Which is the optimal testing method for identifying an AV delay that allows intrinsic conduction?

It is desirable to maintain normal, conducted ventricular activation in patients with dual-chamber pacemakers and preserved atrioventricular (AV) conduction. The shortest AV delay resulting in consistent ventricular inhibition (avoiding ventricular pseudofusion) was determined by a conventional incremental (inside-out) technique vs the alternate decremental (outside-in) technique in 20 such patients. Determinations were made in VDD mode in 20 patients and DDD mode (approximately 10 beats/min faster than the intrinsic rate) in 19. In VDD mode, the shortest AV delay avoiding ventricular pseudofusion was never found during inside-out testing. It was identical with both methods in 10 patients (50%), and shorter by 10-80 ms (mean 20 +/- 20 ms) with the outside-in method in the remaining 10 (P = 0.004). In DDD mode, the shortest AV delay resulting in consistent ventricular inhibition was found only once during inside-out testing. It was the same with both methods in 13 patients (68%), and shorter by 10-20 ms (mean 14 +/- 5 ms) with the outside-in method in the remaining 5 (26%, P = 0.18; Fisher's exact test). The shortest sensed AV delay preventing ventricular pseudofusion is most likely to be found with a decremental method (outside-in). In rare patients, it identifies AV delays resulting in inhibition, while ventricular pacing persists at longer programmable AV delays with the conventional inside-out approach.