Abstracts

Rationale: Seizures can result in impaired cognition and consciousness in epilepsy patients, potentially affecting many aspects of their lives, including work and driving. Clinicians rely on the patient’s self-description of impairment during seizures both for seizure characterization and for making decisions such as driving eligibility, but the reliability of these patient self-reports is unknown. In this study, we aimed to determine the reliability of patient self-reports of their responsiveness, speech, memory, awareness, and consciousness during seizures when asked shortly after the event and at admission.

Methods: We designed a questionnaire asking about the patient’s ability to respond to others, speak, and recall events, along with their awareness and conscious state during their seizures. We administered the questionnaire daily to inpatients undergoing video-EEG monitoring, asking them about their seizures in the past 24-72 hours. In addition, we administered a questionnaire to the same patients upon admission asking them to rate their responsiveness, speech, and awareness during seizures at home. We compared patient answers on both questionnaires to objective behavioral ratings using video review. Concordance of the patients’ subjective ratings and objective ratings was measured using bootstrapped Cohen’s kappa (κ), proportion of positive agreement, and proportion of negative agreement.

Results: We included 86 seizures from 39 patients. For responsiveness, patient answers had moderate concordance with objective ratings (κ = 0.434), and concordance was good for speech (0.679). Concordance for memory was good (0.674-0.743) with high proportions of positive and negative agreement. Concordance of subjective awareness with objective responsiveness and speech was moderate and fair respectively (0.494, 0.291), but concordance with objective memory was good (0.673, 0.774) with high positive and negative agreement. Concordance of subjective consciousness with objective responsiveness, speech, and memory was poor to moderate. In contrast to the daily questionnaire, patient answers to the admission questionnaire had poor concordance with all objective behavioral ratings.

Conclusions: Of all the measured functions, patient subjective report of memory impairment was reliable when compared to objective memory rating on video review, with good concordance and high positive and negative agreement. This suggests that patient self-reports in this domain reflects their true function during seizures. Patient reports when asked soon after their seizures were more reliable than when asked at admission, highlighting the importance of asking the patient about their seizures promptly after the event. Otherwise, determining functional status using both historical information and external observation may be more reliable.

Funding: Please list any funding that was received in support of this abstract.: Mark Loughridge and Michele Williams Foundation, Betsy and Jonathan Blattmachr family, James G. Hirsch, MD medical student research fellowship.