On Fri, 22 Oct 2004 12:22:10 -0600, David Forslund <[EMAIL PROTECTED]> wrote: ... > > Right - of course, we still need a reliable way to > > transfer/synchronize records between different doctors' offices, > > hospitals, etc. MPI is exactly designed to serve that function - but > > there are other ways to do it. > > More than an MPI is needed. Tags as to the origin of the data need > to be provided in some way so that data duplicates can be handled. These > are distinct from the process of identifying the patient.
David, Good point! It seems to me that we really must get together and work on this. Maybe via the hxp effort? http://hxp.sourceforge.net/ ... > > What ends up happening in the absence of a "master" patient index > > (MPI) is that we use record-location, date of birth, SSN, + name to > > serve the record matching function. This is the current state of the > > field. It has advantages and disadvantages - but adopting electronic > > medical records systems (EMR) does not mean we have to adopt MPI. > > Well if every location or system uses a different identification mechanism > it makes the patient identification even harder, in my opinion. You are effectively > adopting an MPI process. Having a relatively more uniform process and > separating out the service explicitly would make it all much easier. I agree. Maybe we can work out a sufficiently uniform process and implement it in our respective software packages? Best regards, Andrew -- Andrew P. Ho, M.D. OIO: Open Infrastructure for Outcomes www.TxOutcome.Org
