We are often asked if the same LOINC® code indicates identical tests. This is a very relevant communications issue for achieving interoperability and meaningful use among laboratories, healthcare providers, and health information exchanges. It is important for the industry to understand that LOINC® does not replace all supporting identification of a laboratory assay.  LOINC® is a universal identifier to be used in HL7 messaging, embedded alongside the local laboratory’s mnemonic, display name, units of measure, reference range, and interpretative data.  The systems should not be asking the LOINC® code to carry the full burden of identification for every regard.

So, as two paths theoretically converge on one side of an interoperability bridge and they need to merge in order to cross, what perspectives are brought along? On the laboratory path, there is a catalog of laboratory tests created by the needs of the providers it serves. They are vetted and investigated by the Medical Director or Pathologist.  Their accuracy and precision are inferred by best practices acknowledged through laboratory accreditations by credentialing agencies.  Some assays additionally are continually participating in external quality assurance programs.  These ongoing processes weigh heavily on the Director’s minds when asked to adopt a harmonizing coding set for sharing lab values with other systems. This shapes their perspectives and expectations of what LOINC® will do.

On the LOINC® coding path, there is a method of evaluating each chartable result field in the laboratory information system (LIS) in which six attributes are used to  “measure up” the assays for identifying a numeric code: component/analyte, property, timing, system, scale and method.  The laboratory domains covered in LOINC® include chemistry, hematology, coagulation, urinalysis, blood bank, microbiology, virology, cytopathology, pathology, molecular pathology, genetics and more. Components include proteins, enzymes, hormones, cellular morphology, therapeutic drugs, antibiotics, etc.   If the assay’s display is only a genus /species name, the LOINC® component evaluation goes further down to the level of organism entity, organism DNA, organism rRNA, IgG or IgM antibody against the organism.  The answer format of each assay is influenced by the abilities/limitations of the kit or instrumentation.  Review of kit packages, reagent or instrumentation documentation along with acceptable specimen types and examination of the actual result entry options will determine the remaining attributes. Answers such as “negative” or “negative <1:20” have different property attributes, and thus have different codes. Units of measure of “mg/dL” and “mmol/L” will have different property attributes and also different LOINC® codes. LOINC® is useful for both identifying similarities and dissimilarities between assays.  

The LOINC® User’s Guide Preface (entire guide available at www.loinc.org) provides examples where “the exchange and pooling of results, such as blood hemoglobin, serum potassium, or vital signs, for clinical care, outcomes management, and research” are facilitated by adoption of the vocabulary standard.  A further example of the theoretical bridge crossing: consider a pilot statewide health information exchange project performed several years ago.  Along the laboratory path came 14,802 local result fields from five test catalogs at unrelated hospital enterprises.  They were mapped to LOINC®, and in the process of crossing the bridge distilled down to 4,051 distinct concepts.  From a viewport of finding identical assays, the team found:

• 25% of the codes had three or more labs reporting the same test/specimen/result format.  
• 23% of the codes were identical reporting formats used by two labs.  
• 52% of the codes were distinct formats across the five catalogs.

From a dissimilarity viewpoint, the team determined ‘Lyme Disease Antibody’ was not the same as ‘Lyme Disease Serology’ from another lab.  [The former was a Borrelia burgdorferi antibody IgM by Western blot, while the latter was a B. burgdorferi antibody IgG titer.]

In order to answer today’s topic question it is necessary to have the right use case in mind. For the purposes of allowing the data stream of verified answers converge from different originating laboratories and have the software automatically manage storage of similar results, they are identical.  The messages retain all information that came from the performing labs for the clinician to have when interpreting the results, and thus retain the units of measure, reference ranges, footnotes, etc.   The assays may not be identical when examining the kit/reagent/instrument they are run on, cost per test, assay calibration, or their CPT codes (which is an entirely different story).  LOINC® coding does not incorporate any of these data elements into the vocabulary standard; thus is not intended as a comparison basis for those use cases.

In conclusion, this blog entry provided the context to view LOINC® codes when comparing tests for downstream collation amongst different facilities. Looking forward to further discussions!