• What are Concrete Domains?

• What is Description Logic (DL)?

• What is the Machine Readable Concept Model (MRCM)?

SNOMED CT is based on relationships between concepts, structured and constrained by the Machine Readable Content Model (MRCM), and represented and distributed using the RF2 release file format. The OWL-based Description Logic (DL) supports the use of classification tools to infer relationships that are not directly stated in concepts, and hence provide additional meaning and context, however there are challenges when concept attributes need to express fixed values.

In the case of numerics - for example, the weight of a particular medicinal substance used in a drug product - it is desirable to be able to represent that amount, and then be able to ask questions like "List all drugs that contain >= 100mg of Paracetamol".

A work-around for this particular attribute was put in place in 2017, namely to create a concept to represent every number needed in the International Drug Model. This solution results in fully correct classification, and in particular has facilitated progression of the Drug Model implementation, however, it also has several drawbacks, particularly when needing to make comparisons and check for equivalencies.

The drawbacks of using concepts as numbers can be resolved by defining concrete values within SNOMED CT in a way that still fully supports classification to reveal inferred relationships, whilst enabling numeric operator comparisons and (subject to appropriate modelling) equivalences between different units to be detected.

Concrete Domains are a means for the SNOMED CT DL to express concrete objects (such as numbers) and predefined predicates on these objects (such as numerical comparisons) when defining concepts, in a way that still supports classification and inferred relationships.

The Authoring Platform (AP) supports the following concrete value data types:

• Integer

• Decimal

• String

The can apply the following types of constraints to concrete values assigned to attributes:

• List (predefined fixed values)

• Range (any numeric value between predefined limits; applicable to Integer and Decimal data types)

Attributes previously using concepts as numbers have the same FSN term when represented using concrete domains, however, the target value box has a green background, and the concrete domain values don't have a semantic tag since they are no longer concepts.

Drag and drop is supported for concrete values, so the "drag handle" arrows icon is still available, however concepts dragged from other fields, search results or taxonomy entries cannot be dropped as the target for a concrete domain attribute value, and likewise a concrete value cannot be dragged and dropped onto a concept target slot. Dropping a concrete domain value onto other concrete domain attributes will also be disabled if the drop value doesn't satisfy any MRCM constraints applicable to the target attribute.

A concrete value is represented by a solid green rectangle with diagonal lines in the corners, and the value centred within the box.

Numeric values are represented in Unicode text expressions preceded by a # (hash) symbol. For example:

• #0.09

• #0.5

• #99.99

• #1

Strings are represented in Unicode text enclosed by double quotes. A backslash is used as an escape character for quotes within the string, and hence any backslashes within the string also need to be escaped by an additional backslash.

• "Product Name"

• "Product Name \"containing\" quotes"

• "Product Name \"containing\" quotes and a backslash (\\) character"

Since concrete domains are simply a form of attribute target value, as well as the above examples from the Task - Concept - Edit Panel they can appear in any feature of the Authoring Platform that presents attributes and their values, for instance:

The Authoring Platform Terminology Server (Snowstorm) maintains SNOMED CT content in a controlled hierarchical structure of project and task branches originating from the mainline (or simply MAIN) branch.

MAIN carries controlled changes to SNOMED CT content in the current authoring cycle, which is periodically used to generate a new SNOMED CT International Release. It contains a set of project child branches, which can promote content for inclusion in the next SNOMED CT International release.

Each project branch contains any number of task child branches, which can promote content to the project level for sharing and synchronizing across the other tasks. Project branches can be synchronized (rebased from MAIN) with content updates from other projects which have promoted their changes to MAIN.

Task branches can be synchronized (rebased from project) with content updates from other tasks which have promoted their changes to the parent project.

The Machine Readable Concept Model (MRCM) represents rules in the SNOMED CT concept model in a form that can be read by a computer and applied to test that concept definitions and expressions comply with the rules. The MRCM may be used for a variety of purposes, including the authoring and validation of SNOMED CT concepts, expressions, expression constraints and queries, Natural Language Processing and terminology binding to support semantic interoperability.

The MRCM may be useful in a number of use cases, including:

• Development of precoordinated terminology content.

• Authoring and validation of SNOMED CT expressions, constraints and queries.

Description Logics (DLs) exist to enable a reduction in content maintenance costs with a concurrent increase in content quality. Processing power and reasoning algorithms have improved such that a more expressive DL is now practical to use in SNOMED CT.

The principle benefit of more DL features is the ability to represent content in a more fully systematised and thus more machine processable way, to enable the following benefits:

• increased efficiency, consistency, precision and quality of authoring through reduced manual effort to organise content

• lower costs to author and maintain content through reduced manual effort

Service Acceptance Criteria (SAC) are used to assure that authored SNOMED CT content meets specific quality measures before the content can be promoted. The purpose of this is to ensure a specific level of quality is met at each promotion level. This in turn assures the quality of the content that is eventually promoted to the code system branch, with the benefit of minimising the quality checks required at that level (in general, the later the stage of the content release process, the greater the amount of work needed to correct any issues, so catching issues earlier in the process has many benefits).

SAC may be either automated or manual.

Automated SAC are checks which the system can sign off automatically without user intervention.

Manual SAC are checks that require sign-off via a particular user interaction, with that interaction governed by role-based access controls.

SAC may also be mandatory or optional.

The SNOMED CT Reporting Platform (RP) is a gateway to reports which can run a range of queries against content created or updated in the active authoring cycle for SNOMED CT.

Queries are structured into categories for ease of definition, navigation and retrieval, such as Ad-Hoc Queries, Release Validation and Quality Improvement.

The results of these queries can support pre-release validation of issues not supported by the Authoring Platform (for instance, MRCM rules such as attribute cardinality).

They can also be used to analyze the structural quality of SNOMED CT, to identify potential issues (for instance, attribute usage for a given sub-hierarchy) which can then be addressed via content change projects in the Authoring Platform.

For content authors, it enables iterative work to define complex QA rules that are eventually used for the Release Validation Framework (RVF) assertions (such as Case Significance).

As a content release planning and tracking tool it can show the number and types of changes made during a SNOMED CT release cycle, and from one release to the next.

• It can reveal trends and particular faceted searches of SNOMED CT content created through the SNOMED CT Authoring Platform that are not easily obtainable via other tools and services.

• It can help with quality assurance review of promoted content prior to publication in the next SNOMED CT release.

• It can help to quickly modify and adapt rules which can enable more rapid SNOMED CT content updates via Authoring Platform authoring projects/tasks.

• Anytime you need to query the content within the current authoring cycle for SNOMED CT.

• In particular, many of the reports are designed to help with content quality assurance reviews before creating the alpha-release version for each SNOMED CT release cycle.

• RP queries run against the content which has been promoted to the mainline branch of the Authoring Platform (ie promoted project content), so some coordination with project content team leaders may be advisable if your report is influenced by or needs to include specific project content for analysis.

• The Reporting Platform is linked from the SNOMED CT Authoring Platform (AP) header drop-down menu (available on all pages within the Authoring Platform).

• The International AP can be bookmarked and opened with:

• The RP can also be bookmarked and opened directly with:

These links will redirect you to the login page if you’re not already logged in within your current browser session. Once your login completes, your browser will be redirected on to the AP Dashboard and RP main page respectively. If already logged in, these pages will load without the login redirect.

Anyone who has a use for the results of queries run against content updates for the current authoring cycle through the SNOMED CT Authoring Platform, such as (but by no means limited to):

• Content authors

• Content project managers

• Content team leaders

• Consultant terminologists

RBAC is an access-control mechanism defined by mapping user-assignable roles to related privileges and features, governed by three core aspects: role assignment, role authorization, and permission authorization.

The Authoring Platform uses RBAC to authorize an authenticated user account with permissions to perform particular actions within the platform, defined by role-group mapping and membership.

The current roles defined on the AP can be considered in three groups:

These roles are assigned to particular access control groups, which are mapped to global or project level content branch permissions on the Terminology Server. The technical support and platform administration, and SNOMED CT release control roles are outside the scope of this AP user guide, but mentioned here for completeness. When an account login is authenticated, it is then assigned AP feature access and permissions according to the roles defined for the groups of which the account is a member.