Paroxysmal phenomena in severe disabled children with refractory seizures. From clinical to long-video-EEG processing data to re-examine suspect events

To provide an estimate of the occurrence of misdiagnosis in paroxysmal events in institutionalized children with severe disabilities and refractory epilepsy. A multi-step diagnostic survey, from observational to long-term video-EEG monitoring was performed in 46 severe disabled children. Multirater Kappa statistic was used to assess agreement between investigators and to individualize children who remained with dubious events. Subsequently, prolonged EEG-video monitoring analysis was performed in selected children to define phenomena due to seizures. A total of 128 video records were performed, 64 routine video-EEG and 27 long-monitoring video-EEG data were screened for detailed analysis. Thirty (21 female, 9 male) children (65%) with dubious seizures were identified by video records (concordance K=0.63). Of these, in 18 children (39%) seizures were excluded by routine video-EEG monitoring (K=0.86). Twelve children (26%) required accurate investigations with long-term video-EEG. In 5 children (11%), 3 symptomatic and 2 cryptogenic, very short and subtle seizures were confirmed by investigators concordance (K=0.83). Distinguishing paroxysmal phenomena is a challenge in children with severe disabilities; its most remarkable consequence is inappropriate pharmacological treatment and social costs. Our data suggest that the frequency of misdiagnosis could have been underestimated. The clinicians who manage children with severe disabilities and refractory epilepsy must remain alert to risk of an incorrect treatment.