Fehmi Keçe

Biophysical parameters and Micro-Embolic Signals 111 4 4.3 Results The median MES count per application was 9 (IQR 1-37). The median MES count was highest for the left superior pulmonary vein (PV; 31 [IQR 6-72]), followed by the left inferior PV (10 [IQR 2-41]), the right superior PV (5 [IQR 1-15]), and the right inferior PV (2 [IQR 0-9]; p<0.001). Electrode proximity, 1-8 and 2-9 electrode interaction, and temperature overshoot were seen in only 1.3%, 0.7%, 0.3%, and 5.0%, respectively, and were excluded from the analysis. Univariable linear mixed models showed that average power (p<0.001), average and total effective energy (p=0.001 and p<0.001, respectively), and average impedance (p=0.039) were positively associated with MES count, while average temperature (p<0.001) and temperature 2 seconds after ablation onset (p<0.001) were negatively associated. Other patient and procedural characteristics were not associated with MES count. Due to collinearity, only 4 variables (average power, total effective energy, average impedance, temperature 2 seconds after ablation start) were included in the final multivariable model. Observed median MES count showed a stepwise increase with every quartile of the multivariable model-based predicted MES count (Figure 1). Figure 1. Observed median MES count within quartiles of the multivariable model-based predicted MES count (fixed-effects predicted values) with confidence bounds.