9.1 Additional attributes of LOINC terms

9.1.1 Names

In addition to the Fully-Specified Name (FSN), which is the 5 or 6 major Parts of a LOINC term concatenated together using the colon character as a separator, each LOINC term has several different names that have been added as LOINC has evolved. Each of these was created for a different purpose and presently, they are generated programmatically with manual curation in special cases. We began with the Short Name and Long Common Name in 2002 and 2009, respectively.

Even after creating the Long Common Name, we continued to receive periodic requests from users to produce “prettier” display names for use in user interfaces, etc. While systematically created LOINC Long Common Names can be guaranteed to be unique (except for truly duplicate concepts), they are sometimes not the most user-friendly. This is also true for our automatically-generated LOINC Short Names. However, mostly what users wished for were names that leave out some of the defining attributes of the term (e.g. Property, Scale, and “default” specimens in the System) because they are unfamiliar to clinicians. This is problematic across all of LOINC because it would create ambiguous and duplicate names. For such reasons, we have always expected that users would link their own local preferred names to LOINC terms for use in reports and displays. Within a given local system context it may be possible to resolve the ambiguities created by names that omit these attributes. For example, if the user display always shows the reported units of measure, then having the Property in the display name may not be necessary.

However, we continued to think about how we could create a more “friendly” name and with the input of the LOINC Community, we developed the Consumer Name and Display Name as described below. We welcome feedback from anyone that has implemented or plans to implement these names.

9.1.1.1 Short Names

In August 2002 we added the Short Name, which is a short, mixed case name for the LOINC concept. All laboratory and clinical LOINC terms have a Short Name. (Many survey LOINC terms do NOT have a Short Name). Our goal was to produce names no longer than 40 characters in order to fit within the space allocated by most laboratory reporting systems. In contrast to the LOINC FSN, case is significant in the LOINC Short Name. When possible, we have used common acronyms and common names rather than the more formal name rules of the FSN. For example, we used the English names of allergens in the short names rather than the formal Latin species names (in part because they were shorter). We have used all upper case to represent acronyms, and mixed case in organism names as specified in naming conventions (e.g., genus is capitalized, species is not). For virus names we used the acronym assigned by Index Virum, where available.

The LOINC Short Names are subject to change and should not be used as identifying keys in any database.

9.1.1.2 Long Common Names

In 2009, after collecting and reviewing display names from several sources, we began releasing the Long Common Name. These names are also created by an algorithmic process and are checked for uniqueness. Most abbreviations and acronyms that are used in the LOINC FSN have been fully spelled out in English. For allergens, the common English names are used instead of the more formal Latin species names. For coagulation, the more commonly used phrases, such as “Prothrombin time”, have been used.

We started creating Long Common Names first for laboratory terms, but are now producing them for all terms. The text strings for the long common names are subject to change over time as we continue to refine the algorithmic process and collect feedback from users. In particular, many of the Long Common Names for clinical terms have not had as intense focus as the laboratory terms have, so we expect these to be refined over time. In LOINC release 2.54, many of the echocardiography and ophthalmology were updated based on input from DICOM and the National Eye Institute, respectively. Through the LOINC 2.65 release, every LOINC Long Common Name was unique. This was true even for deprecated duplicate terms, some of which had additional periods appended to their names in order to make them unique. However, the LOINC Committee recommended that concepts that are truly duplicate (i.e., have the exact same values for each of their major Parts) should have the same Long Common Name, and we have implemented this change starting with the LOINC 2.66 release.

9.1.1.3 Display Name

Beginning with LOINC version 2.64, we began publishing Display Names, which were developed as a prototype for a more “clinician-friendly” set of names compared to the current LOINC Short Name, Long Common Name, and Fully Specified Name. They are created algorithmically from the manually crafted display text for each Part and are generally more concise than the Long Common Name. Currently only LOINC Laboratory terms have Display Names, though we may add Display Names for Clinical terms in the future.

In general, Display Names are unique for a given concept; however, in the case of truly duplicate concepts (i.e., have the exact same values for each of their major Parts), the Display Name will be the same. Unlike the Short Name and Long Common Name, the Display Name does not include an indicator that a term is deprecated.

Some of the high-level rules used to create the Display Names include:

• Components are shortened where possible using common abbreviations and conventions such as including only the first letter of the genus when a Component includes both the genus and species;
• Systems Ser/Plas and ^Patient are implied, and all other Systems are included in parentheses;
  • Exception: for terms with the Property Susc, the System is implied
• For Timing Pt is implied and all other Timing values are specified;
• If the Scale is Qn, the Property value is included and Scale is not;
  • Exception 1 – for the Properties ACnc and Likelihood, the Property is left off and Scale is included
  • Exception 2 – when the Component contains words that make the Property redundant, the Property is left off
• If the Scale is not Qn, the Scale is included instead of the Property;
  • Exception 1 – in a few cases (e.g., ID, Imp) where the Property conveys specific information about what result to expect, the Property is included
  • Exception 2 – Nom is implied

9.1.1.4 Consumer Name

The Consumer Name attribute was originally hand-crafted for a very small subset of LOINC terms. At its inception, these names were labeled as “experimental”. In 2019 we undertook a concerted effort to develop a more comprehensive set of consumer-friendly names, primarily based on the increasing availability of consumer-facing health and wellness apps that take advantage of existing LOINC codes for the underlying data.

Consumer Names are created based on manually crafted consumer name text for each Part. In contrast to the other names, this text may represent a more general concept for the defining Part. These names do not represent each of the defining Parts of a given term, are not unique, and do not include an indicator that a term is deprecated.

Some of the high-level rules used to create the Consumer Names include:

• The specimen is placed at the end of the name and is prefixed by a comma;
• Almost all intravascular specimens are labeled Blood, with a few exceptions such as Dried Blood Spot;
• Property, Scale, and Method values are mostly excluded. This results in many duplicate Consumer Names and is based on the assumption that the result value and/or the units of measure will provide context for the name;
• Abbreviations are not used except in select cases, such as HIV; however, some abbreviations are included in parentheses for education purposes.

The current collection of Consumer Names have a “pre production” maturity level and we welcome user feedback about them. Until they reach production status, we plan to release them in a separate file rather than in the existing Consumer_Name field in the LOINC table.

9.1.1.5 Consumer Name in Linguistic Variants (Alpha)

Beginning in LOINC 2.83, a new ConsumerName column has been added to the linguistic variant accessory files, appended at the end of the existing column set. It is populated with consumer-name translations submitted through the translator portal, beginning with the Dutch (nl_NL) linguistic variant from Nictiz.

This field is Alpha status: provisional, subject to change, and not yet under formal editorial governance. It is not indexed by Search LOINC and is not shown on term detail pages in this release. Consumers who parse the linguistic variant files by column name are unaffected by its addition; consumers who parse by fixed column position should account for the new trailing column.

9.1.2 Classes

We assign each LOINC term to a general category called a Class. These categories are relatively broad and are intended to make it easier to sort and browse the database. They are not intended to be binding definitional characteristics of the term, and we may refine them over time. A more detailed listing of the Classes is presented in Appendix B. Throughout this document many of the naming conventions and approaches are described in reference to a Class of terms. Here we provide a bit of explanation about some of the laboratory term Classes.

The Class of Microbiology includes all tests used to identify microorganisms and evidence for infection by specific organisms as well as cultures direct microscopic exams that identify organisms or prove evidence for present or past infection with specific organisms. Microbiology includes tests for antibodies, antigens, DNA and RNA. The Serology Class does not include measures antibodies or antigens related to microorganisms. Molecular pathology Class does not include RNA or DNA based tests for infectious organisms. (They are all included in Microbiology.)

The Class Blood bank includes all blood bank testing including ABO-Rh testing. Allergy Class includes testing for antibodies to allergens (cat dander, trees, etc.). Serology includes rheumatological, and autoantibodies, and antigen measures not covered by these two classes. Hematology/cell counts exclude coagulation studies that are found in a separate Class. Measures of complement activity are included within Hematology, not Chemistry. Chemistry does not include challenge tests such as Glucose tolerance, ACTH stimulation, etc.; these are in a separate category called Challenge tests.

9.1.3 Categorizing LOINC terms as Order, Observation, or Both

LOINC terms can be matched to local test codes that fulfill different roles. For example, a LOINC code can be used to place an order, to report discrete observations, or in some cases a LOINC term can serve in both ways.

The purpose of the ORDER_OBS field in the LOINC database is to provide users with an idea of the intended use of the term by categorizing it as an order only, observation only, or both. A fourth category, Subset, is used for terms that are subsets of a panel but do not represent a collection of items that is known to be orderable. Our categorization of a term in this way is not a normative or binding resolution. That is, it is not a definitional attribute of the term. Rather, it represents Regenstrief’s best approximation of how the term is used. (If you feel a term should be categorized a different way, please let us know.)

In HL7 messages, the place in the message where the LOINC code is used will indicate what role the LOINC term is fulfilling. If it is in OBX-3, it is being used as an observation. If it is in OBR-4, it is an order.

Examples of each category include:

Order Only

34531-4 Blood type and Crossmatch panel - Blood

This term represents an order for a type and cross that would never carry an answer (i.e., result value). Panel terms typically fall in the Order only category.

Observation Only

51890-2 Factor VIII units given [#]

This term is a discrete observation that would carry a result of the number of units given to the patient, but would never be placed as an order for ‘Factor VIII Units given’.

Both

882-1 ABO and Rh group [Type] in Blood

This term could be used to place an order as well as report the result of the test (e.g., A Positive).

9.1.4 Associated observations

In the LOINC Table, a LOINC term can be linked to one or more optional “associated observations” in the AssociatedObservation field. Associated observations are additional variables that may be reported along with the primary observation. Within LOINC, we can link primary observation terms to other single LOINC terms or panels containing a set of terms.

There are three main use cases for linking LOINC terms to other associated observations:

1. Reporting information related to the primary result that otherwise might be reported using a comment field, such as the volume of urine in a 24-hour urine collection or specimen weight;
2. Including key data in a structured, computable format when sending an otherwise unstructured text report or document;
3. Identifying the sections and entries from HL7 Clinical Document Architecture (CDA) implementation guides that can be reported with certain LOINC document codes.

Associated observations can be used in conjunction with lab results that are typically sent in an unstructured format, such as genetic testing results. We encourage laboratories to report as much data as they can using the discrete variable LOINC codes so that the data are more directly computable. For example, associated observations linked to a BRCA1 gene targeted mutation analysis may include concepts such as gene identifiers, the reason for genetic testing, genetic sequence variation that was found, genomic reference sequence, and specific version of the coding system that was used to report the results. At the same time, laboratories can still retain the flexibility of sending text reports that can be formatted for readability.

Many LOINC document codes can be used with one or more related HL7 CDA implementation guides (IG). To facilitate identifying which sections (and sometimes entry level codes) may be used with a particular document code, we are working to link a panel of associated observations for each IG.

9.1.5 Ask-at-order-entry (AOE)

Ask-at-order-entry (AOE) observations in LOINC are observations obtained from the requester as part of the test order and generally delivered back to the requester as part of the result package. For example, the concentration of inspired oxygen is always an AOE question for blood gas measurements as well as the date of last menstrual period for pap smears.

Within LOINC, we can connect AOE observations to the primary test using a similar structure as described for associated observations. The linkage between the primary observation and AOE observations is stored in the AskAtOrderEntry field of the LOINC Table. We have not implemented AOE observations extensively yet but have the mechanism in place, and we are working with multiple organizations to target specific domains, including infectious diseases and other public health domains such as lead testing.

9.1.6 Related names

The LOINC database includes a field (RELATEDNAMES2) for related names associated with LOINC terms. We provide these associations as a service to LOINC users to help them find the concepts they are searching for. We often colloquially use the word “synonym” when referring to related names, however, most of the terms in the RELATEDNAMES2 field are not true synonyms. Values in this field may include common names, abbreviations, previous nomenclature, common misspellings, allergen codes, and associated diseases. This information is intended to help users locate relevant concepts, but it may also result in “false-positive” search results due to overlap in the related names associated with different concepts. For example, one of the related names for Influenza is Flu, so searching for Flu will return LOINC codes relevant to Influenza. However, Flu is also associated with Fluid, so the same search for Flu will also return LOINC codes with the System of Body fluid, Amniotic fluid, etc.