Evidence basis

• Narrative persuasion. Stories with a relatable character lower counter-arguing and raise self-efficacy (entertainment-education). The reader makes Rosa's choices, so the story answers their objection first.
• Barrier-first design. The branches map to the most-reported reasons people skip colorectal screening: "I feel fine," cost, embarrassment, and prep/time.
• Honest trade-offs. Each test shows its real downsides (yearly repeat; follow-up colonoscopy after a non-normal FIT; prep and a ride). This follows decision-aid principles (IPDAS) instead of pushing one option.
• Implementation intentions. "I will ___ on ___ at ___" plans (Gollwitzer) close the gap between intending and doing. The calendar file turns the plan into a reminder.
• Numbers as natural frequencies with an icon array (1 in 24 / 1 in 26), not percentages.

Behavior Change Techniques (BCT Taxonomy v1)

BCT	Where
5.1 Information about health consequences	Myth card; lifetime risk icon array
1.2 Problem solving	Worry branches (cost, embarrassment, time)
9.1 Credible source	Dr. Lee; plan member services
1.4 Action planning	"I will ___ on ___ at ___" planner
7.1 Prompts/cues	Calendar reminder (.ics)
12.5 Adding objects to the environment	Plan card to print and post

COM-B target

Reflective motivation (myths, risk), psychological capability (what each test involves), and physical opportunity (cost, time, a ride).

How HealthLiteracyCopilot would generate, personalize and assess it

• Generate: a branch graph from the approved screening source content. Each node is checked against the clinical source so no branch drifts from it.
• Personalize: the author picks the lead character, setting, language and which barrier appears first for the audience. Branch order is reviewed with simulated readers (simulated participants, not patients), then with real people in consented research.
• Assess: CDC Clear Communication Index per node, PEMAT actionability, an IPDAS-style balance check (every option shows benefits and harms), and WCAG 2.2 checks on the interaction.

What a pilot would measure

Proximal: completion rate, which barrier members pick (a research signal for the plan), plans made, and calendar downloads. Distal: FIT kit returned or colonoscopy completed within 90 days (the colorectal cancer screening quality measure). Measures a pilot would track, not results or promised effects.