Your response is lucid, but misses the point.

you said... "There is not any new threat here that hasn't been around since before EMR was a buzzword."

Sure there is! A threat is not merely a particular technology, but the popularity of it. Browsers do not have to installed in computer, making the solution that Ignacio is detailing trivial to deploy and use. Trivial means cheap and cheap means pervasive, pervasive means a new threat.

The other problem is with your assumptions. any application where the data cannot be exported at any time and for no cost by the user is evil.

Ignacio is smarter than this. He is not at all arguing this. He knows that most proprietary programs do not have any export functionality or if they do they are in a non-standard format. (that's easy since for the most part, there are no solid standards for health data. Note to readers: do not bother to contradict me with examples of a "standard" that is clearly not "solid"...) What separates hosted solutions from local solutions is that with a local solution you have the right to reverse engineer the data storage format or even the data export format. Ignacio is talking about this "right to reverse engineer" and "access to the data" in the same breath because you have to. Trusting an "export function" does not really help you. (especially since that may be what you are reverse engineering). All of this is wrapped up in his phrase "privately extract data".

You said "And, let's not forget the advantages of having the data reside on servers that are outside of the control of the practitioner."

Totally irrelevant advantages. If the practitioner does offsite backups, there is no risk of data loss. Data thieves will simply steal in some other way, (i.e. print screen button). Viruses can impact anyone, including the datacenter that "hosts" your medical application for "protection". If even if you counted these as advantages, they are irrelevant if the practitioner has lost control of the data.

You said "As long as a "hosted" offering provides tools to export the data and contractual verbage that the customer always "owns" the data, then the data is safe and secure and accessible."

No. In order to truly own the data, the practitioner must have full access not only to the data, but to the source code under a FOSS license (preferably the GPL) for the application that is used to manipulate that data. Again, an "export" is useless unless I can continue to run the practice on the same system using that data, which is impossible without the sourcecode. A MirrorMed or ClearHealth (which are GPL) based ASP with full data export is moral, while a proprietary ASP is not.

Also, data access "in the contract" does not always actually translate to data access. What happens when the company that you have the contract with files bankruptcy? Much better is "data exports backups in practice", where you and the vendor agree that you will download a backup every night to your server!

Ignacio is making an argument and that argument is correct, the fact that he did not give a "full picture" is as irrelevant as it is impossible, I think he made his point.

Trotter