The complexity and incompatibility of clinical information with registration information with billing and reimbursement information is the real heart of the problem and "open standards" will not fix that.
On the contrary, these are just the types of questions that open standards and open source were designed to fix.
All these databases you mention are structured. That is, they have fields which carry specific tags. Personal data in registration, billing codes in reimbursement, diagnoses in clinical.
HL7 is designed to to operate at that intersection. Specifically it tags clinical data to administrative data. It has a flexible design, it's approved by ANSI, and it has an administrative structure in place to handle change.
Getting vendors to commit to HL7, and getting their products to truly interconnect, is at the heart of health IT reform.
If the center is open proprietary tool makers have nothing to lose in supporting it. Getting them to do this is the test.
While most of the pressure for open source in medicine is bottom-up, a thumb placed on the scale from the top-down would accelerate the trend.
None of this is going to be easy. Nothing is going to happen overnight. But the pieces are in place to link all these medical databases, and with proper leadership the work may move with surprising speed.
Open source can create the common ground, and level playing field, to make what you want to happen, happen. I know of no other force in the health IT industry which can make that claim.