When Will Patients Hear “The AI Will See You Now”?
Becoming the Patient Advocate
One of the things you learn in life is that, when a loved one is hospitalized for a serious condition, you need to be an advocate. That can mean being a pest at times, given the large number of people and organizations involved in care and treatment. They don’t always communicate with one another in a timely fashion, and you soon learn that you need to help manage and coordinate things. Even though you may have no medical qualifications, there will be occasions when your pushing for action or better communication can make a positive difference.
Who Sees the Whole Picture?
Another situation occurs when a patient has a rare condition about which little is known with certainty regarding diagnosis or treatment. You are shuttled from specialist to specialist, each offering recommendations and tests from the perspective of a particular specialty. Who functions as a qualified overseer when confusion or miscommunication results? You? Your spouse?
Let’s say you are a member of a group health organization where specialists in different locations and clinics all share access to the same network-based system for scheduling, laboratory results, examinations, prescriptions, and treatment management. Information about an individual patient is immediately available to all authorized users. But how are the data from these different sources—often generated by analysts working in locations remote from the patient—integrated, especially when multiple specialties are involved? How does anyone make sense of all these different data streams? How effectively can individual specialists, often overscheduled and overworked, analyze everything efficiently and thoroughly?
A Supporting Role for AI
In all three of these cases—the patient advocate, the patient with a rare condition going from specialist to specialist, and the harried specialist dealing with information from multiple sources and systems—I believe AI systems could play a role. Systems trained across multiple specialties could help analyze and interpret data from these sources, even when languages, data formats, and physical media differ or are incompatible.
Note that I said “a role.” I did not say turning over prioritization and decision-making authority to an AI. That’s an important distinction. AI tools can make mistakes, as can humans. In life-and-death medical situations, those mistakes can be fatal.
Naturally, there will be situations in which information arrives so quickly that a decision-maker needs software or AI support to respond. One example would be responding quickly to a massive cybersecurity attack on a medical facility.
But that situation is not my focus here. Rather, I’m concerned with the three situations discussed above, all of which I have personally experienced or witnessed. My sense, based on my own experience using tools such as ChatGPT for both professional and personal applications, is that an AI-based advisory service could help in all three instances.
Such a service could review and integrate information from diverse sources, both real-time and historical. It could flag inconsistencies as well as unexpected or unusual findings, and suggest alternative explanations for observed phenomena along with possible courses of action and treatment for clinicians to consider. All of this could potentially be done in a fraction of the time required by a chain of people working without such assistance.
The Expert System Revisited
This is very much what we used to call an “expert system.” The major difference now is that the capabilities for learning, updating, and making recommendations are so much more powerful.
Such systems necessarily come at a cost and would need to be carefully configured to protect privacy and confidential information.
What I Want from My Doctor
For myself, the next time I see a doctor, I want to make sure that he or she has all available resources at hand to diagnose and treat whatever is ailing me quickly and effectively. I want that doctor to be able to draw on knowledge beyond his or her personal experience.
Perhaps we are seeing this already—see Microsoft AI tool beat generalist doctors at step-by-step diagnosis—although that report explicitly acknowledges the limitations of the research.
Still, the next time a family member or I see a doctor, I hope every available resource is brought to bear on diagnosis and treatment.
Copyright (c) 2026 by Dennis D. McDonald
