Re: [Gnumed-devel] Estimated age

[vbanait]

 >Maybe we just add a .dob_is_estimated field next to the .dob field.
Which is maybe a simpler solution than what I came up with - see
screenshot for possible appearance.

I agree with you and the screenshot looks great.

>What would be the data type of this value? Would it be a *date* wherein the 
>user >would, in the case of "estimated 80 years" enter       Jan 1, 2002,      
> or would >they enter Jan 20, 1932 (i.e relative to current date),
> or would they enter only 1932 (which is not a date at all)

Logically it should be only year eg. 1946

>Also, should such a value be represented on the screen (to make clear that the 
>actual dob is not known with certainty) by use of a tilde ~

I think screen should say about approximate age only, when we are not
sure of date, month or the year.

vaibhav



On 1/20/12, Busser, Jim <address@hidden> wrote:
> On 2012-01-19, at 9:15 PM, Vaibhav Banait wrote:
>
>> Hi
>> Many countries will have population which is not aware of dob. Therefore,
>> if we put estimated age as a demographic optional parameter with
>> calculations of further age from estimated year, it should be very useful.
>> If I am not very clear, please have a look at online Openmrs demo.
>> Regards
>> Vaibhav
>
> Instead of creating a
>
>       date of birth variant = year of birth
>
> in the identity table (because in this instance it is of no real use in
> identification) I am thinking we are talking this age proxy as a "clinical
> measurement" to be used with other clinical parameters in things like risk
> prediction.
>
> I am wondering therefore whether this entity could fit into a clinical
> measurement.
>
> If yes, then for example LOINC has a code
>
>       30525-0         Age
>
> which is age in seconds at a point in time. Therefore if at a visit it is
> believed or indicated that a patient is 69 years old with no precision on
> date of birth, the measurement of age of
>
>       69 years
>       x 365.25 days
>       x 24 x 60 x 60
>
>       = age (in seconds) as of the datetimestamp of entry.
>
> In order for this to be useful, any computations in GNUmed would need to
> have a construct
>
>       age =
>
>               if date of birth is non-null
>
>                       then age = current datetime - date of birth
>
>               elif a value for LOINC 30525-0 exists
>
>                       then age = current datetime - (datetime of measurement) 
> + value of entry
>
> I would wonder whether it should be somewhere in the praxis configuration
> that GNUmed could be instructed where to look for a definition of  this
> alternate definition of age.
>
> -- Jim


-- 
Vaibhav S Banait
MD (Med), DNB (Med),
DM (Gastroenterology),
Ph: Vaibhav:09860643436