consent and accepted results. Clients can cache state and queue work, but they cannot independently establish consent or task completion.
Grove supplies reusable application components. The Grove FHIR implementation guides define exchange profiles. Trove adds study management and storage around them.
Storage and Access¶
One measurement can be relevant to several studies. Trove stores its source event once and records each study's authorization separately. Withdrawal from one study does not change another study's evidence or grant it additional access.
FHIR is the exchange format, not a requirement to represent every database row as a FHIR resource. PostgreSQL and TimescaleDB provide transactional and time-series storage; object storage holds larger artifacts.
Researcher Interfaces¶
The dashboard, Swift CLI and planned authoring assistants use the same typed API and permission checks. Visual editing is useful for reviewing consent wording, questionnaire structure and branching. Command-line and assisted workflows reduce repetitive configuration without introducing a separate study model.
Assistance produces reviewable drafts. Publication, participant contact and access to study data require their own permissions. Scientific validity, institutional approval and intervention suitability remain research responsibilities.
Scope Limits¶
Trove is not a general clinical-trial management suite, statistical analysis environment or arbitrary application-hosting service. Research analysis uses authorized exports in separate tools.
Specialized assessments, such as a six-minute walk test or guided ECG capture, remain in study applications. They can use Trove's configuration and result contracts without moving their acquisition UI or scoring into the core platform. A component can move into Grove when multiple applications need the same behavior.
Design Documents¶
- Study Model and Research Workflows: definitions, publication, consent, tasks and research operations.
- Data and Platform Architecture: components, storage, authorization and synchronization.
- Applications and Integrations: mobile clients, wearables, conversation tasks and custom services.
- Engineering and Verification: implementation, testing and deployment requirements.