Epileptic seizure or not? Proportion of correct judgement based only on a video recording of a paroxysmal event.

PURPOSE: Our study was intended to measure the proportion of correct seizure recognition among different medical and non-medical groups based on only a video recording.

METHODS: Video recordings about paroxysmal movements of 15 very young infants (2days - 5 months of age) were displayed for six groups: 159 1st-year medical students, 65 4-5th-year medical students, 52 paediatric residents, 18 paediatric neurologists from different European countries, 43 adult neurologists and 37 parents whose children were treated at our Department. All participants were asked to decide which recording they considered as of epileptic origin or a non-epileptic event. Correct answer rate (CAR) was calculated in each group for every video.

RESULTS: The average CAR was the lowest in the group of 1st-year medical students (36.6%), the best results were reached by paediatric neurologists (67.4%). The CAR was significantly different between the groups of 1st-year medical students and paediatric neurologists (p=0.02), and between the groups of 1st-year medical students and residents (p=0.045). The CAR of the most deceptive epileptic seizure was only 18.2%. The judgement of parents proved to be better than that of the 1st-year medical students.

CONCLUSIONS: Recognising epileptic seizures in very young infants without EEG is extremely inaccurate. Even trained paediatric neurologists were able to judge correctly the different movement types in only 67.4% of the cases. The role of education and experience is clearly indicated by the increase in CAR from 1st-year medical students through well-trained paediatric neurologists.