Smoking status missing from sample patients?

[Farhan Ahmad]

Hi,

Based on the document with sample patients, SMART, Joe (id: 4342010) should have a smoking status (LOINC code 72166-2), but the Observation search does not return it.  It's the same case for the other relevant test patient SMART, Wilma (id: 4342008).  Here is my search...

https://fhir-ehr.sandboxcerner.com/dstu2/0b8a0111-e8e6-4c26-a91c-5069cbc6b1ca/Observation?patient=4342010&_count=200

The above returns about 20 observations total without any sign of a smoking/tobacco status.

Is there another patient that we can use to test the smoking status?

Thanks!

Farhan Ahmad

CTO & Chief of Staff, HealthDecision

http://www.healthdecision.org

[Jenni Syed]

Hi Farhan,

I ran the query above using the Argonaut user and am seeing the smoking status come back (Observation 6309883-q6930400).

Are you signed in as a different user? If so, which one?

~ Jenni

[Farhan Ahmad]

I am using portal/portal credentials that are mentioned on the Developer portal.  I am pasting a screenshot below.

Thanks,

Farhan

[Matthew Henkes]

Hi Farhan, We've updated the privs on the portal/portal user. You should be able to get smoking status back now. Give it a try and let us know.

Thanks,
Matt

[Farhan Ahmad]

Thanks, Matt.  I will test it out again this week.

- Farhan

[jha...@healthdecision.org]

Matt, 

I am a colleague of Farhan. I just ran his query and another that specifically requests the smoking status. I'm still not seeing it returned.

This is the search I just ran: 

Observation?subject%3APatient=4342010&code=http%3A%2F%2Floinc.org%7C3141-9&code=http%3A%2F%2Floinc.org%7C8302-2&code=http%3A%2F%2Floinc.org%7C2093-3&code=http%3A%2F%2Floinc.org%7C2571-8&code=http%3A%2F%2Floinc.org%7C2085-9&code=http%3A%2F%2Floinc.org%7C13457-7&code=http%3A%2F%2Floinc.org%7C55284-4&code=http%3A%2F%2Floinc.org%7C72166-2&_count=50

I'm using the same portal/portal credentials.

Could you take another look at this?

Thanks,

Joseph 

On Thursday, January 5, 2017 at 10:25:44 AM UTC-6, Matthew Henkes wrote:

[Jenni Syed (Cerner)]

Hi Matt,

I'm guessing you don't intend for those codes to be separated by ampersand. In FHIR, this means "AND" whereas comma separation means "OR" [1]. In order to get a list of items matching any of the loincs below, you would have one code parameter with a bunch of comma-separated values. EG:

Observation?subject%3APatient=4342010&code=http%3A%2F%2Floinc.org%7C3141-9,http%3A%2F%2Floinc.org%7C8302-2,http%3A%2F%2Floinc.org%7C2093-3,http%3A%2F%2Floinc.org%7C2571-8,http%3A%2F%2Floinc.org%7C2085-9,http%3A%2F%2Floinc.org%7C13457-7,http%3A%2F%2Floinc.org%7C55284-4,http%3A%2F%2Floinc.org%7C72166-2

Also, you'll need to query by this code for now for tobacco usage answers: http://snomed.info/sct|229819007 (we have an enhancement coming in just to get the usage answer, but it's not there yet)

[1] http://hl7.org/fhir/dstu2/search.html#combining

[Jenni Syed (Cerner)]

That should have been to Joseph - sorry...

[Farhan Ahmad]

Thanks, Jenni.  Joe and I are tag-teaming on this integration so you will hear from both of us.  I see that http://snomed.info/sct|229819007 has 83 different findings (listed in the References/Interprets section under the Concept Details tab).  Does Cerner support all 83 of these or should we be targeting a shorter list?

[Jenni Syed (Cerner)]

We usually have 8 question/answers, I believe, though they don't all get answered by the patient. MU requires at least the one regarding use (loinc 72166-2) to be exposed in the API. The sandbox is currently in a mixed state for Social Histories like this (hence, the snomed at the grouper level but loincs at the lower) - but this is something we're working on improving, as well as enabling social histories to be queried at the question level rather than only the higher grouper. 

A more robust example of a smoking question/answer list can be seen here: https://fhir-open.sandboxcernerpowerchart.com/dstu2/d075cf8b-3261-481d-97e5-ba6c48d3b41f/Observation?patient=1316024&code=http://snomed.info/sct|229819007 (note that some questions have multiple answers).

Currently, the only one we try to ensure that is coded is the use one for meaningful use in sandbox. 

What does the app currently require?

~ Jenni

[Farhan Ahmad]

Currently, most tools require yes/no status, but one tool (Lung Cancer) requires smoking status (current, former, never smoked, and unknown), start and quit dates, and packs/day.  We can work with varying levels of info so these are not required and the goal is to get as much of this data as we possibly can.

Thanks!