Skip to main content

Patients, Encounters & Demographics

Phlox organises clinical data around two concepts: patient profiles (stable, identifying information) and encounters (individual visits/notes). This split keeps demographics stable across visits while letting each encounter carry its own note, transcript, and tasks.

Patient Profiles​

A patient profile holds the stable, identifying information that rarely changes between visits:

  • Name, date of birth, sex
  • UR / MRN number
  • Contact details and address
  • Notes about the patient

Profiles are keyed by UR number. All encounters for the same UR number are grouped together, and history (previous notes, persistent template fields) is carried forward automatically.

Encounters​

An encounter is a single clinical visit. Each encounter contains:

Creating a new note for an existing UR number links the new encounter to that patient's history.

Floating Action Menu​

The Floating Action Menu (FAB) is the vertical action menu on the right side of the encounter workspace. It opens a panel for each common encounter action:

  • Document Upload — upload a PDF, Word, or .txt document to extract text into the current note. See Document Upload.
  • Previous Visit — open the patient's previous encounter for reference.
  • Chat — the reference / agentic chat for this patient.
  • Chart Insights — generate Chart Insights for the encounter. Available after the encounter has been saved.
  • Letter — generate correspondence from the encounter. Available after the encounter has been saved.

Chat, Chart Insights, and Letter are shown when the AI chat is enabled; Chart Insights and Letter additionally require the encounter to be saved first.

Demographics​

Manual entry​

Demographics can be entered or edited via the Demographics modal on the encounter workspace. A specialty-aware field set is used.

Demographics modalDemographics modal

Auto-fill from a dropped document​

You can drag and drop a document (PDF, image, or text file) onto the demographics form. Phlox extracts the demographic fields using the same processing pipeline as document chat:

  • Text-layer extraction first (for digital-native PDFs),
  • falling back to vision (if the model supports it), or
  • OCR (Tesseract, Docker builds only) as a final fallback.

Review the extracted values before saving — AI extraction can make mistakes.

If Require patient consent for ambient scribing is enabled in Settings → User Settings → Advanced, Phlox will prompt for explicit patient consent before starting an ambient recording on a given encounter.

  • The consent prompt appears the first time you start the scribe for that encounter.
  • The consent state is persisted per encounter so you are not re-prompted mid-visit.
  • This is intended to support workflows where documented patient consent is required before recording.

Even with consent captured, you remain responsible for ensuring that making any recording is lawful in your jurisdiction. See Limitations & Warnings.