The heart and soul of Open Source is and always has been Freedom. That is ideally suited to medicine because Doctors need to be able to leverage Health IT to meet their real needs, not just the needs that can be meet by proprietary business models. I have, like you, been focused on what makes a good EHR for years, and I have to admit, I still have no idea. The question itself is unfair, its like asking: "What makes a good car?" The answer will always be: "It depends on how
you want to use it."
Open Source gives doctors the opportunity to get 80% of what they need from a common codebase and then make sure that the 20% that they uniquely need is actually done right. The proprietary alternative is always one-size-fits-all to a certain extent. "Real Profit" in the proprietary vendor business model comes when you can give 1000 doctors exactly the same software, over and over again. This creates a feature-to-funding mapping problem. Proprietary vendors only fund the development that they see will be able to be sold over and over, in a cookie cutter fashion. Features that would improve care, but cannot be copied in this fashion fail to appear.
Consider something as simple as oncology. An EHR in a hospital or a practice is normally designed to help find the diagnosis, but for an oncologist, the diagnosis is already well-understood. The oncologist is looking at the same information seeking the best combination of treatments rather than the diagnosis.
The difference is not just in "what" information is being tracked, that is always slightly different between any given specialty. Rather, it is a whole new way to approach the same information. I know that you could probably come up with 10 different examples of this kind of "non-trivial rethinking" needed for specific issues. For a given doctor, specialty, or even patient the design of the software may need to be turned on its head. That kind of flexibility only comes with source-code access.
You will be keynoting a company that supports VistA. So it is critical that you consider carefully the simple question "Why is VistA good?" It is not a trivial issue.
I am also the primary author of "What is VistA really" on the WorldVistA wiki site
But what are the draw-backs of open source?
It is poorly understood. It is nothing less than the answer for modern Health IT, but important projects continue to struggle for funding. The problem is that doctors do not have the time to understand either software or software licensing. Most doctors operate under a
hand-shake business philosophy. They think: If this deal becomes unfair, I will just leave. It is a privilege of a profession in high-demand. As a result they do not evaluate software licenses, or
understand the implications of software licensing on the software process. Every software contract that a doctor signs should be point-by-point compared against the GPL. When the (new) Medsphere talks about Freedom, it is not a political promise or jargon, it is in the contract. The licenses that they use to release software essentially makes them co-owners of the software with their clients. I am constantly trying to get doctors to understand that the issue is not what you have to pay for the software now, but what does it even mean to "get software". If I offered to sell you a watch for $1000 or to rent you one for $1000, you would immediately focus your attention on the "rent" vs "own" issue. But for some reason the exact same distinction seems slip past most doctors and hospitals (despite my best efforts to make noise about it).