Abstracts

Challenges in Seizure Diagnosis: Patients with Seizures and Seizure Mimics Respond Similarly to Common Diagnostic Questions

Abstract number : 2.126
Submission category : 4. Clinical Epilepsy / 4B. Clinical Diagnosis
Year : 2023
Submission ID : 304
Source : www.aesnet.org
Presentation date : 12/3/2023 12:00:00 AM
Published date :

Authors :
Presenting Author: Ellen Snyder, MD – University of Colorado School of Medicine

Stefan Sillau, PhD – Department of Neurology – University of Colorado School of Medicine; Kelly Knupp, MD, MSCS – Departments of Pediatrics and Neurology – University of Colorado School of Medicine; Jacqueline French, MD – Comprehensive Epilepsy Center – New York University Grossman School of Medicine and NYU Langone Health; Saori Haigo, MD, PhD – Department of Neurology – University of Colorado School of Medicine; Amber Khanna, MD, MS – Department of Cardiology – University of Colorado School of Medicine; Marius Birlea, MD – Department of Neurology – University of Colorado School of Medicine; Kavita Nair, PhD – Departments of Neurology and Pharmacy – University of Colorado School of Medicine; Michelle Sandoval, BS – Department of Neurology – University of Colorado School of Medicine; Jacob Pellinen, MD – Department of Neurology – University of Colorado School of Medicine

Rationale: People with undiagnosed seizures often present to non-specialists in emergency departments or primary care settings, and often experience delays to treatment. Standardized history taking may reduce time to diagnosis. A prior qualitative study assessing themes of epilepsy diagnosis identified common questions used by physicians for identifying seizures vs seizure mimics. In this study, we tested the diagnostic accuracy of questions in people with a known diagnosis of epilepsy and compared responses to people with non-epileptic seizures, migraine, and syncope.

Methods: Two hundred outpatients at the University of Colorado between 10/26/21 and 2/24/23 with established diagnoses of focal epilepsy (n = 50), migraine (n = 50), non-epileptic seizures (n = 50), and syncope (n = 50) completed an online questionnaire about symptoms related to their diagnosis. Eight yes-or-no questions assessed specific characteristics (Figure 1). Responses were examined using two-way frequency tables. Sensitivity and specificity were calculated, and chi-square/Fisher’s exact tests were applied. A weighted score (0-2) was given to responses most supportive of an epilepsy diagnosis based on total number of responses (Y or N), and this was used as an explanatory variable in logistic regression adjusting for age and sex.

Results: The most sensitive question was “do your events happen suddenly or ‘out of the blue’?” (98%), and least sensitive was “are your events always the same?” (46%). Epilepsy was distinguished from migraine when asking about sudden onset of symptoms (98% vs 72%), short duration (76% vs 8%), lack of prolonged episode (82% vs 2%), and stereotyped events (46% vs 24%). Epilepsy was distinguished from syncope with sudden onset of symptoms (98% vs 84%) and post-ictal symptoms (84% vs 52%). After controlling for age and sex, the weighted sum of responses indicating a diagnosis of epilepsy was positively associated with epilepsy ([p< 0.0001], and the area under the model’s ROC curve was 0.80 [95% CI:  0.74, 0.87]), compared to 0.73 for the unweighted model (p=0.03).

Conclusions:
People who experience alterations in consciousness related to neurologic events may struggle when obtaining an accurate diagnosis. Our study shows that with epilepsy and seizure mimics impairing awareness, non-specific diagnostic questions may have limited distinguishing ability. This supports a more nuanced approach to seizure diagnosis utilizing additional questions and information from witnesses, and underscores contribution of confirmatory diagnostic testing, such as video EEG, in epilepsy diagnosis.

Funding: This study was supported by Department of Neurology at the University of Colorado School of Medicine, NIH/NCATS Colorado CTSA Grant Number UL1 TR002535, as well as by the American Epilepsy Society.

Clinical Epilepsy