Definitely an interesting engineering challenge and approach has everything to do with success.
First issue to address is the fact that two completely different operational conditions will exist on the same door(s).
Access control is simple and straight forward enough. You either have access or you do not.
Patient wandering is a completely different approach in that the doors are unlocked until a patient (visitor with a baby) tries to leave thru the doors and they lock and remain locked as long as they are near the doors with a tag attached to the patient/baby.
These two approaches are completely different from each other.
The correct approach will cost more money and the incorrect approach will cost even more money.
Install the two separate systems in tandem on the door(s) and let each one do their respective jobs correctly.
Access Control: Whatever flavor you prefer as they all do the same thing differently (personality) (we all could show off and explain how to manipulate the access control system to do things it was never intended to do, but then you shackle yourself to the site and no one else can ever work on it and that helps no one at all.)
Patient Wandering: Bosch Security Escort is a very solid product with versatility built in. It can lock the door when a tagged patient approaches as well as tell you where there are within the area they are confined to. You can tag equipment too (computers on wheels)
There are others as well, but since you already mentioned those I just added this as an alternate quality product to consider.
If you wish to dig deeper into this topic let me know.