Several recent studies have reported that patient-reported seizure count, a common endpoint used in clinical trials of antiepileptic agents, is highly inaccurate, with patients failing to report a majority of the seizures that occur.

A decade ago, a study of 91 consecutive adult inpatients underwent video encephalographic monitoring to record seizures, with results compared with patient self-reports (Hoppe et al. Arch Neurol 2007;64:1595-1599). Overall, patients did not document 55.5% of all seizures. Accuracy was lowest for nocturnal seizures (83.3% unreported) and highest for simple partial seizures (26% unreported) and seizures while awake (32% unreported).

The group has continued this line of research with two more recent publications. In a survey of 170 adult epilepsy patients, respondents estimated that they were aware of 53% of daytime seizures, and 26% of nocturnal seizures (Blachut et al. Seizure 2015;29:97-103). Respondents who kept a seizure diary reported that their seizure counts were more accurate (87% for daytime and 77% for nocturnal seizures). The researchers noted that seizure unawareness appeared to be largely due to the seizures themselves rather than patients intentionally failing to document seizures.

Patient self-reported seizure frequency was equally inaccurate in an analysis of data for 100 patients enrolled in phase II, III and IV trials (Blachut et al. Epilepsy Behav 2017;67:122-127). Data were compared to the results of the prior study (Blachut 2015). Overall, patients estimated that their accuracy in documenting seizures was 46-53%. Thus, the accuracy of self-reported seizure frequency appears to be no higher in trial subjects. The group concluded that    most clinical trials in epileptology are based on inaccurate measures.