Dec 8, 2023
Dr. Colin Banas, Chief Medical Officer at DrFirst, defines the term clinical-grade AI as he distinguishes it from other forms of AI that may not be trained on robust clinical data. He emphasizes the need for accurate and complete data in healthcare records and the challenges of semantic interoperability and biases when working with AI models. Potential applications for clinical-grade AI include identifying insights about population healthcare, spotting clinical research trends, and addressing burnout by automating administrative tasks, allowing clinicians to focus on patient care.
Colin elaborates, "DrFirst, as many folks know, got its start as an e-prescribing company over 23 years ago. From there, it blossomed into a robust set of solutions around what I call the sweet spot for the company, which is intelligent medication management. This is all the way from writing the prescription to making sure the patient takes the prescription. The care team understands prescription lists and the way that the prescriptions may or may not be taken. And, of course, to do any of this with any level of accuracy and safety, you need data and a lot of it. Fortunately, we have two decades of working with this type of data. We have generated core expertise around medication data and all the various things you can do with it going forward."
"This is an interesting one, especially since, within the last year, we're at a major inflection point for the AI explosion. AI has been with us for quite some time. Even rudimentary things in the EHRs around clinical decision support can often be considered a form of artificial intelligence. And I want to pause there and say I dislike the term artificial intelligence. I would much rather say something like augmented intelligence or even assistive intelligence."
"But the idea behind labeling ourselves as a clinical-grade AI solution stems from what I've seen in the industry and what I've seen in the literature around what I call the five pillars of clinical-grade AI. It is an attempt to distinguish the AI we use in clinical care settings from what we're seeing all over the place right now. If I'm saying it correctly, your ChatGPTs, your Bards, even I think Elon Musk dropped a new one yesterday called GROK. And, of course, you've also seen some of the problems of using that kind of AI in clinical situations. Namely, the AI can hallucinate, the AI may or may not have been trained on robust clinical data, and the AI might not be perfectly reliable, especially if you come to trust the AI without actually checking over it."
@DrFirst #DrFirst #ClinicalGradeAI #PriceTransparency #PopulationHealth #PatientSafety #ClinicianBurnout #HealthcareCosts #HealthcareAnalytics #HealthcareTransformation #HealthcareInnovation
