Where should I go for care?

• Decision support at the moment of need. It turns a vague question (“Where do I go?”) into one clear recommendation plus the reason, instead of a list of all options with equal weight.
• Safety first, always visible. The 911 rule stays at the top of every screen, and red-flag symptoms override every other choice.
• Built for older adults. Large text with a size control, big tap targets, high contrast, plain words, optional read-aloud, and a printable card — following plain-language and senior-friendly web guidance (for example, NIH/NIA guidance on making websites older-adult friendly).
• Cues to action (Health Belief Model). Saving the numbers to the phone and fridge puts the right choice where the decision happens.

Behavior Change Techniques (BCT Taxonomy v1): 4.1 Instruction on how to perform the behavior7.1 Prompts/cues12.5 Adding objects to the environment9.2 Pros and cons5.1 Information about health consequences

COM-B target: Psychological capability (knowing the options), Physical opportunity (numbers at hand, 24/7 options), Automatic motivation (habit of calling the nurse line first).

Persona tuned for: “Dorothy, 74” — Medicare Advantage member, low vision, lives alone, has used the ER for non-emergencies three times this year.

How HealthLiteracyCopilot would make it: HealthLiteracyCopilot would build the navigator from the organization’s published benefit grid (copays come from that grid, never from AI). It would be checked with an adapted PEMAT actionability rubric and the CDC Clear Communication Index, and a clinician signs off on every red-flag rule. It could sit in a welcome series, and Medicare marketing review needs are flagged before release.

What a pilot would measure: nurse-line and video-visit use; avoidable ER visits per 1,000 members; member experience (CAHPS “getting care quickly”). Measures a pilot would track, not results or promised effects.