Style Guide for Diagnostic Archetype Design in openEHR
Introduction and Purpose:
This guide aims to provide a structured approach for designing diagnostic archetypes in openEHR. The purpose is to ensure that archetypes are consistently, accurately, and effectively modeled to represent diagnostic data, particularly in the field of pathology.
Background:
Please find more information why and how this guide was created in the discussion → https://discourse.openehr.org/t/refreshing-archetypes-related-to-pathology-reporting/2393/64
Requirements:
Completeness of data elements
Correct mapping of relations among data elements
Correct mapping of relations among archetypes
Standardization
Background:
Unlike clinical chemistry laboratory data, which primarily consists of name/value pairs, diagnostic pathology data is characterized by its rich semantics. For example, a malignant tumor requires detailed information such as morphological descriptions and specific tumor characteristics.
Structure of the Generic Archetype:
The generic archetype contains standardized core data elements applicable across various examinations. From these, more detailed and specific information is differentiated.
Checklist for Adding New Data Elements:
Generic Element Check: Is the data element generic enough for various examinations?
Multiplicity Check: Determine the occurrence of the data element (once or multiple times).
Grouping/Nesting Check: Decide how to group or nest the data element.
Data Type and Format Rule (Datentypen- und Formatregel): Confirm the data element’s conformity to established types and formats.
Example: Modeling “Malignant Tumor”
Generic Element Check:
Is the data element generic enough for various examinations?
Application: Yes, “malignant tumor” is a generic concept applicable across numerous types of pathological examinations, such as lung, breast, or colon biopsies.
Multiplicity Check:
Determine the occurrence of the data element (once or multiple times).
Application: “Malignant tumor” can occur multiple times within a single examination. For instance, in a lung biopsy, there could be multiple distinct tumors.
Grouping/Nesting Check:
Decide how to group or nest the data element.
Application: Each instance of a “malignant tumor” should be nested within the biopsy report. Further, specific information like tumor type, size, location and ICH results should be nested under each tumor instance.
Data Type and Format Rule:
Confirm the data element’s conformity to established types and formats.
Application: The “malignant tumor” data elements should conform to established medical terminologies and formats. For example, tumor size should be recorded in standardized units (like centimeters), and tumor type should use recognized classifications (like those from SNOMED CT).
Introduction and Purpose:
This guide aims to provide a structured approach for designing diagnostic archetypes in openEHR. The purpose is to ensure that archetypes are consistently, accurately, and effectively modeled to represent diagnostic data, particularly in the field of pathology.
Background:
Please find more information why and how this guide was created in the discussion → Refreshing archetypes related to pathology reporting
Requirements:
Completeness of data elements
Correct mapping of relations among data elements (semantic)
Standardization
Background:
Unlike clinical chemistry laboratory data, which primarily consists of name/value pairs, diagnostic pathology data is characterized by its rich semantics. For example, a malignant tumor requires detailed information such as morphological descriptions and specific tumor characteristics.
Structure of the Generic Archetype:
The generic archetype contains standardized core data elements applicable across various examinations. From these, more detailed and specific information is differentiated.
Checklist for Adding New Data Elements:
Generic Element Check: Is the data element generic enough for various examinations?
Multiplicity Check: Determine the occurrence of the data element (once or multiple times).
Grouping/Nesting Check: Decide how to group or nest the data element.
Data Type and Format Rule (Datentypen- und Formatregel): Confirm the data element’s conformity to established types and formats.
Example: Modeling “Malignant Tumor”
Generic Element Check:
Is the data element generic enough for various examinations?
Application: Yes, “malignant tumor” is a generic concept applicable across numerous types of pathological examinations, such as lung, breast, or colon biopsies.
Multiplicity Check:
Determine the occurrence of the data element (once or multiple times).
Application: “Malignant tumor” can occur multiple times within a single examination. For instance, in a lung biopsy, there could be multiple distinct tumors.
Grouping/Nesting Check:
Decide how to group or nest the data element.
Application: Each instance of a “malignant tumor” should be nested within the biopsy report. Further, specific information like tumor type, size, location and ICH results should be nested under each tumor instance.
Data Type and Format Rule:
Confirm the data element’s conformity to established types and formats.
Application: The “malignant tumor” data elements should conform to established medical terminologies and formats. For example, tumor size should be recorded in standardized units (like centimeters), and tumor type should use recognized classifications (like those from SNOMED CT).
Cross Archetype Relations
Once your archetype is complete you need to make sure if all information of the concept are correctly mapped. For example, radiology typically provides the tumor locations, while microscopy determines if and what type of tumor it is. This means all information for prostate cancer like MRI, microscopic findings, lab results etc. provides us with pieces of the puzzle. To set them into the full picture of prostate cancer we need relationships across archetypes. This is typically done with primary keys. Primary keys are a fundamental concept in data management, used to uniquely identify and link infromation across different data sets or systems. For example in solid tumors, the anatomical site can serve as primary key to correctly assign findings between examinations. A PIRADS lesion can only occur once in a zone.
Requirements of Primary Keys:
Uniqueness:
Each primary key must be distinct within its dataset or database.
No two records can share the same primary key.
Persistence:
Once assigned, the primary key should not change over time.
It must remain constant to maintain data integrity.
Non-null:
A primary key cannot be empty or null; it must always have a value.
This is crucial for the key to serve its purpose of identifying records.
Simplicity (Ideally):
While not a strict requirement, it’s often beneficial for primary keys to be as simple as possible while still maintaining uniqueness.
This can make data management more straightforward.
Relevance to the Data Model:
The primary key should be chosen based on its relevance and importance to the specific data model or system.
Example “Anatomical Site as a Primary Key for solid tumor”:
Uniqueness:
The lesion is located in a specific location e.g. left posterior zone of the prostate.
This location is unique to this specific tumor.
No other lesion or tumor shares the exact same location.
Persistence:
The location of the tumor (left posterior zone of the prostate) remains constant throughout the diagnosis and treatment process.
Even as the patient undergoes various treatments, the original location of the tumor does not change.
Non-null:
The anatomical site (left posterior zone) is a definite, identifiable location.
It is always specified in the patient's medical records, ensuring there is no ambiguity or null value.
Simplicity:
While anatomical locations can be complex, in this case, the designation of the tumor site is straightforward and easily understood by medical professionals (left posterior zone of the prostate).
It's simple enough to be used consistently across different medical records and systems.
Relevance to the Data Model:
In solid tumors, the exact location of the tumor is crucial for diagnosis, treatment planning, and monitoring.
The anatomical site directly correlates with the necessary medical actions, such as the target area for biopsy or localized treatment.