Quote History Quoted:
You are so completely missing the point of how integrated systems work. In a medical context of competing priorities and contraindicated treatments, how did the medical care providers learn to balance their treatments? Formal education, experience, and in some cases- intuition which begs the discussion if intuition is even a valid medical protocol.
In a fully integrated system, interventions could be programmed from existing data and medical knowledge along with the requisite feedback loops to monitor efficacy and monitor treatments. I suspect that a machine could provide even more effective treatment than a human could.
View Quote
Your digging so much deeper based on what your “suspect” it’s just getting more ridiculous.
Protocols, checklists, etc., the type of things that would be programmed in, are often what kill people.
Intuition and gestalt are what often save people from those mistakes.
Computerized, algorithmic EKG interpretations are frequently wrong for one example.
People follow ACLS protocols and kill people instead of doing something different and saving them.
People follow chest pain protocols and kill people instead of skipping one of the meds and having them survive to appropriate care.
At this point, and for the foreseeable future, there is a degree of information that someone can look, see, hear, feel, smell clinically that cannot be measured and programmed in. Most likely, intuition and gestalt are an evaluation of oneself able data that there is not a way to mechanically record and feed into a system. The human body and medical issues are a probability cloud exercise and not a Newtonian calculable function.
A lot of the computerized and evaluated information is wrong. Pulse oximetry is often wrong. Or may read normal in conditions where a human has other observable signs the patient is not oxygenating. Cardiac monitors, like EKGs, often give incorrect rates, rhythms, etc. Arterial lines may not be zeroed properly. Automated blood pressure monitors are frequently significantly wrong.
We are faced with trying to get a computer to manage someone based on information that cannot be observed and fed to them, information that can be measured and fed to them but is frequently wrong, computerized algorithms that are frequently wrong, lab values that are often errors a human would recognize as suspect, and protocols that have many, frequent exceptions.
An automated Tesla vehicle being randomly and incorrectly told it was going twice as fast as it is, that incoming June bugs were on coming automobiles, that its tire pressures were randomly twice or 1/10 the psi, that sidewalks were roads, could not detect bicycles or motorcycles at random periods, etc. would not work.
We don’t even have a way to measure much of the information A human can obtain just from being next to a patient yet.