EMIS IM1 PFS Odd SNOMED CT Codes?

open forum
1

Running some tests on our API’s and our test engine started flagging a series of invalid SNOMED CT codes.

Majority are fine but noticed

Systolic codes coming back as 114311000006111 instead of 72313002 which is what is used in transaction API.
Diastolic comes back as 619931000006119 instead of 1091811000000102

We’re starting to look at automating clinical decision support and so valid UK SNOMED CT is becoming important to us.
We’re also looking at aggregating records from multiple sources … so again clinical coding is an obvious of aggregating data (i.e. to get a complete view of patient blood pressures).

Anyone know what these codes are (SNOMED description id’s?) or just local codes.

2

they exist here: https://datacompass.lshtm.ac.uk/id/eprint/3590/28/codelist_blood_pressure_aurum.txt
in the first column “medcodeid”. No idea what the provenance of that is though. I assume a local or temporary code.

3

Cheers.

I wonder if it’s something from TRUD. Bizarre why neither read code and SNOMED CT code is in the API. Clinical coding is one of the basic requirements in interoperability and has been followed by EMIS for years.

Some of those codes are the same as SNOMED description id - which might be a clue who asked for this in the API.

4

This is a bit confusing

via IM1 Partner I’m getting Observable Entities such as 167272007 Urine protein test not done → So a FHIR Observation

via IM1 PFS the same entry becomes 259802012 Urine protein test not done a situation code → so a FHIR Procedure.

I need to find what code the entry originally was.

p.s. I’m finding these via FHIR Validation connected to NHS England Ontology Server

5

That a great find @riksmithies !!

My take is that the source system is sending back the ‘medCodeId’ and not the SNOMED code

114311000006111 systolic blood pressure 2469 2469.00 72313002 120159016

what you are being sent is a medcodId not a SNOMEd ID (sometimes they are the same.

Is the extract from EMIS being routed via some other service?

6

These are CPRD ‘MedCodes’

7

A little further on.

From UI I can only get top level code

image
image451×328 68.9 KB

In both API’s I only get the component systolic and diastolic codes.

8

Does appear to be some code conversions going on. Another entry for this same (demo) patient has on 21/3/2005

image
image460×298 53.4 KB

IM1 Transaction returns this code but IM1 PFS returns 259802012 a situation code (original is observation entity)

9

Don’t have access to GP Connect, would be interested to see what was done on that API.

(I have both IM1 converted to FHIR, so can compare)

10

A fix is now available on development version of EMIS PFS.

I’d recommend the new snomedCT part of the model being used.

I’ve tested this against using NHS England terminology server (not an official NHS England comment)