Combining shock reduction strategies to enhance ICD therapy: a role for computer modeling.

Objectives: To develop a computer model to test shock reduction strategies such as antitachycardia pacing and shock withholding for supraventricular rhythms, oversensing, and nonsustained ventricular tachycardia.

Background: While the implantable cardioverter defibrillator (ICD) can reduce mortality, inappropriate ICD shocks remain a limitation. Randomized trials provide evidence of efficacy, but they are not always practical. Computer models provide an alternative approach, and are particularly useful when evaluating multiple interventions.

Methods: A computer model was developed using clinical data and validated in a large ICD data set (EMPIRIC). After validation, the model was applied to 736 adjudicated clinical episodes from the ICD arm of Sudden Cardiac Death Heart Failure Trial (SCD-HeFT).

Results: The shock reduction strategies hypothetically reduced the number of VT/VF shocked episodes in SCD-HeFT by an estimated 59% (from 952 observed to 395 modeled shocks, probability of >0.999) at detection duration settings (18 of 24 intervals). The percentage of patients experiencing inappropriate shocks over 5 years was decreased by 15% (23.5-8.4%), and the number of shocks for non-VT/VF episodes was decreased from 423 to 77 (82% reduction). The percentage of patients receiving shocks for VT/VF was reduced from 30.7% (SCD-HeFT) to 26.1% with the addition of ATP. Extended detection (24 of 32 or 30 of 40 intervals) showed modest additional improvement compared to 18 of 24 intervals.

Conclusion: Computer modeling is able to predict the results of a known clinical trial and demonstrate that shock reduction strategies have the potential to significantly reduce inappropriate and unnecessary ICD shocks versus the mandated programming used in SCD-HeFT.