2148 View
5 CommentIn discussionComments enabled
In the category:
Undefined
Hi,
Following a brief discussion that we had April business meetings regarding mapping of these components, I am posting the query as a discussion item looking for feedback: can we consider "Age" as a property?
If we use the "process" alternative we could have two separate processes for "chronological age" and for "gestational age". The processes would be "life" / "the period from birth to death" vs. "gestation". This way we would avoid having to state the starting point or end point but we should also consider the future costs of "technological debt"! If we will eventually need this expressiveness, there will be a major cost adjusting existing content.
That seems tempting in some respects, but in actuality I don't believe that the processes are actually distinct, so modeling them as such could have some potential consequences (as I think you are stating). The continuum is from conception to death, and the difference is in the starting and ending points (there are other related considerations, of course, such as "personhood", which I'm not attempting to comment on here). So I think we would want to try to model it taking into account the different starting and ending points, if that's possible. But I don't recall exactly how you were suggesting it would look if we did that.
Reviewed the two modelling options with examples of use of additional attribute-values:
Quality observable with age as property
| Observable | : | Property | = | New | Age |, | Inheres in | = | Patient | / | Human being | OR | Human oocyte donor (person) | OR | Specimen |, | Precondition | = | New | At time of diagnosis |, | Techniques | = | Calculation Technique|
| Inheres In (similar attribute?) = | Patient | / | Human being | OR | Human oocyte donor (person) | OR | Specimen |,
| Precondition = | New | At time of diagnosis|
| Techniques | = | Calculation Technique|
Examples in SNOMED and LOINC:
Age in patient:
424144002 | Current chronological age (observable entity) |
30525-0 | Age (Age:Time:Pt:^Patient:Qn)
Age in someone other than patient:
443443002 | Age of carer (observable entity) |
68327-6 | Egg donor age (Age:Time:Pt:^Egg donor:Qn )
Age of fetus:
57036006 | Fetal gestational age (observable entity) |
18185-9 | Gestational age (Gestational age:Time:Pt:^Fetus:Qn)
Age of something:
167816006 | Semen sample age (observable entity) |
49546-5 | Age of Blood specimen (Specimen age:Time:Pt:Bld:Qn)
Age at X (X is a precondition):
412726003 | Length of gestation at birth (observable entity) |
445872007 | Length of gestation at time of procedure (observable entity) |
63932-8 | Age at diagnosis (Age at diagnosis:Time:Pt:^Patient:Qn)
Age by Y technique: e.g
123980006 | Bone age (observable entity):
29553-5 | Age calculated (Age:Time:Pt:^Patient:Qn:Calculated)
Age at X (precondition) by Y (technique)
444135009 | Estimated fetal gestational age at delivery (observable entity) |
Discussions:
Use Process Observable model for age ( as well as duration).
Ease of modeling uses cases like gestational age, neonatal age, chronological age and easier to express change from gestation to life
Expand the range of characterizes to include Procedures)
Add Inheres In = Patient when the attribute/value applies.
Fetal age characterizes gestation process and starts from conception up to and terminated by birth. Inheres in Fetal structure. Neonatal age is Age immediately after birth/delivery and characterizes life, which inheres In the patient (baby).
Change "sample degradation process" to "sample aging process"
Add definitions to the added processes (e.g. gestation process)
Chronological is required in the Chronological age to differentiate it from Biological age, bone age, etc.
Deprecate existing concept “105727008 | Age AND/OR growth period (observable entity) |
Remaining item:
What is the difference if any between "At time of birth" and "at time of delivery".
Is a precondition required for existing concept “424144002 | Current chronological age (observable entity) |". Should we remove description "age" from this concept and add it to New | Chronological age |
The range for Inheres In needs to change to include descendants of 125676002 | Person (person)?
Does the same extension applies to 14679004 | Occupation (occupation) | → e.g. Carer in “Age of carer”
OIMP meeting on November 6: 1) Discussed some of the remaining items as well as feedback from CTP call on November 1:
Neonatal age vs. Fetal age:
444135009 | Estimated fetal gestational age at delivery (observable entity) |
412726003 | Length of gestation at birth (observable entity) |
444135009 | Estimated fetal gestational age at delivery (observable entity) |.
What is the difference between "At time of birth" and "at time of delivery" in observable terms related to gestational age: is the difference URU?
Does "estimated" add anything to the meaning for the first concept--considering that gestational age are usually estimated by different methods?
424144002 | Current chronological age (observable entity) |. What does "current" mean? is it required in definition? If yes, a precondition will be required to define it to prevent incorrect subsumption of the concepts that are not "current".
It was suggested that to define these concepts properly we need to look at literature and contact the original submitter (many of these terms seems to be originated from NHS) for definition for these terms. Also ask them to provide a definition and a FSN that is more accurate. The group discussed that when reviewing the existing SNOMED CT concepts for modelling, we should strongly consider retiring some ambiguous observable entity concepts when we encounter them, specifically if we can't get any feedback from the original requester.
2) The group also discussed the extension of the range for Inheres In attribute to include descendants of 125676002 | Person (person), e.g. Mother, father, patient, etc. The group agreed that the extension of the range is required.
5 Comments
Farzaneh Ashrafi
Summary of discussions in OIMP call, August 2017 (To be continued in the next OIMP call):
Two models to consider:
1) Age as "property" (Quantity observable)
2) Age as "process duration" (Process observable, characterizes life process)
Swapna noted that there other LOINC terms that contain age as "age at". We need to consider them as well:
Also consider the following:
Daniel Karlsson
If we use the "process" alternative we could have two separate processes for "chronological age" and for "gestational age". The processes would be "life" / "the period from birth to death" vs. "gestation". This way we would avoid having to state the starting point or end point but we should also consider the future costs of "technological debt"! If we will eventually need this expressiveness, there will be a major cost adjusting existing content.
Rob Hausam
That seems tempting in some respects, but in actuality I don't believe that the processes are actually distinct, so modeling them as such could have some potential consequences (as I think you are stating). The continuum is from conception to death, and the difference is in the starting and ending points (there are other related considerations, of course, such as "personhood", which I'm not attempting to comment on here). So I think we would want to try to model it taking into account the different starting and ending points, if that's possible. But I don't recall exactly how you were suggesting it would look if we did that.
Farzaneh Ashrafi
OIMP meeting in Bratislava:
Reviewed the two modelling options with examples of use of additional attribute-values:
Quality observable with age as property
| Observable | :
| Property | = | New | Age |,
| Inheres in | = | Patient | / | Human being | OR | Human oocyte donor (person) | OR | Specimen |,
| Precondition | = | New | At time of diagnosis |,
| Techniques | = | Calculation Technique|
| Observable | :
| Property | = | New | Duration |,
| Characterizes | = | Life |, |Gestation/Development|, |Sample degradation|
| Inheres In (similar attribute?) = | Patient | / | Human being | OR | Human oocyte donor (person) | OR | Specimen |,
| Precondition = | New | At time of diagnosis|
| Techniques | = | Calculation Technique|
Age in patient:
424144002 | Current chronological age (observable entity) |
30525-0 | Age (Age:Time:Pt:^Patient:Qn)
Age in someone other than patient:
443443002 | Age of carer (observable entity) |
68327-6 | Egg donor age (Age:Time:Pt:^Egg donor:Qn )
Age of fetus:
57036006 | Fetal gestational age (observable entity) |
18185-9 | Gestational age (Gestational age:Time:Pt:^Fetus:Qn)
Age of something:
167816006 | Semen sample age (observable entity) |
49546-5 | Age of Blood specimen (Specimen age:Time:Pt:Bld:Qn)
Age at X (X is a precondition):
412726003 | Length of gestation at birth (observable entity) |
445872007 | Length of gestation at time of procedure (observable entity) |
63932-8 | Age at diagnosis (Age at diagnosis:Time:Pt:^Patient:Qn)
Age by Y technique: e.g
123980006 | Bone age (observable entity):
29553-5 | Age calculated (Age:Time:Pt:^Patient:Qn:Calculated)
Age at X (precondition) by Y (technique)
444135009 | Estimated fetal gestational age at delivery (observable entity) |
Discussions:
Use Process Observable model for age ( as well as duration).
Ease of modeling uses cases like gestational age, neonatal age, chronological age and easier to express change from gestation to life
Expand the range of characterizes to include Procedures)
Add Inheres In = Patient when the attribute/value applies.
Fetal age characterizes gestation process and starts from conception up to and terminated by birth. Inheres in Fetal structure. Neonatal age is Age immediately after birth/delivery and characterizes life, which inheres In the patient (baby).
Change "sample degradation process" to "sample aging process"
Add definitions to the added processes (e.g. gestation process)
Chronological is required in the Chronological age to differentiate it from Biological age, bone age, etc.
Deprecate existing concept “105727008 | Age AND/OR growth period (observable entity) |
Remaining item:
What is the difference if any between "At time of birth" and "at time of delivery".
Is a precondition required for existing concept “424144002 | Current chronological age (observable entity) |". Should we remove description "age" from this concept and add it to New | Chronological age |
The range for Inheres In needs to change to include descendants of 125676002 | Person (person)?
Does the same extension applies to 14679004 | Occupation (occupation) | → e.g. Carer in “Age of carer”
Action Item:
Farzaneh Ashrafi
OIMP meeting on November 6:
1) Discussed some of the remaining items as well as feedback from CTP call on November 1:
It was suggested that to define these concepts properly we need to look at literature and contact the original submitter (many of these terms seems to be originated from NHS) for definition for these terms. Also ask them to provide a definition and a FSN that is more accurate. The group discussed that when reviewing the existing SNOMED CT concepts for modelling, we should strongly consider retiring some ambiguous observable entity concepts when we encounter them, specifically if we can't get any feedback from the original requester.
2) The group also discussed the extension of the range for Inheres In attribute to include descendants of 125676002 | Person (person), e.g. Mother, father, patient, etc. The group agreed that the extension of the range is required.
Farzaneh Ashrafi and Suzanne Santamaria:
CC: Daniel Karlsson: Please consider updating Inception / Elaboration document for the above noted MRCM change