It is not a question of when, AI is here and it was estimated that in 2021, $6.7 billion was spent on AI in healthcare versus only $811 million in 2015. During these times of rapidly increasing costs (labor, prescriptions, and the cost of money) and not so rapidly increasing revenue, we must look to technology to help bridge the gap. Yes, there is uncertainty galore and the ride will, for sure, be bumpy.
It seems that every day you read about a new use of AI (let’s include under this one heading machine learning, deep learning, and natural language processing). The questions for us are, “How are we going to react?” “Can our human interaction with AI lead to improvements in both the quality of care provided and longevity?”, and “Will AI help us become more efficient (provide same or better care for less money) and help improve our traditional methods of billing and collections (revenue cycle)?” [Imagine, a patient is released after a three-day inpatient stay and the bill is ready to transmit (snail mail or other) 24-hours after discharge.]
Within the next five years I can see the majority of my care being managed by AI, I am hopeful. I can see AI coordinating my visits. I was going to say doctor visits, but I’m not sure. AI understands my insurance coverage (more than I can say about most policyholders/members and a lot of healthcare providers) and will act accordingly.
Based on AI’s understanding of my medical history, demographics, and family history, a medical schedule is created (interfaced with my personal calendar) [Predictive Diagnostics] that sets appointments with the right blood labs, testing centers, it enrolls me in a program that my insurer offers that rebates me premium dollars if I agree to submit monthly health data, sets appointments with physicians, if the analysis of my data indicates that the physician is the right caregiver, and to a physician extender (maybe virtual) if the analysis indicates “routine visit”. If AI sets a physician appointment, it sends “findings” to the physician to alert him/her that specific items are of concern.” [Enhanced Diagnostics.]
The physician, or in some cases the physician extender, in cooperation with AI, will develop personalized treatment plans tailored to my unique genetics and circumstances, thereby optimizing outcomes (all in favor of that).
AI becomes that self-driving Tesla, optimizing resource allocation, managing patient flow, predicting patient demands, and even reducing wait times (because it knows based on the model, about how long my visit should take).
I have talked about my fairytale future of healthcare before and with each writing the future becomes clearer. Truthfully, this world still has its detractors. And, rightfully so because we still need to work on the answers to the questions about data confidentiality (you read above where I have AI communicating with my health plan), what feels like the depersonalization of medicine with AI playing a larger role in that conversation, integration with current systems on both the provider and payer sides, regulatory hurdles, and the old “trust” equation (can I really trust something that I can’t touch).
My fairytale world is coming and I would like to talk about it now rather than looking at AI’s taillights as we race to the new healthcare paradigm that answers some of the questions that our current paradigm no longer does. It is up to us to start having those educational conversations about how AI will fit into the future and how AI will augment human capabilities rather than replacing them. As a community, we also need to talk about collaboration and sharing (AI model developed in a hospital setting could potentially benefit a nursing home or rehabilitation facility if shared and adapted accordingly). And, finally, we all need to become advocates for no one will be making this journey alone!
