Abstract
BACKGROUND: There have been limited data examining the temporal relationship between device-derived daily activity and ventricular arrhythmias (VAs). OBJECTIVE: We aimed to assess whether activity predicted VAs or VAs predicted changes in activity. METHODS: The CERTITUDE registry includes over 55,000 implanted devices active on Home Monitoring. Daily data on activity are captured by a 1-axis accelerometer. Temporal changes in activity during treated VAs were analyzed using the first event and 7-day activity windows (baseline, pre-event, and postevent). Baseline period was defined as 31 to 38 days prior to VA. VAs were categorized by heart rate (≤200 beats/min, >200 beats/min) and treatment (shock or antitachycardia pacing). Differences were assessed using the binomial proportion test and case-crossover analysis. RESULTS: A total of 5631 devices (3688 implantable cardioverter-defibrillators, 1943 cardiac resynchronization therapy defibrillators) were analyzed with a cumulative follow-up duration of 18,354 years (5.6 million days). Patients with VA events >200 beats/min with shock (n = 593) had a significant decline in activity post-VA with a median -8.7% reduction (interquartile range -24.6% to 7.3%, P < 0.001). However, there was no reduction in activity before VA events >200 beats/min (P = .690) or before or after VA events >200 beats/min with antitachycardia pacing. However, VA events ≤200 beats/min with shock had reductions in activity following the event (-5.8%, interquartile range -29.5% to 12.3%, P = .003). Case-crossover analyses confirmed lower activity rates following for VA events >200 beats/min with shock. CONCLUSION: In the CERTITUDE registry, we have shown a temporal decline in device-derived activity following VA events >200 beats/min and for VA events <200 beats/min treated with a shock, but we did not find changes in activity preceding a shock event.