Units - Required for Mapping?


We have a terminology server which maps Source Laboratory terms to Target LOINC codes. The Source data does not always populate Units (eg missing in the HL7 lab message) or it is lost somewhere in the data flow before it gets to our system. We are supporting clients in their HEDIS measure implementation, whereby this lab data is considered as Supplemental data by NCQA.

The question is: If we see source data which based on the Code or Description can clearly be matched to a LOINC code, and based on the values for that lab is clearly in a particular Unit, can we map it to the LOINC code for purposes of HEDIS?

It appears that the HEDIS documentation does not specify and would rely on LOINC’s answer here. In the end, it is likely dependent on what a HEDIS auditor would accept, but I wanted to see if LOINC or the LOINC community has a more clear answer.

Thank you!



Hello Benjamin,

I’m sensing two angles of your post. One is mapping LOINC with “spotty” source data, but you do have example answers augmenting. The second is regarding the HEDIS use case.

1st question - of course the ideal situation when adopting LOINC is that a source extract can provide all necessary information. It’s not a perfect world, however, and many of us have several resources at hand while mapping. This definitely includes crossing columns of the extract to find indication of units in a non-unit field, as well as example reports or workstation summaries. I wrote a poster for AMIA 2011 which explains different sources for the six attributes, and how an LIS outperforms the EHR in this source data. [“Mapping the Data Deluge to LOINC - Tsunami, Downpour or Drought?” Not clear how I can attach a jpg or pdf of it in here. If you provide an email address, I’ll send it to you]. Basically, the result display in the LIS often includes information to deduce the LOINC property attribute. Having example answers or reports helps further support the choice.

2nd question - HEDIS use case. Regenstrief confers along with payers on an ongoing HEDIS LOINC advisory group for updates to the HEDIS value sets. Keep in mind that HEDIS Value Sets have their own LOINC panels, helping in choose which LOINC codes pertain. I think that will assist in wondering what a HEDIS auditor would accept. Search in RELMA for ‘HEDIS 2019’. This would be a secondary search AFTER you’ve chosen the LOINC map for the client’s laboratory assay. For instances where the chosen LOINC isn’t listed under the HEDIS LOINC Value Set, have an internal conversation on a) certainty of the mapping and b) is it necessary to change the reported format or lab method in order to comply?

All the best,