Letter to the Editor: Interobserver Variability of Heart-to-Mediastinum Ratio in I-123 MIBG Sympathetic Imaging

One of the significant contributors to the variation of H/ M ratio in cardiac imaging, are the settings of the regions of interest (ROIs) that are used. Currently, as correctly described by Chen et al. [1, 2], there are three main approaches ROI is placed along the anterior part of the LV wall [2]. In our current ongoing study in pediatric patients by van der Palen et al.., we compared both approaches described by Chen et al. (ie, the “whole heart ROI” and the “small anterior wall ROI”). The reason for comparing those instead of the “LV wall ROI” by Gerson et al. is twofold: possible comparison of our results with Chen’s references and potential inaccurate placement of an ROI along the complete LV wall in small pediatric hearts (ie, the child’s heart is often too small for exact placement over the LV wall). Several studies in pediatric cardiology also use this approach [5, 6]. When studying the subsequently calculated H/M ratios we noticed that the interobserver variability when using these methods of ROI placement was only minor, when H/ M ratios were obtained independently by two observers at 4 and 24 h post injection in our population of 30 pediatric patients. Intraclass correlation coefficients (ICCs) were as follows: for the small anterior wall ROI H/M ratio 0.882 (95 % CI: 0.578–0.956) and for the whole heart ROI H/M ratio 0.955 (95 % CI: 0.904–0.979) at 4 h. At 24 h, the ICC was 0.911 (95 % CI: 0.716–0.965) for the small anterior wall ROI H/M ratio and 0.951 (95 % CI: 0.896–0.977) for the whole heart ROI H/M ratio. Therefore, it can be concluded that the use of both ROI definition methods in children results in substantial to almost perfect interobserver B. F. Bulten (*) : L.-F. de Geus-Oei Department of Nuclear Medicine (757), Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands e-mail: [email protected]