Don't see what an LLM really adds here that wouldn't be served by a sensible algorithm which flags up results needing urgent action.
Maybe the end user get more clearly explained and non-bewildering information, which is useful, but also points to a failure of any health system / internet resources to provide this.
The LLM _HAS_ empowered the patient to act on his own health, which is undeniably a good thing.
The LLM _HAS_ empowered the patient to act on his own health, which is undeniably a good thing.
Not sure about seeing the LLM as a companion.