Understand Your Medicine: proposed user-testing guide

Status: protocol only. No participants have been recruited and no results are claimed.

Purpose and sample

Test whether people can explain the mechanism, find supporting evidence and recognize uncertainty. Start with a small exploratory group of patients/caregivers and pharmacists/clinicians, then scientists/analysts. This is a proposed sample, not a representative study or clinical validation.

Introduction to read aloud

“We are testing the explanation, not you. Use an imaginary ibuprofen example. Please do not share health details or other personal information. You may skip any task or stop. We will record task observations only with your consent.”

Tasks (about 20 minutes)

1. From the company home, open the short Patient story. Explain in your own words how ibuprofen may help.

2. Explain whether an intact tablet circulates in blood and whether relief means the cause is resolved.

3. Explain what 30–60 minutes describes. Does it tell you when to take another dose?

4. Find the safety evidence. Explain the difference between severe symptom intensity and a medically serious event. Name one limitation of the percentages.

5. Find a source supporting one explanation, then identify something the prototype cannot establish.

6. Change perspective. Explain what changed and whether the extra detail helps.

7. Share a selected section, reopen the link and verify the same perspective and sentence without autoplay.

8. Repeat a short task with silent captions, keyboard navigation and a narrow screen. Try lightweight mode.

9. Optionally answer the comprehension questions and export local feedback. Inspect the file before deciding whether to share it with the study organizer.

Non-leading prompts

“What do you think this means?” “Where would you look next?” “What is still uncertain?” Avoid explaining the answer until the participant finishes the task.

Record without personal information

Use a session code, task completion, source-finding time, observed navigation errors, and verbatim comments only with consent. Score each comprehension item 0 (incorrect), 1 (partial), or 2 (accurate): dissolved drug rather than intact tablet; mechanism linked to signals; onset distinct from dosing interval; symptom relief distinct from cure; trial-specific risk distinct from personal prediction. Record uncertainty recognition separately. Two reviewers should compare scoring disagreements.

Comparison and decision criteria

For a later comparative study, predefine the outcome rubric and minimum meaningful difference. Counterbalance order between this prototype and the same source information in a conventional text presentation. Keep content and allotted time comparable. Assess comprehension, source findability, uncertainty recognition and burden; do not infer clinical outcomes. Exploratory findings guide revision and do not establish superiority.

Feedback and review checklist

- Local export creates a JSON file; nothing is submitted by the site.

- Do not collect identifying or health information in the free-text field.

- Review unclear or misleading explanations before expanding drug coverage.

- Obtain independent clinical/scientific review and reviewed translations separately.

- Test assistive technology with real users; current keyboard and DOM checks are not accessibility certification.