From Detection to Anticipation: Dr. Ben Steinberg on the Future of Cardiac Monitoring

Q&A with Dr. Benjamin A. Steinberg, MD, MHS, FACP, FACC, FHRS
Chief Medical Officer, PaceMate

Q1: You've spent your career using data to personalize cardiac care. What does that look like in practice?

Envision a 74-year-old with AFib, kidney disease, and a history of falls. You must make a decision about stroke risk: whether to anticoagulate, how aggressively, what tradeoffs they're willing to accept. The guidelines give you a framework, but the framework was built on population averages. The person sitting across from you isn't an average. They have a specific history, a specific physiology, specific things they value that no scoring system captures.

What I’ve spent my career trying to do is narrow that gap, to use data to better personalize the approach for that specific person rather than the statistically reasonable answer for a population they may or may not fit into. That’s what draws me to this work. Not the technology for its own sake, but what it makes possible for the individual patient who deserves more precision than rudimentary scoring systems can offer.

Q2: What's a case or a patient moment that you still think about — one that shaped how you approach this work?

There was a patient early in my career who came in for a routine device check. Everything looked stable on the surface. But when we went back through the transmission data afterward, the signs had been there for weeks: subtle changes that didn't trigger an alert, didn't prompt a call, didn't change anything we did. He ended up in the hospital with something we might have caught earlier. I've thought about that patient a lot. Not because anyone did anything wrong, but because the system wasn't built to surface what it already knew. That's the problem I've spent my career trying to solve.

Q3: Remote monitoring promised to change how we care for patients with implantable devices. In your experience, where has it delivered on that promise — and where has it fallen short?

Where it’s delivered is real and significant and the data are clear — remote monitoring improves care. We can detect changes earlier — arrhythmias, battery changes, lead deterioration — weeks before a patient would have any reason to call their doctor. That’s not incremental. For the right patient at the right moment, it’s the difference between a phone call and an emergency room visit, or worse.

Where it’s fallen short is on the other side of that equation. The data come in, but getting the right information —  and only the actionable information – to the right clinician at the right time has been harder than anyone anticipated. We published data showing that patients couldn't reliably identify when they were in arrhythmia — their symptoms didn't match what the device recorded more often than anyone expected. Integrating all the data points – symptoms and arrhythmia monitoring —  is vital to patient care. Meanwhile, device clinic teams are talented and dedicated, and they are managing enormous volumes with tools that weren’t designed for the scale they’re now operating at. The promise of remote monitoring doesn’t fully materialize until the platform behind it is as good as the underlying technology — filtering what doesn’t need attention so that what does, gets seen immediately. That’s the problem that still needed solving when I came across PaceMate.

Q4: Your work spans clinical practice, research, and AI. How do those pieces fit together for you?

I did my residency on the Osler Service at Johns Hopkins — steeped in tradition, with cutting-edge research all around us. It was an amazing experience, and I loved it. But even then, I noticed the disjointed nature of healthcare information, and the lag between technology development and its impact on medicine. Electrophysiology attracted me because it offers real and consistent opportunities to help patients and cure disease, while sitting at the cutting edge of technology — a lot of what we do today relies on technology that did not even exist when I started my training 20 years ago.

Along the way I kept running into the same problem: the data we had about our patients was far richer than our ability to use it. We had device transmissions, lab values, imaging, clinical history, procedural data — and we were making decisions off a fraction of it because there was no good way to synthesize the rest in real time.

That’s what pushed me toward research, and eventually toward AI. We have these amazing new tools, how can they help us improve care using data we already have? What can we actually know about this patient’s risk if we use everything we have? The research follows the clinical questions, not the other way around.

Q5: You've built your career around rigorous clinical research. What made PaceMate the right fit?

Scale and integrity, in equal measure.

First and foremost, it’s the people and the integrity. Through my own collaborations with PaceMate over the last several years, as well as colleagues who know and respect the PaceMate team, I knew it would be a good fit. PaceMate has done the work — published the evidence, built the infrastructure carefully, stayed close to what device clinics actually need rather than what’s technically impressive. For someone who came up through academic medicine, that commitment to rigor isn’t incidental. It’s the reason the science can be trusted, and it’s the foundation everything else gets built on.

Secondly, the scale is straightforward — PaceMate manages care for millions of patients across health systems nationwide, which means the dataset they’ve built is genuinely unlike anything available in academic research. The questions I’ve spent years trying to answer with carefully assembled research cohorts, PaceMate can approach with a breadth of real-world data, technical expertise, and nimbleness, that changes what’s possible.

Q6: Where does cardiac monitoring go from here?

The direction is toward more personalized care and more active management. That’s a fundamental shift in how medicine has worked for most of its history. For a patient with an implantable device, the promise has always been that someone is watching — but the reality has often been that someone is watching when they have time, when the alert rose to the top of the queue, when nothing else was more urgent. AI changes that calculus. Not because it replaces clinical judgment, but because it makes it possible to add a layer of adjudication at a scale no team of humans could manage alone.

What I find most compelling about where this is going is the move from detection to anticipation. We’re getting better at identifying what’s already happened. The next step is understanding what’s likely to happen, surfacing the patient whose trajectory is changing before the change becomes a crisis. That requires the kind of longitudinal, patient-level data that PaceMate has spent years building. The infrastructure exists. The clinical questions are clear.

What comes next is doing the work carefully enough that the answers can be trusted and making sure that trust translates into better outcomes for the patients. I joined PaceMate because I want to be part of building that.