Mode switch despite undersensing of atrial fibrillation in DDD pacing

Mode switching algorithms are commonly used to protect the ventricles against high rates induced by atrial tachycardia. In the case of atrial fibrillation (AF), the response of these algorithms depends on the quality of atrial sensing. The Chorum 7234 DDDR pacemaker uses a new mode switching algorithm, based on a statistical analysis of the atrial rhythm. It includes two criteria of diagnosis: "high" if more than 28 of 32 cycles are abnormally accelerated; and "low" if more than 36 of 64 cycles are abnormally accelerated.

From a taped database of electrophysiological studies, episodes of AF lasting more than 2 minutes were selected. A tape recorder replayed the atrial signals into an external Chorum device. Each episode was replayed eight times with a programmed atrial sensitivity increasing from 0.4-2.0 mV. For each criterion of diagnosis and each programmed sensitivity, the percentage of atrial sensing, the time to switching, and the mean ventricular rate were measured. Ten episodes of AF from 10 patients (9 men and 1 woman; ages 62 +/- 16 years) were included: 1.95 +/- 0.97 mV and 196 +/- 64 ms. The sensitivity of the algorithm to diagnose atrial tachycardia reached 100%, for an atrial sensitivity set between 0.4 and 1.0 mV. The mean percentages of atrial sensed events were 74% +/- 18% and 46% +/- 9% for the "high" and "low" criteria, respectively. The mean diagnostic times were 28 +/- 26 seconds and 68 +/- 27 seconds, respectively. Sensing of < 23% of AF events resulted in failure to diagnose the arrhythmias by both algorithms. In the event of diagnostic failure, the mean ventricular pacing rate was 79 +/- 9 ppm.

Up to an atrial sensitivity of 1 mV, 100% of AF episodes were diagnosed. The Chorum mode switching algorithms are 100% reliable if > 45% of the AF waves are sensed. In the event of switching failure, the ventricle is protected by an average rate remaining below 80 ppm.