PP14.4 – 2487: Comparison of three spike and wave index quantification methods in patients with electrical status epilepticus in sleep: Variability between observers and methods

Abstract

Objective Different methods have been described in patients with electric status epilepticus during slow sleep (ESES) to calculate the spike-wave index (SWI). We measured the concordance between observers and SWI data using 3 methods. Methods We selected 3/10 SWI calculating methods described to apply in 17 whole night EEG recordings of children with clear ESES: (A1) the classic but long lasting method by Tassinari; (A2) following Aebey et al., 2005; and (A3) by Sanchez et al., 2012. Two experts separately analyzed them using the 3 methods. We used Intraclass Correlation Coefficient (ICC) as a measure of reliability to compare partial and final SWI between methods and between observers. Results SWI between methods: A3>A1>A2. Time consumption: A1 92 min, A2 33 min and A3 8 min. 15/17 patients showed SWI over the classical 85% cutoff values for the 3 methods; 2 patients showed SWI at least over 60%. The inter-observer ICC was Conclusion The SWI is higher in first sleep cycles, and decreases through the night. The SWI value for A3 depends very much on arousals. In spite of having selected unequivocal ESES, some patients did not reach the classical SWI value of 85% but reached the 60%, supporting the actual recommendation of consider ESES with lower indexes. The three quantification methods selected showed high concordance values, so the usefulness of brief methods is of great advantage in terms of time consumption.