Pathology-to-action surveillance

See the result.
Understand the pathway.
Act sooner.

Echo Pathways monitors pathology against approved clinical protocols and alerts the right clinician when action is due — with source, rule and reason attached.

FHIR-native Epic Oracle Health

app.echo-pathways.com / oversight
The Echo Pathways oversight worklist: prioritised recommendations grouped by responsible clinician, each showing severity, source, rule and reason.
De-identified · read-only oversight · grouped by responsible clinician

The problem

The gap isn't the data. It's the delay between result and action.

A result may arrive as a FHIR observation, an HL7 message, a PDF, an image or free text — and vary by laboratory, assay, specimen, unit and wording. The failure mode isn't unstructured pathology. It's clinical latency: the delay between a clinically significant result and the action it should trigger.

Reviewed too late

Overlooked in a busy inbox, or seen after the window in which it mattered.

Hard to compare

Different assay, specimen, unit or scale to the last result for the same patient.

Sent to the wrong clinician

Delayed between systems, or routed to someone who isn't the responsible owner.

Missing when a repeat was due

The absent result — the one nobody is looking at, because it never arrived.

How it works

From result to action in four governed steps.

01 · connect

Retrieve context

A FHIR connection pulls the relevant pathology and patient context from the record system that already holds it.

02 · interpret

Normalise

Reads PDFs, images and free text; normalises units and scales, and flags amended or superseded reports — preserving the source.

03 · apply

Run the protocol

An approved protocol detects trends, missing results and overdue follow-up. This step is deterministic rule execution.

04 · alert

Alert & assign

A prioritised recommendation goes to the responsible clinician, with the reason and full provenance attached.

The role of AI

AI structures the pathology. Approved clinical rules determine whether action is required.

AI's contribution here isn't generating new information — it's making heterogeneous pathology computable. It reads; it does not decide. The decision path is deterministic and replayable, and a human authorises the consequential step.

AI supports

  • Reading different report formats — FHIR, HL7, PDF, image, free text
  • Extracting the relevant value, and identifying test, marker, specimen and assay
  • Normalising units and scales; recognising amended or superseded reports
  • Recognising uncertainty — anything it cannot read confidently goes to a human, never a guess

Approved clinical logic determines

  • Whether a protocol applies, and whether a threshold has been crossed
  • Whether a test is overdue, or a trend requires review
  • Who should be alerted, and the expected review timeframe
  • Every rule is signed off by clinical leaders before it runs

Echo Pathways is not an autonomous treatment-recommendation system.

The system recommends from a bounded vocabulary of actions; a clinician authorises the consequential step. No generative model harmonises a value, decides an event, computes a schedule, or selects an action.

See it in the product

From an overloaded inbox to a single accountable alert.

A rising molecular result looks like every other message in a busy inbox. Echo Pathways applies the approved protocol and surfaces the one result that needs a human — with the reason, and who owns it.

Before — the clinician inbox

Full blood count · result available 09:14
CT chest/abdomen — report 09:02
Discharge summary for review 08:51
NPM1 MRD quantitative · UR-23124 08:47
Medication chart co-sign 08:39
New referral · category 2 08:30
U&E panel · result available 08:12
Blood culture · preliminary 07:44

Row four is a rising npm1_mrd result — clinically urgent, and visually identical to 147 other messages.

After — Echo Pathways oversight

High Molecular relapse trend

NPM1 MRD 1.16% on 2026-07-12 (from an uploaded image) triggered molecular-relapse-trend; log_rise 3.219 vs prior 0.0007 (threshold 0.5).

source · uploaded image
rule · molecular-relapse-trend
reason · log_rise > threshold
assigned · Dr A. Goncalves · patient UR-23124

One recommendation, with source, rule and reason — assigned, and visible until a clinician reviews it.

Explore the surfaces

Three screens, one accountable trail.

oversight
Oversight worklist: consequential recommendations awaiting a human, grouped by responsible clinician.
Oversight worklist — consequential recommendations awaiting a human, grouped by responsible clinician.
patients
Patient timeline: NPM1 MRD plotted on a log scale with treatment phases overlaid and full provenance.
Patient timeline — one patient, one screen: NPM1 MRD on a log scale with treatment phases and full provenance.
protocols
Protocols screen: signed-off detection rules, thresholds and governance history.
Protocols — signed-off detection rules, thresholds and the governance history the engine executes.

Clinical governance

Protocols are governed by clinicians. Echo Pathways makes them operational.

Every active protocol is clinically authored or approved, version-controlled, tested, signed off and traceable — and able to be retired or replaced. A logic change bumps the version and requires re-sign-off before it can run.

Initial use case — AML

The most demanding first validation environment.

AML brings serial molecular monitoring across multiple assays, laboratories, units and reporting conventions — exactly where clinical latency does the most harm. If the platform holds here, it extends across many pathology-driven pathways.

app.echo-pathways.com / patients
NPM1 MRD on a log scale with treatment phases overlaid, showing a relapse trend.
NPM1 MRD · log scale · treatment phases overlaid · relapse trend detected

Clinical accountability

Every significant event reaches a responsible person — and stays visible until reviewed.

Assign

Each recommendation is assigned to the responsible clinician and grouped by ownership — not broadcast to an inbox.

Escalate

Overdue and high-severity items escalate and support follow-up. An absent result is treated as high-severity, not as silence.

Track

Source result, applied rule and reason are recorded — a complete, replayable audit trail.

Integration

FHIR-native. Integration-ready for Epic and Oracle Health.

Echo Pathways is designed to work with existing patient-record systems rather than replace them — preserving provenance and operating within current workflows. FHIR context is read-only.

FHIR R4 HL7 v2 Epic Oracle Health (Cerner)

Research & validation

Retrospective validation. Prospective silent mode. Real-world evidence.

01

Retrospective

Replay historical cohorts against a protocol before it goes live, and measure what it would have caught.

02

Silent mode

Run live without alerting, to measure performance safely in a real service before anyone relies on it.

03

Real-world evidence

Auditable by design, with multi-disease study potential.

Monitor what matters. Act when it matters.

We work with clinical leaders, digital health teams and research partners to bring pathology-to-action surveillance into live services.