problem/diagnosis screening Y/N and SNOMED coding

[Maximilian_Meixner]

Hey everyone,

we are performing a clinical study (pneumological use case) for which the following problems should be recorded by a clinician in a simple Yes/No fashion:
coughing, coughing up sputum, dyspnea, chest tightness, cyanosis, peripheral edema, lung hyperinflation, barrel chest, chest wall retraction.

First, I grouped these in the following two groups:
A, symptom-based: coughing, coughing up sputum, dyspnea, chest tightness
B, inspection-based: cyanosis, peripheral edema, lung hyperinflation, barrel chest, chest wall retraction

I´m currently using symptom_sign_screening.v1 for A and problem_screening.v1 for B, e.g. like this:

where I use a SNOMED Code for the symptom/sign name and Yes/No via “Presence?”

I have several questions to this:

1. Can these screening archetypes generally be used in that way (for clinicians) or are they intended to be answered by patients only?
2. Regarding the simple Yes/No answer, are these archetypes the right choice or would it be better to use problem_diagnosis.v1 and exclusion_specific.v1?
3. when using SNOMED Codes as shown above, is it a problem to mix findings, e.g. 3415004 Cyanosis (finding), with disorders, e.g. 24228002 Barrel chest (disorder), in different clusters of the same archetype?
   My understanding is that a finding is the result of an examination, however here we are not recording the examination itself.
   This also makes me re-think which of the above mentioned problems is really just a symptom and what is already a diagnosis. In our study, the patients have the primary diagnosis of asthma or COPD, which is recorded elsewhere. Here we just want to know if they have these additional problems or not.

Thanks for your comments!
Kind regards,
Max

[heather.leslie]

Hi Max,

I commonly model forms which list a range of parameters such as your example of “coughing, coughing up sputum, dyspnea, chest tightness, cyanosis, peripheral edema, lung hyperinflation, barrel chest, chest wall retraction” and I separate them out into appropriate screening questionnaires as you suggest. This is largely how the evolving family of screening questionnaires has evolved.

@Maximilian_Meixner

Can these screening archetypes generally be used in that way (for clinicians) or are they intended to be answered by patients only?

These archetypes are designed for use in a range of non ‘EHR’ scenarios - patient, investigator, clinician, anyone. They have been used extenively for recording clinician or research investigator responses in secondary use eg disease registries or surveillance where the clinical data remains recorded in the original EHR problem lists or exam findings.

@Maximilian_Meixner

Regarding the simple Yes/No answer, are these archetypes the right choice or would it be better to use problem_diagnosis.v1 and exclusion_specific.v1?

If the use case is for screening and the Yes/No is appropriate use this family of archetypes. If it is to kickstart a Problem list then a User interface with Yes/No could be transformed to data using EVAL.problem_diagnosis +/- exclusions if relevant.

@Maximilian_Meixner

when using SNOMED Codes as shown above, is it a problem to mix findings, e.g. 3415004 Cyanosis (finding), with disorders, e.g. 24228002 Barrel chest (disorder), in different clusters of the same archetype?

This may be a reason to create a new archetype designed to align with the existing published screening questionnaire archetypes but targeting screening for physical findings on exam, which will support correct querying from the correct hierarchies.

[Maximilian_Meixner]

Thank you Heather for your reply and explanations. This answers already a lot of my questions!