Analyte/component subpart(s) and AI

General
1

I’m planning on submitting some new measures for LOINC coding. Currently, there are a few related LOINC codes, but they are ambiguous, IMO. I want to make the new submissions more granular, including methods, which are an important clinical distinctions for the measures I want to submit.

The “measures” I want to submit are for exercise, but some are “estimates” of oxygen uptake. There is a LOINC code for Oxygen “consumption”, but consumption is academically defined as cellular, while uptake is academically defined as whole body. Whole body is for exercise related oxygen uptake.

One current LOINC: 60842-2 Oxygen consumption (VO2) tells nothing about if it is measured, estimated, predicted, or reported - all of which are used clinically. Measured is considered the gold standard, but in clinical practice, estimated is used most often.

LOINC: 103749-8 Oxygen consumption (VO2) --peak, is similarly ambiguous. Peak oxygen “uptake”, not “consumption” is a benchmark for cardiorespiratory fitness assessments in a number of clinical populations.

There’s a lot more to say about these, but I’ll get on to the questions.

How many sub-analyte parts can there be? Is it limited to just 2 subparts as shown in the LOINC manual or can there be more than 2 subparts?

One reason I ask this is because of AI. Can we use more subparts for analytes, or should we, and will that enhance AI tools’ use of LOINC codes?

Some examples of more granular oxygen uptake LOINC codes are these:

oxygen.uptake.gross^during maximal aerobic exercise:VRat:Pt^max:Patient:Qn:measured

Another LOINC would be the same except “net” (minus resting oxygen uptake):

oxygen.uptake.net

Those should also have separate LOINC codes for method “estimated”, “predicted”, or “reported”

1 Like
2

That posted before I could complete it.

As a LOINC coding management consideration, rather than having a separate code for all of the different oxygen uptake methods, for each new LOINC code, it seems that having one Method code with the different methods listed might be better. The requirement with an oxygen uptake LOINC code would be to also attach the method LOINC code with it. Otherwise, the dozens of new LOINC codes I’m interested in submitted will be multiplied by the number of methods that assess oxygen uptake (measured, estimated, predicted, reported).