the big issue i see here is the attempt to install ONE system to cover the health system. Its an enormous task and darnn near impossible.
Having spent a few years involved in a medical software company some years ago, my thoughts are that they should have a fairly generic patient demographic system that linked across hospitals and hospital departments (thats the important point here..read on)
Each hospital department is very very different , both in procedures and in type of care given. Think General Medice vs Stroke Unit. Different data is recorded in each and key info is different in each too.
So a generic demographic on each patient database, with certain common interest elements like what drugs they are currently on is the first think that needs to be in place.
Each department then actually needs a sub-system written for that departments type of care. One for Stroke, one for Oncology, one for Cardio, one for Maternity etc.
That way, each can be deployed independent of each other, at different time scales and budgetary allowances. Rollout and planning can then be much easier. Eventually this builds up to a super system, integrated, but specific care given in each department if private.
Anecdote: I know of one Dr, who, in desperation, got his own system developed and installed, just for his department, trained his staff himself and tweaked it over a few years. When there was a drug recall issue because of side-effects, he simply referenced his system and found all who had been prescribed that one and contacted them all for followup. His colleagues across the same hospital and in other hospital were deep in the site for weeks, trying to compile lists , go over old records, notes etc. Took him 10 mins to have an accurate list. Took others weeks, yes, actually weeks.