codes for new interpretations of susceptibility testing I and SDD

Legacy laboratory-loinc antibiotic-susceptibilities
1

Dear community members,

I was wondering whether some of you have experience with coding the category “I - susceptible, increased exposure” as introduced by EUCAST in 2019 and similar the category “SDD - susceptible-dose dependent” of CLSI.

As far as I know there are codes for susceptible (LA24225-7)/intermediate (LA29303-7)/resistant (LA6676-6) and indetermined (LA11884-6) but not for the above mentioned interpretations. Is someone aware of developments to add theses codes or alternatives?

Thank you in advance,

Benjamin

 

2

Hello,

 

I recently became aware of these new result “responses” too. In the US, these would be mapped to SNOMED CT to be compliant with HL7 ELR and LRI Implementation Guides and information system functionality for these data elements. Are Europeans using LOINC Answers?

Either way, recommend submitting term requests to SNOMED and LOINC if both are needed.

Here’s the submission page: https://loinc.org/submissions/ You can also see what others have requested under the prerelease terms page.

3

Yes, Europeans (like me) are using the LOINC Answers. Why shouldn’t we? We are not aliens.

4

@jozef I was asking as many information systems, especially laboratory information systems and public health systems in the US do not have functionality for supporting LOINC Answers. Rather the requirement for encoding these result values in the US is SNOMED CT.

It sounds like from your response that in Europe, the requirements for mapping these data elements is LOINC Answers? Is that correct?

I know there are diagnostics vendors in this space with both US and European customers (and other countries). To meet encoding requirements for each, multiple sets of codes will be needed to support this new result value. Also, why I recommended requests for both LOINC and SNOMED CT codes to meet each country’s needs for this new term.

It sounds like you do have functionality for LOINC Answers. It would be helpful to understand what information system vendors and types (EHR, PH, LIS, others?) have functionality for LOINC Answers in Europe. Can you shed some light on what you are using and where in the workflow (RWE?)?

5

Hello!

I’m working behind the scenes to see if SNOMED CT has submissions to create content that could be tied to LOINC answers.

More to come! Perhaps if there’s enough information, we can provide an update at the June 15th Lab LOINC telecon.

Pam

6

Hi Pam, Andrea, Jozef and Ben;

Reading down through the thread there are a couple of items I want to respond to.

  1. Do SNOMED CT codes exists for reporting susceptbilities under the EUCAST 2019 format?

Yes, SNOMED contains the following “Degree Finding” codes:

  • 11896004 |Intermediate (qualifier value)|

  • 82334004 |Indeterminate (qualifier value)|

  • 131196009 |Susceptible (qualifier value)|

  • 264841006 |Intermediately susceptible (qualifier value)|

  • 260357007 |Moderately susceptible (qualifier value)|

  • 1255965005 |Susceptible with increased exposure (qualifier value)|

  • 30714006 |Resistant (qualifier value)|

There is a little bit of clean up work that could be done in SNOMED CT as Intermediate is a different part of the qualifier value hierarchy than susceptible and resistant concepts are located, however this should only be an issue when performing data analytics and should not impact data capture.

  1. SNOMED CT versus LOINC answer list (LA codes)

We have to keep in mind the terminological requirements set by the governing body of the country regarding which code system should be used. Additionally, we also need to keep in mind the LOINC answer list binding strength. If the LOINC answer list binding strength is anything other than nominal (i.e. example or preferred) then LOINC answer list codes may be used in conjunction with that particular LOINC code, but there is nothing mandated that the answer list may not be replaced by or extended with content from another code system.

This is where we get into what “needs” to happen at the implementation level in order to support international interoperability. Both HL7 C-CDA and FHIR allow for “Translation” codes to be included in the messaging structure to increase the likely hood that the receiving system can associate the primary code with a local code in their EHR. Looking at the LOINC part mappings, it should be fairly easy to crosswalk the LOINC Answer List values with the associated SNOMED CT concept and vice versa to allow both LOINC answer list codes and SNOMED CT codes to be included in the same message. Ideally, each implementation would include these translation codes with each message. While I understand there may be both technical and human resource limitations at the implementation level that prevent this from happening, I think it is the goal we should be attempting to achieve.

I hope that helps?

Best,

John

7

<span style=“font-family: Lato, Helvetica, Arial, Lucida, sans-serif; font-size: 15.6px;”>Hi Pam, Andrea, Jozef and Ben;</span>

There were a couple of additional items that I attempted to add to the last post, but lost the changes.

When discussing LOINC answer list, we want to keep in mind that the answer lists were designed to support external code systems. Currently, the following external code systems are used in answer lists:

  • HCPCS

  • Unicode

  • SNOMED

  • PHIN/VADS

We also want to keep in mind that, unlike other code systems, there are technically 5 different LOINC codes that can be transmitted under the LOINC OID / URI in HL7 messaging structures. They are:

  • Standard numeric only LOINC codes (These are the most common ones that people know)

  • LOINC Part Code (LP prefixed codes)

  • LOINC List Code (LL Prefixed codes)

  • LOINC Answer List Code (LA Prefixed codes)

  • LOINC Group code (LG Prefixed codes)

To date, the only use case I have run into for transmitting anything other than the LOINC Answer List codes is the LOINC part code. In the absence of a true ontological hierarchy, the LOINC Part code can be assigned as a grouping code to “group” laboratory, cardiology, dental, vital signs, etc. at a very high level. Under Laboratory, a second level of grouping can take place to “group” chemistry, challenge tests, susceptibility. This type of grouping can be very valuable when building an automated grouping of content for display to end-users.

Again, the take away here is that any one of the 5 types of LOINC codes may be validly transmitted in an HL7 message structure using the LOINC OID / URI. Granted, most systems do not recognize the alpha prefixed (i.e. LP, LL, LA, LG) codes, so the sending system should verify with the receiving system that they know how to identify and utilize the codes.