By Padmaja Kandula, MD
Assistant Professor of Neurology and Neuroscience, Comprehensive Epilepsy Center, Weill Cornell Medical College
Dr. Kandula reports no financial relationships relevant to this field of study.
SYNOPSIS: In this study, the investigators show the diagnostic value of video smartphone technology with regard to a seizure diagnosis as well as the ability of smartphone video to help distinguish epileptic seizures from nonepileptic events.
SOURCE: Tatum WO, Hirsch LJ, Gelfand MA, et al. Assessment of the predictive value of outpatient smartphone videos for diagnosis of epileptic seizures. JAMA Neurol 2020;77:593-600.
Semiology, or the study of physical signs, is the cornerstone of seizure diagnosis and epilepsy clinical localization. In recent years, enhancements in smartphone technology have pushed the boundaries of medicine and modernized the study of semiology and diagnostic accuracy of paroxysmal symptoms. The investigators in this study assessed the accuracy of smartphone video in comparison to “gold standard” concurrent video electroencephalographic (EEG) monitoring in making an accurate diagnosis and classification of epilepsy.
Between 2015 and 2018, patients were enrolled in a prospective, multicenter, masked clinical trial across eight Level IV National Association of Epilepsy Centers. Study participants were 18 years of age or older with a suspicion of seizure, recommended for video EEG monitoring by the treating physician, and capable of submitting smartphone video of the target event. Surveys were completed after standard history and physical exam by the treating physician, following smartphone video review (all reviewing physicians), and on reaching a final diagnosis after video EEG monitoring by the treating physician. A choice of epileptic seizure, psychogenic nonepileptic attack, physiologic nonepileptic event, or unknown diagnosis was required of all reviewing physicians after review of the smartphone video clips. Reviewing physicians also submitted a corresponding level of certainty on a scale of 0 to 10 after video smartphone review. Video interpretation and survey completion were performed by 10 epilepsy experts and nine senior neurology residents who were masked to the final video EEG monitoring diagnosis.
The primary aim of the study was to assess the diagnostic accuracy of smartphone video to identify seizures correctly. The secondary aim involved the diagnostic accuracy of identifying psychogenic nonepileptic attacks and the comparison of experts vs. trainees in correctly identifying events with smartphone video. Forty-four eligible patients were identified across multiple centers. All patients completed a full history and physical, smartphone submission of the target event, and final diagnostic inpatient video EEG monitoring. Expert interpretation of smartphone video was accurate in predicting a video EEG-confirmed diagnosis of epileptic seizures 89.1% of the time, with a specificity of 93.3%. Events with a motor component had a higher diagnostic accuracy than events classified as nonconvulsive for epileptic seizures (98.2% vs. 72.4%) and psychogenic nonepileptic attacks (95.7% vs. 71.1%). Trainee or resident responses were less accurate for all metrics involving epileptic seizures and psychogenic nonepileptic attacks, despite greater confidence.
Eleven of the 44 smartphone videos (25.0%) had high diagnostic value, with 100% of reviewing physicians correctly predicting a final diagnosis from the video. All 11 of these videos depicted psychogenic, nonepileptic attacks. When histories and physical examination results were combined with smartphone videos, overall diagnostic accuracy rose from 78.6% to 95.2%. The odds of receiving a correct diagnosis were five times greater using smartphone video alongside patient history and physical exam than with history and physical examination alone.
COMMENTARY
The outcome of this study suggests that adjunctive smartphone video may aid in diagnostic accuracy of all event types. The diagnostic yield of the 20% of patients with unknown or incorrect diagnosis with history and physical exam alone rose to 95% after adjunctive smartphone video EEG review. Smartphone video also added diagnostic value to the existing history and physical evaluation, particularly in those with psychogenic nonepileptic attacks and specifically those with “positive” motor symptoms. Despite the clear advantage of adjunctive smartphone video, diagnostic confidence was consistently higher in the expert vs. trainee group on each studied metric, once again underscoring the need for experience and training for the most accurate epilepsy diagnosis.
However, the pitfall in this study lies in the limited patient sample size and limited semiology of seizure types. In addition, the diagnostic accuracy may not be entirely generalizable to neurology community clinicians at large, since the basis of this study relied on Level IV Center board-certified epileptologists and neurophysiologists for diagnostic accuracy.
Despite these limitations, expert evaluation of smartphone videos can be an effective adjunctive tool to triage and risk stratify those in need of inpatient video EEG monitoring in resource-restricted times or regions. In the era of telemedicine and access to care, smartphone videos also can be an invaluable tool in expert hands, providing immediate semiology-based expert opinion to community physicians and perhaps avoiding a delay in diagnosis and treatment.
