The Person Unique Identifier feature will be included in GNU Health 2.8 to be used in a generic distributed environment, with multiple GNU Health instances across a nation or region.

Adding to the considerations already mentioned in the thread[1] , the following functionality will be default in GNU Health 2.8:

1) The PUID is subject to legislation and nomenclature changes, so, although is unique, we should be able to adapt it in the future

2) A reconciliation process can also take place, which could consist of a O2M field, linking the positive ID with the other instances of that person that was registered / treated as an NN . TBD

3) The party "code" field will be used as the constant, permanent code for the record. It will be used in the Tryton synchronization engine . The code field, by default is composed by the institution name (company) and the python UUID[2] uuid.uuid4 function[3]("institution-uuid.uuid4()") . This "code" field will be always the same, since it's an "internal" id, not bound to any legislation, or reconciliation.

4) When creating the person record, if no PUID is entered, a 9-char random alphanumeric string with the format of NNNXXXNNN will be associated to it. This is the default and can be customized.

5) If the person has no means of a positive ID (see thread[1]) or wishes to remain anonymous, an "NN-" prefix will be appended to the PUID.

6) There will be a validation module that will be used to make sure that the PUID meets the country nomenclature. It can be adapted to meet changes in the legislations or to different countries / regions. The customization would be part of the localization project.

1.- http://lists.gnu.org/archive/html/health-dev/2014-09/msg00074.html

2.- UUID : http://en.wikipedia.org/wiki/Universally_unique_identifier

3.- https://docs.python.org/2/library/uuid.html