Jordan Matthew

04 / Clinical workflow

Trial Coordinator

Can a preliminary trial review stay understandable as observations, follow-ups and public study information change?

Synthetic workflow prototype · TypeScript / React
Private repository ↗Open demo ↗
Trial Coordinator prototype interface
Recorded prototype. Synthetic example. Inspect full desktop capture ↗

Working result

A complete synthetic case loop: criterion review, follow-up, recovery and handoff.

Review status

Workflow checks recorded. Coordinator review remains open.

One consequential decision

Keep unresolved work in the handoff

  1. Criterion + observation
  2. Follow-up + source check
  3. Traceable handoff

A follow-up answer does not silently complete a criterion. Changes to the source require the earlier review to be reconsidered.

The question

Can a preliminary trial review stay understandable as observations, follow-ups and public study information change?

What I built

A workspace for three invented case scenarios and public trial records. A reviewer opens a case, records observations against original criteria, assigns unanswered questions and downloads a handoff. The handoff preserves unfinished review and missing evidence.

How it works

A typed adapter validates ClinicalTrials.gov records. Source snapshots preserve criteria and facility context. Changed source information pauses reuse of earlier observations; restarting archives the prior review.

The primary navigation is Worklist and Case workspace. Study details expand inside the case. Review and Handoff retain the same case selection, progress and save status. Download is available directly from review, reducing required navigation without silently filling in observations.

A handoff requires a fresh source check. An unavailable or changed record prevents download. Answering a follow-up does not automatically resolve its criterion, and there is no aggregate eligibility score.

Browser-local state is validated before restoration. Corruption pauses saving and offers recovery. This is device-local prototype storage, not a shared clinical record or regulated audit log.

Result

The implemented synthetic loop covers evidence entry, follow-up assignment and resolution, reload persistence and a downloaded handoff. All 29 application tests, TypeScript checks and production build passed. Browser checks exercised that workflow, seven fault scenarios and 390/320-width layouts.

Jordan's feedback changed the interface from separate stages into a unified case workspace. This is a concrete design revision, not evidence of measured time savings. Independent coordinator testing remains open; no real patient records or formal eligibility decisions are included.

Repository

Public repository pending. The local release package includes the lockfile, application tests and setup instructions. Browser workflow harnesses remain in the program QA folder and are not portable tests in that package.

What I'd do next

Evaluate whether a reviewer can identify unfinished work and hand it off without losing context. Prioritize observed friction before adding workflow stages.

Next project ↗Entero