Automatic verification of ventricular stimulation: Fusion management algorithm

Background: Ventricular stimulation with automatic control and back-up pulse warrants maximum safety for the patient and increases device longevity. Fusion phenomenon may hinder evoked response (ER) detection and cause unnecessary back-up stimulation. We evaluated an automatic fusion beat management algorithm and its relationship with atrioventricular (AV) interval programming in a DDD/R pacemaker. Methods: We analyzed 45 Holter registrations of patients implanted with an Insignia Ultra DR. Fusion beat classification, as performed automatically by the Fusion Management Algorithm, was compared to visual inspection by the analyzing physician. Fusions were classified as loss of capture (LOC), true fusions, or fusions classified as captured beats. Patients were divided into two groups according to AV interval programming: long AV delay (AVl)(>150 ms) short AV delay (AVs)(≤150 ms). Results: Percentage of paced beats was on average 65%. Fusion beats resulted in 26% of the total. No and LOC was detected as a fusion beat by the algorithm. No useless back-up pulses were delivered because of fusion. Misclassification of fusion beats as captured beats occurred at different rates depending on programmed AV delay. The accuracy in detecting true fusion was greatly improved by programming a long AV delay (P < 0.001 vs AVs). Conclusions: Performance of fusion management algorithm (FMA) is improved by programming a long AV interval or with automatic search of intrinsic activity. Programming the AV interval to enhance intrinsic conduction does not hinder FMA performance. © 2008, The Authors.