Third-generation antitachycardia devices (ATDs) have the ability to be programmed with multiple rate zones for detection of ventricular tachycardia (VT) and ventricular fibrillation (VF) and to deliver three therapeutic choices: antitachycardia pacing and cardioversion for VT, and defibrillation for VF. In ATD research, much attention is devoted to the sensitivity and specificity of VT/VF detection versus benign rhythms. However, little information is available concerning the accuracy of separation of VT from VF in order to derive appropriate treatment. Two studies report defibrillation was only delivered to true VF in 10 - 21% of the shock episodes [ , ]. Validation of detection schemes currently implemented in ATDs will identify strengths and weaknesses, particularly with regard to excessive therapy for VT.

Methods: This study assessed the programmable parameters for proper distinction between VT and VF in three commercially approved devices. Detection schemes from the Cardiac Pacemakers, Inc. Ventak PrX 1700 (PRX), Medtronic, Inc. PCD Jewel 7219 (PCD), and Ventritex, Inc. Cadence V-100 (CAD) were simulated using custom software and tested with 63 VTs (355 + 79 ms, range: 227 -562 ms) and 18 VFs (234 + 49 ms, range: 142-334 ms) (total 71 patients). Thirty-two separate detection settings were analyzed to determine results for VT and VF detection. Detection intervals were programmed to 450 ms for VT and incrementally (by 20 ms) from 200 to 340 ms for the fibrillation detection interval (FDI). Counters were varied incrementally (by 3) from 6 to 21 for VF and from 4 to 19 for VT. Sensing was simulated using an autoadjusting threshold trigger (sensitivity 0.75, time constant 0.5 s) with refractory periods coincident with each device.

Results: Most device settings elicited 100% detection of VF for all patients. However, specificity (detection of VT) was dramatically affected by the FDI settings. (See figure.) For decreases in FDI, the specificity increased to 85% for a setting of 200 ms. However, actual devices do not allow FDI settings below 270 ms (240 ms for the PCD) in order to obviate possible misdiagnoses of VF due to sensing failures. These restrictions limit maximum specificity to 60%. Counter variations had little effect on specificity.

Conclusion: In order to accommodate sensing failures during VF, current ATD detection algorithms provide a buffer zone to ensure 100% VF detection at the expense of accurate VT detection. Despite sophisticated zones and tiered therapy, VT misdiagnoses from commercially-available ATDs do not provide optimal utilization of limited battery power. Further signal analysis using additional features may be necessary for improved VT/VF discrimination.