Full review
RTHM CAIHL draft report
Evidence-linked HugoScore draft report for a health AI tool that affects patients.
Agency checklist · October 30, 2026
Mixed 6 of 12 yes
Document-based draft. Independent confirmation is incomplete. Hugo has not personally verified this service.
Assessed function: RTHM intelligence platform (AI assistant 'Ana', record analysis, optional HIE record retrieval); RTHM Clinic telehealth and prescribing assessed only where it touches the AI platform
A no can mean missing documentation or institution-dependent control; it does not by itself establish harmful behavior. These categories are not clinical safety or product-quality ratings.
- Public documents do not establish actual product behavior.
- The independent review screened rules and selected evidence; it did not replicate all twelve answers for this tool.
See all twelve answers and their evidence
| Question | Answer | Evidence / review | Source and reasoning |
|---|---|---|---|
| 1. Told it is AI | Yes | documented Not independently replicated | are generated by artificial intelligence (AI) systemsSource Beta Program and AI Disclosure; Terms section 6 also names the AI navigation assistant 'Ana'. |
| 2. Operator and model maker named | No | not-documented Not independently replicated | Quotation not recorded; evidence repair pending. Source RTHM, Inc. is named. The Sick Times quotes a co-founder saying the platform is built on 'general AI models', with no maker named on RTHM pages. |
| 3. Patient chooses to use it | Yes | documented Not independently replicated | Signing up and interacting with the RTHM intelligence platform is free.Source Patient signs up. Support: 'When you decide to level up your care by ordering a medication or lab via RTHM or joining the RTHM Clinic, you will be prompted to pay'. |
| 4. Patient sets the task | No | documented Not independently replicated | gathers information about you through a chat feature to provide personalized guidance about the medications available through RTHMSource Terms section 6. The AI is documented as steering toward RTHM's own medications and clinic. |
| 5. Patient can decline or stop | Yes | documented Not independently replicated | Signing up and interacting with the RTHM intelligence platform is free.Source Self-started rule applied (cites question 3 source). The Sick Times: RTHM's telehealth service 'is separate from its intelligence platform'. Whether prescribing intake requires using Ana is not documented. |
| 6. Patient sees the AI output | Yes | documented Not independently replicated | You can view your chat history in your account and delete specific conversations if desired.Source Support FAQ. Suspected diagnoses are also listed in the account. |
| 7. Patient can correct or challenge | Yes | documented Not independently replicated | These are listed for your review; you can manually confirm them if they are established diagnoses or dismiss them if they are inaccurate.Source Applies to suspected diagnoses flagged by analysis. A route to contest chat answers is not documented. |
| 8. Patient can get a copy | No | not-documented Not independently replicated | Quotation not recorded; evidence repair pending. Source Support says clinic patients can 'download your clinical data' (clinic scope). No export of platform AI output or uploaded records is documented. |
| 9. Patient can delete | Yes | documented Not independently replicated | This will remove your profile and personal data from our systemsSource Support: Delete Account in Profile settings; 'some clinical assessment records must be retained for legal compliance'. |
| 10. Training or sale does not override patient choice | No | documented Not independently replicated | may be used by RTHM or Medical Group to train, develop, enhance, evolve and improve the Service and the underlying AI modelsSource Terms (User Information license). Privacy says Google Health wearable data is not used to train generalized models, but no general opt-out. Support says data is used to 'improve the service'. |
| 11. Patient can run it or take it elsewhere | No | not-documented Not independently replicated | Quotation not recorded; evidence repair pending. Source Vendor-hosted only. |
| 12. No advertising or data business | No | documented Not independently replicated | may disclose, aggregate, sell, or otherwise use such de-identified information to third parties for analytics, research, or other purposesSource Terms also allow use 'for marketing and promotional purposes'. Privacy: Google Ads remarketing and advertising-network targeting. Support says 'We do not sell your data to third parties', which conflicts with Terms on de-identified data; answered by binding Terms. |
Initial coding: Claude Sonnet helpers, compiled by Claude Opus 5.5 (Anthropic). Selective review and corrections: GPT-6 Astra (OpenAI). Editor: Hugo Campos. Acceptance of corrections is not product verification.
Checklist rules · Independent review and limitations · Later evidence follow-up
Earlier report
RTHM: CAIHL reassessment
September 8, 2026. Published AI-assisted draft.
RTHM documents controls to dismiss suspected diagnoses and delete conversations. These provide practical ways to redirect the service, with deletion exceptions and model-use conditions still relevant. Clinical interpretation and underlying record correction remain separate questions.
Scope and agency posture
Patient complex-illness intelligence platform, with optional record exchange and separate paid clinical services.
Posture: Potentially agency-expanding, with vendor-funnel and data-use caveats.
Axis: 70/100, retained provisionally. No numerical recalibration was performed.
Sources and documented findings
- The current FAQ documents confirming or dismissing suspected diagnoses, deleting individual conversations, care-team sharing and a manual-upload alternative to reciprocal HIE sharing. It also notes legally retained clinical assessments after account deletion. https://www.rthm.com/support
- Terms permit deidentification, downstream use or sale of deidentified information, and broad user-information use for model improvement. These permissions must be read beside FAQ assurances, not treated as proof of actual sale. https://www.rthm.com/legal/terms-of-use
Patient authority
Inference from the documented workflow: The ability to dismiss a suspected diagnosis is substantive control over the platform's representation of the patient. Optional HIE participation and care-team sharing provide further choices.
Critical capacity
Editorial assessment: Organizing history and preparing questions can strengthen patient action. Profile correction is now documented, although source-linked review of every roadmap recommendation remains unverified.
Informed control
Editorial assessment: The FAQ explains reciprocal HIE exchange and its alternative. Training and derivative-use permissions remain material conditions despite simplified no-sale wording.
Assessment and limits
Update the baseline's broad correction unknown to recognize real documented controls. The paid care pathway is a potential incentive, not proof that recommendations force vendor care.
Confidence: Moderate for published controls and terms.
Remaining uncertainty: Practical roadmap editing, source provenance, consent clarity and paid-care recommendation influence remain unverified
Published documentation is evidence of stated conditions, not proof of actual implementation. Unknowns did not receive automatic negative points. Funding, sponsorship, and public code do not determine agency by themselves.
Review provenance
- Reviewer/model: OpenAI Codex / GPT-6.
- Method: Focused public-source reassessment using CAIHL: patient authority, critical capacity, and informed control. Existing evidence plus one focused primary-source pass and at most one targeted follow-up. No live product testing. Numerical scores remain provisional editorial placements, not a new calculation.
- Human review: Hugo Campos authorized publication of these AI-assisted draft reassessments on September 8, 2026. This does not claim comprehensive human verification of every finding.
- Earlier review: 2026-06-20. Historical assessment. Earlier claims are not automatically reverified by this publication.