Diagnosing Infantile Epileptic Spasms: Can Smartphone Videos Replace EEG in Resource-limited Settings?
Christine Shrock1, Margaret Savage2, Lauren Sham2, Jacqueline French3, Orrin Devinsky4, Jo Wilmshurst5, Archana Patel2
1Neurology, Boston Children's Hospital, 2Boston Children's Hospital, 3NYU Comprehensive Epilepsy Ctr, 4NYU Epilepsy Center, 5Pediatric Neurology, Red Cross War Memorial Children’s Hospital
Objective:
To determine the utility of using smartphone videos in diagnosing infantile spasms (IS), specifically the impact of brief IS education and clinical history on accuracy of diagnosis and comfort treating without EEG.
Background:

Infantile spasms (IS) are one of the most concerning seizure types globally. Delayed diagnosis and treatment of IS can significantly impact outcomes. There is an ongoing need to expedite diagnosis, especially in low-resource settings with limited access to electroencephalogram (EEG). The use of smartphone video to aid in diagnosing epileptic seizures has yielded promising results in adult populations but has not been evaluated in IS. If video diagnosis is highly accurate, this could be employed in settings where EEG is not readily available to prompt earlier empiric treatment of IS.

Design/Methods:

Home videos of episodes concerning for IS were solicited in children 3-24 months in two tertiary care referral centers. Of the obtained videos, 9 (6 spasms, 3 non-epileptic) were deemed sufficient quality for use, with diagnosis confirmed by EEG. Participants watched videos before and after IS teaching and then with clinical history and answered questions on diagnosis and management.

Results:

Out of 112 participants, 63% on average gave accurate diagnoses based on video alone, which improved to 68% after IS teaching and 73% after clinical history (p<0.05). Overall confidence in diagnosis improved after teaching and clinical history, as did comfort in treating for seizures without EEG. Correlation between comfort in treating and accuracy increased after teaching and clinical history (from r=0.55 to r=0.87).

Conclusions:

Smartphone videos of abnormal movements are a potentially helpful diagnostic tool for identifying and treating IS early, especially in resource-limited settings where EEG is not readily available. Brief IS teaching and clinical history improve diagnostic accuracy and comfort in treating without EEG.

10.1212/WNL.0000000000204228