Re: [Health-dev] Should PUID be a compulsory field when creating new patients?

[Luis Falcon]

Hi Andrew !

Thanks for your mail ! Let me try to explain the current status and
upcoming developments on this matter.

On Sun, 21 Sep 2014 12:33:35 +0100 (BST)
"address@hidden" <address@hidden> wrote:

> Hi Luis
> 
> I have just noticed that when creating a new "patient" in GNU Health 
> that the PUID field is optional. I think that it should be made a
> compulsory field otherwise it will be possible to create duplicated
> records for the same patient. And we all know what havoc that
> creates...
> 
In version 2.6, the PUID is not required, so we can have unidentified
("NN"[1]) patients, that are unable or unwilling to show any valid /
credible identification means. Some scenarios are :

- The patient wishes to remain anonymous 
- Unconscious patients
- Major disasters

Once the institution is able to find a positive ID, or the patient does
not want to be anonymous anymore, then, we can set an ID (new or
existing).

So, currently GNU Health allows empty IDs. At the moment you
put any ID on the PUID, this must be unique.

For 2.8, and especially thinking about distributed / health
institution networks, and the unique/universal person identifier across
the public health system, we have to consider the following situations.

Scenario A : The NN person has no previous record .

Scenario B : The NN person already exists in the system. This implies
demographics, medical history, etc.. is already in the system.


Steps to take in 2.8 series, to come up with a generic and scalable
approach to NN .

1) The field PUID will be required. The PUID will be used in
synchronization processes in distributed environments / national health
networks as the person ID. So, whether is a "NN-" or a positive ID, the
creation of the person will have this value. Also, this patient
might be derived to another institution, and the initial
evaluation and procedures will be available upon arrival, with the
unique ID (again, regardless is NN or not).

2) If we set the "unidentified" attribute, then the "NN-"
prefix will be set.

3) We have to be extremely careful with "reconciliation" of NN to
positive / valid IDs. For example, the patient might have been
transferred to another institution and will be recognized by the unique
NN number. Another situation is that some tests might have been ordered
with that unique NN number. 

4) Upon a positive ID, the "NN" code will be set on the list of
"alternative IDs".

5) The field "code" at party model (person) will have the institution
name as a prefix. This has to do again with the synchronization model
when having multiple health institutions and a UNIQUE / UNIVERSAL
person record and patient history.

Some of these features are already in place and I will push them
during the weekend to the development branch. I have now a 12-hour
flight ahead, so I will work a bit on it until Morpheus shows up :) 

Your comments, thoughts are most welcome ! :)


All the best
1 .- http://en.wikipedia.org/wiki/Nomen_nescio


> Regards
> 
> Andrew
> 
>