Technically, terminology, vocabulary, and code system are not synonyms, but the measure development community often uses these phrases interchangeably. The Blueprint content preferentially uses the term code system to describe the managed concept collections from which to draw value set content.

Code systems are a collection of concepts (ideas) with unique identifiers that exist in some sort of structure. The code system structure should provide each concept with a code-system-specific meaning, a concept identifier (a code), and a string description (the name, and a definition of the concept meaning). Code systems should ensure meaning permanence for all the concepts in the code system (Cimino, 1998). For example, if the meaning of the concept changes, the code system may need to retire the old concept and introduce one or more new ones to better characterize the meaning. This provides consistency in data analysis and retrieval over time. Some local environments define their own code systems, making sharing outside the local institution difficult. Successful interoperability is dependent on either using common code systems for data capture or through mapping the local content to an interoperable code system.

Code systems may have

• a language-dependent description
• very specific concepts or more general concepts
• complex ideas that include multiple, nuanced sub-elements such as the International Classification of Diseases, 10th Revision, Clinical Modification (ICD-10-CM)
• internal hierarchies built upon increasing specificity and may also include relationships among the concepts (e.g., caused-by or finding-site)
• a broad scope (e.g., SNOMED CT)
• a focus on a specific domain (e.g., laboratory tests for Logical Observation Identifiers Names and Codes [LOINC], medications for RxNorm)

Many code systems overlap in coverage (e.g., ICD-10-CM and SNOMED CT). When they do, the overlap may not result in simple one-to-one mapping between the concepts. Most code systems have an area of focused use that tends to shape crafting of the concepts and the relationships among these concepts. For example, the focus of ICD-10-CM is on disorders that cause mortality and morbidity. ICD-10-CM categorizes the disorders into unique groupings such that any single disorder will always be associated with only one ICD code and this categorization is useful for health care billing. Other code systems are multi-hierarchical such that the concepts capture multiple nuances and serve multiple purposes.

A code system authority, such as SNOMED International for SNOMED CT, should manage a code system. The code system authority is responsible for ongoing maintenance such as updates and corrections, and for content coherence and consistency.