By James Allred, MD, FACC, FHRS and Amber Seiler, NP, FHRS
A study we co-authored with Wake Forest School of Medicine, published in the Journal of Cardiovascular Electrophysiology, shows that guideline-based reprogramming of CIED alert parameters reduced non-actionable alerts by 74% over 90 days, with no increase in stroke or sudden cardiac death over a median 1.8-year follow-up.
Remote monitoring is essential to modern cardiac device care. It lets clinicians follow patients between visits, catch meaningful changes sooner, and manage more patients outside the clinic walls.
But anyone who has worked in a device clinic knows the other side of that progress. More data does not always mean better care. Device teams sort through a growing volume of alerts every day. Some demand immediate clinical attention. Many do not. The question isn’t whether remote monitoring works. It does. The question is whether the right alerts are rising to the surface. This study helps answer that question with data.
What the Study Found
The study included 536 patients with cardiac implantable electronic devices. In 413 of them, alert settings were reprogrammed using a clinically-rational protocol designed to silence alerts unlikely to change care. The censored categories were:
- AF episodes lasting less than 5.5 hours (supported by the TRENDS trial)
- Persistent AF episodes lasting more than one month, in patients who had already generated an alert
- AF episodes under 24 hours in patients already on anticoagulation or with prior left atrial appendage closure
- Non-sustained ventricular tachycardia on defibrillator platforms
The findings, after 90 days of follow-up on alert volume and a median 1.8 years on clinical outcomes:
- 74.2% overall reduction in non-actionable alerts across all device types
- 80.9% reduction for ILRs, the highest of any device class
- 73.3% reduction for ICDs
- 66.3% reduction for pacemakers
- Stroke rates in AF patients were essentially identical between reprogrammed (5.2%) and control (5.4%) groups
- Sudden cardiac death rates in NSVT patients trended lower in the reprogrammed group (2.3% vs 9.3%)
After adjustment for age, sex, race, CHA₂DS₂-VASc score, anticoagulation status, and prior stroke history, there was no statistically significant difference in stroke risk between groups.
Why This Matters Clinically
This isn’t about turning alerts off to make the workload easier. It’s about designing a smarter system around the patient, the device, and the clinical evidence.
Manufacturer default settings aren’t always aligned with current evidence or with how a specific patient should be managed. A patient already anticoagulated for paroxysmal AF doesn’t need an alert every time a six-hour episode is detected. The clinical action doesn’t change. A patient with a closed left atrial appendage and a known AF history doesn’t need the same alert profile as someone with new-onset AF and no anticoagulation. Evidence from TRENDS, ASSERT, ARTESIA, and NOAH-AFNET all supports a more nuanced approach to short-duration subclinical AF than what default settings deliver.
When alerts aren’t thoughtfully managed, device teams spend hours each week reviewing events that won’t change a patient’s care. Over time, that creates fatigue, slows workflows, and makes it harder to catch the alerts that genuinely need attention. The signal gets buried.
Is Remote Monitoring a Technology Function or a Clinical One?
Remote monitoring is a clinical operations function, not a technology function.
The distinction matters. Treating it as a technology workflow leads to one-size-fits-all alert profiles, vendor default settings carried forward indefinitely, and a steady drift toward higher non-actionable burden. Treating it as a clinical operations function means alert settings reflect the patient, the device platform, the comorbidities, and the current evidence — and that those settings get reviewed and adjusted as those factors change.
What Device Clinics Can Take From This
A few takeaways stand out from the data.
Protocols beat defaults. Clinics with clear, evidence-based protocols for which alerts to act on, monitor, or censor consistently outperform clinics relying on manufacturer defaults. This study reinforces what the field has been moving toward.
Customization is part of safety, not opposed to it. The most important finding here may be the absence of harm. Censoring alerts in clinically appropriate populations did not increase adverse outcomes, even across 1.8 years of follow-up. For a field that has reasonably been cautious about reducing alert sensitivity, that’s the data point to anchor on.
The work is ongoing, not one-time. Patients change. Devices change. Protocols need maintenance and review. A single reprogramming event is not a strategy.
What Comes Next
The next phase of remote monitoring isn’t about collecting more data. It’s about making the data we already have usable.
For device clinics, that means workflows that reduce unnecessary noise, protect clinical attention, and surface the transmissions that genuinely need review. It means recognizing that alert optimization isn’t a software setting. It takes clinical judgment, ongoing management, and the experience to know which patients fit which protocol.
Done well, alert optimization improves efficiency, reduces burnout, and keeps clinical focus where it belongs: on the patients who need it most.
Study citation: Mostafa MA, Bodziock G, Cotten L, Schaich CL, Seiler A, Dillon J, Brock J, Hansen R, Kozak P, Simmons T, Bradford N, Allred J, Whalen P, Bhave PD. Exploring Optimal Cardiac Electronic Implantable Devices Programming: Reducing Non-Actionable Alerts and Assessing Clinical Outcomes. Journal of Cardiovascular Electrophysiology. 2025;36(9):2226-2232. doi: 10.1111/jce.16789.
Read the full open-access study →
If your device clinic is overwhelmed by alert volume or struggling to separate meaningful transmissions from non-actionable noise, CV Remote Solutions can help evaluate your current workflows, optimize alert programming strategies, and provide clinical support tailored to your team. Schedule a consultation with a clinician.