Real-time detection of neonatal seizures improves with on demand EEG interpretation

•Concurrent Continuous conventional EEG (cEEG) and amplitude integrated EEG (aEEG) monitoring improved correct seizure identification in patients at risk for seizures.•Concurrent cEEG and aEEG decreased seizure over-diagnosis, and prevented medication overuse in patients at risk for seizures.•Concurrent cEEG and aEEG monitoring should be further evaluated for optimizing patient care. Under- and overtreatment of seizures may result in adverse outcomes; thus, early, reliable seizure identification is crucial. Continuous conventional ElectroEncephaloGram (cEEG) is the gold standard, but amplitude integrated EEG (aEEG) is most widely-used in the Neonatal Intensive Care Unit (NICU). We aimed to determine whether a novel pathway combining cEEG and aEEG for seizure detection would improve real-time seizure identification. A single-center, prospective quality improvement project cohort. Patients at-risk of seizures were monitored by cEEG and aEEG concurrently, with the option for the neonatologist to contact a neurologist for real-time cEEG interpretation. The primary outcome was correct identification of seizures by the new combined pathway compared to aEEG alone. Seizure detection using aEEG had a sensitivity of 46.2%; specificity of 93.3%; PPV of 66.7%; and NPV of 85.7%. Utilizing the new on-demand, real-time cEEG interpretation by a neurologist, correct seizure identification increased by 27% (95%CI: 10–56%). Over-diagnosis of seizures was avoided in 33.3% (95% CI: 12.1–64.6%) and misuse of anti-seizure medication was prevented. Combining aEEG with on-demand cEEG interpretation improved accurate seizure detection in a high-risk NICU population, with the potential to avoid over-treatment. We describe a novel combined EEG monitoring pathway to improve seizure detection, and prevent unnecessary treatment.