FHIR Chat · socialhistory · implementers

Stream: implementers

Topic: socialhistory


view this post on Zulip Mounika (Jul 19 2017 at 06:29):

Hi all, In our ipad app we want to capture social history of a patient.In social history we are capturing smoking status. So we are using Observation resource. In our screen we have fields like when the patient quits smoking and counselling for minutes are unable to match. So please help me.Thank you

view this post on Zulip Vadim Peretokin (Jul 19 2017 at 07:37):

Have a look at http://hl7.org/fhir/us/core/2017Jan/StructureDefinition-us-core-smokingstatus.html, that could help you with the smoking status

view this post on Zulip Mounika (Jul 19 2017 at 07:41):

@Vadim Peretokin I had already gone through the content in the link you have mentioned above but couldn't find.

view this post on Zulip Vadim Peretokin (Jul 19 2017 at 08:09):

Sorry - couldn't find what?

If you mean to record _when_ they quit smoking, you could record the status as 8517006 and an effectiveDateTime for when they did?

view this post on Zulip Mounika (Jul 19 2017 at 09:40):

@Vadim Peretokin We want to capture the fields counselling, for minutes.Those are not found

view this post on Zulip Vadim Peretokin (Jul 19 2017 at 09:41):

Yes, I don't think that would make sense to go into an Observation

view this post on Zulip Vadim Peretokin (Jul 19 2017 at 09:42):

I think you'd like something in the Encounter for that - see Encounter.length

view this post on Zulip Mounika (Jul 19 2017 at 09:46):

Okay @Vadim Peretokin Thank you

view this post on Zulip Lloyd McKenzie (Jul 19 2017 at 14:28):

Counselling is actually a Procedure (that takes place during an Encounter)

view this post on Zulip Mounika (Jul 20 2017 at 08:56):

Thank you @Lloyd McKenzie

view this post on Zulip Mounika (Jul 28 2017 at 13:28):

Hi @Lloyd McKenzie In social history observation, previously you said that Counselling is a procedure. But we would like to calculate counselling for minutes. In Performance element we have PerformedDateTime, PerformedPeriod...I want to record time for min/hrs. So which datatype should I use? Please suggest me. Thank you

view this post on Zulip Lloyd McKenzie (Jul 28 2017 at 15:54):

Are you looking at it from a billing perspective or from a clinical perspective? If the former, you'd use BillableItem. If the latter, then right now you'd need an extension or to calculate it from the start and end of performedPeriod. However, it seems appropriate to submit a change request for this, as a pattern of "Patient had 45 minute councelling session on Feb. 3" without indicating a start time is reasonably common.

view this post on Zulip Eric Haas (Jul 29 2017 at 18:45):

A change request for what exactly?

view this post on Zulip Lloyd McKenzie (Jul 29 2017 at 19:11):

To add duration to Procedure

view this post on Zulip Mounika (Jul 31 2017 at 05:50):

Okay @Lloyd McKenzie Thank you.
In our Ipad screen, we are capturing smoking status of a patient. So we have fileds like smoking, Pack per day, pack years, chewing tobacco, etc.. These all are mapped with Observation.component resource. When coming to counselling and for mins.We are using Procedure. My Question is... Wheather we should link procedure to observation resource or Can we use individually?

view this post on Zulip Mounika (Jul 31 2017 at 07:09):

Hi all, We are trying to retrieve alcohol Intake observation of a patient. We have fields like alcohol use, age started, type, daily amount, etc.. for those fields we are using observation resource also asking type of yes/no questions as..
sought treatement for alcohol abuse, involved in a 12 step program? So how to map these type questions to observation resource?

view this post on Zulip Lloyd McKenzie (Jul 31 2017 at 14:19):

It's certainly possible that both the Observations and the councelling would be linked to the same Encounter resource. Are there other linkages that would be meaningful?

view this post on Zulip Lloyd McKenzie (Jul 31 2017 at 14:20):

In terms of questions, you can put the question in Observation.code and the value in Observation.value[x]. If you're actually doing a proper Questionnaire, you can also look at QuestionnaireResponse.

view this post on Zulip Mounika (Aug 01 2017 at 05:10):

Okay @Lloyd McKenzie But, we are asking multiple questions. Shall I add slices to Observation.code?

view this post on Zulip Lloyd McKenzie (Aug 01 2017 at 05:15):

For separate questions, you'd typically need separate profiles on Observation - unless the questions were all so tightly coupled that they'd need to appear as components of a single observation. (Ask yourself the question of whether it'd typically be useful for someone to look at the answers to one of the questions without looking at the answers to all of the other questions.)

view this post on Zulip Mounika (Aug 01 2017 at 06:12):

@Lloyd McKenzie we are asking yes/no type of questions For example, sought treatement for alchol abuse, involved in a 12 step program, have you had withdrawl problem,seizures or blackout from alcohol or drugs, emergency medical attention required due to intoxication.For these type of questions Either we need to use Observation.component? or QuestionnarieResponse resource

view this post on Zulip Lloyd McKenzie (Aug 01 2017 at 06:42):

Each of those seem like independent questions. I can interpret the answer "Have you been involved in a 12 step program" completely independently from "have you had seasures or blackout from alcohol or drugs". So those should be separate Observation instances, not Observation.components. Observation.component is for observations that are tightly intertwined. For example "sensitivity=high" "antibiotic=amoxicillin". You can't really get any value from either of those unless they're together.

view this post on Zulip Lloyd McKenzie (Aug 01 2017 at 06:42):

So I'd say you need either a Bundle of Observation instances or a QuestionnaireResponse

view this post on Zulip Mounika (Aug 01 2017 at 06:57):

Okay @Lloyd McKenzie Thank you

view this post on Zulip Michelle (Moseman) Miller (Aug 01 2017 at 12:57):

FYI -- here are some related discussions about using Observation for Social History:
https://chat.fhir.org/#narrow/stream/implementers/topic/social.20history
https://chat.fhir.org/#narrow/stream/implementers/topic/Social.20History.20Observation.20example

Our implementation uses Observation.related to link together all of the related Observations. Thus, we have one overall Observation for Tobacco, which has multiple related Observations for each of the individual questions.

Here is our DSTU2 example of tobacco social history (with related observations for tobacco type, age started, readiness to change, etc.): http://fhir.cerner.com/millennium/dstu2/diagnostic/observation/

view this post on Zulip Mounika (Aug 03 2017 at 09:08):

Hello all, I had created a sample profile of Questionnarie for asking questions in social history. I have sliced the Questionnaaire.item element and added questions in Questionnaaire.item.text Please check it onceMyQuestionnaire.structuredefinition.xml

view this post on Zulip Mounika (Aug 03 2017 at 09:27):

Sexual History Profile : In our Ipad app, We want to capture sexual history observation of a patient. In that screen we have fields like Marital status, Sexual Orientation, Sexually Active, Number of sexual partners. So, I found loinc codes for all the above fields. Shall I add them as a components in obseravtion resource. Can anyone suggest me..

view this post on Zulip Eric Haas (Aug 03 2017 at 14:23):

using components instead related means these Observations can not be exchanged separately. The intent of components is for multipart observations that need to looked at together to understand the context of the measure and be able to make decisions based on it. Those observations sound like should be separate and grouped using .related in some sort of panel to me.

view this post on Zulip Mounika (Aug 04 2017 at 05:30):

Okay @Eric Haas Thank you. For asking questions, shall I use which resource Either QuestionnaireResponse or observation resource ?

view this post on Zulip Lloyd McKenzie (Aug 04 2017 at 15:44):

Questionnaire gives you control over how the questions are presented, how they're phrased, what order they're asked in and dependency controls such that answers to some questions can trigger or exclude the asking of other questions. However, the data gathered isn't terribly useful for subsequent querying and analysis. Observation makes your data fully comparible with other data of the same type, regardless of how that data was captured. It's actually common to use both - capture the data using QuestionnaireResponse to ensure tight control over data capture, then extract the information and put it into Observations for subsequent retrieval and analysis.

view this post on Zulip Mounika (Aug 07 2017 at 13:24):

Okay @Lloyd McKenzie Thank you. Can you please explain me any real time example.

view this post on Zulip Lloyd McKenzie (Aug 07 2017 at 22:14):

I'm not sure I'm understanding the question

view this post on Zulip Mounika (Aug 09 2017 at 12:56):

Hi all, Whenever we are asking number of questions, to capturing questions we are using Questionnarie resource. So I want to capture answers for those questions. For that we would use Either QuestionnarieResponse or Observation resource ?

view this post on Zulip Lloyd McKenzie (Aug 09 2017 at 17:50):

Responses to Questionnaires are intended to be captured using QuestionnaireResponse. It's possible that answers in the QuestionnaireResponse might also be propagated to Observation instances.

view this post on Zulip Mounika (Aug 10 2017 at 09:20):

Okay @Lloyd McKenzie We are capturing answers using QuestionnaireResponse How can an observation instance be obtained from a QuestionnaireResponse

view this post on Zulip Lloyd McKenzie (Aug 10 2017 at 15:06):

You'd have to map the questions to specific observation codes. You could use StructureMap to define a transformation between a QuestionnaireResponse and a Bundle of Observations (mapping based on the linkIds), or you could just write code.

view this post on Zulip Mounika (Aug 11 2017 at 10:34):

I have gone through the content of StructureMap resource.But unable to understand the procedure of how to use the resource for transforming QuestionnarieResponse to Observation.Please explain me. Thank you

view this post on Zulip Lloyd McKenzie (Aug 11 2017 at 22:06):

It's probably best to look at the StructureMap examples. You'll also need to get familiar with the FHIRPath language. It would be a good idea for us to include an example like that in the spec. Feel free to submit a change request proposing that

view this post on Zulip Mounika (Aug 16 2017 at 13:06):

Okay @Lloyd McKenzie Thank you

view this post on Zulip Mounika (Aug 17 2017 at 10:47):

Hi all, We want to capture Domestic Violence history of a patient. In that observation we want to retreive Offender, Type of child abuse. So can we add them as components to an Observation resource. Give me suggetions. Thank you.

view this post on Zulip Lloyd McKenzie (Aug 17 2017 at 14:52):

Using components seems approriate there. The additional properties you're capturing are qualifying the base observation and can't really stand alone

view this post on Zulip Mounika (Aug 19 2017 at 10:32):

Thank you @Lloyd McKenzie
We are capturing smoking status of a patient using Obseravtion resource. Here we want to capture How Long a patient is smoking in years for example 2-10 yrs, 10+years and Quit smoking(in years). please give me any suggestions. Thanks

view this post on Zulip Lloyd McKenzie (Aug 19 2017 at 13:39):

Give you suggestions about what @Mounika K? I'm not clear on where you're needing help.

view this post on Zulip Mounika (Aug 19 2017 at 13:46):

Okay @Lloyd McKenzie I mean we want to capture some fileds as "How long" and When he was "Quit" the smoking

view this post on Zulip Mounika (Aug 19 2017 at 13:47):

when we don't know th exact date. I want to mention only years

view this post on Zulip Lloyd McKenzie (Aug 19 2017 at 14:01):

So it sounds like you've got two questions, and thus two Observations. One is "How long have you smoked/how long did you smoke?" and the other is "How long ago did you stop smoking?". Is that accurate?

view this post on Zulip Mounika (Aug 21 2017 at 07:20):

Yes @Lloyd McKenzie I am creating a profile for capturing Quit smoking. I found snomed code for "stopped smoking" only. But we have valusets like 2-10 yrs, 10+ years etc..I did'nt find codes for these fields. Give me suggestions please

view this post on Zulip Mounika (Aug 21 2017 at 09:57):

Can anyone tell me how do we know, In which scenario we use Observation resource and where should we use Questionnarie resource. For example we want to capture history of suicidal thoughts, homicidal thoughts, diagonised with psychiatric problem, Consent for communicational between psychiatric and PCP.. please clarify me.

view this post on Zulip Grahame Grieve (Aug 21 2017 at 11:44):

the difference between Questionnaire and Observation is not about the semantics of the data, but about how you are managing the information. If you want a user input device, then you'd use the questionnaire approach. If you want to manage the data and filter it and apply understanding etc, then you'd use Observation. If you want to both things - then you'd probably use both

view this post on Zulip Mounika (Aug 21 2017 at 13:10):

@Grahame Grieve Thank you

view this post on Zulip Lloyd McKenzie (Aug 21 2017 at 15:45):

If you can't find a standard code in SNOMED or LOINC that matches what you need, then you may need to define your own code system.

view this post on Zulip Mounika (Aug 22 2017 at 06:04):

Okay @Lloyd McKenzie Thank you

view this post on Zulip Mounika (Aug 22 2017 at 06:12):

Hello all, I am creating a profile for capturing Alcohol intake of a patient. I want to know when he was "last Drunk". But I found snomed code of "Time of last oral fluid intake (observable entity)". Shall I use this code for last drink?

view this post on Zulip Jose Costa Teixeira (Aug 22 2017 at 07:38):

Do you want to capture when he last drank, or when he last got intoxicated ("drunk")?

view this post on Zulip Jose Costa Teixeira (Aug 22 2017 at 07:40):

in any case "last oral fluid intake" seems to be time that a person last drank anything, relevant for all sorts of procedures, and I think it would not make sense to profile it for alcohol intake. There may be a conflict there.

view this post on Zulip Mounika (Aug 22 2017 at 13:31):

I need to create a profile for Psychitric history using Observation and have to capture some yes/no type questions like History of psychiatric thoughts, homocidal thoughts, diagnonised problem and some text fileds like psychiatric name, phone number, Therapist name, phone number. Please suggest me. Thank you

view this post on Zulip Lloyd McKenzie (Aug 22 2017 at 14:01):

History of psychiatric thoughts or homocidal thoughts would be Observations. Diagnosed problem would be Condition. Psychiatrist and Therapist information would be on Practitioner. (Though you may want to look at capturing all of those things using a Questionnaire.)

view this post on Zulip Elliot Silver (Aug 22 2017 at 22:28):

Interesting, I've usually pictured using Practitioner for practitioners within your organization or that your organization has some sort of relationship with. Here, you're suggesting to use it to record an arbitrary practitioner from a patient's statement, potentially with incomplete information. It seems to me that in this case, unless it is a known practitioner, you want to create a contained Practitioner (although what it is contained in, I'm not sure). Having this practitioner show up in a practitioner search is probably not what you want in most cases.

view this post on Zulip Lloyd McKenzie (Aug 23 2017 at 04:00):

Whether it's contained or part of a document or message Bundle or registered as stand-alone using REST will depend on the context and how much information is available. If you've got a name and phone number, that's in principle sufficiently unique and resolvable to post as a stand-alone instance, though I agree the amount of information is a little light. So a contained resource is most likely.

view this post on Zulip Mounika (Aug 23 2017 at 05:20):

Okay @Lloyd McKenzie Thank you. But we want to capture both Psychiatric and Therapist information. So how to use Practitioner resource.

view this post on Zulip Mounika (Aug 23 2017 at 09:47):

When I am creating a profile for Alcohol History. I found some example in Simplifiet.Net. There is a valueset for alcoholUseStatus. http://decor.nictiz.nl/fhir/ValueSet/2.16.840.1.113883.2.4.3.11.60.40.2.7.3.2--20150401000000
Shall I use this valueset as reference to my profile directly. Suggest me

view this post on Zulip Lloyd McKenzie (Aug 23 2017 at 13:26):

The patient would have a CareTeam with multiple Practitioners each with a separate role. In terms of using a value set published by someone else, you certainly can. It's encouraged, in fact. However, make sure you're comfortable with them having custodianship and possibly changing the value set over time.

view this post on Zulip Brian Postlethwaite (Aug 27 2017 at 10:07):

The mapping from the Questionnaire to the resources is done using the Element mapping isn't it @Lloyd McKenzie ?
http://hl7.org/fhir/questionnaire.html#2.38.5.4

view this post on Zulip Lloyd McKenzie (Aug 27 2017 at 14:29):

There are a few options. Questionnaire.item.definition can point directly to the StructureDefinition.snapshot.element of the resource. Or it can point to a logical model element or data element which has mappings to the resource. Or you can have a StructureMap that defines the mapping from QuestionnaireResponse to a resource or Bundle of resources.

view this post on Zulip Mounika (Aug 28 2017 at 10:05):

Hi all. I had created a profile for Sexual History and created group of Observation profiles.
Shall I add the individual Observations to the"related" element by slicing it? Thank you.

view this post on Zulip Mounika (Aug 28 2017 at 13:28):

The patient would have a CareTeam with multiple Practitioners each with a separate role.

@Lloyd McKenzie How can I create a profile for Psychiatric History using Practitioner resource. Please suggest me

view this post on Zulip Lloyd McKenzie (Aug 28 2017 at 15:21):

Well, to start, you need to decide what your purpose is. If it's data capture, you might want to just use QuestionnaireResponse (where you don't need a profile, just a Questionnaire). If you're wanting the data to be queriable and manipulatable and integrated with all of the patient's other history, then you might look at submission of all of the information as a FHIR document or message or other form of Bundle. Once you've decided that, you'll know if you need to profile and what the base of your profile will be. Before creating the profile, it's often helpful to mock up a sample instance example showing all of the data you're wanting to share and how it will be organized - including interrelationships. From that you can then start building your profiles using a tool like Forge.

view this post on Zulip Mounika (Aug 30 2017 at 05:41):

@ Lloyd McKenzie
History of psychiatric thoughts or homocidal thoughts would be Observations. Diagnosed problem would be Condition. Psychiatrist and Therapist information would be on Practitioner.

how to map this careTeam to psychiatric History observation?

view this post on Zulip Mounika (Aug 30 2017 at 06:51):

Hi all, I want to capture Social History and It's having profiles of Smoking History, Alcohol History, Drug History, Sexual History, Psychiatric History and Domestic Violence History. So when compared to VitalSign Profile we take it as main Observation likewise Can we take Social History as main Observation or take it as Document. Calrify me

view this post on Zulip Mounika (Aug 30 2017 at 07:00):

(deleted)

view this post on Zulip Lloyd McKenzie (Aug 30 2017 at 18:39):

If you're creating a DiagnosticReport for "Psychiatric History", the only way I can think to link them is to have the CareTeam and the Observations both tied to an EpisodeOfCare representing psychiatric care.

view this post on Zulip Lloyd McKenzie (Aug 30 2017 at 18:40):

No need to introduce documents to capture SocialHistory - it can also be a grouping Observation (or DiagnosticReport)

view this post on Zulip Mounika (Aug 31 2017 at 08:41):

Okay @Lloyd McKenzie Thank you

view this post on Zulip Mounika (Sep 06 2017 at 09:38):

Hi all, I am creating a profile for Smoking History using "StructureDefinition-us-core-smokingstatus". So I want to add components as Amount, Type and Pack years to the us core profile. Is this correct ?
I am capturing Age start, Age cessation using Observation.effectivePeriod. We also want to capture Total time (months/years). So how to calculate Total time.

view this post on Zulip Mounika (Sep 06 2017 at 13:41):

I am creating a profile Alcohol History using Observation resource.Alcohol Intake as Observation.code and taking components as Alcohol types like Beer intake, Hard liquor intake, wine intake. Is this Correct?
Also want to calculate "Last Use" of Alcohol Use. How to capture that one.

view this post on Zulip Lloyd McKenzie (Sep 06 2017 at 14:04):

I don't think "Last Use" is calculatable - it'll have to be entered. That could be a component of any of the others. However, beer, hard liquor and wine all seem like separate observations. You can understand them without looking at the others. If total intake is actually a calculated observation, then it could have a "derived-from" relationship with the others.

view this post on Zulip Mounika (Sep 07 2017 at 06:22):

Thank you @Lloyd McKenziewhat data type we need to choose to ask last use of Alcohol?
We also want to ask some questions about AlcoholUse. So we took them as Observations they would comes under which type either "derived-from" or "has-member" type. Please clarify me.
And can you give me clarify the previous question of Smoking History.

view this post on Zulip Lloyd McKenzie (Sep 07 2017 at 13:08):

I was hoping @Brett Marquard or @Eric Haas who have been more involved with the US core profiles could respond to the smoking question. For last use, it depends on what type of answers you want. You could use date (which allows YYYY-MM-DD, YYYY-MM or YYYY. You could also use Quantity to capture a duration of days/weeks/months/years. Or you could allow string for free-text.

has-member is for things like panels. For example, a CBC (complete blood count) would have members capturing white blood cell count, red blood cell count, hemoglobin, etc.

derived-from is where one observation or measurement allows the determination of another. This might be as a calculation or simply supporting evidence.

view this post on Zulip Mounika (Sep 07 2017 at 13:15):

Thank you @Lloyd McKenzie

view this post on Zulip Michelle (Moseman) Miller (Sep 07 2017 at 20:11):

@Mounika K I had provided Cerner's Social History tobacco example earlier in this discussion thread: http://fhir.cerner.com/millennium/dstu2/diagnostic/observation/
Essentially, our implementation uses Observation.related to link together all of the related Observations within one category of SHx (such as alcohol or tobacco). For example, we have one overall Observation for Tobacco (Risk) with related "has-member" observations for the other questions, such as tobacco type, readiness to change, previous treatment, started at age, etc.

view this post on Zulip Michelle (Moseman) Miller (Sep 07 2017 at 20:20):

US-Core IG has a profile for tobacco use (http://hl7.org/fhir/us/core/ValueSet-us-core-observation-ccdasmokingstatus.html), which manifests in our implementation as one of the related observations associated with the overall tobacco (risk) observation. I don't believe the US-Core profiles specify how to represent the "age started". Cerner doesn't use effectivePeriod for that. Instead, we would have a related Observation with
Observation.code = SHX Tobacco started at age
Observation.value = 22 Years

view this post on Zulip Mounika (Sep 08 2017 at 05:06):

Okay @Michelle (Moseman) Miller
what about the server performance ? If we create these many Observations.

view this post on Zulip Mounika (Sep 08 2017 at 05:09):

Effectiveperiod:
In simplifier.net(https://simplifier.net/NictizSTU3/zib-TobaccoUse) they have taken effective peroid as start and quit date of smoking.

view this post on Zulip Mounika (Sep 08 2017 at 05:31):

Previously you mentioned about Cerner's "Social History tobacco example". Is there any profile for Social History

view this post on Zulip Brett Marquard (Sep 08 2017 at 13:56):

The US Core Smoking Status profiles is just a snapshot of smoking status at a specific time. It doesn't provide guidance for age started. This would require a different observation

view this post on Zulip Eric Haas (Sep 08 2017 at 23:57):

(deleted)

view this post on Zulip Mounika (Sep 09 2017 at 13:13):

@Lloyd McKenzie
No need to introduce documents to capture SocialHistory - it can also be a grouping Observation (or DiagnosticReport)
May I know the reason why documents are not needed to capture Social History? Thank you.

view this post on Zulip Lloyd McKenzie (Sep 09 2017 at 17:34):

In FHIR, documents are one of four exchange paradigms. (The others are REST, messaging and services.) You can choose to put anything you wish in a document, but you don't need to use documents. Documents add strict control over the order in which data is presented. However, all of the essential relationships between information is there whether you have documents or not. So it's perfectly possible to capture social history as a DiagnosticReport or parent Observation with associated Observations and not introduce documents at all.

view this post on Zulip Mounika (Sep 11 2017 at 13:28):

Thank you @Lloyd McKenzie

view this post on Zulip Mounika (Sep 12 2017 at 09:27):

shall I create an observation profile for each question individually?
main profile : Alcohol intake observation profile
childs(related) :
ex.1.observation profile - Have you ever felt Guilty about drinking?
2. observation profile -Have people Annoyed you by criticizing your drinking?
and etc.

view this post on Zulip Mounika (Sep 14 2017 at 05:41):

Hi, I have created profiles for smoking history, alcohol history, drug history, sexual history, psychiatric history individually. Do I need to create a main profile as Social History like vitalsigns profile?

view this post on Zulip Lloyd McKenzie (Sep 14 2017 at 05:54):

It depends on whether you want to require the observations from those other profiles to be organized under another observation or DiagnosticReport.

view this post on Zulip Mounika (Sep 14 2017 at 05:56):

ok. But which one is the best way?

view this post on Zulip Jayashree Surnar (Sep 14 2017 at 06:14):

hello all,
i'm creating the profile for HitoryOfDrugUse using Observation Resource.Main profile has the child profiles namely
1. details of drug(oxycode,vicodin,Xanax,ritalin etc).
* type,method,Age at start, age at cessation,frequency, total time
2.family history of drug use : yes/no
3.use drugs because of
*options:anger,stress,depression etc *
4.what is the longest time you were completly sober and drug-free in your life?
5.how long were you completly sober and drug-free in the last 6months(approximately)?

here my doubt is how to create a profile for 1. details of drug, as it has ~50 drugs so do i need to create 50 individual profiles here? or is there any other way to do this?

Thank you.

view this post on Zulip Lloyd McKenzie (Sep 14 2017 at 06:22):

@Mounika K I'm not sure there is a best - it depends what you're wanting to use. You can group them - if you have a use for a group. If not, no point creating a group.

view this post on Zulip Lloyd McKenzie (Sep 14 2017 at 06:23):

@Jayashree Surnar That sounds like either a QuestionnnaireResponse or a set of Observations grouped under a parent Observation (which you use is driven by what you want to do with the data

view this post on Zulip Jayashree Surnar (Sep 14 2017 at 06:34):

@Lloyd McKenzie , That sounds like either a QuestionnnaireResponse or a set of Observations grouped under a parent Observation (which you use is driven by what you want to do with the data
like vital signs profile?

view this post on Zulip Mounika (Sep 14 2017 at 09:39):

@Lloyd McKenzie
Each of those seem like independent questions. I can interpret the answer "Have you been involved in a 12 step program" completely independently from "have you had seasures or blackout from alcohol or drugs". So those should be separate Observation instances
Shall I need to create individual Observation instances or Observation profile. Please clarify me.

view this post on Zulip Lloyd McKenzie (Sep 14 2017 at 18:36):

Yes, those will be separate Observations and profiles - just as for vital signs

view this post on Zulip Jayashree Surnar (Sep 15 2017 at 07:18):

@Jayashree Surnar That sounds like either a QuestionnnaireResponse or a set of Observations grouped under a parent Observation (which you use is driven by what you want to do with the data)
sorry i didn't get That sounds like either a QuestionnnaireResponse.
here QuestionnnaireResponse ? or Questionnnaire?

view this post on Zulip Jayashree Surnar (Sep 15 2017 at 13:09):

@Lloyd McKenzie
Shall i create a generic profile(like: https://simplifier.net/NictizDSTU2/zib-DrugUse) for details of drug.which has

  • type,method,Age at start, age at cessation,frequency, total time as the components? and use. so that no need to create a profile for each drug type separately. Is it correct?

view this post on Zulip Lloyd McKenzie (Sep 15 2017 at 14:30):

Correct. One profile that allows indication of the drug is fine. However, you'll need to think about whether those observations should be components or separate observations. They each seem like they could be understood on their own. And I don't think it's true to say they are almost always captured and reviewed together. So what you'd likely have is an Observation that said "At what age did drug use start?" as the code, and components for drug and method. Then a separate Observation of "At what age did drug use end?" with the same components, and the same sort of thing for frequency and total time. You could choose to create a grouping Observation based on drug as well if you wished (e.g. called "Drug usage") with the same components, but there'd be no need to do that.

view this post on Zulip Lloyd McKenzie (Sep 15 2017 at 14:31):

Components are intended for data that isn't really intended to be looked at by itself or is rarely looked at or captured by itself.

view this post on Zulip Mounika (Oct 18 2017 at 06:13):

Hi all, I had created a profile for DrugUse. I want to capture some Drug or medication codes. So I added the MedicationStatement as an extension to the Observation resource for capturing medication codes. Is this a right process? Can anyone tell me.

view this post on Zulip Jose Costa Teixeira (Oct 18 2017 at 07:27):

That is an interesting point. IMO, medicationStatement is kind of an "observation" on its own, so normally you would not need to add it to an observation.

view this post on Zulip Jose Costa Teixeira (Oct 18 2017 at 07:28):

Could you not use the medicationstatement alone? Why do you need the observation? (note that I'm just asking, because I am not sure of your scope and intent)

view this post on Zulip Eric Haas (Oct 18 2017 at 17:18):

agree that you would not use Observation linked to MedStatement. That is indirect way of doing it. Instead of using Observation to say ' here is the Observation for drug use' just use MedStatement.

view this post on Zulip Mounika (Oct 30 2017 at 09:59):

Hello all, I had created a profile for Smoking status using Observation resource and I want to use Procedure resource to capture Counselling . So how to link the Procedure to Observation resource.

view this post on Zulip Lloyd McKenzie (Oct 30 2017 at 18:45):

Procedure.reasonReference can point to the Observation

view this post on Zulip Mounika (Oct 31 2017 at 05:28):

Thank you @Lloyd McKenzie


Last updated: Apr 12 2022 at 19:14 UTC