Appearance
The PFA Race
Every PFA platform has safety problems. This page shows what FDA adverse event data reveals when you estimate per-procedure rates with the best available denominators.
The reporting rate gap matters. VARIPULSE generates 87 FDA adverse event reports per 1,000 procedures. FARAPULSE generates 12. That 7x difference — driven by post-recall reporting — inflates every VARIPULSE per-procedure metric. Keep it in mind for every comparison below.
FARAPULSE (Boston Scientific)
Market leader. 200,000+ procedures. $1B+ year-one revenue. Approved for paroxysmal and persistent AFib. And 91 deaths in FDA filings.
91 Deaths: The Investigation Gap
| Count | % of deaths | |
|---|---|---|
| Uninvestigated (no device returned, no root cause) | 43 | 47% |
| Software design flaw | 4 | 4% |
| Cardiac tamponade | 5+ | |
| Cardiac arrest | 3+ | |
| Other / investigated | 36+ |
So what: 91 deaths across 200,000+ procedures is ~0.05% — consistent with BSC's published rate. The rate isn't the story. The story is that 47% of deaths have no root cause, and 4 patients were killed by a software bug (80% fatality from a product defect, not a procedural complication).
FARAPULSE vs VARIPULSE: Estimated Rates
| Per 10,000 procedures | FARAPULSE | VARIPULSE | Gap |
|---|---|---|---|
| Deaths | 4.5 | 10.0 | FARAPULSE 2x better (scale) |
| Mfg defects | 14 | 40 | FARAPULSE 3x better (scale) |
| Design flaws | 5 | 103 | See note |
| FDA reporting rate (per 1K) | 12 | 87 | 7x gap = reporting bias |
VARIPULSE rates are unreliable at ~3,000 procedures and inflated by 7x reporting intensity. FARAPULSE's denominator advantage dominates all per-procedure comparisons.
What holds up regardless of sample size:
- 47% of FARAPULSE deaths uninvestigated vs 0% of VARIPULSE's
- 4 FARAPULSE software deaths (product defect)
- J&J paused voluntarily (Class I). BSC didn't pause for the same stroke signal (Class II).
FARAWAVE Recall
January 2025: FARAWAVE recalled for neurovascular events and cracked electrode bands. Class II. VARIPULSE got Class I for the same stroke signal — because J&J voluntarily paused sales. Same problem. Different response.
Affera / Sphere-9 (Medtronic)
The strategic threat. Not because of safety — because it changes the buying decision. When Affera enters a lab, there's no separate CARTO sale. Medtronic is doubling PFA revenue from $1B to $2B on this strategy.
Safety Profile (Volume Unknown)
| Affera | VARIPULSE | |
|---|---|---|
| FDA adverse events | 391 | 260 |
| Deaths | 5 | 3 |
| Design flaws | 24 | 31 |
| Electrical failures | 63 | — |
Per-procedure rates cannot be estimated — Medtronic hasn't disclosed Affera volumes. The 63 electrical failures (most common failure mode) reflect the risk of integrated mapping/ablation: when one function fails, both are lost.
Where Each Wins
- Affera: Simple AF PVI. No capital equipment. Faster setup. Lower per-case cost.
- CARTO + VARIPULSE: Complex anatomy, redo procedures, VT (QDOT MICRO), training, multi-catheter workflows.
PulseSelect (Medtronic)
First PFA system approved (Dec 2023). 10,000+ procedures globally. Worst manufacturing quality in PFA.
| Per 10,000 procedures | PulseSelect | FARAPULSE | VARIPULSE |
|---|---|---|---|
| Deaths | 11.0 | 4.5 | 10.0 |
| Mfg defects | 133 | 14 | 40 |
So what: 133 mfg defects per 10K is 10x FARAPULSE and 3x VARIPULSE. No recall to explain the inflation. PulseSelect is also dying internally — Q2→Q4 2025 volume dropped 46% as Medtronic's own customers switch to Affera.
Abbott (Volt PFA)
FDA approved December 22, 2025. Too early for FDA adverse event data. The differentiator: conscious sedation instead of general anesthesia. If it works in practice, Abbott has an ASC-native advantage nobody else has.
Abbott also has TactiFlex (RF, solid) and EnSite X (mapping, zero deaths). No longer the "no PFA" competitor. Battle cards need updating.
Adoption Trends
FARAPULSE (blue) | PulseSelect (orange) | Affera (green) | VARIPULSE (red)
So what:
- FARAPULSE peaked Q3, dropped 28% in Q4. Growth stalled.
- PulseSelect collapsed 46% Q2→Q4. Medtronic customers switching to Affera.
- Affera grew 77% across the year. Fastest-growing PFA platform.
- VARIPULSE grew 90% (39→74) but from a tiny base. 1/6th Affera's volume.
The trajectory favors Affera.
The Investigation Gap
This is the competitive data point that doesn't depend on procedure volume, sample size, or reporting rate. Almost half of FARAPULSE's deaths have no root cause. VARIPULSE: 0 of 3.
FDA adverse event data, company-reported procedure volumes, MANIFEST-US registry (41,968 patients). See Method.