In recent Observables Group calls we have discussed classification of the LOINC Terms with "body fluid" and if the hierarches such as below are to be expected:
Information from the section "Body fluid" in the LOINC Users Guide is at the end of the message. As a follow up to recent discussions, we asked the Regenstrief Institute if Terms with "Body fluid" are abstract classes and if these hierarchies to be expected. They responded that:
"Body fluid concepts in LOINC should not be parent concepts for similar codes with different specimens, but rather should be at the same level. The body fluid codes should be used for tests run on less common specimens for which a more specific code does not exist. So using your serine example, all of the following terms would be at the same level, and 32270-1 could be used to report serine in synovial fluid, for example, because this would be an unusual case for which there is no existing LOINC code.
"
We need to discuss the implications of this and any potential changes in modeling that are needed.
------
Information from the section "Body fluid" in the LOINC Users Guide:
The name for the LOINC Part “Body fld” is actually “Body fluid, unsp.” According to the LOINC Users Guide, it is used as a grouper for when the fluid type is not explicated stated in the database, as well as for when a test is not standardized based on location. Excerpt from the Guide: “For many chemistry tests we have included in the LOINC database a test name for identifying miscellaneous types of body fluid (Body fld), to provide a way to distinguish tests that are performed on fluid types that are not explicitly represented in the database. We use “XXX” in the system to identify a material that is unknown or not specified — it could be solid or fluid, for example. XXX is also used in the system when the material is specified, but recorded elsewhere in the HL7/ASTM message. Yet, using XXX as a system can be problematic (see also “Special issues related to XXX as a system” below).”
When should we lump a variety of specimen types under the nonspecific “Body fld” and when we should give a body material its own unique name for a given component? The decision depends upon the degree to which laboratories have reported the system-component pair as a separate “result” and the degree to which the normal ranges for a given component-system have been standardized. By this rule, we will always define different tests for serum and for urine, when a component can be measured in both. We define sweat sodium as a distinct test because it is a standardized test used to diagnose cystic fibrosis. We did not define duodenal fluid sodium as a separate LOINC code because this measure has not been standardized. This does not mean that the specifics about the system would be ignored. It just means that this information would be recorded in another field of the message (the specimen field of the HL7 OBR segment), not in the name. Generally, we will specify the type of system to distinguish at least among blood, urine, cerebrospinal fluid, pleural fluid, synovial fluid, and peritoneal fluid.”