Amy Pierre:
I had the pleasure attending our opening panel presentation entitled “Understanding the Benefits and Pitfalls of Artificial Intelligence in Oncology.” The panel comprised of incredibly knowledgeable experts in the field of AI ranging from APs and MDs with a strong understanding of the basics and progression of AI models, to caregivers who were skilled in using AI to their advantage in difficult clinical situations. Given the myriad of perspectives, it was really a well-rounded panel discussion. And the discussion first began about ChatGPT. A quick poll of the audience of APs showed that the majority were familiar and actually used ChatGPT. It was astounding to see that, even though most of us feel that AI is new, the panel explained that AI has actually been around since the 1950s. And the progression and progress has now developed models so sophisticated that it can now go beyond just text and now incorporate audio and visuals.
A question was asked to the panel as to how AI has become the most beneficial to them in clinic. And they spoke of how using ambient AI, which is just recording your visit, allows the clinician to be 100% tuned into the patient and how your documentation is pretty much complete and how that's a real huge time saver. This type of AI can have the potential to document more granular information from the patient visit that leads to additional billing codes that you may not necessarily have thought of. The panelists also spoke of AI having the potential to have a deep understanding of a patient's real world journey through cancer care and generate actual real world outcomes so that policymakers can understand the totality of the patient's experience. These benefits are really exciting, and I'm curious to learn more about this in the future.
But what I found really compelling was the panelists who was a caregiver for her son who has cancer. She explained that a cancer diagnosis, as we know, is a very traumatic event that pretty much permeates every aspect of the patient and the patient's family's life. And that the executive functioning of an oncology caregiver is extensive, and that AI is really poised to address those functions as an organization and almost serve as a partner in distilling medical information. Patients can take complex medical reports, such as molecular profiling reports, and feed it into AI to generate an easier interpretation of the results and what should be most meaningful to know as a patient. They're essentially asking AI how to accurately make clinical decisions as a patient without a clinical background and help them assist in understanding the pros and cons of certain treatments.
So that really hit me. Patients are going further than just looking at websites and blogs to get their medical information. They're actually utilizing these AI tools to assist in decision making in their cancer care. And we as APs, we need to evolve as our patients have evolved, we need to collaborate with them to help tease out what is just data and what is actual true knowledge and intellect. In fact, the panelists went over an extensive AI checklist. They talked about, as a clinician utilizing these models you need to understand what data was used to train the model, because as we all have heard, if it's garbage going in, it's garbage coming out. So we need to ask ourselves, do we understand how the AI is arriving at its recommendations? And what is the true risk if the AI is wrong, what are the privacy concerns? What are the potential biases that could be occurring?
And most importantly, the AI is your copilot, it does not replace a skilled clinician human being. You are the clinical expert. Don't forget that it's still only a machine. So be aware of how it's influencing your clinical judgment and make sure you are still critically thinking.
All in all, I think my take-home message is that we are really in an exciting time with the development and usage of AI. These models really have the potential to transform oncology care. But as we remain intuitive and critically think about how we care for oncology patients, we need to have that same line of thinking with AI.
