Skip to main content

Create and maintain clinical cases

Access: case edit permission. Open Moderation > Case templates and create a case or open an editable revision. Existing published content is not overwritten merely by opening it.

Build the scenario​

  1. Set a recognizable internal name and authoring notes.
  2. Choose the relevant cardinal symptoms and describe the scenario.
  3. Configure age, gender, history, vital signs, and laboratory values. Use fixed values for a standardized case and ranges/probabilities for a variable scenario.
  4. Write the learner-facing vignette, patient conversation context, and completion/evaluation vignette for their distinct audiences.
  5. Configure findings, permitted requests, and expected solutions where applicable.

The structured editor handles the supported case fields. Use its fixed/range controls rather than manually rebuilding JSON. Advanced JSON remains useful for supported content transfer or unusual fields, not a prerequisite for ordinary authoring.

Findings, images, and written reports​

Add static reports to the intended request item/type. Upload approved, de-identified raster images or use approved secure image sources. Where a written learner report is enabled, provide the teaching solution and check the report/save/reveal flow.

Review every image, caption, filename, and vignette for identifiers before publication. The application does not automatically remove identifying information from image pixels.

Patient conversation​

Set patient-specific context and any configured conversation tabs/prompts. Test the sandbox before publication. Conversation quality depends on the scenario and the installation's AI configuration; do not promise a response is medically authoritative.

Preview, then publish​

  • Generated-data preview uses the current draft without saving a patient session.
  • Play preview uses the published case and is excluded from usage statistics.
  • JSON download is a portable case-data format, not an export of learner records.

Check several generated patients for variable scenarios: one plausible sample does not validate every combination. Review the request restrictions and result text together, then follow review and publication.

Changing the source case does not retroactively update an already published sequence snapshot.