[
Lists Home |
Date Index |
Thread Index
]
Bullard, Claude L (Len) wrote:
>Precisely. And there are the aliases as well. And
>is Smith, John J. (1) born on Aug. 12 1954 the same as
>Smith, John J. (2) born on Aug 12, 1954 and are either
>of them lieing about that.
>
>
Oh, those are just the well known problems with names as identifiers :-)
Apparently there are names used by several ?South American? tribes,
which have components which are only to be shared with family members
etc. -- hence the need for a <given
confidentiality="foo">Fribbet</given> attribute. Names and birthdates
are not always good unique identifiers, particularly when you can easily
have *multiple* John Smiths on the same patient floor etc. which is why
hospitals use their own unique IDs -- social security numbers are not
entirely without problems, in two situations:
1) SSNs are *reused*, ostensibly when people die, but nonetheless.
2) Illegal immigrants often "borrow" a SSN from a relative. Not great
when they don't also share the same medical conditions, allergies,
medications etc.
In cases where it is really important to know who we are dealing with,
we attach a photo to the chart.
But I digress...
In any case to deal with aliases in specific is why I have the "type"
attribute on the <person.name> element. e.g.
<patient>
<person.name>
<given>John</given>
<given>William</given>
<family>Smith</family>
<suffix>Jr.</suffix>
</person.name>
<person.name type="alias">
<given>Jimmie</given>
<family>Buffet</family>
</person.name>
<id authority="U.S. Social Security Administration"
type="SSN">111-22-3333</id>
...
</patient>
In any case, that is perhaps more detail than needed for many
applications, but for me its everyday stuff :-)
Jonathan
|
