The problem is that this type of overthinking goes on too much in IT, without consideration for the exigencies of their daily jobs. What if the doctor that has to 'sign off' is gone or unreachable? Things change! Paper systems had the FLEXIBILITY to be adapted to changing circumstances.
Any workable system has to be UNDER-designed. It should start with a simple scanning of paper records for electronic sharing, together with a system to make sure new papers are scanned. For many applications, that would probably be way better than having any 'data model' at all. The developers cannot and will never be able to predict the variety of data and instructions and random information that a doctor should put into a patient's record. The system needs to be open-ended and lack a rigid structure that ties the hands of the actual practitioners.
For example, in the described system, he said that allergies had to be associated with specific causes, which is limiting and may not be true. He said that drugs had to be separately chosen and were limited to a specific list! This is ridiculous. New drugs appear all the time. Who knows what sort of drug a random doctor may prescribe? Docs need the flexibility to write anything at all into the record, and find it again when they check it.
They probably also need the ability to quickly jot something down for the record and take off to deal with someone else who's dying. We have the answer - it's PAPER. If you want to have a tablet on every bed, fine, splurge, but you should have a paper chart or at least a pad as well in case of loss of power or network connectivity.