Terms: Coded Elements

A term is a coded element. Every coded data property in HIEBus is represented by a term. For example, elements like gender, race, patient class, and patient type are stored as terms in HIEBus.

A term is identified by a code, a namespace, and its context. A specific code in a context within a given namespace is assumed to represent a unique clinical concept. I.e. the same code in a namespace should not be reused to have different meanings in the same context. This does not preclude, however, the code in the namespace having a different meaning in a different context. For example, a race code of “B” would mean something completely different than a gender code of “B.”

HIEBus utilizes term mappings to map terms to their canonical or standard equivalents. In many cases, a new term will automatically map to its equivalent based on conventions. For example, a new gender term of “M” will be automatically mapped to the equivalent standard term representing male. In some cases, terms will have to be mapped manually, however; a gender term of “I” for instance cannot map automatically to any of the available standard genders.

HIEBus also utilizes term subsets to allow grouping of terms that represent clinical conditions. For this reason, it is important that attention is paid to the term namespaces in use by the HL7 interface.

The HIEBus term namespace is used to provide scoping to terms. It is equivalent to the “coding system” used in the HL7 CNE and CWE data types. Term namespaces in HIEBus may be canonical such as ICD10, RxNorm, or SNOMED-CT, or they may represent custom or vendor-specific coding sets.

Term Namespaces

Namespaces scope terms to ensure that a code received from one source is not equivalent to the same code received from a different source. Term mappings are used to map terms to their canonical equivalents. All HL7 drivers have a default namespace code set up for them, which will be used when no other namespace is available. For cases where a single HL7 feed is aggregating messages from multiple sources, the MSH-4 Sending Facility can also be incorporated into the default namespace code to ensure that codes do not clash from two sources in the same feed.

HL7 also provides the ability for sources to specify the coding system in use for a given code. When a coding system is specified, the coding system is prefixed by the default namespace code before being persisted. An exception is made for the coding system value of “L” which represents a local coding system.

HL7 coded entry mapping to HIEBus term

Coded entries of the type CNE and CWE are mapped to HIEBus terms as follows:

Example

Consider the following message:

MSH|^~\&|HL7REG|UH|HL7LAB|CH|200102280700||ADT^A08^ADT_A08|MSGID002|P|2.7|||AL|NE
PID|1||PATID1234^5^M11^ADT1^MR^GOOD HEALTH HOSPITAL~123456789^^^USSSA^SS||EVERYMAN^ADAM^A^III||19610615|M||2106-3|2222 HOME STREET^^GREENSBORO^NC^27401-1020|||||||ENC1234
PV1|1|I|2000^2012^01||||004777^ATTEND^AARON^A|||SUR||||DR^PHYSICIAN OR SELF REFERRAL^^PR^Physician Referral(Or Self)^L^^X^Referred by PCP^1^Physician Referral^USH|A0|

If the message is coming through an HIE sourced from multiple source systems, then namespaces will be prefixed with both configured term namespace prefix for the HIE, “USHIE” in this example, and the source facility value in MSH 4.1. The encounter will be given the following admit source values: