NEONATAL SOCIETY ABSTRACTS
The predictive value of EEG-Monitoring for subsequent intracerebral complications during ventilation in preterm infants
Presented at the Neonatal Society 2003 Summer Meeting (programme).
The Eppendorf Center for Pediatric Neurology, Hamburg, Germany
Objectives: Only little data exist on bioelectrical activity of the brain in preterm infants during the periods of increased risk of intracerebral impairment during the first days of life.
Study design: In order to evaluate parameters influencing the EEG-activity, polygraphic recordings including EEG, tcpO2/pCO2, SaO2, and blood pressure (RR), in 37 ventilated preterm infants (gestational age (GA) 25 to 34 weeks, χ: 29.6, birthweight 535 to 2640g, χ: 1364) and 7 spontaneously breathing infants (GA 29 to 32 weeks, *:31.1, birthweight 645 to 1850g, χ: 1336) were performed during the first weeks of life. In the obtained 24 hour recordings, mean power (FFT), mean burst power, duration of bursts and interburst intervals, and power of interburst intervals within the discontinuous EEG patterns were calculated by means of computerized quantitative analysis including FFT, Hartley and Hilbert Transformation. The study was performed during 1991 and 1995 at the University of Hamburg, Children's Hospital.
Results: The discontinuous EEG pattern with periodically appearing burstlike EEG activation (theta/beta) followed by a flattening of the EEG is the predominating pattern in preterm infants. With brain development during the first months of life the interburst intervals are filled with EEG activity of higher amplitude until the EEG bursts seem to vanish (n-rem sleep). Such EEG bursts arise from thalamic areas but can also be evoked by arousal activation in brainstem areas (reticular formation) appearing like a pace maker of bioelectrical brain activity.
In contrast to spontaneously breathing infants, a drastic EEG activity suppression, characterised by increased interburst intervals and profound activity suppression during the interburst intervals was found in all ventilated infants on the first 10 days of life. In addition to gestational age, several other factors affected EEG activity significantly: 1) Application of surfactant correlated with an EEG power decrease and interburst-interval prolongation especially during periods of hyperoxygenation. 2) Moreover periods of even milder hyperventilation, 3) sedation and 4) low blood pressure levels correlated with an interburst-interval prolongation and decrease of interval EEG-activity (Wilcoxon Rank Test, p<0.05). After an initial major EEG activity suppression in ventilated infants during the first day of life a high incidence of pathological ultrasound findings was encountered (Mann Whitney Rank Sum Test, p<0.05) already during the following 5 days.
Conclusions: Changes of homeostasis of blood pressure and blood gases during ventilation significantly diminish EEG activity immediately. Suppression of bioelectrical brain activity might be related to a decreased metabolism as well as decreased or impaired autoregulation of cerebral perfusion and thus by an increased risk for structural brain lesions which can be monitored "online" by this new method before brain lesions can be detected by other methods.
Ackowledgements: This study was supported by the Deutsche Forschungsgemeinschaft WU 171 1-1.