Longitudinal outcome measurement
Repeat administration of GAD-7, PHQ-9, DASS-21, WHO-5, the Copenhagen Burnout Inventory and a PSQI-based sleep screener, tracked per person and aggregated per cohort.
Research & clinical governance
If we can't measure it, we don't claim it
Digital mental health has a credibility problem: plenty of apps, very little evidence. We built Saantvana around validated instruments administered repeatedly, so improvement is something we observe rather than something we assert.
This page explains what we measure, how we govern the data behind it, and the boundaries we keep between research and the private material people trust us with.
- Validated instruments only
- De-identified datasets
- Cohort minimums enforced
How the evidence layer works
Effectiveness analysis
We compare movement across care pathways — companion use, journeys, peer circles, expert sessions — to see which combinations actually shift scores.
Quality improvement loop
Findings feed a standing governance review that changes journey content, companion guardrails, matching logic and panel composition.
Safety surveillance
Crisis events, escalations and moderation actions are reviewed as a safety signal, not just an operational metric.
What research data can and cannot contain
- Direct identifiers are removed before any dataset leaves the clinical layer.
- Cohort minimums are enforced, so no cell can be small enough to re-identify a person.
- Journal entries, companion conversations and session notes are excluded entirely.
- Exports are logged to an audit trail with the requesting purpose recorded.
- Employers and organisations never receive record-level data under any arrangement.
Instruments behind our reporting
Every effectiveness claim we make traces back to one of these.
- GAD-7 — generalised anxiety severity
- PHQ-9 — depression severity, including risk screening
- DASS-21 — depression, anxiety and stress subscales
- WHO-5 — positive wellbeing index
- Copenhagen Burnout Inventory — personal, work-related and client-related burnout
- PSQI-based screener — sleep quality and daytime dysfunction
Research — common questions
- Does Saantvana use my data for research?
- Only in de-identified, aggregated form, and only where our governance rules permit it. Research datasets strip direct identifiers, enforce cohort minimums, and never include journal text, companion conversations or session notes.
- How do you measure whether the platform works?
- Through repeat administration of validated instruments — GAD-7, PHQ-9, DASS-21, WHO-5, CBI and our sleep screener — so we can report change over time rather than satisfaction scores.
- What is clinical governance here?
- A standing process that reviews outcome data, crisis events, moderation actions and expert verification, and feeds findings back into the product, the companion's guardrails and the expert panel.
- Are your experts' credentials independently checked?
- Yes. Identity and qualifications are verified for every expert, and regulated professions have their registration confirmed with the issuing authority before listing.
- Can researchers or institutions collaborate with you?
- Yes. We work with academic and institutional partners on multilingual mental health delivery in India. Contact us with your proposal and ethics context.
- Do you publish your outcomes?
- We report aggregate effectiveness internally today and are building toward external publication. We do not claim outcomes we have not measured.
Related on Saantvana
Working on multilingual mental health in India?
We collaborate with academic and institutional partners on delivery, measurement and outcomes research.