Inter‐rater reliability and clinical utility of the <scp>BASED</scp> score in infantile epileptic spasms syndrome

OBJECTIVE: Hypsarrhythmia is the classical EEG pattern of children with infantile epileptic spasms syndrome (IESS). Multifocal spikes, slow waves of large amplitude, and chaoticity are its main characteristics, but these lack clear definitions, and the interrater reliability (IRR) is poor. The aim of this study was to substantiate the use of the Burden of Amplitudes and Epileptiform Discharges (BASED) score for children with hypsarrhythmia on initial EEG and to compare the IRR to traditional scoring. METHODS: We identified before and after treatment EEGs from 20 consecutive patients from a local database. Clinical remission was defined as the remission of spasms. Three experienced Danish neurophysiologists, blinded to clinical information, rated each EEG separately using traditional visual analysis and the BASED score. A BASED score ≥4 and any hypsarrhythmic pattern were considered consistent with an epileptic encephalopathy. IRR was estimated using Gwet's AC1. RESULTS: IRR for scoring an epileptic encephalopathy was significantly better when using the BASED score compared with the traditional method (0.88 vs. 0.62, p = 0.01). Complete agreement was observed in 36 out of 40 EEGs using the BASED score vs. 28 out of 40 using the traditional method (Fisher's exact test: p = 0.048). IRR was near perfect or substantial in all but one of the individual components of the BASED score. Clinical remission was seen in 14 out of 19 patients, and the clinical course was consistent with EEG in 18 out of 19 patients using the BASED score and in 16 out of 19 patients using the traditional method (Fisher's exact test: p = 0.60). SIGNIFICANCE: The BASED score demonstrated very high interrater reliability, exceeding that of traditional hypsarrhythmia scoring. It also corresponded closely with the clinical course, supporting its validity.