ABSTRACT
Purpose: A government funded;interactive cloud storage platform (www.vcreate.tv/neuro) allowing patients and carers to upload video and linked metadata for neurological diagnosis was established during the Covid-19 pandemic. We describe the utility for epilepsy and paroxysmal disorders in 16 centres with the first centre active from 01/05/2020. Method: Users are invited to register and utilise a password and passcode for access. Videos are uploaded with a structured history. The clinician classifies seizure type, syndrome, aetiology or other diagnosis using drop-down menus. Users and clinicians complete online evaluations. Postcode allows linkage to user index of deprivation score. Consents for teaching by the local clinical team and research within a national neurology video research database with research ethics approval are optional. All data, except the video file, transfer to the electronic patient record. Result: To 24/03/2021, 4582 video uploads (4024 paediatric, 558 adult), 1889 patients (1594 paediatric, 295 adult). 400-600 new videos per month. 323 physician and nurse users. Deprivation scores indicate equitable use across socio-economic groups. Paediatric classification: non-epileptic 55%, epileptic (36.5%), unknown (8.5%). Adult: non-epileptic 73.5% (34% dissociative, 41% movement disorders), epileptic 11%, unknown 15.5%. Paediatric seizure types include: focal impaired awareness (19%), generalised tonic clonic (18%), focal clonic (17%), epileptic spasms (13%). Non-epileptic events: tics (13%), normal behaviour (12%), sleep myoclonus (10%) gratification (8%), dissociative (5%). >95% carers ranked the system positively. Clinicians report video prevented face-to-face review in 57%, investigations in 44% and reduced time to diagnosis in 97%. Median time to review video and classify was 5 minutes. Conclusions: Remote care is facilitated, investigations prevented or prioritised, with rapid diagnosis and efficiencies in the patient pathway. A rapidly growing teaching resource and research database for semiology and machine learning diagnostics for paroxysmal disorders has been established. We plan to establish the system in low-income countries without cost.