Cardiology leaders are split on where the hype outpaces the clinical reality.
Becker’s asked four cardiovascular leaders the same question: What’s a trend in cardiology today that you think is overhyped?
Editor’s note: Responses have been lightly edited for length and clarity.
Johnathan Klaus. Consulting Director of Cardiovascular Services at South Shore Hospital (Weymouth, Mass.): I think some of the hype has died down, but there’s still an undercurrent of irrational exuberance with AI in cardiology. I believe there’s a lot of promise, but there’s also a lot of ongoing validation that needs to be confirmed in order to gain the trust of clinicians and patients. There’s a lot of promise in diagnostic support, but we still need to prove that we aren’t allowing biases to sneak into the models, whether through biased training data or ongoing biases in the population. For instance, if we train a model on data from a health system that has a more liberal diagnosis pattern, will the model continue to have that pattern? Will the model have that pattern only for that hospital based on its diagnostic history, or will it change its pattern based on the diagnostic history of the hospital where it’s deployed?
Not to be totally down on diagnostic abilities — there’s huge potential to notice subtle changes and shifts that bring a physician’s attention to something the human eye can’t detect, or that gets missed in the mountain of information in the chart.
At present, I don’t think the diagnostic abilities of AI have been proven or validated to a point where clinicians trust them enough, and given the public’s skepticism of AI at large, if there’s no clinician support, there will probably be little public support.
We need to be comfortable with getting better, but not perfect. The models are going to incrementally help, but they likely won’t solve every problem. I think the hype comes in where we’re sold AI as a solution that will solve all the problems, but we probably need to lower our expectations and understand that we’ll put in a lot of effort and solve many, if not most, of the problems — but not all. As that relates to cardiology, we’ll probably be able to [help] most heart failure patients, but we won’t completely solve the problem, and it will take a lot of work to get to that point too.
Jeffrey Marshall, MD. Retired Chief of Cardiology at Northside Hospital Heart Institute (Atlanta): AI is being overhyped by the media, and it’s never going to replace the work of physicians, nurse practitioners, PAs or any other healthcare professionals. They’re really tools that physicians and healthcare providers should be using to make healthcare more efficient and help patients understand what’s happening to them. That’s the hype side.
Kevin McIntyre. Senior Director, Physician Practices at Baptist Health-Cardiovascular Associates and Baptist Health Medical Group (Montgomery, Ala.): One that seems to be getting overhyped is outsourcing device clinics. When a patient has an implantable cardioverter-defibrillator or pacemaker, those patients need to continue to be followed, both virtually and in person, and I keep seeing a lot of companies that are offering to do this for practices — a lot of practices are moving in that direction. I think it’s probably a little bit better from a cost-saving or full-time equivalent standpoint to outsource these types of services, but from what we’ve seen in our own practice, the cost savings don’t always supersede the clinical benefits of having your own staff who work with your physicians daily on how to interrogate device checks and remote monitors, and how to best interpret those. It seems to be a trend I keep seeing and getting contacted about, but I think it’s a little overhyped from a clinical quality standpoint. I just don’t think the juice is worth the squeeze there.
Karen Walker, MSN, RN. Enterprise Director for Cardiovascular Services at Santa Clara Valley Health Systems (San Jose, Calif.): I think AI is the big elephant in the room — it’s what everyone is talking about right now. I think it’s really overhyped because organizations are trying to move forward with AI and integrate it into cardiology as a standalone solution, and that’s not going to automatically improve outcomes, nor will it reduce costs. It certainly has the potential, but you have to make sure you have the right team at the table to build the algorithm. Whether it’s the C-suite coming up with the algorithm, or a technology vendor coming up with an algorithm they think will work for a particular institution, that’s not going to help an organization become more efficient. It can’t repair a fragmented workflow. You have to make sure you pull the entire team together in order to effectively make an impact — you have to have everyone on the front end, because if you try to integrate an AI algorithm without having those frontline people, managers and leaders involved, you’re not going to reduce costs, you’re not going to solve any workforce challenges, you’re going to be inefficient, and it certainly won’t improve outcomes.
At the Becker’s 32nd Annual Meeting: The Business and Operations of ASCs, taking place October 29-31 in Chicago, ASC leaders, surgeons and healthcare executives will explore strategies to drive growth, enhance operational performance, navigate reimbursement challenges and prepare for the future of ambulatory surgery. Apply for complimentary registration now.
