Limitations with HL7 in the past regarding an anti-pattern called sub-classing by inheritance. FHIR is the next iteration to contain and alignment best practices around clinical data statements. KNARTs also came about to help document clinical knowledge artifacts, but still limited to multiple possible implementations which can differ and is subjective across individual informaticist/clinicians.
Construct redundancy vs construct overload - overlord is bad, but polymorphism is beneficial with in a certain degree
CIMI tries to work with an idea of a clinical statement. This idea is the essence in a reproducible representation and query process.
statements have topics (a good starting point for ANF)
ANF is an extension of CIMI geared towards forcing/enabling a reconciliation of how to capture a clinical statement across multiple data source from different
Originally there were 4 statements to capture medical data (Observation, Finding, outcomes, Request)
This was reduced through collaboration efforts to Performance and Request clinical statements
E.g., Performance is that I lost 30 pounds over the next three months which could be a result of a request to loose 30 pounds over the next three months
The topic is still the same across both statement types, weight loss of loosing 30 pounds over the next three months
The narrative statement is geared towards individuals who aren’t utilizing performance and request statement types
Turning coded data elements into ANF topics
Discussed past collaboration outcomes of what is the technique of taking blood pressure measure while feet are flat on the floor
Action items
Linda Wedemeyer Send latest and greatest BPMN source and pdf files
Andrew Sills Disseminate previous ANF presentations
Penni Hernandez Provide updated content for review with Dr. Gianola