David, thanks for the details on MD3. Comments embedded below jon -- Jon Patrick Chair of Language Technology Australian R&D Centre for Health Informatics School of Information Technologies University of Sydney
Quoting David Guest <[EMAIL PROTECTED]>: > Jon / Tim > > I think this is a great idea, although I am not really sure what an API > (http://en.wikipedia.org/wiki/Api) is. ;-) > > As I understand it, HCN is happy for you to access the information in > your databases. Their license says, "All supplied databases are > copyright protected. None of them may be copied, or used outside of the > Medical Director software program (Medical Director), or reproduced in > any way or by any process, without written permission from the copyright > owner." I interpret this to mean their database layout and stored > procedures are all copyright but accessing the data contained within > them is permissible. In fact, they provide a read only authentication > mechanism called hcnreader, that allows you to access the SQL data > directly. I have recently revisited this using OOo Base and their ODBC > connector as we talked about a few months ago > (http://ozdoc.mine.nu/index.php/Medical_Director). I wish people would use their terminology a little more precisely.Interpreting your interpretation I take the license to mean they claim copyright over their Database Management System (DBMS) and the schema design for their application - fair enough if they can show they indeed were the first programmers to use those names and tables for that type of data - did they build their own DBMS or are they using something propriety like Oracle, MS SQL Server, or Sybase? Does the hcnreader give you access to an SQL command line prompt or do you pass it SQL formatted statements from a program as with an ODBC driver. > > Accessing the data in MD3 is an area of increasing interest to many > Australian GPs. As you will have seen in last month's budget > (http://www.npcc.com.au/Media%20Releases/Budget_release.pdf) , the Feds > have enhanced funding for the NPCC (http://www.npcc.com.au/). In > addition to Cardiab, other commercial groups like the Canning Division > of General Practice > (http://www.canningdivision.com.au/lnk_downloads.html) have developed > data extraction tools for pulling relevant clinical data from the > various EHRs including MD2 and MD3. The areas they are focusing on are > IHD, DM and asthma. We are testing a prototype at the RPAH of an idea we have for a "Generalised Data Analytics Language". Its aim is to allow queries to be framed in a restricted natural language so as to to express a much wider range of queries than possible in "simple" SQL. It will also allow for the use of SNOMED expressions in retrievals, and contains notions of locality and calender - an underlying technology for this function is the Text to SNOMED converter I've talked about on the list previously, and ultimately the system will allow for SNOMED terminology searching of text fields. PArt of the heavy work in such a project is to have a translator that maps a pseudo natural language query into a complex set of SQL retrievals from the underlying application database. If this tool can be made usable on MD then it would help the many non-SQL literate doctors to express a richer set of analyses about their own data. This has all been applied to bolt on to the ICU IS known as CareVue. Do you think we get engagement from any interested parties to support funding to get it bolted on to other systems, even if onyl for experiemtnal purposes. > > I've also been puddling around in MS SQL in the last few days. The > appended scripts do fun stuff like identify those patients that (1) are > overdue for a DM SIP consult and (2) have IHD. (Kuangie, did I get it > right?) > > Writing a script to pull out data for a particular patient should not be > too difficult and this facility is already available through the MD3 GUI > with output in XML format. I presume this would be ideal for further > manipulation into other formats including SQLite > (http://www.sqlite.org/) for Google gears. > > I suspect that HCN would not be too keen on any "API" that would not > permit the writing of data directly to the database. I don't have too > much of a problem with this since data can be incorporated via the > existing pathology import facility into letters, investigations or > documents. I am unsure why you say this - surely HCN would NOT want a 3rd party code to be writing to their databases. > > One restriction at the moment with the HCN database is that progress > notes and letters are encrypted. This will be a problem for Jon since > his interest is in NLP (nosological linguistic programming :-D. Again I > understand the decrypter is available from HCN but for obvious reasons > they are a bit particular about who gets access. Yes this would be an issues for us as we are interested in analysing the notes. > > I'd certainly be interested in working with others to improve health > care for the Australian community by unlocking the potential held within > our databases. Perhaps Jon would be the best person to co-ordinate this > effort. I'm happy to serve a role here but I don't know the key players. cheers jon ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program. _______________________________________________ Gpcg_talk mailing list [email protected] http://ozdocit.org/cgi-bin/mailman/listinfo/gpcg_talk
