← Free Tools · MedLegal AI

The Medical Chronology, Built From the EHR

A source-linked timeline assembled directly from structured records — that auto-flags entries changed after the fact.

FHIR R4 / structured EHR Epic · Oracle-Cerner · athenahealth · +aggregators Patient-consented access Amendment / late-entry flagging
Synthetic demo — no PHI. Everything below is the tool's real output on a fabricated test case ("Jane Synthetic"), the same fact pattern used to validate the engine. In production, the patient authorizes access to their own records and the chronology is built from their real EHR data.

The case

A 54-year-old presents to the ED on Sept 2 with pleuritic chest pain, shortness of breath, a heart rate of 112 and an O₂ sat of 93%. She is diagnosed with musculoskeletal chest pain, given ibuprofen, and discharged. Four days later she returns by ambulance with syncope — a saddle pulmonary embolism. The printed chart looks routine. The metadata does not.

The chronology the tool produces

WhenCategoryEntry
2025-09-02 14:10EncounterEmergency — Pleuritic chest pain, shortness of breath
2025-09-02 14:25ObservationHeart rate: 112 bpm
2025-09-02 14:25ObservationOxygen saturation: 93 %
2025-09-02 14:26ObservationRespiratory rate: 22 breaths/min
2025-09-02 16:40DiagnosisMusculoskeletal chest pain (active)  ⚠ entered/changed 20 days later (2025-09-22)
2025-09-02 16:45MedicationIbuprofen 600 mg PO PRN
2025-09-02 16:50DocumentED discharge instructions — return if worse
— 4 days —
2025-09-06 08:05EncounterEmergency — Syncope, severe dyspnea
2025-09-06 08:40ObservationTroponin I: 0.41 ng/mL
2025-09-06 08:42ObservationD-dimer: 4200 ng/mL FEU
2025-09-06 09:30ProcedureCT pulmonary angiography
2025-09-06 10:05Diagnostic ReportCT angiogram, chest — Saddle pulmonary embolism with right heart strain.
2025-09-06 10:10DiagnosisAcute pulmonary embolism (saddle) (active)
2025-09-06 10:15MedicationHeparin IV infusion
2025-09-22 11:00AuthorshipProvenance — Recorded by Dr. A. Resident — Amended

What the metadata caught

The "musculoskeletal chest pain" diagnosis is dated Sept 2 — the day of the first visit. But its system last-modified timestamp is Sept 22: twenty days later, and over two weeks after the patient came back with a saddle PE. The Provenance record names who touched it last.

On the printed chart, that diagnosis looks contemporaneous. The tool flags the gap automatically — pointing you exactly where to aim an audit-trail discovery request.

Two lanes, by design

Structured EHR access (FHIR) carries authorship and last-modified metadata — enough to flag a suspicious entry. It does not carry the deep forensic audit trail (chart-open durations, access logs, full edit history); that exists only behind discovery. The product uses both:

2 · Discovery lane (live tool)

The flag tells you where to look. The audit-trail discovery request gets the forensic proof — who entered or changed each note, the exact timestamps, who accessed the chart.

Generate an EHR audit-trail request →

Want this on a real case?

The consent-connect chronology is in active development and we're piloting it with a small number of plaintiff firms. See it on a sample, or put it on one of your cases.

Request a pilot Try the audit-trail tool

Synthetic demonstration for attorneys — not legal or medical advice, and not a real patient record. A last-modified timestamp later than the clinical date is a signal to investigate, not proof of wrongdoing (it can reflect a legitimate addendum or a data migration). Confirm timing and authorship through the audit trail and discovery in your venue.