The Search for Logical Models and the place of the openEHR Reference Model

Published on by Sam HeardLeave a comment

Sam Heard, April 2026 PDF download

PDF download?

Table of Contents

The search for a comprehensive set of logical models for health care has a long history. My experience began as a key investigator in the Good European Health Record project in the early 90s. It became clear during this project that 1) an all-inclusive single model of health data was not achievable, and 2) the data specifications had to meet the needs of the people who wanted to use the data.

I have always been interested in a single life-long health record that could grow over time and be available throughout a person’s life. Clearly it would not be appropriate to base this on a particular technology, terminology or other feature that is bound to evolve over time. This is considered to be the role of a logical model. Let us consider then what logical models are and why people are interested in them.

What is a logical model?

A logical model is a structured representation of data, showing entities, attributes, and relationships, independent of any specific database technology. It acts as a blueprint that defines how data is organized conceptually and ensures consistency, integrity, and clarity before moving to physical implementation.

Key Characteristics of a Logical Model

  • Technology-independent: Unlike physical models, logical models don’t depend on a particular database system.
  • Entities: Represent real-world objects or concepts (e.g., Customer, Product, Order).
  • Attributes: Describe properties of entities (e.g., Customer Name, Order Date).
  • Relationships: Define how entities connect (e.g., Customer places Order).
  • Normalization: Organizes data to reduce redundancy and improve integrity.
  • Semantic layer: Provides a business-friendly view of data, bridging raw data and meaningful insights.

Why Logical Models Matter

Standards bodies like the Centre for European Standards (CEN) have been creating logical models for more than 30 years in an attempt to standardise health information.

Logical models potentially offer:

  • Blueprint for databases: Guides the transition from conceptual ideas to physical database design.
  • Business alignment: Ensures data structures reflect business rules and requirements.
  • Data integrity: Prevents anomalies (like duplication or inconsistent updates).
  •         Scalability: Supports analytical applications by offering a consistent framework across systems.

Table 1: Comparison of data models

Model typeFocusExample use caseDependency on technology
ConceptualHigh-level business conceptsDefine what data exists (e.g., “Students enroll in Courses”)No
LogicalDetailed structure of entities, attributes, relationshipsDefine how data relates (e.g., “Student has StudentID, Course has CourseCode”)No
PhysicalActual implementation in a DBMSCreate tables, indexes, constraints in SQL Server, Oracle, etc.Yes

A Simple Example

Imagine designing a IT system for a university, using an SQL database. The different models developed might include:

  • Conceptual model: Students, Courses, Professors.
  • Logical model:
    • Entity: Student → Attributes: StudentID (PK), Name, Email
    • Entity: Course → Attributes: CourseCode (PK), Title, Credits
    • Relationship: Student enrolls in Course (many-to-many)
  • Physical model: Actual SQL tables with primary/foreign keys, indexes, and constraints.

 In short, a logical model is the bridge between abstract business concepts and concrete database implementation, ensuring clarity, consistency, and scalability.

 openEHR Reference Model and Logical Models

A design philosophy of openEHR is to provide a single logical model. The openEHR reference model can be frustrating for developers who want to see how to implement the specifications. What is the role of archetypes? The combination of a logical reference model (to ensure consistent representation of recurrent concepts) and archetypes (to define how to represent ever evolving clinical data using that reference model) is considered one of the most elegant aspects of the design.

Let’s consider how the reference model and archetypes relate to the idea of logical models.

First, the Reference Model (RM) defines the core entities, attributes, and relationships (e.g., Composition, Entry, AuditInfo).[1] This highly specified model captures data and relationships that:

  • Ensure data provenance, integrity, and consistency across implementations; and
  • Is technology-agnostic: it doesn’t care if the backend is relational, document-based, or graph-oriented.

This is essentially the logical model layer, specifying what data is and how it relates, independent of physical storage.

Second, the Archetypes which are built on top of the RM and constrain and specialize the generic structures into clinical concepts (e.g., “Blood Pressure Measurement,” “Medication Order”). Archetypes are:

  • Reusable, shareable definitions that map clinical semantics onto the logical backbone; and
  •  Acting as domain-specific logical models, expressing how to use the RM entities to represent real-world healthcare data, and requiring no change to the underlying database implementation.

How do these two models relate to Logical Models?

  • In database design, a logical model defines entities, attributes, and relationships without tying them to a physical schema.
  • In openEHR:
    • The Reference Model is the foundational logical model — the abstract schema of health record data. This provides all the entities and attributes required to ensure the integrity of a life-long health record.
    • Archetypes are applied logical models — they specify how to represent particular clinical concepts using the RM, and can be used to capture, query and validate data. These provide the changing, context specific, evolving and often critical data specifications for personal health care – at times fractal in nature.
  • Both layers remain agnostic to technology and DBMS, just like a logical model in traditional data modeling.

Table 2: Analogous equivalence in standard data modelling

Layer in openEHREquivalent in Data ModellingRole
Reference ModelLogical Data ModelDefines abstract entities, attributes, relationships necessary for data integrity and provenance of a life-long health record
ArchetypesDomain-specific constraints expressions independent of the RM implementationConstrains the logical RM model to provide the diverse and ever evolving clinical concepts necessary for health care
Physical SchemaPhysical data modelTables, indexes, Json docs of the reference model

There are a number of features of this two-level modelling that are a good fit for personal health data, which is almost infinitely complex and certainly evolving in nature. These include as key features:

  • Interoperability: Logical models (like openEHR RM) ensure that data can be exchanged across systems without loss of meaning.
  • Flexibility: Archetypes allow clinical communities to define semantics without changing the underlying logical backbone.
  • Future-proofing: Because both RM and archetypes are DB-agnostic, they can be implemented in SQL, NoSQL, or even distributed ledger systems without redesign.

In summary, openEHR’s Reference Model provides a generic, technology-independent information model for representing health record data. It defines the structural and lifecycle semantics of clinical information but does not represent specific clinical concepts.

Archetypes specialise the Reference Model by expressing domain-specific constraints that define clinical concepts such as observations, diagnoses, or procedures. Terminology bindings within archetypes further define the meaning of coded data.

In this openEHR architecture, the Reference Model enables structural interoperability between systems, while archetypes and terminology bindings enable semantic interoperability of clinical data.

A detailed description of the openEHR RM as a logical model

Introduction

The openEHR Reference Model (RM) is a foundational, formal information model that defines the structure, semantics, and behaviour of electronic health record (EHR) data. It is designed to enable semantic interoperability, robust data provenance, and clinical integrity across healthcare systems and over time. The RM achieves this by providing a stable, extensible set of information models and data types that serve as the backbone for all openEHR-based systems. Through its integration with archetypes and templates, the RM supports the consistent, high-fidelity representation of clinical data, facilitating safe data exchange, querying, and long-term record-keeping.

This section provides a comprehensive overview of the openEHR RM, detailing its key entities, attributes, and packages. It explains the role and utility of each major component—such as Composition, Entry, Section, Cluster, Element, and others—in the context of a longitudinal EHR. There is an emphasis on how the RM supports data provenance, clinical integrity, and semantic interoperability, and how it interacts with archetypes and templates to enable consistent clinical data representation across systems.

Overview of the openEHR Reference Model

Purpose and Scope

The openEHR RM defines a logical, interoperable architecture for EHRs. Its primary goals are to:

  • Enable consistent representation of clinical data across systems and over time.
  • Support semantic interoperability by providing a stable, well-defined set of structures and data types.
  • Ensure data provenance and clinical integrity through robust versioning, audit trails, and validation mechanisms.
  • Facilitate integration with archetypes and templates, allowing domain experts to define and constrain clinical content without altering the underlying technical model.

Package Structure

The openEHR RM documentation is organized into several interrelated packages, each responsible for a specific aspect of the EHR:

  • ehr: Defines the top-level EHR structure, including access control, status, compositions, folders, and contributions.
  • composition: Specifies the COMPOSITION class, the primary container for clinical content.
  • content: Contains navigation (SECTION) and entry (ENTRY and its subtypes) packages.
  • data_structures: Provides generic data structures such as ITEM_TREE, ITEM_LIST, ITEM_TABLE, and HISTORY.
  • data_types: Defines core data types (DV_TEXT, DV_CODED_TEXT, DV_QUANTITY, etc.).
  • common: Contains shared concepts like LOCATABLE, versioning, and participation.
  • support: Provides supporting types such as identifiers and URIs.
  • ehr_extract: Defines the EHR Extract model for data exchange.
  • integration: Supports integration with legacy and external systems.
  • (demographic: Models parties, roles, and demographic entities but is not included in this document as this package is not required for openEHR health record implementations).

Core RM Entities: Structure and Longitudinal Utility

The following Entities are the major components and provide the structure and organisation of health related data, which itself conforms to archetypes.

EHR

The EHR object is the root container for all health record content for a patient. It references all compositions, folders, status, access control, and contributions. The EHR object supports longitudinal record-keeping by maintaining a complete, versioned history of all clinical and administrative data for a subject. It is the central access point for querying, auditing, and managing the patient’s health information.

EHR-Level Structures: EHR_STATUS, EHR_ACCESS, Folder, Director

  • EHR_STATUS: Contains metadata about the EHR, such as subject, queryability, and modifiability.
  • EHR_ACCESS: Defines access control policies and settings.
  • FOLDER: Organizes compositions into a directory structure, supporting thematic or temporal grouping.
  • Directory: Optional hierarchical folder structure for advanced organization.

These structures support privacy, security, and efficient retrieval of clinical content.

Composition

A COMPOSITION represents a unit of information committed to the EHR, typically corresponding to a clinical document or event (e.g., a discharge summary, progress note, or lab report). Each composition is versioned, allowing for full audit trails and rollback. Compositions are categorized as:

  • Persistent: Represent long-term patient state (e.g., problem list, medication list).
  • Event: Capture time-specific clinical encounters.
  • Episodic: Support episode-based care (e.g., pregnancy episode).

Compositions encapsulate context (EVENT_CONTEXT), content (Sections and Entries), and metadata (language, territory, category, composer).

Event Context and Clinical Context

The EVENT_CONTEXT class captures the context of a clinical event, generally a composition, including:

•             start_time, end_time: Time interval of the event.

•             location: Physical location.

•             setting: Clinical setting (coded).

•             health_care_facility: Facility where the event occurred.

•             participations: List of parties involved.

EVENT_CONTEXT enables understanding of the circumstances surrounding clinical data, supporting provenance and contextualization.

Section

A SECTION is an organizational structure within a composition, used to group entries under logical headings (e.g., “History”, “Examination”, “Medications”). Sections can be nested, allowing for hierarchical organization of clinical content. While sections themselves do not carry clinical meaning, they provide structure for consistent documentation and facilitate navigation and querying over time.

Entry (and subtypes)

The ENTRY class is the abstract superclass for all clinical statements in the EHR. It is further specialized into several subtypes, each representing a distinct type of clinical information:

  • OBSERVATION: Records measurements, findings, and test results and data with specific timing. It utilises a time-series structure (see History below) to support repeated measurements or data streaming (e.g., vital signs, lab results).
  • EVALUATION: Captures clinical assessments, diagnoses, and opinions. Represents the outcome of clinical reasoning. The timing is drawn from the date in the archetyped data (e.g. date of onset) or from the composition.
  • INSTRUCTION: Specifies intended future actions, such as orders, prescriptions, or care plans. Supports workflow and care planning.
  • ACTION: Documents actions that have been performed, which may be in response to instructions (e.g., medication administration, procedures).
  • ADMIN_ENTRY: Records administrative or non-clinical data (e.g., admission details, social services information).

Each ENTRY subtype supports longitudinal tracking of clinical reasoning, interventions, and administrative events, enabling a comprehensive view of patient care over time.

Item Structures: Element, Cluster, Item_Tree, Item_List, Item_Table

Item structures provide the building blocks for organizing clinical data within entries:

  • ELEMENT: The atomic unit of data, holding a single value (e.g., blood pressure reading). Supports null_flavour for missing data semantics.
  • CLUSTER: Groups related elements or nested clusters, enabling modular and reusable data structures (e.g., a device cluster containing make and model elements).
  • ITEM_TREE: Represents hierarchical data, allowing for complex nested structures and is used in archetypes at the root of all data collections, although any structure is allowed in the reference model. The following ITEM_LIST and ITEM_TABLE are not used widely as they limit evolution of the data.
  • ITEM_LIST: Represents ordered lists of elements (e.g., problem lists).
  • ITEM_TABLE: Models tabular data with named columns and rows (e.g., lab result panels).

These structures enable flexible, archetypable organisation of clinical information, supporting both simple and complex data patterns.

History and Time-Series Modelling: History, Event, Point_Event, Interval_Event

The HISTORY structure models time-series data, essential for representing longitudinal observations such as vital signs or glucose tolerance tests. It contains:

  • origin: The starting point of the time series.
  • events: A list of EVENT instances, each representing a data point.

EVENT is the abstract superclass for time-stamped data points, with two main subtypes:

  • POINT_EVENT: Represents an instantaneous observation (e.g., a single blood pressure reading).
  • INTERVAL_EVENT: Represents data aggregated over a time interval (e.g., average blood pressure over 5 minutes), with attributes for width (duration), math_function (e.g., mean, max, min, change etc.), and sample_count.

The HISTORY structure supports both periodic and aperiodic sampling, efficient representation of fine-grained device data, and inclusion of summary data. It enables detailed temporal modeling of clinical phenomena, supporting longitudinal analysis and trend detection.

Data Types in the RM

The RM defines a comprehensive set of data types, all inheriting from the abstract DATA_VALUE class. Key data types include:

  • DV_TEXT: Free text with optional language and informal encoding.
  • DV_CODED_TEXT: Text with associated coded terminology, supporting terminology bindings and semantic interoperability. As this is a subclass of DV_TEXT, it enables any text to be formally coded when appropriate, linking the text and code from the terminology source.
  • DV_QUANTITY: Numeric values with units (e.g., 120 mmHg).
  • DV_DATE_TIME: Date and time values, supporting temporal precision.
  • DV_MULTIMEDIA: Encapsulated multimedia content (e.g., images, audio).
  • DV_BOOLEAN, DV_COUNT, DV_PROPORTION, DV_ORDINAL: Support for boolean, count, ratio, and ordinal data (name: value pair e.g. Apgar scoring).

These data types enable precise, semantically meaningful representation of clinical information, supporting both human readability and machine processability.

Versioning, Change-control, and Audit

The RM implements a robust versioning and change-control mechanism to ensure data provenance, integrity, and traceability:

  • VERSIONED_OBJECT<T>: A container for all versions of a top-level object (e.g., Composition).
  • VERSION<T>: Represents a single version, with audit details and digital signature.
  • CONTRIBUTION: Groups one or more versions committed together, representing a logical change-set.
  • AUDIT_DETAILS: Records metadata about each commit (who, when, what, why).
  • ATTESTATION: Supports legal attestation and digital signing of content.

This mechanism ensures that all changes are indelible (no physical deletion), fully auditable, and reconstructible, supporting medico-legal requirements and forensic examination of past states.

Feeder Audit and Provenance

The FEEDER_AUDIT class captures metadata about the origin of data imported from external systems. It includes:

  • originating_system_audit: Audit details from the source system.
  • feeder_system_audit: Audit details from any intermediate system.
  • original_content: Optional inclusion of the original content.

FEEDER_AUDIT enables traceability, trust, and integration of data from non-openEHR sources, supporting data provenance and duplicate detection.

Locatable, Identifiers, and Paths

The LOCATABLE class is the base for all archetypable objects in the RM. It provides:

  • uid: Unique identifier for the object.
  • archetype_node_id: Identifier linking the object to its archetype definition.
  • name: Human-readable name.
  • archetype_details: Metadata about the archetype and template used.
  • feeder_audit: Provenance information for imported data.

LOCATABLE supports path-based querying using Xpath-compatible syntax such as openEHR’s AQL, enabling precise referencing and navigation of any node within the EHR. This is essential for semantic querying, linking data, data extraction, and interoperability.

Interaction with Archetypes and Templates

openEHR employs a two-level modeling approach involving use of two models:

  1. Reference Model (RM): Provides a stable, technical foundation for data structures and types.
  2. Archetype Model (AM): Allows domain experts to define reusable, constraint-based models (archetypes) for clinical concepts (e.g., blood pressure, lab result).

The actual clinical data is specified as archetypes, which are statements of how to use the reference model to express specific data instances. The archetypes, often expressed as ADL, conform to the archetype model.

Templates combine multiple archetypes to define forms, documents, or messages for specific use cases (e.g., discharge summary, lab report).

Archetype and Template Binding

  • Archetypes constrain RM structures, specifying allowed data elements, value sets, and terminology bindings.
  • Templates further constrain archetypes for particular contexts, setting occurrence constraints, default values, and hiding irrelevant elements.

At runtime, archetype_node_id and archetype_details in LOCATABLE instances link data nodes to their generating archetypes and templates. This enables:

  • Validation: Ensuring data conforms to clinical and technical constraints.
  • Semantic querying: Using archetype paths and codes for precise data retrieval.
  • Consistent modification: Ensuring updates respect original constraints.

Archetypes and templates are stored separately from data, supporting reuse and independent evolution of clinical models and technical infrastructure.

Common features

  • Archetype-root points: ENTRY subtypes (OBSERVATION, EVALUATION, etc.) are typical root points for archetypes.
  • Templates: Used to define structures for common documents (e.g., discharge summaries, lab results, care plans).
    • Example structures: OGTT (oral glucose tolerance test), asthma management plan, multi-drug therapy.

Templates can be viewed and edited using tools such as the openEHR Template Designer, and numerous real-world examples are available in the openEHR Clinical Knowledge Manager (CKM).

Semantic Interoperability

Terminology Binding

The RM and archetypes support semantic interoperability through:

  • DV_CODED_TEXT and CODE_PHRASE: Allowing data values to be coded using standard terminologies (e.g., SNOMED CT, LOINC, ICD).
  • Archetype term definitions: Each archetype may contain an internal terminology for node definitions and value sets which can be directly linked to multiple terminologies through:
    • External bindings: Archetypes can bind internal codes to external terminologies, enabling mapping and querying across systems.

This approach ensures that clinical data retains its meaning across different systems, languages, and contexts, supporting consistent querying, data sharing, and decision support.

Path-Based Querying

Every data point can be accessed directly using path based queries.

  • LOCATABLE paths: providing Xpath-compatible syntax enables precise referencing of any node.
  • Archetype paths: Combine RM attribute names and archetype codes for semantic navigation.
  • AQL (Archetype Query Language): Supports expressive, semantically rich queries over archetyped data.

Semantic interoperability is further enhanced by the use of archetype constraints, which enforce consistent use of codes, value sets, and data structures across systems and allow tight data validation in forms or at the time of committing the data.

Clinical Integrity and Data Quality

The quality of data is increasingly important with more automatic processing of health data through AI and other mechanisms. It is worth summarising what openEHR offers in this domain.

Data Validation and Constraints

  • Archetype constraints: Define permissible values, structures, and terminologies for each data element.
  • Templates: Further constrain archetypes for specific contexts, ensuring only relevant data is captured.
  • Null_flavour: Indicates reasons for missing data (e.g., unknown, not applicable, masked), supporting accurate interpretation and data quality assessment.
  • Data validation: Implementations in all technologies are able to validate data against archetype and template constraints at data entry and modification time.

Attestation and Audit

Clinical accountability and medico-legal provenance is important for longitudinal health records. openEHR provides many features to support this, as well as versioning of all contributions to the health record.

  • ATTESTATION: Supports digital signing and legal attestation of content, ensuring accountability and trust.
  • AUDIT_DETAILS: Records who made changes, when, and why, supporting traceability and medico-legal compliance.

These mechanisms ensure that clinical data is accurate, complete, and trustworthy, supporting safe patient care and regulatory compliance.

Interoperability and Exchange

With the future of health care in mind, openEHR provides a means of extracting all or part of a health record and providing it to a new health record instance, while maintaining the provenance of the data, as well as the state of each health record instance at any point in time. This is considered to be critical for health care professionals.

EHR Extract Model

The EHR Extract model enables standardized exchange of health record content between systems, supporting:

  • Full, simplified, and synchronization extracts: For ad hoc queries, batch transfers, migrations, and system synchronization.
  • Compatibility with external standards: Such as ISO 13606, HL7 CDA, and FHIR.
  • Preservation of versioning and audit trails: Ensuring data integrity and provenance during exchange.

The extract model uses X_VERSIONED_OBJECT and related classes to serialize versioned content for lossless transmission, supporting a wide range of interoperability scenarios.

Integration with Other Standards

  • GENERIC_ENTRY: Supports integration with legacy and non-openEHR systems.
  • Terminology bindings: Enable mapping to external vocabularies and value sets.
  • Canonical and simplified serializations: Support XML, JSON, and other formats for data exchange and integration.

These features ensure that openEHR-based systems can interoperate with a diverse ecosystem of health IT solutions, supporting data liquidity and patient mobility.

Knowledge Engineering and Ontologies

There are many specialists who understand a great deal about the structure and organisation of knowledge, and the meaning of the words we use and the data we collect. Thomas Beale brought an ontological perspective to the design of the openEHR reference model – which represented the consistent features of health data, provenance, timing, language, and many other aspects.

Archetypes., although designed to work with any (and multiple) terminologies, were firmly based on clinical requirements and needed no semantic basis, at least at the outset, as clinicians already shared this very large data space that was common; straddling language, culture, time and space. Organising the archetypes to maximise interoperability and reuse was the challenge and it is a challenge that persists. Some data collections, such as laboratory results, were so well established that following the widespread pattern and using LOINC terms aided implementation.

The great benefits that openEHR offers are consistency of representation provided by the reference model and the community of practice to create the archetypes which together provide the context for the words and phrases in terminologies, and limit the set to a small number in many situations. All this makes meaning manageable; and automatic processing more resilient.

Conclusion

The openEHR Reference Model provides a robust, semantically rich, and longitudinally consistent framework for representing electronic health records. Its modular architecture, comprehensive set of entities and data types, and integration with archetypes and templates enable high-quality, interoperable clinical data across systems and over time. Through its support for versioning, audit, provenance, and semantic interoperability, the RM ensures clinical integrity, data quality, and safe data exchange in diverse healthcare environment.

The RM is explicitly designed to support longitudinal health records, enabling comprehensive tracking of patient data over time:

  • Persistent Compositions: Maintain single sources of truth for long-term patient state (e.g., problem list, medication list).
  • Event Compositions: Capture discrete clinical encounters or events.
  • Episodic Compositions: Support episode-based care (e.g., pregnancy, hospital admission).
  • HISTORY and EVENT structures: Model time-series data, supporting trend analysis and temporal reasoning.
  • Versioning and audit: Enable reconstruction of past informational states, supporting medico-legal requirements and forensic analysis.
  • Folders and directories: Organize data thematically or temporally for efficient retrieval and navigation.

This longitudinal utility is critical for supporting chronic disease management, population health analytics, and continuity of care across organizational boundaries.

By separating the technical foundation (RM) from clinical content modeling (archetypes and templates), openEHR empowers domain experts to define and evolve clinical models independently of technical infrastructure, fostering innovation, adaptability, and future-proof health information systems.

Appendix A: Major Reference Model classes, purpose and use

EntityMain AttributesPurposeLongitudinal Utility
The features of the entire longitudinal EHR and how it is organised
EHRsystem_id, ehr_id, contributions, ehr_status, ehr_access, compositions, foldersTop-level container for a patient’s health recordProvides the longitudinal anchor for all clinical data; ensures continuity across systems and time
EHR statusis_modifiable, is_queryable, subject, other_detailsDefines modifiability and queryability of the EHRControls lifecycle and accessibility of the record over time
EHR Accesssettings, policiesDefines access control policies for the EHRSupports longitudinal governance and security of patient data
Foldername, itemsLogical grouping of CompositionsSupports organization of longitudinal record across episodes or domains
The components of the EHR
Compositionlanguage, territory, category, context, composer, contentTop-level container for clinical content; represents a single clinical encounter or documentAnchors entries to time, place, and author; provides audit trail and context for longitudinal record
Sectionname, further sections or entriesOrganizational grouping within a CompositionSupports hierarchical structuring of content across encounters; aids navigation and retrieval and human readability
Entrylanguage, encoding, subject, provider, other_participations, data (ItemStructure)Abstract superclass for all clinical and administrative statementsProvides consistent framework for observations, evaluations, instructions, and actions across time
-Observationdata (history), state, protocolCaptures measured or observed dataSupports time-series data (e.g., vitals, labs) enabling longitudinal tracking and trend analysis
–Historyorigin (time of first event), events or serial measurements of fixed intervalTime-structured series of data pointsEnables single point in time or longitudinal representation of repeated measures (e.g., lab series). Longitudinal measures include maximum value, minimum value, average, range, or sum.
—Event (abstract)time, data and state informationSingle point in a History series. Events are either point in time or an interval of timeCaptures discrete longitudinal data points within a time series
—Point eventtime, data and state informationRepresents a single point in timeCaptures instantaneous observations
—Interval eventtime, width, math_function, dataRepresents an interval of timeCaptures aggregated or interval-based observations
-Evaluationdata, protocolRepresents clinical interpretation, diagnosis, or assessmentPreserves evolving clinical judgments and assessments over time
-Instructionnarrative, activities, workflow plan, expiry_time, wf_definitionPrescribes actions to be carried outCaptures intended care plans and orders; supports continuity of care across encounters
-Actiondescription, time, instruction details, ism_transition (state transition such as from prescribed to dispensed for a medication)Documents actions performed in response to instructionsProvides record of what was actually done; supports audit and outcome tracking
-Admin Entrydata (item_structure)Records administrative dataSupports non-clinical longitudinal data
Clusteritems, archetype_node_idStructure for nested or complex data – may be archetyped for reuse in other EntriesSupports detailed modeling of panels, imaging, or structured measurements across time
Elementname, valueLeaf node holding a single data valueBasic unit of clinical data; enables fine-grained longitudinal comparison (e.g., systolic BP values)
The features which allow the health record to be coherent, safe and useful
AuditInfocommitter, time_committed, change_typeMetadata for provenance and versioningEnsures traceability, accountability, and integrity across the evolving longitudinal record
Partyidentifiers, name, rolesRepresents people or organizations involvedSupports continuity by linking data to patients, clinicians, institutions across time
Locatablearchetype_node_id, nameBase class for all RM structuresProvides consistent identification and semantic anchoring across longitudinal data and enables a single data point to be queried and linked
Versionuid, lifecycle_state, contributionRepresents a version of an RM objectSupports version control and audit trail across longitudinal record, independent of the system it may be stored within
Contributionuid, versions, auditGroups versions committed togetherEnsures atomic commits and provenance across longitudinal updates
Archetypedarchetype_id, template_idLinks RM structures to archetypes/templatesProvides semantic binding for longitudinal consistency
Linkmeaning, type, targetRepresents relationships between RM objectsSupports semantic connections across longitudinal record
Attestationattester, proof, reason, itemsLegal attestation of contentSupports medico-legal integrity

[1] See table on page 16
ChatGPT was used to summarise the reference model specification and provided the many of the headings in this paper.

You touch me

Published on by Sam HeardLeave a comment

I was walking down the corridor in a clinic and a middle-aged Aboriginal man called out to me from a colleague’s room, “Hey Doc, I want to see you!” I explained that the very proficient and careful younger doctor that was attending to him was as good as me. He said, “I know, but you touch me.” It was not something I had been told before, and I contemplated a range of interpretations: did I connect to him emotionally; was I more intimate; or did I examine him, perhaps more thoroughly?

There is a wealth of literature on this subject perhaps best summarised in a TED talk by Dr Abraham Verghese “A doctor’s touch“. Every family has stories that demonstrate this fundamentally important aspect of medical care and there are popular medical aphorisms such as “if you don’t put your finger in it, you’ll put your foot in it”. Health practitioners know it is important, but….

There are a number of forces rising against medical examination that offer relief to the uncomfortable doctor or nurse and their patient, both concerned about social boundaries that may be crossed, or the possible experience of emotional or physical discomfort. First is the rise of ‘tests’ which provide detailed internal images which can be wholly reassuring if normal or biochemical measurements are incompatible with disease, thus avoiding the need to see and feel what might be going on. Second, the increase in a general ‘hypochondriasis’ which involves less than perfect wellness, where examination is redundant as only esoteric genetic, enzyme or protein tests will reveal the minor malady and enable (largely) dietary adjustments. How ‘nice’ to be at a comfortable distance and really get to the core issue while avoiding any embarrassment. It is a dangerous collusion.

One person I saw recently really brought to light this issue. A middle-aged man with right sided chest pain had been seen on a number of occasions in various settings over a period of more than a year. His records and the way he presented with yet further pain, despite two normal CT scans done over a year apart (the second one a few weeks ago), suggested that he was seeking pain medication, perhaps even for illicit use. I checked his story, and his tests. “Well,” I thought, “I owe it to him to examine him carefully if I am going to deny him pain relief”.

Sitting one to two meters away, I asked him to take his shirt off. There in the middle of the right side of his chest was a large lump – perhaps 10cm in diameter. I did not need a CT or any other test to know this was grossly abnormal. I am not an expert at reading CT Scans but I looked up the films. There it was, very obvious to the novice radiologist, and yet the CT Scan was reported as normal! How could that be? I rang the radiologist, who concurred – and then checking the previous CT scan revealed that it was present, although smaller, over a year before. No excuses. Both scans were reported as normal. Here I was, looking from a distance of a little over a meter (with my glasses on admittedly) and detecting something that was not picked up on two expensive scans and perhaps as many as 6 attendances for nursing and medical assessment.

Examining him further, he had widespread sounds in his right lung; evidence that the pain he was experiencing was preventing him breathing fully. He had had a couple of chest infections as a result. With a little advocacy he is now getting the care he needs.

I have many experiences where people want to be examined but have found it difficult to get someone to do so. I do wonder if the increase in post-graduate training for doctors, where older medical students already socialised as adults find it more difficult to cross the major social boundaries necessary to be an effective doctor. Added to this are the concerns of litigation for inappropriate intimate examinations and the rise of gender preference; all both excuse and press for avoiding examination. The sexualisation of children has led to even young prepubertal girls feeling uncomfortable having their chests examined. We can only wonder what it was that led to nobody exposing this man’s chest during examination.

On reflection I am proud that I touch people and want other doctors to feel the same comfort with doing a job well. Our patients want high quality care, even if a little reluctant at times to go through some embarrassment. A young woman with an urgent health concern passing through Alice Springs recently had been to the emergency department and wanted to be sure what was wrong. When I explained that examining her would help sort out what was wrong, and asked if she wanted me to examine her, she replied in a relieved voice, “Please”. An old Aboriginal man with limited English, when I explained to be sure about his prostate I would need to put my finger in his bottom, looked at me and my raised index finger and said while looking at my finger, “I want that one”.

I hope young doctors and nurses will realise how important it is to examine people, and I hope people seeking health care will also expect to be examined thoroughly, even demand it if there is clear value. Clothes need to come off to see the region and even the surface changes. Breast and genital examination are not required in many instances and always need careful explanation and consent if not requested. When the need is high, even major cultural barriers will come down to enable adequate health care. The rise of point of care ultrasound and testing in the hands of primary care practitioners will mean more people will have access to tests, but the eyes, ears, nose and especially touch of the practitioner are all rapid information gatherers of very significant value.

Alright, sure, certainly, indeed, by all means

Published on by Sam HeardLeave a comment

I was at a meeting in Alice Springs about 30 years ago in a session led by Ada Parry and Peter Thomsen, two pioneer Aboriginal cultural educators, who set a task for two large groups of doctors in a big meeting room. One group was told they were on an island which was about to be visited by a powerful nation. They had to erect a monument to show their own prowess and demonstrate the importance of their leader. The other group were told of the poverty and lack of education of the island people and were to go to the island to help. They were told to mingle and find out what the population really thought was the best thing for them to offer. My colleagues present will remember the insight we all gained into the difficulties that arise when people negotiate from different positions.

The resonance with Australia’s current experience with ‘the voice to parliament’ is extraordinary. This is not a new experience and one of which every Indigenous leader is acutely aware and it is potentially devastating.

All Australians can understand the “Uluru statement from the heart” – it is a strong request for recognition in the constitution and finishes with “In 1967 we were counted, in 2017 we seek to be heard.” It was the result of a massive national effort to consult with all Australian First Nations’ leaders and communities culminating in a meeting at Uluru and the creation of this historical document. It was delivered to the Prime Minister, Malcolm Turnbull, with pomp and grandeur, but it was not accepted using excuses that still echo today. Mr Turnbull is now advocating for Aboriginal and Torres Strait Islanders to be heard, demonstrating that it takes time for all of us to really appreciate the importance of this small generous step. Changing our minds is always required for progress.

I grew up in a small country town in South Australia with Tony Dodd and Bobby Wanganeen in my class and their brothers and sisters in my school. Two proud families from a South Australian mission given passes to live in our town and work for local industry. I was 12 when these families were accepted as citizens and they moved on, back to their country. My ancestors settled the fertile land in regions of South Australia and my grandmother remembered only one older Aboriginal man working in the region during the second world war. My father, soon to be 100, remembers as a child, a man escaping to Adelaide on horseback to establish an alibi after killing an Aboriginal man. That’s it – no history, no story, no place in the region. But I can still sit on the abundant middens along the limestone coast, find flint knives and imagine the Booandik people feasting on cockles around big fires. Most First Nations’ people of that region died before Federation, without being counted, or heard.

Constitutional change is necessary to guarantee that our Parliament will consider the impact of its new laws on our First Nations people. Why? Because this sovereign group will forever be a minority in their own land.

We can be more generous to the people who have a connection with this country going back millennia. We can see people who live in remote Australia or our islands in the north, not as pioneers but as keeping what is uniquely Australian alive. We can see the imposition of law to stop the cattle being killed as understandable from a British perspective but complex and foreign to open range hunters. We can see the idea that a person can come with some wooden stakes and bits of paper that gives them the right to dig up country and blow up ancient monuments as strange and somehow unfair. Farmers are now experiencing this with fracking and are duly concerned. This ‘voice’ we are voting for (or against), might actually be good for all Australians who have established non-mining interests in rural and remote parts of our country. In the most urbanised country in the world, this voice may also speak for them.

This vote for ‘the voice’ has strong international interest in the press around the world. Our country is under scrutiny. It is seen as a sign of our inclusiveness, our maturity as a nation. I have never had anyone speak out that they voted against the 1967 referendum and opponents have quietly disappeared. This vote will also be seen as a turning point in our nation’s history – towards the future or away from it. It is about inclusion, safety and well-being. It is simply a recognition of First Nations people in our constitution and a requirement that their views be heard by Parliament.

We all have a choice on the 14th October 2023, approaching 250yrs since Federation. If you are uncertain about how you will vote, I hope that you might consider a generous response. One for the history of all our nations.

On air

Published on by Sam HeardLeave a comment

Been on a pocast…Dr Luke Crantock, Everyday Medicine

Episode 100 : Working at Central Australian Aboriginal Congress

Cheeky Docs’ “No Cola, Just fun” tour of Central Australia

Published on by Sam HeardLeave a comment
Cheeky Docs 2019 tour

A NT band of musical general practitioners toured Alice Springs and communities to the west serviced by Central Australian Aboriginal Congress. They traveled with Prof. Mike Lean who is pioneering reversing diabetes on a large scale in the UK through dietary interventions – the Direct Study.

Why have we got smooth muscle around our bronchi (tubes in our lungs)?

Published on by Sam Heard2 Comments

I am a general practitioner who hits “retirement age” in 3 days. I do feel qualified to ask this question but not really to answer it. It has been bugging me my whole career, but I have realised most people never contemplate this backwater of physiology.

I always expected to ask someone in the know and get a clear answer. I asked professors of paediatrics, as these muscles seem to cause a lot of trouble in young people. Nothing. I moved on to professors of respiratory medicine, as a lot of their work relates to this biological machinery that can narrow our airways. They smiled, looked around quizzically and got on with their work. More recently I trapped a respiratory professor in a pleasant restaurant, but he was not really interested. He proposed that the muscles were embryonic vestiges of bowel muscle.

You may not be interested in my explanation either. But it is an important question. These muscles are in a critical location in our lungs. The inappropriate contraction of these muscles, called asthma, is a significant cause of death in adults. It tragically happened to one of my otherwise healthy young patients and has been a near miss on a number of occasions. During that huge thunderstorm in Melbourne a few years ago, these muscles went on a rampage. We have drugs to relax these muscles that are contained in a blue inhaler and used by many people around the world. We know a lot about these muscles and how they work.

“These smooth muscle cells have muscarinic M3 receptors on their membrane. The activation of these receptors by acetylcholine will activate an intracellular G protein, that in turn will activate the phospholipase C pathway, that will end in an increase of intracellular calcium concentrations and therefore contraction of the smooth muscle cell. The muscle contraction will cause the diameter of the bronchus to decrease, therefore increasing its resistance to airflow.”

https://en.wikipedia.org/wiki/Bronchoconstriction

Searching the internet, has provided no explanation. So, I will have a shot, excused by unrequited inquisitiveness and my age.

My only real experience of these muscles in action is wheeze, the symptom that is caused by their contraction. The noise is generated by the air we breath going through a narrow tube. If you take a deep breath with your mouth open you will hear a soft noise, if you do it again with your mouth shut, you will get a lot more noise in your nose. The same amount of air will make higher pitched noises as the tube narrows. When the tubes get narrow enough they start to hum and whistle and we can hear a “wheeze”.

Living in London for a decade, I noticed that children wheezed a lot. It was usually with a viral infection and was almost always worse at night. It seemed relatively harmless for most, but a few had difficulty. I treated them with the medications we had available, and noticed that most of them settled down as they got older.

I imagined that the viral infection was causing some swelling of the tubes and narrowing them somewhat. I could explain the wheezes disappearing as children got older because the tubes of their lungs were getting bigger. The deterioration at night, something that was particularly evident with croup, was not so easy to explain.

Traveling piqued my interest in this phenomenon. In 1981, I got off a train at Linkoping in Sweden at 1am in the morning and the temperature was minus 30 degrees. The hairs in my nostrils froze in a snap and I began to wheeze gently for the first time in my life. It didn’t last long as we walked knee deep in powder snow to our friend’s apartment. In the early 90s when we returned from London to live in Darwin our wheezy child stopped wheezing altogether. The air was cleaner and it was a lot warmer, especially at night.

Someone told me that the surface area of our lungs, as an adult, is the size of a tennis court. There is a tiny membrane between our blood, pulsing at 37 degrees, and the air we breathe. How do we possibly stay alive and not die of hypothermia? Especially in Sweden! The solution is that our body heats and moisturises the air between entering our nose and delivering the oxygen to our blood in the tiny sacks at the end of the muscular tubes of our lungs. When we breath hard, we bypass our nose, but we are generally exercising when we do this.

Could these muscles have a role to play in temperature regulation? It seemed quite possible. To warm air, it needs to be in contact with a surface, so the thinner a tube the more likely it is to warm during its passage. But the same amount of air passing through a thin tube would travel faster and have less time to warm. What if we didn’t need so much air? That is the case when we are not exercising, especially when we are asleep. Could this be useful when we are at rest, particularly when we are asleep?

The other effect of reducing the diameter of the tubes is to reduce the volume of the tubes. It is dramatic. If you halve the width of the tube, the amount of air being drawn in is a quarter. So for a small reduction, perhaps to 3/4 width, only half as much air would need to be drawn in through our nose to get the same amount of oxygen to our lungs. This sounds handy, less air, more efficient warming and same oxygen.

It seemed to explain why children get wheezier at night. I was told by my Swedish friends that all Swedish footballers wheeze for the first few minutes of the game in winter. Could this explain exercise induced asthma, where the airways constrict until the person is generating enough heat from exercise? Cold air is a well known trigger to wheeze. A person probably also has some swelling in the lining of the tubes when this happens; allergy or a viral infection. Clearly things go horribly wrong with this mechanism at times, causing severe asthma, but a lot of wheeze can be explained physiologically.

So what is my answer to what has become a rhetorical question?

The muscles around your bronchi protect you from excessive heat loss when you are not exercising, and particularly when you are asleep. You can get by at those times with a lot less air coming into your lungs, and so can conserve heat efficiently. You can survive and sleep in cold weather and cold nights without becoming hypothermic.

Now that sounds like evolution.

Whether or not you like the answer, you can start asking experts the question. I think you will be as bewildered as I am by their inability to answer. You should at least consider why that is so and what it says about our profession? No doubt some of you will search the internet. Please comment if you find the answer out there.

Looking after your own

Published on by Sam HeardLeave a comment

We are living in a moment when there is evident risk to every soul on earth. We worry for the people we love if not for ourselves. My good friend, young and healthy, has just spent a week in hospital and 4 days in intensive care in a personal battle with this virus. It is a depressingly passive experience lying with plastic lines and tubes, depending on oxygen, coughing discretely when others are around. All the while alone, removed from family and friends. Dying was a real possibility. We talked about it on the phone. The virus spread and wreaked havoc in her family. Her three children had both parents in hospital for a few days, all separated and unable to help each other. This was in Australia in a street near you.

Behind the face shields in intensive care, others hovered, their eyes concerned, concentrating but creasing kindly when looking into another. A gloved tap on the arm, and patience with the endless shuffling of bedpans required by a person whose dignity lives only in facing this dark unmeasured threat to their life. It is a kind of love. These caring people matter immensely. To all of us. Time will inevitably take quite a few of us to the edge of this slippery chasm held back by special people decorated only with gowns and gloves and masks. There will be no one else.

Meanwhile the children danced and played with some delight with a visiting aunt who, recovering from the disease, found herself taking care of energised children who were not allowed out. Their young bodies unaffected and one, despite coughing a little like the rest of the family, could not manage to give a positive swab to the persistent masked visitors in gowns. Like other children around the world, they were somehow protected from its power; the younger, the stronger even if afflicted with other conditions. Cryptonite of youth.

It is a wonder of this condition that the deaths that occur mirror almost perfectly the ‘natural’ rates of death in the community. It is for this reason that despite reaching 4,000 deaths in Italy, this virus had not taken a single life under the age of 30. The problem for all of us is that this ‘natural’ curve of human attrition is delivered over a period of 14 days, and the other health problems facing us do not go away. It is a massive impost on health care and society and a disaster facing us all. But it still has this compassionate element – to take we older folk in stark preference. We, who want nothing more than for our children to outlive us.

Health care workers, paramedics and aged care workers are now the farmers in the drought, the police on Saturday night and the firefighters of the summer. We are needed desperately, our attention, our care and compassion. We will be needed for sometime and it could be dangerous. Some are going to catch the virus. It is going to be tough, but it is going to be much tougher for the older members of our professions. The haunting image of an aged care home in Spain, abandoned by carers who may themselves have been sick, must guide our future.

I hear some of my colleagues saying they did not sign up for this, or they have a young family. This is stark. The younger you are, the safer you are. Society needs young health professionals to step forward and engage. Your older colleagues will be there with you but there will be no equality when the virus hits. This is a real test of civility, of commitment to society.

We sit under a cloud of adversity, wondering what will become of us, of our loved ones. Will we come through this, will we all survive? But we are your own. Please take care of us.

Could my death go viral?

Published on by Sam Heard1 Comment

Image by Meredith Reardon

I am afraid. My eyes fill a little, mostly at night when I consider my future, our future. I have lived a full life, more focussed on the world around me than my family. My lovely wife has channelled her energy to family and friends. Together we have stood easy in our space, even when apart.

We share some sadness; a fatal crash, estranged loved ones and health has become more fragile. We have a wealth of offspring who pluck our sense of being. We have always loved to live in the villages and spaces we moved in, with little concern for the future. We have had good parents, and siblings who have found their feet. My parents still go about their lives, playing bowls and music as they approach their centenary. My body is likely strong.

I have always anticipated my death a little, and felt vaguely cosy with the tears that I would share with my cherished sons and daughters and those grandchildren who had enjoyed exploring the opaque and cheeky thoughts of elders. Even the regrets that we didn’t do more together, that I didn’t show my love enough, or wasn’t there for some when I should have been, could be resolved in that final reckoning. If I was younger my one love in life would carry me, my family would lift my mood and surround me. Perhaps even some of the people I worked with would join the throng and replay forgotten moments. If I was older, the sheer number of offspring would sustain me and I would go easily.

With the news of this virus and my work amongst potential victims a new contagion has begun to haunt my dreams. It is bleak. What if my wife or I were to catch it and get sick? How would it be to need increasing support and finally, to know one of us could not recover? What then?

Would I die in an expensive shiny intensive care bed, my tubes and vital signs attended by masked and selfless strangers, fading into a world where a distant and sunless window occasionally showed a fleeting familiar face? Or less alone in a ward of others suffocating slowly while sharing our stoic denial or whispered acknowledgement of our very possible future? Perhaps I will be amongst a hoard in a warehouse, given what sustenance the conscripted carers can find. They may be the untidy recovered or people out to make some money to pay off recent debts.

Whatever the stage, whatever the course, we seem deemed to tackle this grim reaper alone. Any peace cut by sirens of the tiny oxygen meters on our fingers or the soon futile wailing of ambulances arriving at the barred doors below. Or the creaking of lines of makeshift beds and gasping neighbours. Alone in a sea of drowning strangers.

That is not what I want for me, for my family or anyone. Let me have a little of my own future, and the people I love. Even one to be with me will make me smile and wonder at what might have been and see inside another’s eyes. I think we can and we should.

But how?

Let households be the victim of this wanton infection, strong together within familiar walls. Let resilient children play with their snotty parents, and lovers lie with their fevered friends. They will joke and laugh and cough even as they feel a shared fear of the future. Take out the elders if they wish and test them, returning immediately if they have the little beast within and later if not. The children will wonder what the fuss is about and happiness will replace fear as each recovers, cared for by ones affected not at all or in a little way.

The home while coughing will need help to bring supplies. They will wave together through clean blue windows and call gratefully to the helpful neighbour who turns away. They might go outside in their garden or play music and sing in a balcony choir. Every now and then one will get very sick, and if there is somewhere better to be with a ventilator or new treatment, they can go. They can be visited there and kissed and hugged by their recovered family. And if all fails or they are making a slow recovery, they can go home if they wish to be cherished.

They say things will change with this king of germs. Perhaps it is a chance to discard the notion of health and even being as an individual experience.

Who knows who blows?

Published on by Sam HeardLeave a comment
Could we greet each other in a way that meant we kept 1.5m apart?
Photo by Guilherme Stecanella on Unsplash

Few of us blame the Government for this virus pandemic but on a dreadful day when Australia has twice as many new cases as China (and a 60th of the population) we have to ponder whether we have done enough. Peter Wener in the Atlantic points to a major issues in the USA such as, “… the decision to test too few people, the delay in expanding testing to labs outside the Centers for Disease Control and Prevention, and problems in the supply chain. These mistakes have left us blind and badly behind the curve, and, for a few crucial weeks, they created a false sense of security. What we now know is that the coronavirus silently spread for several weeks, without us being aware of it and while we were doing nothing to stop it. Containment and mitigation efforts could have significantly slowed its spread at an early, critical point, but we frittered away that opportunity.” It sounds a lot like us. Five hundred people getting together at the moment is NOT safe. It is very hard to get tested, and we are being told it may be more difficult.

We have seen TV presenters, film stars and politicians come down with the virus. These people certainly mix with more people than we do day to day, but there are not many of them. So if we extrapolate back from their numbers and determine the rate of infection we will arrive at figures many orders of magnitude above that being reported. Let’s face it:

It is infectious, it is deadly and we need to stop it.

Economic predictions of impact are unlikely to be accurate and should not be listened to. Minimising the spread of the virus will have the most positive economic impact. Short, sharp, severe. Norman Swan has been getting air time to promote more health of the nation approaches. If we stop the growth, we can reopen schools and kindergartens. We can probably keep schools open now for children of emergency workers, health workers and other services that are understaffed.

Most of all we need to consider testing. At the moment we are testing a lot of worried people and groups of people who are in contact with an infected case. If we take a community view rather than an individual view and extreme social distancing is in place, we can check one member of a potentially infected household with symptoms and declare it positive or negative. That household should then be supported at home. Children are likely to be infected very quickly and have no difficulty. The disruption of treating individuals in this situation is massive and quite likely to lead to asymptomatic people thinking they are clear. Household members over 70 in infected households can be removed if asymptomatic and offered alternative accommodation if they do not have the virus after brief quarantine and testing.

Small towns with limited movement in and out can test the first locals with fever, and no travel or contact, as sentinel cases. If the sentinel febrile person is negative then we can assume safely that other fevers in that community are almost certainly negative for a period – perhaps the next seven days. This will be especially important when we add the usual winter viruses to the mix. Obviously any new arrivals with symptoms must be tested.

Regions, such as Central Australia, Cape York and Northern Western Australia (and many others) lend themselves to a regional community approach with community-based testing. The experts on our nation’s 60,000 year old cultures are at grave risk and require extreme measures of preservation. This requires checking all arrivals into the region (which is often one or two roads and an airport or two) and testing any of these who are symptomatic or develop symptoms over the next 14 days. We can tell them when we check them at the regional boundaries. In two weeks they become part of the “viral control” community. We will need sentinel testing in communities, perhaps one or two tests per week. The extraordinary thing in Central Australia is that this will require less than 100 tests a week. If we can “bank” the saved tests, we will be able to carry out contact tracing in the way Singapore has done.

If we are short of tests, lets think about how to use them effectively. It is time for extreme social distancing. Whatever we are putting in place, we need to know where the virus has spread.

The Authority of Colour

Published on by Sam HeardLeave a comment

I have been moved by the wonderful Toni Morrison, one of my thought mothers, to consider racism from the oppressors point of view. I am a white man, tall, middle class and educated. I work as a doctor of medicine, a powerful role in our country. I have been rewarded with an Order of Australia Medal, probably by people a lot like me (but I do not know). I am ‘in the driving seat’, so to speak, but I have chosen, with the relentless encouragement of Mimi, my ‘specialist in life’, to drive off the track.

I find myself in a position where Aboriginal people employ me. I have two levels of Aboriginal executive above me (who are both women) and the organisation is controlled by an Aboriginal board. I mention this as it is not a common environment for someone like me to work within. It is a very well run organisation, the best I have experienced, very motivating and with great working relationships. It is challenging and I feel good about it.

So Toni has asked for people ‘on the other side’ to describe what racism is. My first piece on this difficult topic was with Waiting for the world to change. I am interested in how the presence of white people changes things in my setting and if this is how institutional racism works. This is not about how redneck we are, or how vile. Rather, how there is an accepted code that gives white people more authority in social settings. There are a couple of subtle ways I have seen this being played out.

Aboriginal people in the Northern Territory use words in everyday discussion that are not so common or acceptable in mainstream culture. I did not notice this when I first came to work in Central Australia but now I do. I have used the ubiquitous terms ‘gammon’ and ‘budju’ for sometime having raised a family in the NT. I now use swear words a lot more at work than I did. I have become reasonably comfortable with the work “cunt”, not when it is used in hate, but in many contexts. Budju is also a word for vagina but is now used by all Territorians to describe very attractive people of both sexes. I am sure this word is acceptable in any setting.

Like my Aboriginal friends and colleagues, I do not say things that might offend when there are too many white fellas about. I wondered about this; about why Aboriginal people change their language when talking to non-Aboriginal people? Why Aboriginal people might stop being Aboriginal when a white person comes close?

Toni Morrison wrote:

“I have always myself felt most alive, most alert, and most sterling among my own people. All of my creative energy comes from there. My stimulation for any artistic effort at all originates there. The compulsion to write, even to be, begins with my consciousness of, experience with, and even my awe of black people and the quality of our lives as lived (not as perceived). And all of my instincts tell me that both as a writer and as a person any total surrender to another culture would destroy me. And the danger is not always from indifference; it is also from acceptance. It is sometimes called the fear of absorption, the horror of cultural embrace. But at the heart of the horror for me is what I know about what the history of the culture that pervades this country have been.”

Toni Morrison “Hard, True and Lasting” in A Mouth Full of Blood.

Lukas Williams, the wonderful young change advocate from Gan’na Healing, told me a story about one meeting he attended in a desert community where he was living. He was talking and projecting slides onto a make-do chipboard screen nailed to a post. The large group of local men who attended these meetings were chatting around the fires for warmth. Lukas put up a drawing of Captain Cook landing on Australia’s shore and waited until people grew silent. He started to talk about sovereignty. As the talking slowed to quiet, a large rock came flying through the air and smashed through the screen leaving a dark hole.

There was an immediate hush: uncomfortable silence. A snigger echoed in the far back of the group. Then more, some now laughing. Then, with the explosive urgency of freedom, a massive outburst of laughter filled the air. It rang and rang and echoed off the nearby mountains.