Diagnostic accuracy of a structured interview for nocturnal frontal lobe epilepsy (SINFLE): a proposal for developing diagnostic criteria.

OBJECTIVES: To measure the accuracy of anamnestic features collected during clinical history for the diagnosis of nocturnal frontal lobe epilepsy (NFLE). METHODS: A case-control diagnostic study. Participants included a case group of people with ascertained target disease (NFLE group) and a control group of people with sleep disorders potentially confounding for NFLS (NOT-NFLE group), defined by means of a consensus procedure among experts (panel diagnosis as reference standard). Two major clinical patterns defining the semeiology of the epileptic event (i.e. dystonic, DP, and/or hyperkinetic pattern, HP), and 13 additional minor features were identified, formulated as questions, and telephonically administered to NFLE and NOT-NFLE groups by a trained doctor blinded to the final diagnosis. The diagnostic accuracy of each characteristic was tested against the reference standard. RESULTS: Out of 262 selected subjects, 101 were recruited; 42 were NFLE and 59 NOT-NFLE. A positive history of DP or HP had a sensitivity of 59.5% and a specificity of 91.5%, irrespective of the other minor anamnestic features. The anamnestic model improved, with a sensitivity of 59.5% and specificity of 96.6%, if at least one of the following four minor anamnestic features was added: (a) duration less than two minutes, (b) unstructured vocalization during the episode, (c) experience of an aura preceding the motor attack, and (d) a history of tonic-clonic seizures during sleep. CONCLUSIONS: The present study disclosed two major anamnestic patterns and four minor features that we called SINFLE, with unsatisfactory sensitivity but high specificity. These patterns could be the basis for developing future NFLE diagnostic criteria and to quantify the diagnostic accuracy of elements usually collected in the clinical history.