Every LOINC code is a composition of six axes, and knowing the axes at a glance is the shortcut from "I searched for hemoglobin and got twenty results" to "here is the right code and here is why." The axes are stable, small, and worth memorizing. Terminology teams that internalize them move faster than teams that treat each LOINC code as a black box.
Naming the axes is the discipline. For related walkthroughs, related FHIR walkthroughs collects the surrounding material.
The Six Axes
LOINC organizes every code along six axes:
- Component: what is measured (hemoglobin, sodium, blood pressure).
- Property: how it is measured (mass concentration, count, ratio, percent).
- Time aspect: point in time or duration.
- System: specimen or body location (blood, urine, arterial line).
- Scale: quantitative, ordinal, nominal.
- Method: the analytical technique (immunoassay, electrophoresis, calculated).
Each axis has a controlled vocabulary. A LOINC code is the intersection of one value from each axis. A pass through the site's LOINC picker surfaces the axis values inline with the common codes.
Property Is the First Axis That Disambiguates
Between two codes with the same component (hemoglobin), the property axis is usually what differs. MCnc (mass concentration in mg/dL) is different from NCnc (number concentration in cells/mL) and from MFr (mass fraction, percent).
Reading the property first narrows almost every ambiguous search. For the wider search framing, starting a LOINC search when the requester speaks clinical, not codes covers the entry pathway.
System Names the Specimen
The system axis names where the measurement was taken. Blood, urine, serum, plasma, cerebrospinal fluid. Some codes carry "Bld" for whole blood; others "Ser" for serum; the values are not interchangeable.
Sizing plans that confuse blood versus serum codes produce results that flow through to the wrong reference range and confuse decision support downstream.
Method Distinguishes Techniques
Method matters when the same measurement can be performed multiple ways with different clinical implications. HbA1c by HPLC differs from HbA1c by immunoassay in reference ranges and in interference patterns.
Method is often omitted (leaving the LOINC method-agnostic). When method is present, treat it as a constraint on the accepted lab platform.
Time Aspect Is Frequently Overlooked
Time aspect (Pt for point in time, 24H for over twenty-four hours) is small but sometimes decisive. Twenty-four-hour urine sodium is a different code from spot urine sodium, and mixing them silently corrupts trend analysis.
The Parts Enable Hierarchy Navigation
The parts model is what makes LOINC hierarchies useful. Codes that share five of six parts sit near each other in the hierarchy; codes that share fewer parts sit farther apart. Walking up the hierarchy releases specificity along one axis at a time.
For the hierarchy framing, using LOINC hierarchies to broaden a match without losing precision covers the pattern.
Search Optimization Uses the Parts
Terminology tools that index codes by parts return searches faster than tools that only index by name. Every LOINC integration that runs at scale eventually indexes by parts. For the specific performance framing, keeping LOINC lookups fast as your value set grows covers the approach.
The Cheat Sheet Habit
Every terminology team benefits from a wall poster with the six axes and their common values. Internalizing them takes weeks; the payoff lasts years.
Naming the parts model is the small discipline that turns LOINC from a lookup service into a searchable structure.
Sources
- LOINC canonical multiaxial-hierarchy users-guide chapter - LOINC canonical multiaxial-hierarchy users-guide chapter defining the six parts model
