Perioperative monitoring in oesophageal atresia and congenital diaphragmatic hernia patients - exploring the impact of diagnosis on neurocardiovascular coupling

Abstract Neonates undergoing major surgical procedures are at risk for neurodevelopmental impairment. Previous research in patients with congenital diaphragmatic hernia (CDH) showed that signal interaction graphs, which combine data on near-infrared spectroscopy (NIRS), amplitude integrated electroencephalography (aEEG) and clinical monitoring, are a promising way of investigating cerebral physiology. Yet, variability in treatment and disease severity, inherent to CDH, complicated analysis. In contrast, neonates with isolated oesophageal atresia (OA) undergo major surgery but are typically hemodynamically stable and pathophysiologically homogeneous. We aimed to evaluate and compare the effects of anaesthesia and surgery on perioperative cerebral physiology in patients with OA and CDH. For this observational cohort study, we included neonates with CDH or OA who underwent surgical treatment between 2018 and 2022. Exclusion criteria were major cardiac or chromosomal anomalies and syndromes associated with major neurodevelopmental impairment. Patients underwent NIRS and aEEG monitoring alongside standard perioperative monitoring. Signal interaction graphs and connectivity measures (neurovascular coupling, the coupling between neuronal activity and cerebral blood flow (NVC), cerebral autoregulation, baroreflex and heart rate passivity) were computed using an established data pipeline. Significantly stronger NVC intraoperatively was found in CDH patients compared to OA patients. A higher median dosage sevoflurane was associated with weaker NVC. Surgical technique (open versus minimally invasive) did not affect any of the connectivity measures. Except for NVC, perioperative connectivity measures do not differ between patients with CDH and OA. Future research should validate these integrative monitoring models in healthy populations and connect them to long-term outcomes, to support development of an interpretable bedside tool for risk stratification and personalized interventions. Trial registration: NL-OMON49066, URL: https://onderzoekmetmensen.nl/en/trial/49066 .