Ambulatory ECG analysis
VE variability was significantly higher in the RVOT cohort; mean CoV 1.09 ± 0.51 versus 0.42 ± 0.15 (p < 0.0001). The lowest number of VEs in a single hour was also significantly lower in the RVOT cohort, with any hourly total of < 50 VE observed in 22/23 (96%) of the RVOT cohort versus 1/17 (6%) of the LVOT cohort (p < 0.001). (Figure 1). Mean hourly combined VE was significantly lower in the RVOT cohort (253 ± 242) than the LVOT cohort (473 ± 191), p = 0.004. (Table 2).
Similarly, the total number of bigeminy or trigeminy episodes was significantly lower in the RVOT cohort (852 ± 855 versus 1572 ± 935, p = 0.02), but the hourly variability in this parameter was significantly higher in the RVOT cohort (CoV 1.72 ± 1.20 versus 0.69 ± 0.36, p = 0.001).
There was a wide range in total percentage VE burden (0.50 - 40.9%) but no significant difference between the RVOT and LVOT cohorts (16 ± 11.3% versus 24.2 ± 10.2% respectively, p = 0.3).
ROC curve analysis revealed that a CoV ≥ 0.7 predicted RVOT SOO with a sensitivity of 78% and a specificity of 94% (AUC 0.91, p < 0.0001), whilst any hour with < 50 VEs predicted a RVOT SOO with a sensitivity of 96% and a specificity of 94% (AUC 0.96, p < 0.0001). (Figure 2). ROC curve analysis of other 24-hour Holter based parameters which were inferior to CoV ≥ 0.7 and any hour with < 50 VEs are shown in the supplementary data Figure 1.