Whether you have been following the buzz related to the ARRA meaningful use criteria or you are involved in clinical information exchange, you have likely run across the acronym LOINC recently. A significant chunk of my career has been spent in the lab system environment, so I thought it might be useful for me to share my thoughts and experience on the lab domain and LOINC.
What is LOINC?
LOINC stands for Logical Observation Identifiers Names and Codes. It is maintained by Regenstrief Institute and, according to their website, it’s purpose is “to facilitate the exchange and pooling of clinical results for clinical care, outcomes management, and research by providing a set of universal codes and names to identify laboratory and other clinical observations.”
What LOINC is most commonly known for is lab test-related codes. In order to understand LOINC, you must first have a basic understanding of the Lab domain.
An Overview of the Lab Domain
When it comes to structured terminology, there are a handful of concept types in the lab domain that a healthcare information architect needs to worry about.
Lab orders are the codes that one entity sends to another when it is requesting that a test, or group of tests (also known as a Panel or Battery), be performed. An order is not a clinical entity, it is an administrative entity. Other than seeing which orders might be pending, no clinician cares about an order. You can think of an order as an expectation of clinical data.
Typically, a price is associated with an order. As a result, additional specificity may be required in the expression of an order so that it can be billed appropriately. For example, a test done using an automatic method is likely less expensive than one done manually. In that case, expressing the method in the term allows us to assign a higher price to the manual test, even if it does not make a difference clinically.
A panel, or battery, order represents a group of individual lab orders. Some of these, like a CBC or Complete Blood Count, can be fairly standard. Consider a typical CBC or Complete Blood Count panel:
[Order] Complete blood count (hemogram) panel in Blood by Automated count
[Result] Leukocytes [#/volume] in Blood by Automated count
[Result] Erythrocytes [#/volume] in Blood by Automated count
[Result] Hemoglobin [Mass/volume] in Blood
[Result] Hematocrit [Volume Fraction] of Blood by Automated count
[Result] Erythrocyte mean corpuscular volume [Entitic volume] by Automated count
[Result] Erythrocyte mean corpuscular hemoglobin [Entitic mass] by Automated count
[Result] Erythrocyte mean corpuscular hemoglobin concentration [Mass/volume] by Automated count
[Result] Platelets [#/volume] in Blood by Automated count
Sometimes a panel can represent a custom group of tests that are associated with a special group price. Very often, panels represent tests that are performed on the same instrument. It is always good to know if a given lab order represents a single order or multiple orders and which lab results you can expect to receive if the order is fulfilled.
Lab results are the codes that represent an analytical result and are returned when an order is fulfilled. A lab result is a clinical entity. It represents a piece of information about the patient that is important to the delivery of care.
It is important to point out a difference between lab results and other types of clinical terms. When you consider other common clinical terminologies, like current medications, allergies, procedures and problems, they represent a finite list of terms that tend to be holistic in nature (the term stands by itself). Lab results are different. The entity associated with the lab result is incomplete without the result value and unit. The actual ‘lab result’ is a dynamic composite comprised of the result code (which is a structured term from a finite list) and the result value, which may or may not have a unit of measure. The result code can be used to identify a result and group like results, but, by itself, it is not very useful.
There are many types of lab results and their structures vary accordingly. The most typical lab result is a simple analyte which consists of a result code, result value, result unit, reference ranges and abnormal flags.
Below are some examples of common analytical lab results.
These results were extracted from actual, anonymous HL7 lab result transaction records. The lab result names were the actual names represented by the result codes in the record.
A real lab result, encountered in transactions today, is comprised of the following parts:
- Lab result code or mnemonic
- Lab result term
- Lab result value
- Lab result unit of measure
- Lab result applied reference range
- Lab result abnormal flag
Of these, the Lab result term is the most dynamic in how it manifests itself in local and standard lab result terminologies. The lab result term is also the item that gets mapped to a LOINC term.
A lab result term is what is displayed or printed on a lab report when the lab results are presented to the patient or provider. The lab result term may imply a number of things about the context of the lab result. Some common attributes implied by a lab result term are as follows:
|Analyte||This is the specific characteristic that a lab result is measuring. This attribute is always represented in a lab result code in some form or another.|
|Analyte Context||Some lab tests may include words that apply additional context to the analyte. For example a result may have a peak and trough value. This context does not change the nature of the result but impact how it is interpreted.|
|Specimen Type||This is the type of specimen that was used to perform the test. This is often not explicitly stated in local lab terminologies if it is commonly known.|
|Time Specific Data||Some tests may be performed as a timed series or have time specific collection characteristics. If this is the case the lab result term may include applicable time information.|
|Result Type||This is the nature of the result itself. For example, a specific analyte may fall into a category where it can be measured either quantitatively or qualitatively. This may also be represented by a specific word like ‘ratio’.
Like the specimen type, this is typically omitted if the type is generally accepted for a given analyte. This can also be inferred from the result unit associated with the lab result.
|Testing Method||This is the method that was used to perform the test. In some cases the method can alter how the result is interpreted; in others it may provide a degree of certainty in the results validity. In most cases this information is not represented in local lab result terminologies.|
In the grid below you can see a number of actual lab result terms and how they break down into the attributes mentioned above.
I should point out that these attributes are slightly different than the attributes used in LOINC. We will discuss in more detail later. However, this is the way that I thought about lab codes before LOINC became the standard that it is today. You will also note that what is being described here is simple analytes, not microbiology, radiology or other types of results. This is because they are fundamentally different and have their own, in some cases more complex, frameworks and attributes.
In the next article, I will write about the history of result codes and why exchanging them, in general, is difficult.
Lingua freaka– a ‘Battery’ of tests
Every now and then you come across a word and it makes you go “hm?”. When that happens to me, I like to chase it down and figure out what caused the blip on my cognitive radar. As I was writing this article, my radar was ‘pinged’ by the term ‘Battery’ as it refers to a group of lab tests.
In typical American discourse, the term most commonly refers to a portable power cell that you buy at the store and load neatly into your 5-year-olds “I’m ready – SpongeBob talking action figure”, but they can also be used for good.
So what does a portable power cell have to do with a grouping of related tests? The answer is an excellent example of how the meaning of language can change based on the irresistible force of common use.
According to Wiktionary, the origin of the word is the Old French baterie, which means “the action of beating.” While this is oddly appropriate, considering the effect that the SpongeBob talking action figure has on my brain (and its use in ‘assault and battery’), it actually adds to the confusion.
It would seem that in the 1500s, the term shifted to also mean a grouping of artillery units as in a “gun battery”. So this notion of grouping tools in an arsenal is the likely source of a “battery of tests”.
But why do we call a portable cell a battery? This is because of Benjamin Franklin. In 1748 he described multiple Leyden Jars (early capacitors) by analogy to a battery of cannons.
So, if you want to be a smarty pants, the next time you are at a dinner party and someone refers to a ‘Energizer AA’ as a battery, you can explain that since it only consists of one cell, it is not technically a battery.
I know what your thinking… why do we also call a group of tests a “panel”? If you know the answer, leave it in the comments.