Conference Coverage

Outpatient Videos May Help Diagnose Psychogenic Nonepileptic Seizures

Review of patient-generated videos correctly differentiated epilepsy from PNES in 68% of videos evaluated by experts and in 58% of videos assessed by residents.


WASHINGTON, DC—Smartphone videos may help to support the clinical diagnosis of psychogenic nonepileptic seizures (PNES), according to research presented at the 71st Annual Meeting of the American Epilepsy Society. However, smartphone videos do not replace the need for video-EEG monitoring, researchers noted. “Smartphone videos are a complementary addition to medical history and physical examinations in the outpatient epilepsy clinic and can help triage hospital admission for video EEG-monitoring,” said William O. Tatum, DO, Professor of Neurology at the Mayo Clinic College of Medicine in Jacksonville, Florida, and colleagues. “Most patients bringing smart videos to the clinic are ultimately diagnosed with PNES and nearly 70% are women.”

William O. Tatum, DO

Video-EEG monitoring is the standard technique for a definitive diagnosis in patients with suspected seizures. Approximately 20% to 30% of patients admitted to the video-EEG monitoring unit are misdiagnosed with epilepsy, however. In addition, expertise availability, cost, and resource utilization of video-EEG monitoring is limited. Patient-generated videos can potentially address these limitations.

A Multicenter Blinded Trial

Dr. Tatum and colleagues conducted a prospective, multicenter, blinded trial to determine the usefulness of outpatient smartphone videos in epilepsy evaluation. Investigators evaluated 41 consecutive patients with uncontrolled seizures (13 participants were male; mean age was 43.7). Patients were excluded if they were younger than 18, had an incomplete medical history and physical examination, had an atypical event, had an inadequate smartphone video, had an unconfirmed video-EEG monitoring diagnosis, or if they did not consent.

Medical history and physical examinations, smartphone videos, and video-EEG monitoring were performed from July 2014 to November 2017. Treating physicians reached a final clinical diagnosis of epilepsy, PNES, or physiologic non-epileptic events using a degree of certainty (scale: 0–10). Ten epileptologists and eight general neurology residents without a special interest in epilepsy were surveyed for a blinded smartphone video diagnosis.

Researchers shared data via HIPPA-protected data transfer utilizing web-based software. The history and physical exams, smartphone videos, and video-EEG monitoring results were obtained using survey forms and were then compared. Finally, sensitivity, specificity, and positive and negative predictive values were analyzed.

Most Smartphone Videos Showed Nonconvulsive Seizures

Epileptologists performed 310 smartphone video reads and residents performed 230 smartphone video reads. Smartphone videos were reviewed in 2.15 minutes compared with 60 minutes for medical history and physicals and 2.54 days for video-EEG monitoring. Most semiology was convulsive and most epilepsy was nonconvulsive. Physicians made a final diagnosis of PNES in 26 patients, epilepsy in 11 patients, physiologic non-epileptic events in three patients, and a PNES with physiologic non-epileptic events in one patient. Medical history and physical examination predicted a definitive diagnosis by video-EEG monitoring in 31 patients.

The median correct response for a smart phone video was 71.4% for epileptologists and 66.7% for residents. The level of confidence was similar between experts and residents, but those who made a correct diagnosis were slightly more confident. Using a level of confidence of at least five, 78% of epileptologists provided correct identification versus 68% of residents. “This suggests that a gap exists in training relative to viewing semiology for diagnostic implications and supports the ongoing need for education in patients with ’events‘”, said Dr. Tatum and colleagues.

The overall quality of smartphone videos was considered adequate for interpretation in 78% of patients. Inter-subject differences were present mainly based upon technical limitations as opposed to video quality. The primary technical limitation was lack of focus on the area of the interest. Researchers concluded that the “secure uploading, exchange, and analysis of smartphone video data in patients with paroxysmal neurological events is feasible.” No safety concerns or complications of taking smartphone videos were reported.

—Erica Tricarico

Next Article:

Life Expectancy Varies by Epilepsy Type

Related Articles