In Response

We agree with Ender and Swanevelder1 that the results from Mukherjee et al.2 put into question our results3 regarding the significant difference in prolapse localization between 2-dimensional (2D) and 3-dimensional (3D) transesophageal echocardiography (TEE). Although lacking significance, the studies cited by Ender and Swanevelder also detected an advantage for 3D TEE in P2 prolapse localization.4–6 The results by Ben Zekry et al.4 and Grewal et al.5 show significant superiority of 3D TEE for various other scallops. Despite superiority, the level of significance was not reached for the scallop P2. However, and what is also indicative of an advantage of spatial imaging for this specific location is that Ben Zekry et al.4 found a significant difference when detection of P2 flail instead of prolapse was investigated. From data extracted from Manda et al.,6 a respective 2D sensitivity of 80% and specificity of 62% can be derived for the detection of P2 prolapses. This yields lower accuracy than that found in our patient population. Therefore, we conclude that our 2D images contained the same, or more, information about P2 than the 2D images in comparable studies. It may be however that the controversy claimed between studies in detecting significant differences for pathologies at specific locations is due to their inadequate power. Mukherjee et al.2 together with Ender and Swanevelder have performed and published a quasi power analysis in 2011 analyzing results from 50 patients with mitral regurgitation scheduled for surgery to calculate the number of patients necessary to establish significant results for differences in affected scallops. The number of involved pathologies and their respective sites were not reported. Assuming that this calculation was transferable and accurate, no study has yet been performed that included a sufficient number of patients to obtain significant results for any scallop (this would have required 220 patients for P3 and more than 700 patients for any other locations). To date, numerous studies from various institutions have compared intraoperative 2D and real-time 3D TEE for the localization of mitral valve pathologies.3–8 These studies included between 18 and 73 patients and all agree regarding the superiority of 3D TEE accuracy in prolapse localization for several specific investigated sites. Consequently, we do not see a disagreement between our data and the current literature. We doubt the generalization of the conflicting results and conclusions by Mukherjee et al.2 that more than 220 patients are required to establish significance for P3 or more than 700 patients for any other possibly affected scallop. However, we agree with Ender and Swanevelder1 that multicenter studies comparing data from different institutions, incorporating image acquisition and image evaluation are needed. These studies should also address the question of reproducibility and multirater agreement. Christian Rosendal, MD Klinik Beau-Site Hirslanden Clinics Berne Bern, Switzerland Maximilian D. Hien, MD Department of Pediatrics Research Training Group 1126 University of Heidelberg Heidelberg, Germany [email protected]