Abstracts

Rationale: Introduction: Amplitude integrated EEG has become a commonly-used bedside monitoring tool in many neonatal intensive care units (NICUs). aEEG has been used to predict neurodevelopmental outcome, diagnose clinical and subclinical seizures and evaluate cerebral recovery after hypoxic-ischemic insults. However, the impact of aEEG use on neonatal neurological care remains unclear. Objective: To investigate the relationship between introduction of aEEG at a single tertiary care hospital NICU and the: 1. Frequency and timing of conventional EEG recordings. 2. Frequency of neurology consultation. 3. Frequency of diagnosis of neonatal seizures. Methods: This was a retrospective observational study conducted at The Hospital for Sick Children, Toronto, Canada. The study population included all patients admitted to the NICU over a six-year period from July 2001 to June 2007. The sample was divided into two epochs: * Epoch 1 (July 2001 to June 2004), during which only conventional EEG was available. * Epoch 2 (July 2004 to June 2007), during which both aEEG and conventional EEG were available. The frequency and timing of conventional EEG and aEEG recordings, neurology consultation and diagnosis of neonatal seizures were obtained. Data was abstracted from a Health Records database, NICU aEEG database and Neurophysiology EEG database. Statistical comparisons were made using Fisher’s exact test. Results: A total of 2219 neonates were admitted during Epoch 1, and 2202 during Epoch 2. The rates of neurological diagnoses, length of stay and mortality did not differ between Epochs 1 and 2. The percentage of admitted neonates receiving conventional EEG recordings did not differ between Epoch 1 (18.0%) and Epoch 2 (18.5%). The percentage of neonates receiving aEEG recordings during Epoch 2 (18.4%) approached that of conventional EEGs. The percentage of conventional EEG recordings initiated after-hours did not change following introduction of aEEG (Epoch 1: 7%, Epoch 2: 11%; p=0.08). However, far more aEEGs (74%) were initiated after-hours. Following the introduction of aEEG, there was a significant reduction in the rate of repeated EEGs from 30% in Epoch 1 to 20% in Epoch 2 (p=0.003). In addition, there was a 5% decline in the percentage of newborns diagnosed with seizures (Epoch 1: 36%, Epoch 2: 31%; p=0.047). However, the number of Neurology consultations to the NICU did not change (Epoch 1: 14.7%, Epoch 2: 12.4%; p=0.20). Conclusions: aEEG monitoring has become common practice in this NICU, with usage now approaching that of conventional EEG. Whereas most conventional EEGs were initiated during the daytime, most aEEGs were initiated after-hours, reflecting the 24/7 in-house availability of required personnel. The introduction of aEEG was associated with a significant reduction in the rate of repeated EEGs and, unexpectedly, a small but significant reduction in the percentage of newborns diagnosed with seizures. This was a population level study; the impact of aEEG on individual patient care and outcomes remains to be determined.