‘The silent pandemic’ — AMR: Collaboration is key with AI, physicians and other experts

Artificial intelligence (AI) and large language model computing holds tremendous potential to help, possibly with discovery or testing of new antimicrobials, and for the diagnostic and treatment processes that doctors use with patients. Those tools won’t replace human physicians and other clinicians, and if fact will make collaboration among human experts more important than ever. Jason G. Newland, MD, MEd, division chief of infectious diseases for Nationwide Children’s Hospital in Columbus, Ohio, describes some possibilities for the future of AMR and AI.

This transcript has been edited for length and clarity.

Medical Economics: Cutting edge technology: Will artificial intelligence (AI) help research about use and creation of new antimicrobials?

Jason G. Newland, MD, MEd: I think it could be tremendous, right? I mean, I think large language models, AI can help us, and we have to realize that. And I think there's, I haven't looked specifically more like, can I find new antibiotics? But you can see how that can matter, right? With all the literature, all the genetic stuff, all of this sort of stuff, it's going to be there, and now you have things that can identify it quicker, and then maybe you can get it into that research. I think that's super exciting. It's going to be another tool, and it's going to be a tool that, as we know like any tool, sometimes can be problematic. It's still going to require human knowledge, human interface, to take that really impressive tool to the next step to whether it's developing a new antibiotic or in developing maybe an algorithm that's helpful in determining maybe I use an antibiotic or not. With all the latest information, maybe you're bringing in some of the great epidemiologic stuff that's going on in a current situation that allows you to decide, do I give an antibiotic or not? Well, maybe I know that there's tons of mycoplasma circulating. Oh, well, maybe I have to rethink this. But now you know, because it's at hand with a large language model. And then you have these large health records, right? And then we're starting to comb and see, oh, now I'm learning how long I should treat, right? I don't need to treat for two weeks. Actually, if I look at this data, and I take the AI and we look, we actually do better. I think there's such great opportunity we just have to embrace and be thoughtful, know that junk in means junk out, so making sure that we're putting in the right information. All of these things will have to be thought of, and that means the informaticists of the world, and infectious diseases of the world, microbiologists world, that, in the end, the collaboration — and I'm going to stress this, collaboration is key for advancement in our profession, and it's going to be key going forward, especially when we talk about AI and large language models and such. But I'm tremendously excited about what we're going to see out of this.