Copy Assemble creates a unique patient graph from a patient's EHR for use in precision CDS Create an Assemble patient graph
post
https://cg-api-prod.system.com Production API chevron-down https://cg-api-prod.system.com https://cg-api-stage.system.com https://cg-api-dev.system.com http://localhost:80
/v1/assemble copy Assemble a patient graph from FHIR bundle.
Body
application/json chevron-down application/json
object · AssembleIn Optional
Assemble API Schema Documentation
The Assemble Schema provides a comprehensive data model for representing patient healthcare information as a graph structure. It supports various clinical entities including conditions, observations, interventions, allergies, and immunizations, along with their relationships.
Table of Contents
AssembledPatient
The root model representing a complete patient record with all associated clinical data.
Fields:
Field
Type
Required
Description
List of all clinical entities (conditions, observations, etc.)
Graph edges connecting related nodes
Patient demographic information
Example:
Represents a clinical entity in the patient's healthcare graph.
Fields:
Field
Type
Required
Description
Example
Discriminator for the node subtype (condition, observation, intervention, etc.)
FHIR ID from the original FHIR bundle
Whether this node is notable/important
System Interoperable Variable ID
Human-readable name for this entity
"Type 2 Diabetes Mellitus"
Additional System Interoperable Variable IDs
External coding references (ICD10, LOINC, etc.)
Example:
Represents a relationship between two nodes in the healthcare graph.
Fields:
Field
Type
Required
Description
system_id of the source node
system_id of the target node
Example:
Supporting Models
PatientDemographics
Patient demographic information.
Fields:
Field
Type
Required
Description
Example
External coding system reference (ICD10, LOINC, SNOMED, etc.).
Fields:
Field
Type
Required
Description
Example
Human-readable description
"Type 2 diabetes mellitus without complications"
Container for all possible metadata types for a node. Only one metadata field should be populated based on the node type.
Fields:
Metadata for condition nodes
Metadata for observation nodes
Metadata for intervention nodes
Metadata for allergy nodes
Metadata specific to disease or condition nodes.
Fields:
Field
Type
Required
Description
Current clinical status of the condition
True if this is a family history condition
Example:
Metadata specific to observation nodes (lab results, vital signs, etc.).
Fields:
Field
Type
Required
Description
ObservationInterpretationEnum
Clinical interpretation (e.g., High, Low, Normal)
Risk bucket (optimal, low, normal, high)
When the observation was recorded
Example:
Metadata specific to intervention nodes (medications, procedures).
Fields:
Field
Type
Required
Description
Type of intervention (medication, procedure, immunization)
Current status of the intervention
When the intervention was last updated
Example:
Metadata specific to allergy nodes.
Fields:
Field
Type
Required
Description
Clinical status (active, inactive, resolved)
Verification status (confirmed, unconfirmed, etc.)
Criticality level (low, high, unable-to-assess)
Example:
ClinicalStatusEnum
Clinical status values for conditions and allergies.
Values:
Condition is currently active
Condition is in remission
Condition has been resolved
Risk level categorization.
Values:
RiskDirectionEnum
Directionality of risk for observations.
Values:
Risk is from very high values
Risk is from very low values
ObservationInterpretationEnum
Clinical interpretation codes for observations. Based on HL7 standards.
Common Values (for a full list of values, see https://hl7.org/fhir/R4/valueset-observation-interpretation.htmlarrow-up-right ):
InterventionTypeEnum
Type of clinical intervention.
Values:
Immunization intervention
InterventionStatusEnum
Status of an intervention.
Values:
VerificationStatusEnum
Verification status for allergies.
Values:
CriticalityEnum
Criticality level for allergies.
Values:
Unable to assess criticality
The Assemble API does not store customer-submitted input data or any generated patient graphs. All patient identification and data retention are the responsibility of the customer, and must be managed within the customer’s own systems.
The Assemble API processes only the FHIR fields defined in the schema above and will ignore any fields outside that schema. However, customers should avoid transmitting unnecessary PHI and are strongly encouraged to remove direct identifiers (e.g., patient name, address, phone number, email) prior to sending data to the API.