LETTER TO THE EDITOR To the Editor: In an article published in the September 2006 issue of Heart Rhythm, Mainigi et al 1 address an important issue of elevated defibrillation thresholds (DFTs) among recipients of a biventricular implantable cardioverter-defibrillator (ICD). However, the study design had some serious flaws, as acknowledged by the authors. DFT testing was left to the discretion of individual im- planters as long as patients undergoing the procedure met the criteria of two successful shocks with a 10-J safety margin. This methodology is inherently flawed because it can yield erroneously high DFT among such patients. For example, if an individual implanter checked the defibrilla- tion efficacy at 25 J in a 35-J output device and was successful twice, the estimated DFT would be 25 J. It is quite probable that the higher mortality associated with higher DFTs, as noted in other studies, was not reproducible in the present study because the higher DFT group did not truly have high DFT. Because DFT testing was the focus of the study, the protocol for DFT testing should have been more rigorous, using any one of the previously published protocols. In addition, the authors did not consider renal function as a variable in their analysis, which could adversely affect DFTs. We reported a study that showed renal function to be an independent predictor of high DFT among all ICD re- cipients, and proportionately more patients died of arrhyth- mia as renal function worsened. 2 This is in line with a preliminary post hoc analysis of patients in SCD-HeFT (Sudden Cardiac Death in Heart Failure Trial), in which renal insufficiency was also associated with higher mortality among ICD recipients. 3 The effect on sudden cardiac death reduction was less robust compared with the nonrenal in- sufficiency group. In another recently published study, 4 mortality was higher among ICD recipients with renal in- sufficiency compared to patients with normal renal function. Therefore, it is quite conceivable that properly performed DFT testing and inclusion of renal function as a variable would have altered the results of the study. Abdul Wase, MD abdul.wase@wright.edu Division of Cardiology Wright State University Boonshoft School of Medicine Dayton, Ohio References 1. Mainigi SK, Cooper JM, Russo AM, Nayak HM, Lin D, Dixit S, Gerstenfeld EP, Hsia HH, Callans DJ, Marchlinski FE, Verdino RJ. Elevated defibrillation thresholds in patients undergoing biventricular defibrillator implantation: inci- dence and predictors. Heart Rhythm 2006;3:1010 –1016. 2. Wase A, Basit A, Nazir R, Jamal A, Shah S, Khan T, Mohiuddin I, White C, Saklayen M, McCullough PA. Impact of chronic kidney disease upon survival among implantable cardioverter-defibrillator recipients. J Interv Card Electro- physiol 2004;11:199 –204. 3. Dickinson MG, Ip JH, Hellkamp AS, Anderson J, Poole JE, Sharma AD, Johnson GW, Freudenberger RS, Mark DB, Lee KL, Bardy GH, and The SCD-HeFT Investigators. Decreased impact of implantable defibrillators on sudden cardiac death in patients with renal insufficiency and heart failure: an analysis of the SCD-HEFT. Heart Rhythm 2006;3:S38. 4. Eckart RE, Gula LJ, Reynolds MR, Shry EA, Maisel WH. Mortality following defibrillator implantation in patients with renal insufficiency. J Cardiovasc Electrophysiol 2006;17:940 –943. To the Editor—Response: We appreciate Dr. Wase’s thoughtful comments on our article in the September 2006 issue of Heart Rhythm. 1 As reported in our article, implanters in our retrospective in- vestigation used a defibrillation efficacy strategy in which patients who did not respond to defibrillation 10 J below the maximum output of the device were defined as having an elevated defibrillation threshold (DFT). This technique has been shown to predict implantable cardioverter-defibril- lator (ICD) success to a high degree 2 and is a clinically more efficient testing method, albeit one that does not define the precise DFT. We disagree with the criticism that this strategy would lead to erroneously high DFTs among pa- tients. A patient with a 35-J device who had two successful shocks at 25 J was categorized in our study as not having an elevated DFT. However, a similar patient who failed at 25 J but subsequently was successful between 26 and 35 J would have been included in the elevated DFT group. Al- though it is true that the actual DFT might well be lower than 25 J, we did not assess DFT as a continuous variable but rather one with a binary value— elevated or not ele- vated. Thus, only patients with a safety margin 10 J were categorized as having high DFT, and such patients would be equally identified with either a DFT or a defibrillation efficacy strategy. This 10-J safety margin was chosen be- cause of its common acceptance in clinical practice and its support from early evaluation. 3 A significantly more rigor- ous DFT testing strategy would not have more accurately defined patients with DFT 10 J below the maximum output of the device. Regarding the impact of renal function on DFT, we agree that this certainly would be an interesting variable to inves- tigate in our cohort, although having precise DFT data for all patients would make such an analysis more meaningful. One must be cautious, however, when making a correlation between a higher DFT during life and the mode of death. The implication that patients with renal failure have a higher arrhythmic death rate because of extreme DFT elevation and failed ICD shocks is questionable. Defining the mode of death in any patient with an ICD is challenging, as patients 1547-5271/$ -see front matter © 2007 Heart Rhythm Society. All rights reserved.