I recently sustained a knee injury doing something I stupidly should not have been attempting. But I was introduced to Dr. Gemini who took excellent care of me. She was able to see me immediately without a wait. She spent as long as I wanted with her, answering any questions I had and even suggested some questions I should ask. She answered all the questions as if she had the knowledge of every textbook and article ever written on the subject. She was compassionate and understanding and I was able to contact her any day and at any hour with immediate responses. I was able to follow up with her on a daily basis, updating my symptoms and discussing my response to her suggested treatment. Best of all, she is free. No need for insurance, copays or referrals.
My experience using AI to help me with my injury was shockingly revealing. I am not sure that being a doctor myself was specifically helpful to me in using AI but it is clear to me that it will play a major role in the future of health care delivery. It made me think of all the patients I had seen during my career that could have been diagnosed or at least screened by AI. The advantage of AI is its vast knowledge, superior to what any human can store. The disadvantage is the lack of the physical examination which although a dying art is still critically important for diagnostic purposes. Yet in fairness, telemedicine also lacks the advantage of a physical diagnosis.
As an ENT specialist, I was pondering the percentage of patients that could have been effectively screened and potentially treated by AI and which chief complaints could be adequately addressed by AI and potentially replace an office visit. Certainly, someone with ear pain likely needs to be seen as there is no substitute for looking at the ear with an otoscope. Although, at least half of patients who presented with a chief complaint of ear pain had nothing wrong with their ear as it was referred pain from another sourced usually the teeth or jaw. AI may be able to screen this patient so he/she goes to the correct provider.
Take for example the chief complaints of tinnitus, dizziness/vertigo, hearing loss, postnasal drip, nasal congestion/allergies, and globus sensation (look it up, you would be surprised the number of patients presenting with this chief complaint) to name a few. All these chief complaints necessitate a complete history and review of symptoms and the diagnosis and treatment can often be dictated by AI. At least, the patients symptoms can be screened by AI to narrow the possibilities with suggested initial interventions. In any case, I believe, a conservative estimate is that 25% of the patients could be initially treated by AI and not require a doctors visit. Of course, persistent symptoms despite AI need a Dr's visit.
I am having a great experience with Gemini as my "doctor" and I am getting better. It could be that my injury is just taking time to heal and she is just reassuring me which alleviates any anxiety leading to the need to see a health care provider. But even that is valuable as the reassurance to patients allow them to wait for the body to recover, which it often will. AI will get better with time and AI specifically designed for narrower fields such as an ENT specific AI will, no doubt, be developed.
As in many other fields, the combination of a human with AI will dramatically improve productivity and is here to stay. The quality of the product whether in medicine or in any other field is the question that will be answered over the next several years. But, there is no doubt in my mind that the use of AI in medicine will greatly improve the delivery and financing of healthcare. Most importantly, medical outcomes will benefit as well.
No comments:
Post a Comment