PREDICTORS OF DEFIBRILLATION EFFICACY IN PATIENTS UNDERGOING EPICARDIAL DEFIBRILLATOR IMPLANTATION

Objectives. The objective of this study was to identify predictors of defibrillation threshold in patients undergoing epicardial defibrillat or implantation.Background. Factors that predict epicardial defibrilla tion efficacy are poorly defined. Methods. The data from 375 consecuti ve adult patients were reviewed. After exclusion of 137 patients in wh om defibrillation threshold was not obtained, 238 patients (32 women a nd 206 men) with a mean age of 58.9 +/- 13.3 years formed the study gr oup. Coronary heart disease was present in 175 patients and the mean l eft ventricular ejection fraction was 35.8 +/- 15.4%. At device implan tation, three epicardial patch sizes were available and shocks could b e delivered over one current pathway (two patches) or over two current pathways (three patches with simultaneous or sequential shocks). Defi brillation threshold was defined as the lowest programmed energy that successfully defibrillated the heart, provided there had been an unsuc cessful shock at a lower energy level or successful defibrillation at less-than-or-equal-to 5 J. Results. The mean defibrillation threshold was 8.6 +/- 5.3 J. With univariate analysis, female gender, sequential shocks with three patches, higher left ventricular ejection fraction and lower New York Heart Association functional class predicted a lowe r defibrillation threshold. In the multivariate analysis, female gende r (coefficient -3.9; 95% confidence interval [CI] -1.9 to -5.0 J), eje ction fraction (coefficient -0.6; CI -0.1 to -1.0 J/decile) and sequen tial shocks (coefficient -2.5; CI -1.0 to -4.0 J) were independently a ssociated with a lower defibrillation threshold. Total epicardial patc h conductive surface area normalized to body surface area reached bord erline significance (coefficient 0.004; CI 0 to 0.01; p = 0.10). Antia rrhythmic drug use, including amiodarone, did not predict defibrillati on threshold. Conclusions. Female gender, high left ventricular ejecti on fraction and the use of sequential pulse shocks were important dete rminants of improved defibrillation efficacy.