The mean ventricular fibrillation cycle length: a potentially useful parameter for programming implantable cardioverter defibrillators.

In programming the implantable cardioverter defibrillator (ICD), the ventricular tachycardia (VT) detection cycle length (CL) is based on the CL of the documented tachycardia but the ventricular fibrillation (VF) detection CL is set arbitrarily. Appropriate programming of VF detection may not only reduce the incidence of inappropriate ICD shocks for non-VF rhythms but can also avoid the fatal underdetection of VF. The mean VFCL may provide a useful parameter for optimal ICD programming for VF detection if it is reproducible. This study examined the intrapatient reproducibility and interpatient variation of the mean VFCL in 30 ICD patients (25 men and 5 women, mean age 63 +/- 13 years). A total of 210 VF episodes (7 +/- 4 per patient, range 3-17) induced by T-wave shocks (166) or AC (44) at the ICD implant (30 patients) and the predischarge test (12 of 30 patients) were analyzed. The mean VFCL was calculated from the stored V-V intervals in the ICDs. Although the mean VFCL varied significantly from 171 +/- 6 to 263 +/- 11 ms (P < 0.01) among different patients, it was reproducible among different VF episodes in an individual patient (maximal variation 4-50 ms, P > 0.05). The mean VFCL was not significantly different between patients with and without antiarrhythmic drugs (210 +/- 32 vs 210 +/- 23 ms, P > 0.05) and was correlated with the ventricular effective refractory period (r = 0.5, P < 0.05). The mean VFCL varies greatly among different patients but remains reproducible in an individual patient, suggesting that the mean VFCL may serve as a reference for ICD programming of VF detection.