Dear colleagues, In many ways, I view the OSHCA meeting as an extension of this very productive mailing list. Thus, I am necessarily biased by the joy of meeting some of you in person, many for the first time. With that bias in mind, I hope to share what I learned so that we can build greater understanding of each other's work - leading to even quicker achievement of our shared goals. Since Joseph already summarized the overall success of the meeting, I will simply agree with his congratulations to the organizers and proceed to other details. (BTW, the food was quite good.) 1. For those who could not physically attend the meeting, streaming video of the meeting will be available (edited and hosted by Mary Kratz via Internet2). This makes it possible to verify what I am about to write :-). Also, even those at the meeting (including myself) missed nearly half of the meeting due to division into two tracks. I look forward to viewing the half that I missed. 2. Many prominent absences were noted. FreeMed, FreePM, OpenEMed, and LinuxMedNews to name just a few. In the interest of balanced reporting, it would be nice to learn what they have been upto. 3. David Chan (OSCAR) kicked off the meeting with the "cautionary tale" of the MUFFIN/OSCAR project. It appears that centrally-developed free software (MUFFIN), even with local user groups, is not adequate to support successful dissemination. OSCAR is a new public-funded project (=web-based MUFFIN) which should have entered production last week. Hopefully, lessons from MUFFIN will move OSCAR and other free software projects down a different path. 4. David Pepper showed MedMapper (www.medmapper.com) in TkFp. It seems like a promising user interface for structured data entry. A text note is also automatically generated when a path through a clinical algorithm is traversed. 5. Horst Herb from GNUMed (www.gnumed.org) and GNotary made an impressive demo of text encryption and trusted-third-party signature. GNUMed seems to be progressing rapidly with over 200 active contributors according to Horst. 6. Simion Pruna (BSTD at http://www.telemed.ro/web_bstd/Trio_new.htm) described his diabetes registry project based on GEHR-1 (not GEHR with archetypes). It is unclear how hard it would be to re-purpose/extend the software for anything other than diabetes. [I just discovered the link above. The software is now available for download. It seems to be written in C++ and runs on Microsoft Windows and requires Microsoft Access database.] It is indexed by OIO Library here: http://www.txoutcome.org/scripts/zope/library/files/browse/show_contents/objectid-235 7. Stan Huff (LOINC) described the goals and structure of LOINC (http://www.regenstrief.org/loinc/loinc_information.html). According to Philippe Ameline (and I concur), we are nowhere close to universal semantic interoperability. LOINC, however, could be quite useful as a free and readily accessible common ground. 8. I demo'ed OIO. Bob Mayes told me afterwards that he worked with large (Medicare reporting) systems that used similar plug-and-play metadata approach. Mary Kratz told me she finally understood what OIO is trying to do. This made me quite happy. 9. SMARTIE (http://www.mobi-dev.arakne.it/) - this is an interesting kind of project. They are funded by EU to make free "calculator-sized" software for PDA devices and web-clients. For example, electrolyte/fluid replacement applications. The catch is that the development tool will be proprietary. Nothing will be available until late 2003 (2 years from now). It's really really possible that SMARTIE will not be nearly so smart in two years. Second day... 10. I missed quite a few projects because of parallel track. It will be great if someone can share what they learned from these presentations: Physionet, Medzope, Odyssee, OpenEHR, Medal, and ??? 11. Physionet (Isaac Henry - www.physionet.org) is a repository of EEG and other biological signals. The idea is similar to Genebank where reseachers can deposit their original data for replication or data pooling/re-analysis. There is also a collection of software at the site for handling physiologic data (under GPL). 12. Medzope (Jon Edwards - www.medzope.org) is web-portal software for clinics/hospitals. Implemented using Zope and based in U.K. OSHCA2001 is the launch event for MedZope! The software is available for download under GPL. A mailing list is also active. 13. Odyssee (Philippe Ameline - www.nautilus-info.com) [I had problem accessing the site just now. Don't know if the URL is current.] This is a semantic tree constructing and utilizing system for medical records and decision support. I don't know what's new with this project but I am very interested to get an update. 14. Open source in resource-constrained countries. I went to this session because I am curious whether there is any difference in open source opportunities between Los Angeles and other resource-constrained countries. It seems that we are all in the same boat. The difference may come down to Bob Mayes who is looking to fund the development of an open source laboratory information system for Africa but not for Los Angeles :-). 15. VistA (the VA medical record system) is big, powerful, well-tested, free, and established in many parts of U.S. and Europe. The big question is why we don't all use it??? Perhaps if we build some good customization tools and make it work like (or with) GNUMed, GEHR(with archetype), and OIO - we will have the ultimate tool??? 16. Ed Hammond (AMIA president-elect) outlined a proposal for a comprehensive open source PDA software project. Still quite preliminary and in the brain-storming stage. It is also unclear whether a centralized funding/project management model will undermine the open source development method. 17. HL7 - there was a good debate about access to HL7 specs. It seems that payment is required to view the HL7 specs - which gives the impression that HL7 is proprietary technology (when it is really a public domain spec according the HL9's own by-laws). Another example of behaving open vs. saying it is open. 18. EuSPIRIT workshop - I spent several hours with the Minoru group trying to figure out what services SPIRIT will be providing and how best to interface with them. First of all, Joseph is no longer with Minoru according to Brian. This is unfortunate from my perspective since I have heard Joseph speak twice (AMIA2000 and OSHCA2001) and he has a great understanding of the open source methodology. Hopefully, he will continue to help. The following is what I learned from Minoru/EuSpirit: a. EuSpirit included plans for a "Hall of Fame". They were considering kicking this off at OSHCA2001 - but decided not to because LinuxMedNews announced an award on September 5. The EuSpirit Hall of Fame will simply link to LinuxMedNews' award. b. CD-ROM that includes easy to install open source software will be compiled in the next few months. Disc-image file will be available for free download. c. The success of EuSpirit is being judged by annual surveys and traffic on openhealth list, web-sites (Minoru, Openhealth, Euspirit). (Did you know there are about 250 subscribers to openhealth list?) d. My impression of Minoru is that they are a well-meaning group that is rightly concerned about the financial viability of their enterprise. Their business model appears to be 1) fulfilling EUSpirit contract and 2) establishing a consulting business for open source software. This means there are certain things that they will not do - for example (per Brian), providing in-depth reviews/test reports of open source software for free. Any volunteers? :-) 19. OSHCA2002 in Los Angeles. Since it is clear that competition for hosting OSHCA will quickly heat up over the next few years, Mike McCoy (CIO of UCLA Medical Center) is truly a visionary. That's it for now. Best regards, Andrew --- Andrew P. Ho, M.D. OIO: Open Infrastructure for Outcomes TxOutcome.Org (hosting OIO Library #1) Assistant Clinical Professor Department of Psychiatry, Harbor-UCLA Medical Center University of California, Los Angeles
