NIAD-3217 - Send Observation (Test Result) when they do not have a Specimen attached to them
#963
Conversation
| "resource": { | ||
| "resourceType": "Specimen", | ||
| "id": "NON-SIGNIFICANT-SPECIMEN", | ||
| "meta": { | ||
| "profile": [ | ||
| "https://fhir.nhs.uk/STU3/StructureDefinition/CareConnect-GPC-Specimen-1" | ||
| ] | ||
| }, | ||
| "identifier": [ | ||
| { | ||
| "system": "https://wiremock/", | ||
| "value": "NON-SIGNIFICANT-SPECIMEN" | ||
| } | ||
| ], | ||
| "subject": { | ||
| "reference": "Patient/DAED5527-1985-45D9-993E-C5FF51F36828" | ||
| }, | ||
| "receivedTime": "2001-02-03T10:00:00+00:00", | ||
| "collection": { | ||
| "collectedDateTime": "2001-02-03T11:00:00+00:00" | ||
| } | ||
| } | ||
| }, |
There was a problem hiding this comment.
Apologies to the reviewer - I have formatted the bundle which has introduced a bunch of formatting changes which made the git changes difficult to find, just wanted to highlight this was added.
| { | ||
| "resource": { | ||
| "resourceType": "Observation", | ||
| "id": "TEST-RESULT-NO-SPECIMEN", | ||
| "meta": { | ||
| "profile": [ | ||
| "https://fhir.nhs.uk/STU3/StructureDefinition/CareConnect-GPC-Observation-1" | ||
| ] | ||
| }, | ||
| "identifier": [ | ||
| { | ||
| "system": "https://wiremock/", | ||
| "value": "TEST-RESULT-NO-SPECIMEN" | ||
| } | ||
| ], | ||
| "status": "final", | ||
| "code": { | ||
| "coding": [ | ||
| { | ||
| "system": "http://read.info/readv2", | ||
| "code": "42R5.00", | ||
| "display": "Serum TIBC", | ||
| "userSelected": true | ||
| }, | ||
| { | ||
| "extension": [ | ||
| { | ||
| "url": "https://fhir.nhs.uk/STU3/StructureDefinition/Extension-coding-sctdescid", | ||
| "extension": [ | ||
| { | ||
| "url": "descriptionId", | ||
| "valueId": "2580111000000118" | ||
| } | ||
| ] | ||
| } | ||
| ], | ||
| "system": "http://snomed.info/sct", | ||
| "code": "1015451000000101", | ||
| "display": "Serum total iron binding capacity" | ||
| } | ||
| ] | ||
| }, | ||
| "subject": { | ||
| "reference": "Patient/DAED5527-1985-45D9-993E-C5FF51F36828" | ||
| }, | ||
| "effectiveDateTime": "2001-02-03T12:00:00+00:00", | ||
| "issued": "2001-02-03T13:00:00+00:00", | ||
| "valueQuantity": { | ||
| "value": 42.000, | ||
| "unit": "umol/L" | ||
| } | ||
| } | ||
| }, |
There was a problem hiding this comment.
Apologies to the reviewer - I have formatted the bundle which has introduced a bunch of formatting changes which made the git changes difficult to find, just wanted to highlight this was added.
Everything else remains the same.
Scenario 1: Files: diagnostic-report-with-one-specimen-and-one-unrelated-observation Diagnostic Report linked to Specimen Observation linked to Diagnostic Report Specimen linked to Diagnostic report (Observation is not linked to Specimen) ---- Scenario 2: Files: diagnostic-report-with-one-specimen-one-linked-observation-and-one-unlinked-observation Diagnostic Report linked to Specimen Observation linked to Diagnostic Report Specimen linked to both Diagnostic Report and Observation Observation linked to only Diagnostic Report (This creates an Observation unlinked to a Specimen)
stevenmccullaghmadetech
force-pushed
the
NIAD-3217
branch
from
fa79e1c
to
df9bd62
Compare
...pper/diagnosticreport/diagnostic-report-with-one-specimen-and-one-unrelated-observation.json
Outdated
Show resolved
Hide resolved
.../diagnostic-report-with-one-specimen-one-linked-observation-and-one-unlinked-observation.xml
Outdated
Show resolved
Hide resolved
This is not complete as it: a) does no ID checking for the dummy specimen b) we do not yet verify that a result has a link to specimen, we just check that there is a specimen and result in a diagnostic report
Added placeholder for new linking between specimen and observation, and removed the old one. Added test for the second scenario and fixed the test data.
.../test/java/uk/nhs/adaptors/gp2gp/ehr/mapper/diagnosticreport/DiagnosticReportMapperTest.java
Outdated
Show resolved
Hide resolved
.../src/test/resources/ehr/request/fhir/output/expected-xml-with-observations-inside-report.xml
Outdated
Show resolved
Hide resolved
| <id root="test-id-3" /> | ||
| <code code="109341000000100" displayName="GP to GP communication transaction" codeSystem="2.16.840.1.113883.2.1.3.2.4.15"/> | ||
| <statusCode code="COMPLETE" /> | ||
| <component typeCode="COMP" contextConductionInd="true"> |
There was a problem hiding this comment.
This change looks sus. Looking at the JSON, the DiagnosticReport has 1 Specimen, and zero test results.
We'd expect that the Mapper would generate 1 Specimen, and no Dummy, but it appears to be generating both.
There was a problem hiding this comment.
From looking at the XML, we can see two instances of <specimen typeCode="SPC">, where the JSON suggests we should only have one, which agrees with what you've said.
However, that specific diagnostic report has one specimen and zero observations. A diagnostic report must have an observation. Therefore, a dummy observation is made, and linked to a dummy specimen. This accounts for the two specimens.
There was a problem hiding this comment.
Hypothesis as to what's going on:
- When it grabs Specimens, it knows that there is one already.
- When it grabs the Observations, it identifies there isn't one, so creates one.
- Because this dummy Observation that has been created doesn't have a Specimen it also generates a Dummy specimen to go along with it.
Hypothesis as to what was happening previously before our change:
- When it grabs Specimens, it knows that there is one already.
- When it grabs the Observations, it identifies there isn't one, so creates one with a NULL specimen.
- When the DiagnosticReportMapper iterates over each Specimen, the Dummy Observation never gets output as it doesn't belong to any Specimen.
There was a problem hiding this comment.
So while I don't think this change in behaviour is desirable, it isn't bad enough to be a blocker. We can come back and remove the superfluous Specimen at a later date.
service/src/test/resources/ehr/mapper/encountercomponents/expected-components-1.xml
Outdated
Show resolved
Hide resolved
…t to diagnostic report mapper
.../test/java/uk/nhs/adaptors/gp2gp/ehr/mapper/diagnosticreport/DiagnosticReportMapperTest.java
Show resolved
Hide resolved
|
|
||
| List<Specimen> specimens = new ArrayList<>(); | ||
|
|
||
| // At least one specimen is required to exist for any DiagnosticReport, according to the mim |
There was a problem hiding this comment.
Where in the mim did you see that?
| private boolean hasObservationsWithoutSpecimen(List<Observation> observations) { | ||
| return observations | ||
| .stream() | ||
| .filter(observation -> !isFilingComment(observation)) |
There was a problem hiding this comment.
I don't think this line is super well tested.
.../src/main/java/uk/nhs/adaptors/gp2gp/ehr/mapper/diagnosticreport/DiagnosticReportMapper.java
Outdated
Show resolved
Hide resolved
|
|
What
Please include a summary of the changes and the related issue
Why
Please include details of the reasoning for these changes
Type of change
Please delete options that are not relevant.
Checklist: