Stream: implementers
Topic: Hapi FHIR STU3 practitioners -> encounter with no tie to pat
Garth Schneider (Apr 07 2017 at 18:25):
for patient 20226 when I do $everything I get practitioner coming back with encounters that neither have ties back to patient. Also get MedicationStatements that are attached to a List that references a patient, but the MedicationStatements do not reference patient.
Garth Schneider (Apr 07 2017 at 18:27):
So are the encounters that are not attached really encounters of that patient? Second question: Is it normal for MedicationStatement to always just be attached to List is pointing to patient?
Garth Schneider (Apr 07 2017 at 18:33):
The patient points to the practitioner, but more than one patient would see a practitioner? Or is practitioner (by patient instance) being pointed to by the patient creating a List of sorts and anything on that Practitioner is of that patient?
Rick Geimer (Apr 07 2017 at 20:11):
If I were to paraphrase Garth's concern, it is that we state that context does not conduct in FHIR and that resources must explicitly declare their context (i.e. the patient that they apply to), but his example using http://fhirtest.uhn.ca/baseDstu3/Patient/20226/$everything is definitely returning resources without direct links to the patient. Is this correct behavior? If so, do we have a problem in the spec (i.e .context does not conduct except when it does)?
David Hay (Apr 07 2017 at 20:23):
ev.png just for fun, I imported it into clinFHIR (sorry the image is so small)...
David Hay (Apr 07 2017 at 20:36):
my guess is that this is a hapi issue...
David Hay (Apr 07 2017 at 20:44):
actually - hold on - those encounters do have a reference to the patient... - theres a CF bug. bother.
David Hay (Apr 07 2017 at 20:45):
it's using encounter.patient not encounter.subject...
Garth Schneider (Apr 07 2017 at 20:46):
Do not see that:
Garth Schneider (Apr 07 2017 at 20:46):
{
"resourceType": "Encounter",
"id": "20244",
"meta": {
"versionId": "1",
"lastUpdated": "2017-01-23T00:09:13.668-05:00"
},
"status": "finished",
"type": [
{
"text": "GP"
}
],
"participant": [
{
"individual": {
"reference": "Practitioner/5355"
}
}
],
"period": {
"start": "2017-01-18T05:09:13.498Z"
},
"reason": [
{
"text": "headache"
}
]
}
Garth Schneider (Apr 07 2017 at 20:47):
Just to practitioner, which does not have a reference to patient
David Hay (Apr 07 2017 at 20:48):
hmm..
Garth Schneider (Apr 07 2017 at 20:49):
The one Rick shared has a bunch like that. Also MedicationStatement connected to List. The list points to patient
Garth Schneider (Apr 07 2017 at 20:49):
MedicationStatement does not reference anything
David Hay (Apr 07 2017 at 20:53):
true... The Encounter link was an artifact (well, bug) in the CF importer. sigh.
David Hay (Apr 07 2017 at 20:55):
Here's the Medication statements: (with no patient links) Screen-Shot-2017-04-08-at-8.54.07-am.png
David Hay (Apr 07 2017 at 20:55):
I've not included the orphan encounters we already know about...
Garth Schneider (Apr 07 2017 at 20:58):
Using clinfhir I have never been able to get the patient in it online.
Garth Schneider (Apr 07 2017 at 20:59):
It even says "Note that the Patient is no displayed in this graph"
David Hay (Apr 07 2017 at 21:17):
How are you importing the bundle?
Garth Schneider (Apr 07 2017 at 21:18):
This is going right against the HAPI FHIR STU3 test server
Garth Schneider (Apr 07 2017 at 21:18):
It was already on server
David Hay (Apr 07 2017 at 21:18):
I do see the patient, and can select them (Though CF is helpfully adding the patient to the resources!)
David Hay (Apr 07 2017 at 21:19):
So I did the $everything call (setting _count to 100 to get everything), then copied the json bundle, created a new scenario in CF and imported the bundle using the 'import resource' link in the navbar...
David Hay (Apr 07 2017 at 21:20):
Apart from the addition of the link to patient where that didn't exist (& I 'll fix that), it seemed to work OK...
Garth Schneider (Apr 07 2017 at 21:20):
Screenshot-from-2017-04-07-16-20-04.png
Garth Schneider (Apr 07 2017 at 21:21):
There is no patient there
David Hay (Apr 07 2017 at 21:21):
interesting...
Garth Schneider (Apr 07 2017 at 21:22):
See the note at the bottom saying patient not included
David Hay (Apr 07 2017 at 21:22):
This was the original call to hapi: http://fhirtest.uhn.ca/baseDstu3/Patient/20226/$everything?_count=100 - yes?
Garth Schneider (Apr 07 2017 at 21:22):
yes
David Hay (Apr 07 2017 at 21:23):
58 resources in the bundle, the first being Patient?
Garth Schneider (Apr 07 2017 at 21:23):
yes indeed
Garth Schneider (Apr 07 2017 at 21:24):
How do you get patient in your graph?
Garth Schneider (Apr 07 2017 at 21:25):
should MedicationStatement also always have a subject or patient reference?
David Hay (Apr 07 2017 at 21:25):
It's just there when I import it. I did use my local copy of CF - I'll try again using the the web version - hold on... (We have veered off topic - but helpful to me!)
Garth Schneider (Apr 07 2017 at 21:25):
or is hanging off of a list that is connected to a patient good enough?
Garth Schneider (Apr 07 2017 at 21:25):
Sorry
Garth Schneider (Apr 07 2017 at 21:26):
I will build a local copy
Garth Schneider (Apr 07 2017 at 21:26):
of cf
David Hay (Apr 07 2017 at 21:26):
No - there does need to be an explicit reference...
David Hay (Apr 07 2017 at 21:27):
Are you using clinfhir.com?
Garth Schneider (Apr 07 2017 at 21:27):
Ok, excellent. I am building a UI off of fhir and just need to know all of this. Thanks
Garth Schneider (Apr 07 2017 at 21:28):
yes clinfhir.com
Garth Schneider (Apr 07 2017 at 21:28):
looking at 20226 never see a patient
Garth Schneider (Apr 07 2017 at 21:29):
on the web one
David Hay (Apr 07 2017 at 21:29):
it works for me... Let me look into this and I'll carry on the conversation in the clinFHIR stream. Always good to find (and squash) bugs though!
Garth Schneider (Apr 07 2017 at 21:31):
Yea, you can see from my png what I have. Thank you for letting me know how it is suppose to work. I will write to that spec
Anton Selin (Apr 10 2017 at 17:58):
Getting bunch of Maven pom loading problems when import HAPI maven into eclipse. How to fix it? here is a error below:
Project build error: Non-resolvable parent POM for ca.uhn.hapi.fhir:hapi-fhir:2.4-SNAPSHOT: Failure to transfer org.sonatype.oss:oss-parent:pom:9 from https://jitpack.io was cached in the local repository, resolution will not be reattempted until the update interval of jitpack.io has elapsed or updates are forced. Original error: Could not transfer artifact org.sonatype.oss:oss-parent:pom:9 from/to jitpack.io (https://jitpack.io): sun.security.validator.ValidatorException: PKIX path building failed: sun.security.provider.certpath.SunCertPathBuilderException: unable to find valid certification path to requested target and 'parent.relativePath' points at wrong local POM
Last updated: Apr 12 2022 at 19:14 UTC
