ValueSet

The primary function of a ValueSet is to act as a binding between a data element in a questionnaire or profile and a specific set of allowed codes. This ensures that when a clinician selects a diagnosis or a lab test, the code is drawn from a controlled, authoritative list. The binding strength can be required (must use codes from the set), extensible (must use codes from the set if applicable, but can add local codes), or example (a suggestion only). This mechanism is the cornerstone of semantic interoperability in FHIR.

ValueSets are not simple flat lists; they are built using powerful compositional rules. A ValueSet can be defined by:

• Include/Exclude Filters: Explicitly including or excluding specific codes or entire hierarchies from a code system like SNOMED CT.
• Intensional Definition: Using logical rules, such as 'all codes that are descendants of the concept Cardiovascular Disease'.
• Extensional Definition: Explicitly listing every single code in the set. This allows for the creation of both simple, static lists and complex, dynamically-resolved sets of clinical concepts.

A single ValueSet can aggregate codes from multiple, disparate code systems to form a single, unified list for a clinical context. For example, a ValueSet for 'Smoking Status' might include codes from SNOMED CT for clinical findings, LOINC for the assessment question, and HL7 v3 administrative codes. This agnosticism allows implementers to create a single binding point that harmonizes different terminological standards without forcing a single code system on all use cases.

ValueSets are designed to be consumed and processed by a FHIR TerminologyService. A client application can send a ValueSet's canonical URL to a terminology server and request operations like $expand (to get the full, resolved list of codes at that moment) or $validate-code (to check if a specific code is a member of the set). This server-side processing is critical for handling intensional ValueSets that may resolve to thousands of codes, offloading the computational logic from the client application.

ValueSets are versioned resources with a canonical URL, allowing for strict governance over their lifecycle. When a medical guideline changes, a new version of the ValueSet can be published, and systems can be updated to bind to the new version. This prevents the 'drift' that occurs when code lists are managed in spreadsheets. The ValueSet.expansion element can even contain a timestamped, point-in-time snapshot of the resolved codes, ensuring auditability and reproducibility of the code set used for a specific clinical decision or research query.

While ValueSets can be authored in raw JSON or XML, the FHIR Shorthand (FSH) language provides a far more concise and human-readable syntax. A complex intensional ValueSet can be defined in a few lines of FSH, which is then compiled into the full FHIR resource. This dramatically reduces the barrier to creating and maintaining high-quality terminology artifacts, enabling clinical informaticists to define precise code sets without needing deep expertise in the underlying JSON structure.