Boston Scientific Resonate Adverse Events: 5,542 MAUDE Reports, 37 Deaths, and a Defibrillation-Lead Recall

The FDA's MAUDE database holds 5,542 adverse-event reports tied to Boston Scientific's Resonate family of implantable defibrillators since January 2019. Of those, 37 are coded as deaths and 3,011 as injuries. These are reports, not rates. A report coded as a death does not establish that the device caused the death, and MAUDE carries no denominator for how many Resonate devices clinicians implanted in a given year.

The Resonate adverse events record carries a clear signature. The reports cluster around sensing and shock behavior: over-sensing, inappropriate shocks, high lead impedance, and pacing problems. Deaths stay low and flat year to year, while the injury and malfunction volume climbs. The Resonate high-voltage generators have never drawn a device-specific Class I recall. The only recalls that name the device are three Class II firmware corrections from 2017. The Class I safety action that reaches these patients belongs to a defibrillation lead rather than the generator, a distinction the raw data can obscure.

#What the Resonate Is and Why Its Record Looks the Way It Does

Boston Scientific received FDA approval for the Resonate family of high-voltage devices on May 9, 2017. The line spans implantable cardioverter defibrillators (ICDs) and cardiac resynchronization therapy defibrillators (CRT-Ds), and the approval introduced the HeartLogic heart failure diagnostic alongside the company's EnduraLife battery technology. These are implanted, high-voltage devices that monitor the heart's rhythm and deliver pacing or a defibrillating shock when they detect a dangerous arrhythmia.

That job description explains the shape of the MAUDE record. A device that watches every heartbeat and decides whether to shock generates reports about two failure directions: acting when it should not (inappropriate shocks, over-sensing) and failing to act as intended (under-sensing, failure to capture). The Resonate X4 CRT-D carries the largest share of the reports, followed by the extended-longevity ICD models.

#Reports by component

Component	Reports
RESONATE™ X4 CRT-D	1,989
RESONATE™ EL ICD DR	1,652
RESONATE™ EL ICD VR	778
RESONATE™ HF CRT-D	518
RESONATE™ HF ICD DR	463
RESONATE™ HF ICD VR	142

#Reports Climbed Sevenfold While Deaths Stayed Flat

Resonate reports rose from 193 in 2019 to 1,482 in 2025. The 2026 record holds 713 reports through late May. Death-coded reports moved in a narrower band, between 2 and 9 per year across the whole window. The growth sits in injuries and device malfunctions, and mortality stays put.

#Reports by year

Year	Reports	Deaths
2019	193	3
2020	288	7
2021	339	5
2022	458	2
2023	870	2
2024	1,199	5
2025	1,482	9
2026	713	4

A rising report count on a device that has sold since 2017 reflects an installed base that grows each year and accumulates device-years at risk. You would expect this pattern from a widely implanted device with a long service life. The mortality line stayed flat against a sevenfold rise in total reports, which is the comparison that matters.

#The Failure-Mode Signature: Sensing and Shocks

Sensing and shock behavior dominate the product-problem breakdown. Over-sensing leads the specific mechanisms at 1,253 reports, followed by pacing problems at 1,188 and inappropriate or inadequate shock or stimulation at 925. High impedance appears in 780 reports and signal artifact or noise in 628. The single largest bucket, 1,490 reports, is the non-specific "Adverse Event Without Identified Device or Use Problem," a reminder that a large fraction of MAUDE entries never resolve to a concrete device fault.

#Most reported product problems

Product problem	Reports
Adverse Event Without Identified Device or Use Problem	1,490
Over-Sensing	1,253
Pacing Problem	1,188
Inappropriate/Inadequate Shock/Stimulation	925
High impedance	780
Signal Artifact/Noise	628
Under-Sensing	373
Failure to Capture	296
Inaccurate Synchronization	267
Failure to Read Input Signal	267
Premature Discharge of Battery	263
High Capture Threshold	212

The patient-problem coding runs parallel. The largest category, 2,416 reports, records no clinical signs or symptoms, which fits a population whose clinicians caught the device malfunction at interrogation rather than at a cardiac event. Infection is the next theme: 1,286 reports of unspecified infection, 224 of sepsis, and 107 of pocket erosion. Shock-related patient harm appears as 589 reports of shock from patient leads and 343 of electric shock.

#Most reported patient problems

Patient problem	Reports
No Clinical Signs, Symptoms or Conditions	2,416
Unspecified Infection	1,286
Shock from Patient Lead(s)	589
Electric Shock	343
Sepsis	224
Ventricular Fibrillation	163
Arrhythmia	154
No Code Available	134
Pocket Erosion	107
No Known Impact Or Consequence To Patient	104

Grouping the narratives into failure modes sharpens the picture. Lead integrity alerts, inappropriate shocks from noisy signals or T-wave over-sensing, and high shock-impedance measurements recur across hundreds of reports. Two of the ten largest modes point at the lead-tissue interface rather than the generator: high shock impedance and lead integrity alerts.

#Failure-mode clusters

Failure mode	Reports	Emerging
Lead Integrity Alert Triggered	230	no
Inappropriate Shock from Noisy Signals	158	no
High Shock Impedance Measurements	135	no
High Shock Impedance Measurements	132	no
Inappropriate Shock from T-Wave Oversensing	125	no
Premature Battery Depletion in CRT Device	115	no
Inappropriate Shocks and ATP Delivered	104	no
CRT Device Explanted Due to Infection	102	no
RV Lead Connector Insertion Issue	94	no
Pocket Infection with Erosion	94	no

#The Recall Picture: Read the Device Name, Not the Product Code

Resonate's own FDA recall history is thin and old. The three recalls that name the device are Class II corrections issued on October 4, 2017 for an incorrect firmware configuration, all since terminated.

#FDA recalls

Recall	Class	Date	Status	Reason
Z-0081-2018	Class II	2017-10-04	Terminated	The devices have an incorrect firmware configuration.
Z-0078-2018	Class II	2017-10-04	Terminated	The devices have an incorrect firmware configuration.
Z-0077-2018	Class II	2017-10-04	Terminated	The devices have an incorrect firmware configuration.

The Class I action patients with a Resonate device should know about targets the defibrillation lead, not the generator. In July 2025 Boston Scientific issued an advisory on its RELIANCE and Endotak defibrillation leads with expanded-polytetrafluoroethylene coated shock coils. The Heart Rhythm Society summarized the mechanism: the coating can allow debris and minerals to enter and calcify over time, producing "gradually rising low-voltage shock impedance" that "could reduce shock efficacy, and instances of failed shock therapy have been reported." The advisory estimated roughly 354,000 affected leads still in service, 250,000 of them in the United States, and attributed ten deaths over the prior decade to failure to convert a sustained ventricular arrhythmia.

The lead advisory maps onto the Resonate MAUDE signature. High impedance appears in 780 of these reports, and high shock-impedance measurements form two of the largest narrative clusters. A generator wired to a calcifying lead measures rising impedance and flags reduced shock efficacy, which is what these reports describe. The reports sit in the Resonate record because that is the device the clinician was interrogating, even though the root cause lives in the lead.

MAUDE and the FDA recall database can mislead you here. Both index devices by FDA product code, a class-level label that Boston Scientific's Resonate, Emblem, and Endotak devices share with each other and with Medtronic, Abbott, and Biotronik hardware. Matching a recall to a device by product code alone stamps another manufacturer's recall onto the wrong product. The recalls above are the ones whose FDA description names the Resonate device itself.

#The Comparator: Battery Longevity Is the Resonate's Strong Suit

Boston Scientific built the Resonate marketing case on EnduraLife battery longevity, and the independent evidence supports it. A 2025 real-world analysis by Nishino and colleagues in the Journal of Arrhythmia found manufacturer to be an independent predictor of battery depletion. Six-year replacement-free survival for CRT-Ds was 93 percent for Boston Scientific devices, 80 percent for Abbott, and 21 percent for Medtronic. For ICDs the gap narrowed, with Boston Scientific and Abbott both near 100 percent and Medtronic at 91 percent.

The reports keep that advantage in proportion. Premature battery discharge accounts for 263 Resonate reports, and premature battery depletion in a CRT device forms a 115-report cluster. A longevity edge across a population does not mean zero battery failures in the field, and the reports that do arrive are a minority of the Resonate record, which remains dominated by sensing and shock events rather than battery events.

#How to Read the Disproportionality Signals

Our pipeline carries active disproportionality signals across the Resonate line. The extended-longevity ICD VR model shows the highest relative excess reporting at an EBGM of 46.3, the EL ICD DR sits at 29.7, the HF ICD models at 25.1 and 25.5, and the X4 CRT-D at 17.6. EBGM measures how much a device's reporting exceeds the background rate across all FDA-regulated devices. A high value flags relative excess reporting, not absolute danger, and a widely implanted defibrillator with a recognized lead issue is the kind of device that produces one. We made the broader argument about why clinicians and reporters misread these scores in Signal Detection in Post-Market Surveillance Is Broken.

The limits run deeper than any single score. MAUDE is the public surveillance record, not a registry. Reports are not rates, the denominator is not in the data, and reporting volume rises with an installed base regardless of device performance. If you want to reproduce or extend this analysis, How to Search the FDA MAUDE Database walks through the query mechanics, and our overview of post-market surveillance in the United States covers what the system can and cannot catch. For a comparison against another widely reported Boston Scientific device, see our analysis of the Watchman left atrial appendage closure device.

#For Patients and Clinicians

For a clinician following a patient with a Resonate ICD or CRT-D, the thread to follow in this data is impedance. Rising shock impedance is the earliest sign of the defibrillation-lead calcification issue, and the Heart Rhythm Society advisory frames trend monitoring at interrogation as the practical response. The sensing and shock reports that dominate the record are the events a device interrogation catches.

For patients, a large report count on a high-volume defibrillator is not a reason to act on your own. The mortality signal in this data is small and flat, and the device is doing continuous work that a report count cannot summarize. Bring questions to the clinician who manages your device. You can review the full reporting history on the Resonate device detail page or run the same query yourself with the public MAUDE lookup tool.

#Methodology and Caveats

The figures here come from a Claripulse query over FDA MAUDE data, attributing reports to the six Resonate ICD and CRT-D components at the device level rather than by shared product code. We deduplicate reports on the FDA report key. Counts of deaths and injuries reflect the reporter-assigned event type, which does not establish device causation. We attribute a recall only where the FDA recall description names the Resonate device, because the product codes these devices carry are shared across multiple brands and manufacturers. External claims about the 2017 firmware recall, the 2025 defibrillation-lead advisory, the 2017 approval, and comparative battery longevity carry source links at the point they appear.

This analysis is based on publicly available FDA MAUDE data and published literature. It does not constitute medical advice. MAUDE reports are submitted under FDA regulation, are de-identified, and may be incomplete or unverified. No attempt should be made to re-identify any patient.