Condition Suspecting & Recapture | Metriport

Metriport raises $26 million

Every suspect arrives with its evidence.

Discover conditions that should be presently diagnosed, but are not, based on a patient's longitudinal medical record.

Longitudinal record

Date PHQ-9 Score Observation Medication Request
2026-07-14 14
2026-05-02 Lithium carbonate, active fill
2026-03-19 Hemoglobin A1C 6.9 %
2026-01-08 GAD-7 score 11

Suspected conditions

Condition Code Status
Major depressive disorder, recurrent F33.1 unconfirmed
Bipolar II disorder F31.81 unconfirmed
Type 2 diabetes, uncomplicated E11.9 unconfirmed
Generalized anxiety disorder F41.1 unconfirmed

Join the teams finding what the problem list missed in the history they already hold.

Candidates for review, with the evidence attached.

Condition Verification Status Evidence Code
Major depressive disorder, recurrent unconfirmed PHQ-9 score 14 · 2 encounters F33.1
Bipolar II disorder unconfirmed Lithium carbonate active fill F31.81
Type 2 diabetes, uncomplicated unconfirmed A1C 6.9% · metformin fill E11.9
Generalized anxiety disorder unconfirmed GAD-7 score 11 · buspirone F41.1
Vascular disease unconfirmed Imaging impression · 1 encounter I73.9

The record, screened against clinical definitions

The longitudinal history is screened for conditions that meet documented clinical definitions but appear nowhere on the problem list.

Condition Needs review Code Verification Status Evidence
Type 2 diabetes Yes E11.9 unconfirmed A1C 6.9% · metformin fill

Never an automatic diagnosis

Candidates arrive as unconfirmed FHIR Conditions for review. The problem list never changes on its own.

Evidence you can open

Each candidate carries its supporting Observations, Procedures, and MedicationRequests, so review starts from the proof.

Detection modes: Two Suspect

Suspect, recapture, and chronicity

New candidates, previously documented conditions due for review, and a separate chronic / non-chronic / unspecified enrichment.

Benefits

An empty chart is not an empty history.

Suspecting reads what you retrieved and says what to review.

Find the conditions your charts never mention.

A connected chart is not the same as a complete one. Screening the retrieved history surfaces the patients whose records exist elsewhere, and gives care teams something to treat against.

Problem list

Your chart
Migraine without aura
Seasonal allergic rhinitis
Osteopenia
Type 2 diabetes, uncomplicated
A1C 6.9% · metformin fill · retrieved history

Your clinicians stay the decision-makers.

Every output is an unconfirmed candidate with an ICD-10 code and its evidence. Review confirms or rejects it; nothing writes to the problem list automatically.

Risk work starts from more usable data.

Suspecting runs on the harmonized record, so candidates are computed from history that has already been coded and deduplicated, not from one source's fragment.

Coded once, mapped to risk

Code Condition HCC
F33.1 Major depressive disorder, recurrent 155
F31.81 Bipolar II disorder 154
E11.9 Type 2 diabetes, uncomplicated 19
F41.1 Generalized anxiety disorder Unmapped

How it works

Suspecting only argues. The problem list stays yours.

  1. Evaluate the record against clinical definitions.
  2. Receive unconfirmed candidates.
  3. Query the consolidated record behind them.

FAQs

  1. Does condition suspecting automatically add diagnoses?
    • No. It returns unconfirmed condition candidates for clinical or coding review and does not change the problem list automatically.
  2. Which resources can provide supporting evidence?
    • The documented evidence resources are Observations, Procedures, and MedicationRequests.
  3. What is the difference between suspect and recapture?
    • Suspect identifies a condition that may not have been documented, while recapture evaluates a condition documented previously that may need review again.
  4. How is chronicity classification different?
    • Chronicity classification is a separate enrichment that labels coded conditions as chronic, non-chronic, or unspecified. It is not a confidence score or diagnosis.