Aktuelle Folge
Automated EEG Background Analysis and 2-Year Outcomes in Neonatal Hypoxic-Ischemic Encephalopathy Citation:
Cornet MC, Numis AL, Wusthoff CJ, Bernardo D, Mietzsch U, Thomas C, et al. Automated EEG Background Analysis and 2-Year Outcomes in Neonatal Hypoxic-Ischemic Encephalopathy. JAMA Netw Open. 2025;8(12):e2548321. doi:10.1001/jamanetworkopen.2025
Länge: 11:57
Citation:
Cornet MC, Numis AL, Wusthoff CJ, Bernardo D, Mietzsch U, Thomas C, et al. Automated EEG Background Analysis and 2-Year Outcomes in Neonatal Hypoxic-Ischemic Encephalopathy. JAMA Netw Open. 2025;8(12):e2548321. doi:10.1001/jamanetworkopen.2025.48321.
Study at a glance:
- Design and setting: Multicenter prognostic study (secondary analysis of the HEAL RCT) in 8 US academic NICUs; generalized linear mixed models with site-level random effects and leave-one-site-out cross-validation.
- Participants: 203 cooled infants ≥36 weeks with moderate–severe HIE, continuous EEG started 24 h of life and lasting ≥24 h, BSN available, and 2-year follow-up; 49/203 (24%) had death or severe neurodevelopmental impairment (NDI).
- Predictors and comparators: Automated Brain State of the Newborn (BSN) scores (median at onset and at 24, 48, 72 h and overall), simple clinical variables (Apgar at 5 min, HIE severity), and expert visual EEG background classification (normal, excessively discontinuous, severely abnormal) in a 127-infant subset.
- Primary outcome: Composite of death or severe NDI at ~2 years (Bayley-III cognitive score 70 and/or moderate–severe motor impairment, or death); secondary 5-level ordinal outcome (no/mild/moderate/severe NDI or death).
- Prognostic performance: Median BSN at onset alone discriminated death/severe NDI with AUROC 0.85 (95% CI 0.79–0.92); BSN at all epochs alone AUROC 0.92 (0.87–0.97). Clinical variables alone had AUROC 0.79 (0.70–0.87); adding overall BSN improved AUROC to 0.90 (0.84–0.97), and adding BSN at all epochs to clinical variables reached AUROC 0.93 (0.88–0.98). Expert EEG background alone achieved AUROC 0.90 (0.81–0.98).
- Clinical signal: Very low early BSN was strongly associated with poor outcome (median BSN 25 at onset: 17/18 [94%] death/severe NDI; BSN ≥75 at onset: 3/77 [4%] death/severe NDI).
- Limitations and risk of bias: PROBAST overall high risk of bias—performance reported only from internal cross-validation (no calibration metrics or external validation), BSN requires specialized continuous EEG and analysis pipeline in high-resource NICUs, so transportability and real-world impact remain uncertain.
GUID: 8cf0cf72-6d3c-4144-a9f3-45ad13379af1
Erscheinungsdatum: 5.1.2026, 00:00:00
