Question (5 May 2025):
What will be the battery life of the Abbott Aveir VR leadless pacemaker in a 70-year-old patient with a base rate of 50 bpm who exercises ~700 minutes per week, given the following interrogation data?
• Implant (8 May 2024) – Capture 1.5 V @ 0.4 ms, Sense 2.7 mV, Impedance 340 Ω
• Follow-up (1 Apr 2025) – Capture 3.0 V @ 0.4 ms, Sense 4.0 mV, Impedance 550 Ω
Programming: Mode VVI, Sensor Passive, Base Rate 50 bpm, Max Sensor Rate 130 bpm.
Assumptions (Abbott IFU) | Service-life BOS → RRT |
---|---|
50 bpm, 2.5 V, 0.4 ms, 100 % pacing, 500 Ω | ≈ 10.8 yr |
Same, 50 % pacing | ≈ 15.9 yr |
Battery drain is proportional to V² × PW ⁄ R
.
Increasing output from 2.5 V to 3.5 V multiplies pulse energy by (3.5/2.5)2 ≈ 1.96; the rise in impedance from 340 Ω to 550 Ω reduces it by ÷1.62, so net energy per paced beat is ~1.2× higher.
With a 243 mAh cell, total drain at 3.5 V and 100 % pacing is ~3.0 µA, giving about 9 years. Halving the pacing percentage to 50 % yields roughly 13–17 years. The device’s internal algorithm, however, observes an actual pacing burden high enough to produce the 4½-year print-out.
Factor | Impact on life | Actionable? |
---|---|---|
Capture threshold / Programmed amplitude | +0.5 V costs ~15–25 % life | Ask EP to re-test; lower output if safe |
Pacing % | Each inhibited beat saves energy | Maintain fitness; avoid excess bradycardia meds |
Pulse width (0.4 ms) | Halving PW adds ~40 % | Only if consistent capture at 0.24 ms |
Telemetry sessions | Each follow-up costs hours | Space out checks; avoid “peeking” weekly |
If nothing changes—output ~3.5 V, similar night-time bradycardia, same follow-up cadence—the 4–5 year projection remains realistic. If a threshold test shows capture at 2.0–2.5 V, re-programming to that level could roughly double remaining longevity.
What will be the battery life of the Abbott Aveir VR leadless pacemaker implanted to a 70-year-old patient with a heart rate of 50 bpm, who exercises an average of 700 minutes weekly, given these interrogation data?
Programming parameters: Mode VVI • Sensor Passive • Base Rate 50 bpm • Max Sensor Rate 130 bpm
Programming assumptions (Abbott IFU — model LSP112V) | Projected service life (BOS → RRT) |
---|---|
50 bpm, 2.5 V, 0.4 ms, 100 % pacing, 500 Ω | ≈ 10.8 yr |
Same, 50 % pacing | ≈ 15.9 yr |
Energy per pulse is proportional to V² × PW ⁄ R
.
Raising output from 2.5 V to 3.5 V multiplies pulse energy by
(3.5 / 2.5)² ≈ 1.96
. The rise in impedance from 340 Ω to 550 Ω partly offsets
this (÷ 1.62), so net energy per paced beat is about 1.2-fold higher.
Housekeeping current (sensing, clocks, memory, telemetry) is ≈ 0.8 µA even when the device is
inhibited. At 2.5 V and 100 % pacing Abbott lists total drain 2.3 µA; scaling to 3.5 V gives
≈ 3.0 µA. With a 243 mAh cell, that provides
243 mAh ⁄ 3.0 µA ≈ 9.3 years
at 100 % pacing. Halving the pacing percentage to
50 % raises longevity to roughly 13–17 years.
The Aveir’s on-board algorithm tracks actual pacing percentage since the last follow-up, the programmed capture margin, diagnostic tally pulses, and every telemetry session. That real-world drain is higher than nominal, so the projection is shorter. A common culprit is night-time bradycardia < 50 bpm — each missed beat triggers a 3.5 V pulse.
Factor | What it does to battery life | Can you influence it? |
---|---|---|
Capture-threshold & programmed amplitude | Doubling V ≈ quadruples pulse energy; each extra 0.5 V costs 15–25 % life | Ask your EP to run a threshold test; if safe, lower output to keep a 2× margin |
Pacing percentage | Each inhibited beat saves energy; heavy exercise (> 700 min wk⁻¹) helps | Maintain conditioning; avoid medications that induce excess bradycardia |
Pulse width (0.4 ms) | Halving PW adds ≈ 40 % life | Only if reliable capture at 0.24 ms |
Rate-responsive sensor (OFF) | Turning it ON raises average rate → higher drain | Usually left OFF when base rate 50 bpm and patient active |
Telemetry sessions | Each interrogation costs several hours of life | Cluster downloads; avoid weekly “peeks” unless clinically needed |
If nothing changes (output ≈ 3.5 V, same night-time bradycardia, similar interrogation cadence) the 4–5 year projection remains realistic. If a threshold test shows capture at ≤ 2.5 V, re-programming to that level could roughly double the remaining longevity.
Ask your electrophysiologist to (1) perform a formal threshold test at rest and during deep breathing, (2) review whether 3.5 V output is still necessary, and (3) print the device’s detailed current-drain report. Those three steps usually yield the biggest, safest gain in projected service life.
```