[florenceheinkelein-ops]

Why

In the patient sessions, people found the opening questionnaire too restrictive. Single-choice answers couldn't capture overlapping or changing symptoms. Questions assumed pain was the main symptom. Timeframes were unclear, MRI terms were hard to recognise, the body map needed guidance, patients with both back and leg symptoms were confused by similar questions in the two modules, and there was nowhere to describe a past spinal surgery.

The study is already running, so no question changes its meaning or answer codes. Only wording, multi-select, instructions and one free-text follow-up change.

What changes (questionnaire 1.6.0 → 1.7.0)

• Multi-select: 1.1 (symptoms), 6.6 (worse sitting/standing/walking), 6.10 (worse side) and 7.2 (numbness) now accept every answer that applies. "About the same" and "No" are exclusive. Module routing already checks for any matching answer, so no logic changes.
• Pain or symptoms:
  • The 1.1 option, the whole back module (6-intro, 6.1–6.10 and the 6.7 options) and the 7.11 walking options now say "pain or symptoms". The 7.11 options depart from the original Oswestry wording ("Pain prevents me…").
  • The leg module asks about "leg symptoms" (7.4, 7.7, 7.11).
  • Unchanged on purpose: the red-flag questions (1.3a–c, 1.4a, 1.4c, 1.5b, 1.6b) and the advice pages, which ask about a specific pain pattern, and the pain body map (7.1), which has its own numbness map.
• Timeframe: both modules open with an instruction to answer for a typical day over the past month, and to select every answer that applies when symptoms change or switch sides.
• Overlap: patients who get both modules see a note at the start of the leg module asking them to answer only about their legs.
• Body map: an instruction before 7.1 explains how to open, draw, zoom, undo and finish, using the editor's own controls.
• Diagnoses: options in 1.7 and 7.12 carry a plain-language explanation, e.g. "Lumbar stenosis (narrowing of the spinal canal in the low back)".
• Spinal surgery: 1.9 is now "Have you ever had spinal surgery?" (Yes / No / Not sure, same codes). A new free-text question, 1.9a, asks which surgery and when, and invites a best guess. It only appears after Yes or Not sure. The old open-choice item only offered a hidden "Other" field, so patients never saw a place to write.
• Red-flag routing is unchanged.

For the analysis

• 1.1 used to mean "pick one" and now means "select all that apply". Use the questionnaire version (1.7.0) to split responses.
• From 1.7.0 on, surgery details are in 1.9a instead of an "Other" answer on 1.9.

Tests

• New backAndLegSymptomsTogetherOpenBothModules test: selecting back and leg symptoms opens both modules, and the overlap note shows only then.
• New surgeryDetailsAreAskedOnlyAfterYesOrNotSure test. To support it, ScreeningAnswers now evaluates conditions on a selected option.
• The UI test selects the new 1.1 label.
• swift test --filter SpineAI: 24 tests pass. The UI test and the simulator walkthrough haven't been run.

Before merging

• IRB modification for the reworded text and the new free-text question, if required

🤖 Generated with Claude Code

[coderabbitai]

Important

• 🔍 Trigger review

This repository does not receive automatic reviews because it has fewer than 10 stars.

⚙️ Run configuration

• Configuration used: Organization UI
• Review profile: CHILL
• Plan: Advanced
• Run ID: 9e74acc5-5066-4cd0-8259-2794531677ae

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[florenceheinkelein-ops]

@PSchmiedmayer could you review this? It updates the SpineAI opening questionnaire based on the feedback from the patient sessions (multi-select answers, "pain or symptoms" wording, timeframe and body-map instructions, and a text box for prior spinal surgery). The red-flag logic is unchanged.
The failing Build and Test and Periphery checks aren't from this PR