Synthetic Objects

Synthetic objects represent an extra level of transformation or refinement of data from the natural object level. Measures typically consist of some combination of natural objects and synthetic objects. For example, “Professional Billing Line Item” is a natural object, whereas “Professional Billing Diagnosis” is a synthetic object. Or, “Medication Orders” is a natural object, whereas “Diabetes Medication Orders” is a synthetic object. It’s possible to build synthetic objects in the synthetic object workshop, but in practice many of these are bespoke SQL objects.

The three most common categories of synthetic objects:

• Aggregators – collect all records associated with a concept, such as an encounter. For example, all documents with a qualifying HCPCS/CPT code indicating primary care services were performed.
• Encounters – create a single record representing the final encounter-level properties of an event. For example, the encounter for hospital inpatient admissions will aggregate the total plan paid amount, determine the principal discharge diagnosis, and provider encounter start and end dates representing the duration of the patient’s time in the hospital.
• Timelines - one row represents a single, unambiguous state of the entity on any given date. Common examples would be the patient’s attributed PCP, plan membership, or active chronic conditions.