Introduction: Defibrillation testing of the implantable cardioverter-defibrillator (ICD) is considered a standard and required practice at the time of implantation. How much testing, if any in some cases, should be performed, however, remains unknown.
Methods and results: Included in this retrospective analysis were 835 patients (77% men; age 65 +/- 13 years) who received transvenous ICDs between January 1996 and December 2003. One hundred twenty-nine (15.5%) had intraoperative defibrillation threshold (DFT) testing, 503 (60.2%) had limited defibrillation safety margin testing, and 203 (24.3%) had no defibrillation testing. We compared the outcome (success of ICD therapies against spontaneous VT/VF events and survival) of the three groups of patients, who in some respects had important clinical differences. The success of the first delivered shocks against VT/VF was similar for DFT (91%), safety margin testing (91%), and no-testing (92%) groups; and the second shocks terminated the remaining episodes in all three groups. Sudden-death-free survival rates were similar in the three groups, however, the overall long-term survival rate was significantly lower in the no-testing group (58%) than in the DFT (74%) and safety margin testing (69%) groups (P < 0.0005). Multivariate analysis found no strong predictors of sudden death, but there were several independent predictors of overall mortality including lack of ICD testing (HR: 2.031, CI: 1.253-3.290, P = 0.004).
Conclusion: In this select patient cohort, success of ICD therapies and sudden-death-free survival were similar in patients who had DFT, safety margin testing, and no testing, but overall survival was significantly lower in the no-testing group. Thus in the absence of prospective mortality data, a minimum of safety margin ICD testing should remain standard practice.