Understanding LOINC®

Part I: History

LOINC® (Logical Observation Identifiers Names and Codes) has actually been around for a while. It was initiated in 1994 by the Regenstrief Institute (in Indianapolis, Indiana) and the LOINC committee.

The LOINC database provides a universal code system for identifying laboratory and clinical test results. These codes are used in electronic messages such as Health Level Seven (HL7), so that when hospitals, physicians, health maintenance organizations, pharmaceutical manufacturers, researchers, and public health departments receive such messages from multiple sources, they can automatically file the results in the right slots of their medical records, research, and/or public health systems. LOINC facilitates the exchange and pooling of results such as blood hemoglobin, serum potassium, or vital signs for clinical care, outcomes management, and research.

On March 21, 2003, the Departments of Health and Human Services, Defense, and Veterans Affairs announced the first set of five uniform standards for the electronic exchange of clinical health information to be adopted across all US federal agencies. LOINC was adopted “to standardize the electronic exchange of clinical laboratory results.”

LOINC has been endorsed by the American Clinical Laboratory Association and the College of American Pathologists. It has been adopted as an alternate test reporting code by large commercial laboratories including Quest, LabCorp, Mayo Medical Laboratories, and MDS Labs; large HMOs including Kaiser Permanente and Aetna; government organizations including the CDC, DOD, VA, and NLM; and as part of the Health Insurance Portability and Accountability Act (HIPAA) attachment proposal.[1]

The LOINC result code is not intended to transmit all possible information about a test or observation. The LOINC code’s purpose is to identify the test result or clinical observation. Other fields in the electronic message are used to identify the source laboratory and any special details about the sample. (For instance, the result code may identify a blood culture, but the message source code can be more specific and identify the sample as pump blood.) The level of detail in the LOINC definitions is designed to universally define a test at the lowest level of reportable result data.

Most laboratories identify tests in HL7 messages by means of their internal (and idiosyncratic) code values. Receiving medical informatics systems cannot fully “understand” the results they receive, unless they either adopt the producer’s laboratory codes (which is impossible if the informatics system receives results from multiple source laboratories, e.g., the hospital lab, the physician office lab, and the local commercial lab), or invest in the work to map each laboratory’s coding system to their internal code system.

Because there are over 58,000 entries in the database, the Regenstrief Institute has also developed RELMA® (the Regenstrief LOINC Mapping Assistant), a tool you may download or use online to support browsing and searching of the LOINC database to help you map your local test codes to LOINC codes. Users cannot modify official codes, but they can modify the table to extend them locally or for their own use.

Part II: The Rule – Why We Need LOINC

If you intend to demonstrate meaningful use and receive the incentive payments, you will need to be familiar with LOINC®. Standards such as LOINC are included in the ONC final rule. From page 44599 of 45 CFR Part 170, “Health Information Technology: Initial Set of Standards, Implementation Specifications, and Certification Criteria for Electronic Health Record Technology,” the final rule states:

“As previously discussed in the Interim Final Rule, we adopted several minimum code set standards. This approach will permit Complete EHRs and EHR Modules to be tested and certified, to, ‘at a minimum,’ the version of the standard that has been adopted or a more current or subsequently released version. We would note that consistent with this approach the Secretary has proactively identified and deemed acceptable newer versions of the following adopted ‘‘minimum standard’’ code sets: (1) LOINC version 2.3, released on February 26, 2010”

The HHS/Office of the National Coordinator for Health Information Technology (ONC) has defined the standards, implementation guides, and certification criteria to be used for certification.

Test procedures to evaluate conformance of EHR technology to ONC’s requirements are defined by the National Institute of Standards and Technology (NIST), including the NIST test procedure for §170.302 (h) Incorporate Laboratory Test Results, which contains the following test data elements:

  • Test Result
  • Test Report Date
  • Test Type
  • LOINC Code
  • Test Name (and Normal Range)
  • Test Result Value
  • Test Result Unit of Measure
  • Test Specimen Source
  • Condition/Disposition of Specimen

EHR vendors need to receive LOINC if they want to be certified. Currently, for Stage 1, they just need to demonstrate they can receive the code. Stage 2 and 3 for meaningful use will focus more on interoperability, so more and more vendors will be requiring the codes from the LIS. It is the responsibility of the EHR to translate the codes into their human readable form.

Part III: Understanding and Using the LOINC Database

The overall organization of the database is divided first into four categories: “lab,” “clinical,” “attachments,” and “surveys” (this split is recorded in CLASSTYPE). The laboratory portion is further divided into the usual categories of chemistry, hematology, serology, microbiology (which includes parasitology and virology), and toxicology. Antibiotic susceptibilities have their own category.

Each LOINC record (fully specified name) corresponds to a single test result by method or panel. While each LOINC is unique, the codes were not designed, nor intended, to be unique by testing location. For example, your lab may use the same methodology as your reference lab. If so, both results would use the same LOINC. A formal, distinct, and unique six-part name is given to each term for test or observation identity. The database currently has over 58,000 observation terms that can be accessed and understood universally. Each database record includes six fields for the unique specification of each identified single test, observation, or measurement:

  • Component (analyte): What is measured, e.g., potassium or hemoglobin.
  • Property Measured: How it is measured, e.g., a mass concentration or enzyme activity (catalytic rate).
  • Timing: Whether the measurement is an observation at a moment of time, or an observation integrated over an extended duration of time, e.g., 24-hour urine.
  • Sample Type: The type of sample, e.g., urine or blood.
  • Scale Type: The type of scale, e.g., whether the measurement is quantitative (a true measurement), ordinal (a ranked set of options), nominal (e.g., E. coli, Staphylococcus aureus), or narrative (e.g., dictation results from x-rays).
  • Method: The method used to produce the result or other observation, where relevant.

The analyte is sodium, measured by substance concentration, done at one point in time, on either serum or plasma, and the result is quantitative. In this case, the method used is not included.

Certain parameters and descriptions pertaining to test performance are specifically excluded from the database. These parameters will typically be reported in separate fields (attributes) of a test/observation report message, not as part of the observation name. Attributes that are explicitly excluded from the fully specified name are:

  • The instrument used in testing.
  • Fine details about the sample or the site of collection, e.g., “right antecubital fossa.”
  • The priority of the testing, e.g., whether STAT or routine.
  • Who verified the result.
  • The size of the sample collected.
  • The place of testing, e.g., home, bedside, clinical lab.

Also, become familiar with the online search tool, RELMA®. To use it, go to www.search.loinc.org, and agree to the disclaimer. Enter your test name into the search field. Knowing the database structure will aid your search. In the example above, searching on “glucose” brings 851 hits, but searching on glucose and fasting reduced the number to 19. You can sort on any of the columns. Scroll to the right to the “Long Name” column and find the one that best fits your lab. To make it easier, we offer:

  • A list of the Long Common Names
  • The LOINC Manual

Late in 2009, a new set of codes at the order level was introduced. Still in the experimental stage, approximately 300 have been published. The intent of the order-level codes is to simplify communication between disparate systems at the order level to complement the existing database, which focuses on the result level. We will be hearing more on this.

Part IV: Populating Your LIS with LOINC

Now that you know all about LOINC, the million dollar question becomes: “How do I get this information in my LIS?” It would be great if your LIS vendor could do this for you, but they can’t. It’s not that the vendors don’t want to help. The reality is that there are far too many variables in each of your test menu dictionaries.

Consider a simple glucose test. A glucose is a glucose is a glucose, right? Not quite. Is it fasting, is it timed, is it following an oral challenge, and if so, how many hours? While the glucose test is the same on the analyzer, how many variations of glucose do you use at your facility? Is it serum, urine, CSF, or another fluid? Which method is used on your analyzer? Do you use reagents manufactured by the instrument vendor, or are you using another vendor’s reagents? The method may vary between the instrument vendor’s and the third party’s reagents. I think you get the picture. Multiply that by the number of clients for each LIS vendor, and it becomes an impossible task to simply import the codes. A search of the LOINC database for glucose brings up 851 different codes. How would your LIS vendor possibly know which one to map to yours? The appropriate code depends on the combination of the instrument and reagent vendors.

Don’t despair, though. There are ways to minimize the scope of entering the codes in your LIS. Your instrument vendor knows the methodology better than anyone, so call them and/or reagent vendor(s) and ask them what LOINC is appropriate for the method used on your analyzer, as the method on one analyzer in a series may differ from the next generation of that analyzer from the same manufacturer. Remember, too, that your LIS vendor is not a client of your instrument manufacturer, so they will not have access to the instrument or reagent vendor technical support groups, but you may access these groups through your instrument and reagent vendors for further assistance. Fortunately, some vendors are beginning to include the LOINC codes in their package inserts to make things easier.

Once the codes are populated into the correct field of the LIS, it will include them in interface transmissions to the EMR, and you’ll be ready to go!

For additional information and last minute news regarding LOINC, be sure to visit our blog, colLABorate.

[1] McDonald CJ, Huff SM, Suico JG et al. LOINC, a universal standard for identifying laboratory observations: a 5-year update. Clin Chem. 2003 Apr;49(4):624-33.