The problem of artifacts in using the amplitude-integrated electroencephalogram (AIE) to assess cortical function in premature infants in the NICU were studied at Weill Cornell Medical College, New York, NY. A pair of standard EEG electrodes were attached to scalp frontotemporal areas of 10 infants. Impedance was maintained at <10 kohms. Continuous AIE recordings were performed for at least 60 min repeatedly in the first month. Artifacts were identified as large amplitude difference between jagged wave peaks and troughs. When the AIE tracing spikes upward in amplitude, the accompanying raw EEG during these segments was classified as artifact. Of 1683 segments of 48 recordings analyzed, 31% were normal brain waves, 60% were artifacts, and 8% indeterminate. No clinical or electrographic seizures were noted. Artifact related to muscle activity and electrode placement detracts from the value of the AIE in assessing cortical function in premature infants. 
COMMENT. Amplitude-integrated EEG (aEEG) is a sensitive monitor of background cerebral activity in the early prediction of outcome after perinatal asphyxia in term infants. Abnormal aEEG is used in selection of patients for hypothermia. The above study introduces a note of caution in the use of the aEEG as an indicator of cortical function in premature infants in the NICU. The concurrent use of the conventional EEG and careful attention to electrode placement are suggested.