Appendix J
Appendix J — Genus Capability and Evidence Matrix
Make the claim no larger than the verified record
Use with: Chapters 8, 9, and 13–15.
This appendix separates capability and deployment from evidence of benefit. The matrix below is a verification tool, not a live product inventory. Current entries require the relevant product and partner owners to confirm them.
Genus supplies technology; partners retain their services, staff, programs, and professional responsibilities. Another organization can apply the same matrix to its own suppliers.
J1. Record one experience or claim
Field | What must be established |
|---|---|
Claim and intended audience | The exact proposed statement and where it will appear. |
Capability and source | What the tool does, the supporting record, its date, and what was actually inspected. |
Availability status | The documented label: generally available, configurable, pilot, proposed, or future concept. Record scope rather than forcing overlapping conditions into one label. |
Implementation and use | What is built, tested, and actually deployed; exact version, setting, and date where verified. |
Local conditions | Configuration, integration, permissions, equipment, training, and staffing needed. |
Human and partner responsibility | Who reviews, approves, responds, delivers care, and maintains the arrangement. |
Evaluation evidence | What was tested, with whom, against what comparison, on which outcome, and with which limitations. Link Appendix A. |
Not established | Missing current verification, unavailable functions, untested outcomes, and any claim requiring narrower wording. |
Approval and review | Product owner, relevant partner approval, permitted wording, next review date, and reason to withdraw the claim. |
“Implemented,” “tested,” and “deployed” answer different questions. “Configurable” does not establish that the particular organization has enabled and tested the required configuration. Nor does deployment establish better health or lower workload.
J2. A dated-source example—not a current capability verdict
The August 31, 2026 export says that family notes require human approval before publication. It also states that updates do not automatically return to a referring practice. Those descriptions can inform what a reviewer checks; they do not establish today’s configuration at a particular partner.
For that entry, record the documented function, then leave current local verification and outcome evidence unresolved until appropriate records are reviewed. Do not convert “supports a reviewed note” into “reduces caregiver strain” without relevant evaluation.
A financial or funding statement also needs its own current evidence and responsible review. Service counts, clinical outcomes, costs, and payment eligibility are not interchangeable facts.
J3. Dated implementation record: what the example can support
The narrative chapters use these examples to make design choices concrete. Their details are kept here so the principles do not depend on a frozen product version. This record does not establish effectiveness. The source is identified below by its title, schema, and export date; it is an author-held organizational record.
Dated item | What the August 31, 2026 record establishes—and what to verify locally |
|---|---|
Record status | Genus Day Centers Inspire, schema genus.dayCenterInspire.aiExport.v2, exported August 31, 2026. This is an organizational content/configuration record, not an independent product audit or current availability guarantee. |
Intake and profile | The record describes communication preferences, updated intake information, and history. Verify actual field access and updating. The recorded intake date is not a separate referral date; source-of-referral free text is not a closed referral record. |
Bookings and attendance | Reservations include approval, waitlisting or decline. A booked day is not an attended day. Closeout describes a day or Monday-to-Friday week, with participant detail inside it; arbitrary per-person date-range reporting was not described as available. |
Reviewed family notes | Four editable sections are described: To talk about tonight; Ideas for home; Caring for yourself; For your next visit. A person approves publication. Verify this gate and every disclosure route; the caregiver section does not authorize use of private assessment answers. |
Draft persistence | The export says some family-report drafts and day narratives remain session-only until publication. Test saving and recovery; do not imply that an unpublished draft has been persistently stored. |
Feedback and referring practices | Tile feedback is saved and notifies the center; comments depend on configuration. A human still reads and responds. Automatic return of information to the referring practice is not established. |
Education and training | Family guides and educational material can be delivered/exported; reading completion is not captured. Staff training records are per learner; a center-wide completion roll-up is described as planned. Neither delivery nor a course pass establishes competence for every care task. |
Practice and group entry | Paired huddles, branching practice scenarios, and group activity entry with individually correctable records are described. Private huddle commitments are not collected or compared. Check the actual design, burden, and professional requirements locally. |
Sensitive caregiver information | The record describes a four-item intake baseline and a caregiver-private lane in development. This does not establish equivalence to C4C instruments or tested separation from family-facing outputs. Use a suitably restricted process until required controls are verified. |
Program fit and clinical roles | A dementia diagnosis is described as an eligibility condition in this particular configuration. It is not a universal Day Center rule. A support-planning badge is non-diagnostic; clinical staging and treatment belong to qualified clinicians. |
Local measurement | Documentation time is measured locally, not automatically by the software. The incident record has owners and due dates but no completed-at timestamp for closure-duration analysis. Intake lacks demographic fields needed for some proposed comparisons. |
Content authoring and shared learning | Some common course/story content requires a code change and redeployment; limited local supplements are separate. Broader authoring, provenance and cross-center contribution are described as planned. The proposed improvement register is maintained outside the product. |
Respite and program allowances | The described system does not meter respite allowances. Any allowance is held by the responsible partner program; verify current program rules and local records rather than infer available hours from bookings. |
Continuing verification | Retain the version, owner, evidence, test date, permitted use, and unresolved limits. Test privacy at collection, retrieval, drafts, exports, notifications, print, and publication. A newer version must be checked rather than presumed equivalent. |
J4. Verify the caregiver pathway and the 4Ms claim
The August 31, 2026 Day Centers Inspire record described caregiver-only separation as still in development while also stating the intended privacy boundary. That is a dated requirement and capability gap, not proof that today’s configuration enforces it. Record and test actual local permissions across intake, AI retrieval, meeting drafts, printing, exports, and family summaries before collecting sensitive answers in that pathway.
A four-item local intake is not established as the full C4C assessment battery. An AI instruction is not a separate clinical record. A reviewed family note is not a private caregiver consultation. Leave each implementation claim unresolved until the relevant records and tests support it.
Use 4Ms-informed for this book’s proposed Day Center mapping. Document which assessments and actions the service can actually deliver, which qualified professionals are responsible, and which require partners. Cite the Hartford Foundation and IHI; do not claim an Age-Friendly designation, affiliation, or endorsement unless the particular site has the required evidence. Public guidance and recognition information are in R1, CG4–CG7 and CG14.
Decision this tool supports: What can we truthfully promise in this setting, and what still needs to be established?
Use the Tools, Then Return to the People
You do not need every appendix for every task. A first conversation may need only selected questions from B. A handoff may need D and G. A team preparing an AI-assisted resource may use A, C, H, and I, with a small measurement plan in F.
Start with one meaningful need. Adapt the relevant tool with the people who will use it. Provide the time, resources, and authority behind the next action. Then ask what happened.
Keep what helps. Revise what creates confusion. Remove duplication. Return to contributors with what changed, and preserve the learning another team can responsibly use.
The final question is not whether the worksheet is complete. It is whether someone is better able to understand a choice, receive support, share a responsibility, or enjoy an ordinary part of life.
The tools belong behind that experience.
More Room to Be a Family.