Implementing a Digital Neurocritical Care Unit for Neonates in Brazil: A 4-Year Experience

This study aimed to describe neuromonitoring findings and short-term outcomes after the implementation of a digital health strategy comprising continuous, real-time, tele-based video-aEEG/EEG monitoring in a publicly funded NICU in Brazil.Prospective, observational cohort study conducted between July 2017 and June 2021, analyzing neuromonitoring data of high-risk newborns and correlating it with clinical and imaging outcomes.A total of 116 newborns, with a median gestational age of 37 weeks (interquartile range [IQR]: 32^2/7-39^2/7) and a median birth weight of 2,800 g (IQR: 1,472-3,305), were enrolled with more than 8,000 hours of monitoring. The main indication was suspected seizure (<i>n</i> = 49, 42.2%). A total of 43 (37.1%) neonates presented pathological background activity, and sleep-wake cycle (SWC) was absent in 68 (58.6%). Seizures were identified in 36 (31.0%) neonates, predominantly within the first 12 hours of life (<i>n</i> = 14, 38.9%), electrographic-only (<i>n</i> = 29, 80.6%), and repetitive (<i>n</i> = 24, 66.7%). A total of 47 (40.5%) neonates received antiseizure medications, with phenobarbital being the most frequently used (46; 97.9%). Only one patient (2.1%) was discharged receiving antiseizure medication. Cranial ultrasound (cUS) was performed in 94 (81.0%) infants, with abnormal findings in 34 (36.2%) infants. Pathological background activity, absence of SWC, and seizures were significantly associated with severe abnormalities on cUS, and increased risk of death before discharge.The implementation of a digital health strategy incorporating real-time and continuous video-aEEG/EEG monitoring demonstrated potential to improve diagnostic accuracy for electrographic seizures, optimize antiseizure medication stewardship, and inform early neuroprotective interventions. · Brain monitoring improves seizure diagnosis.. · aEEG/EEG supported antiseizure medication discontinuation.. · Abnormal aEEG/EEG findings are associated with poor outcomes.. · Remote aEEG/EEG monitoring is feasible in LMIC..