Computers are incredibly useful in recalling vast databases of information reliably and quickly. Knowing this, should we be “teaching” computers medicine? If so, how should we utilize them in the process of working up a patient’s symptoms?
I don’t think computers will ever substitute a physician’s ability to “heal” the patient, nor do I believe primary care doctors are in jeopardy of being replaced (although I do feel nurse practitioners will take on a greater role in the initial workup of patients). But what I do know is that primary care will be foolish not to incorporate a computer’s suggestions in the future.
The concept that patients should/must go through an “escalation tree” of non-physician evaluations, beginning with a trained worker plugging symptoms into a computer program that also has the patient’s history loaded is inevitable. For example, all physicians are trained in at least minor surgery. Why does one need a surgically trained MD to diagnose crotch rot or an ear infection? Diagnostic experts can be trained without the comprehensive four year med school followed by years of residency paradigm. At some point in any “tree of escalation” there can be way points where an operator can receive a suggestion to refer to a physician or someone with a higher level of training.
Even if one ultimately winds up in the hands of a specialist physician, the history and suggested diagnoses that come beforehand in the “escalation tree” cannot help but make the visit with him/her more productive in less time.
The ONLY surgical training needed by the average doc is what he/she experiences with a med school cadaver. One does not need to be a neurologist to diagnose knee problems. This is, of course, proven by the existence of podiatrists.