Overview
Suspecting automatically identifies patients who may have undiagnosed or under-coded conditions based on clinical evidence in their medical records. The system analyzes prior medical history to find patterns that suggest a condition may be present but has not yet been formally coded. Suspected conditions are represented as FHIR Condition resources with averificationStatus
of unconfirmed. Each suspected condition includes the ICD-10 code, supporting evidence, and
the clinical rules (definitions) that triggered the suspicion.
When clinical data becomes available in our system for a patient, suspecting runs on the next
analytics cycle (typically within about 30 minutes). Data may arrive through a
Network Query or
real-time patient notifications such as ADTs. Later cycles keep re-evaluating your
patient population to surface coding opportunities and improve risk adjustment accuracy.
How It Works
The system uses two detection modes:- Suspect — The patient has clinical evidence suggesting a condition but no existing diagnosis code on file. For example, two fasting glucose readings above 126 mg/dL without a diabetes diagnosis.
- Recapture — The patient had a diagnosis code in a prior year but it has not been captured in the current year, and there is supporting clinical evidence.
Available Condition Groups
The following condition groups are currently supported.More condition groups are under active development. If
there is a specific condition you’d like added, get in
touch with us directly.
Mental Health
Cardiovascular
Metabolic
Respiratory
Renal / Hepatic
Other
Getting Started
Once suspecting is enabled for your account, suspected conditions are available through the API and Dashboard as each analytics cycle completes.Accessing via API
Suspect results are available through the API:- List Patient Suspects — paginated suspected conditions for one patient.
Bundle that includes a Patient resource, the suspected Condition, and
supporting evidence resources (Observations, Procedures, MedicationRequests).