Stream: SANER-Connectathon
Topic: Test Data
Gino Canessa (Sep 10 2020 at 01:54):
Howdy all, just wanted to post a note that the data I received from @David Pyke and @Keith Boone is available on the server: http://prototype-erp-fhir.azurewebsites.net/ for the Connectathon. It's open, so please be courteous and don't modify those resources.
For convenience, I've tagged everything with https://github.com/microsoft-healthcare-madison/projects|saner. I also added the test data security label: http://terminology.hl7.org/CodeSystem/v3-ActReason|HTEST, so there should be no confusion.
Since I'm unlikely to make any of the meetings, I thought I'd post here. That said, if the data needs to be reloaded just ping me. Cheers!
Keith Boone (Sep 10 2020 at 06:59):
New data available at https://github.com/HL7/fhir-saner/raw/master/cthon25-data.zip
Heath Dinkins (Sep 10 2020 at 13:07):
wow, 3AM... Keith was kickin' ass into the wee hours
David Pyke (Sep 10 2020 at 13:07):
That's our Keith
Keith Boone (Sep 10 2020 at 13:10):
I can hear you.
Keith Boone (Sep 10 2020 at 13:10):
I've been having audio problems since the start of this at 9
Keith Boone (Sep 10 2020 at 13:10):
CHAT is disabled in whova.
Austin Kreisler (Sep 10 2020 at 16:01):
@Keith Boone @David Pyke The new data Keith posted this morning is missing three Observation resources that were in the previous set:
Observation-Observation12c
Observation-Observation22c
Observation-Observation32c
Now that I look closer, it looks like any resource for Test Case 2c is missing (Patient, Encounter, Observation, etc.)
David Pyke (Sep 10 2020 at 16:02):
Okay, I'll work on adding them back
David Pyke (Sep 10 2020 at 16:03):
TestCase2c.fsh is an accidental copy of Testcase3a.fsh
Austin Kreisler (Sep 10 2020 at 16:12):
You will be able to restore the original testcase2c? One thing Keith was updating last night was to add the missing Observation values.
David Pyke (Sep 10 2020 at 16:12):
I"m copying all the 2c conent back into the testcase2c and running a sushi build. I'll find the missing observation values but I though those were in the patient9s
David Pyke (Sep 10 2020 at 16:14):
I"m going through the spreadsheet to find what might be missing
Austin Kreisler (Sep 10 2020 at 16:18):
For the two observations for testcase2c, the first has a value of 260415000 Not Detected (SNOMED), and the second observation should have 260373001 Detected (SNOMED)
David Pyke (Sep 10 2020 at 16:22):
Okay, I'll update them
David Pyke (Sep 10 2020 at 16:23):
I'll run the sushi build now and upload the new zip of the 2c resources here
David Pyke (Sep 10 2020 at 16:26):
Testcase2c.zip Updated case 2c
Austin Kreisler (Sep 10 2020 at 16:26):
Thanks!
Austin Kreisler (Sep 10 2020 at 16:33):
Still missing observation values in the first two observations
David Pyke (Sep 10 2020 at 16:57):
Testcase2c.zip Revised
Abbie Watson (Sep 10 2020 at 17:19):
(deleted)
Austin Kreisler (Sep 10 2020 at 19:23):
@Keith Boone The single most used LOINC test code used in our data set is 94759-8 SARS-CoV-2 (COVID-19) RNA [Presence] in Nasopharynx by NAA with probe detection. This isn't listed as a member of the SANER COVID-19 Labs Value Set (http://build.fhir.org/ig/HL7/fhir-saner/ValueSet-Covid19Labs.html) that's used in the computable measure. Was this the value set change that was going to change the counts on day 2 of the connectathon? It's pretty clearly impacting my manual counts for day 1.
David Pyke (Sep 10 2020 at 19:24):
We use MITRE's set. It doesn't include that one so we would need to extend ours
David Pyke (Sep 10 2020 at 19:26):
None of theirs use nasopharynx they only have respiratory or unspecified
Austin Kreisler (Sep 10 2020 at 19:28):
Much of our test case data is built around that one being included in the valueset. We may need to change the test day yet again to use a different valid test code.
David Pyke (Sep 10 2020 at 19:28):
loincs list is now 84 items long. We need to update it
David Pyke (Sep 10 2020 at 19:30):
IF someone can go to https://loinc.org/sars-cov-2-and-covid-19/ , log in and download the test list, I can update the codesystem
David Pyke (Sep 10 2020 at 19:47):
Nevermind I've copied them off the website. I'll update the codesystem now
Austin Kreisler (Sep 10 2020 at 20:44):
On to the next QA issues with the test data, or in this case the Patient COVID 19 Group measure. In this case I'm trying to evaluate the following against the test data:
// The reason or diagnosis associated with the encounter is COVID-19
( Encounter.reasonCode | Condition.code ).memberOf(%SuspectedOrConfirmedCOVID19Diagnoses.url),
The value set SuspectedOrConfirmedCOVID19Diagnoses found here:
http://build.fhir.org/ig/HL7/fhir-saner/ValueSet-SuspectedOrConfirmedCOVID19Diagnoses.html
is empty. It has no logical definition, and its expansion has no members.
There is a second value set that might fit the bill here:
http://build.fhir.org/ig/HL7/fhir-saner/ValueSet-ConfirmedCOVID19Diagnoses.html
The issue with that one is that the expansion doesn't include the ICD-10 codes, even though they are identified in the logical description. It does include a code for suspected COVID.
There is a separate value set for Suspected Covid found here:
http://build.fhir.org/ig/HL7/fhir-saner/ValueSet-SuspectedCOVID19Diagnoses.html
We can discuss at our 5pm Touchpoint session
Austin Kreisler (Sep 10 2020 at 22:19):
The test data is using 882784691000119100 Pneumonia caused by Severe acute respiratory syndrome coronavirus 2 (disorder) (SNOMED) in a number of cases. That needs to be in the confirmed covid SNOMED value set.
David Pyke (Sep 10 2020 at 23:50):
I'll fix it
Keith Boone (May 17 2021 at 21:09):
See https://confluence.hl7.org/display/FHIR/2021-05+Public+Health+--+Situation+Awareness+for+Novel+Epidemic+Response+Track#id-202105PublicHealthSituationAwarenessforNovelEpidemicResponseTrack-TestData to read the following on the SANER Track in the HL7 Confluence page. Below is a copy of that text without all of the pretty formatting:
Test Data
The Endpoint http://test.sanerproject.org/fhir has test data for patients from multiple organizations.
All test data for Connectathon 27 tests used for scalability testing is marked with meta.tag=
{
"system": "http://sanerproject.org/tags",
"code": "MTEST",
"display": "Measure Test Data"
}
Test patients exist for the following Organizations. Each test patient will reference the organization in Patient.managingOrganization, thus, this server contains data for multiple hospitals.
Location Organization Number of Patients Organization Name Total Resources
Loc-X140008 X140008 164 CLARKE MEMORIAL HOSPITAL 123180
Loc-X140010 X140010 493 VINGE UNIVERSITY HOSPITAL 380213
Loc-X140048 X140048 128 VALENTINE HOSPITAL 98723
Loc-X140049 X140049 110 JEMISIN MEDICAL CENTER 78772
Loc-X140054 X140054 268 WILHELM HOSPITAL 207397
Loc-X140062 X140062 249 GIBSON COMMUNITY HOSPITAL 197149
Loc-X140063 X140063 70 ROBERT GALBRAITH HOSPITAL 55172
Loc-X140065 X140065 116 BESTER MEMORIAL HOSPITAL 92083
Loc-X140068 X140068 80 WILLIS COMMUNITY HOSPITAL 61194
Loc-X140080 X140080 161 SAINT LOIS HOSPITAL 126112
Loc-X140082 X140082 121 PAUL FRENCH MEMORIAL HOSPITAL 91189
Loc-X140083 X140083 111 FARMERS HOSPITAL 85541
Loc-X140088 X140088 380 DAHL UNIVERSITY MEDICAL CENTER 303314
Loc-X140095 X140095 99 ST ROBERT HOSPITAL 75124
Loc-X140103 X140103 60 LABRADOR HOSPITAL 48636
Encounters for this test data set occur from December through February and are meant for testing the production of MeasureReport resources for the period from January 1, 2021 through January 31, 2021 for each of the facilities above.
See notes from NHSN for Connectathon 26 on use of Location and Organization identifiers and mappings.
Keith Boone (May 17 2021 at 21:09):
https://confluence.hl7.org/pages/viewpage.action?pageId=97470616#SANERTestDataSets-NHSNFacilityOrgLocationID%27s is the link to NHSN Notes
Keith Boone (May 17 2021 at 21:11):
I'm working on producing example MeasureReports reports for X140049 (Jemison Medical Center) b/c they are among the first four previously supported for mapping by NHSN.
Keith Boone (May 17 2021 at 21:12):
I have to tweak a few things b/c my MeasureComputer is not set up to query from a server containing data for multiple hospitals.
Keith Boone (May 18 2021 at 15:16):
http://test.sanerproject.org/fhir/Bundle has 31 bundles containing MeasureReport and Supplemental data resources for the month of January for one Hospital. I'm still working on figuring out why the MeasureReport resources themselves aren't being separately uploaded as they should be.
Keith Boone (May 18 2021 at 15:19):
This will get you the latest MeasureReport resources: http://test.sanerproject.org/fhir/MeasureReport?_lastUpdated=gt2021-05-18T00:00:00-04:00
Mark Deaton (Jun 23 2021 at 00:13):
@Keith Boone @David Pyke I'm trying to package up our importer for other groups potentially to use. It's the Organization resource I want to query to get the dots on our map; but I see that each organization has multiple records. I'd like to use a filter that only pulls records with a lat/lon address extension, but I'm having trouble applying a filter.
My basic approach was to try something like this: http://test.sanerproject.org/fhir/Organization?_format=json&_count=200&active=true&_filter=address.extension.extension.url%20co%20latitude
But it just doesn't seem to be the right format. Is this something you know about doing off the top of your head?
Thanks!
Last updated: Apr 12 2022 at 19:14 UTC
