Continuous EEG–SEP monitoring of severely brain injured patients in NICU: methods and feasibility

Abstract

Aims To evaluate the feasibility of a continuous neurophysiologic monitoring (electroencephalography (EEG)–somatosensory evoked potentials (SEPs)) in the neuro-intensive care unit (NICU), taking into account both the technical and medical aspects that are specific of this environment. Methods We used an extension of the recording software that is routinely used in our unit of clinical neurophysiology. It performs cycles of alternate EEG and SEP recordings. Raw traces and trends are simultaneously displayed. Patient head and stimulator box are placed behind the bed and linked to the ICU monitoring terminal through optic fibers. The NICU staff has been trained to note directly clinical events, main artefacts and therapeutic changes. The hospital local area network (LAN) enables remote monitoring survey. Results Continuous EEG (CEEG)–SEP monitoring was performed in 44 patients. Problems of needle detachment were seldomly encountered, thanks to the use of a sterile plastic dressing, which covers needles. We never had infection or skin lesions due to needles or the electrical stimulator. The frequent administration of sedative at high doses prevented us from having a clinically valuable EEG in several cases but SEPs were always monitorable, independently of the level of EEG suppression. The diagnosis of seizures and non-epileptic status was based on raw EEG, while quantitative EEG (QEEG) was used to quantify ictal activity as a guide to treatment. Conclusions EEG and EP waveforms collected in NICU were of comparable quality to routine clinical measurements and contained the same clinical information. A continuous SEP monitoring in a comatose and sedated patient in NICU is not technically more difficult and potentially less useful than in operating room. This monitoring appears to be feasible provided the observance of some requirement regarding setting, electrodes, montages, personnel integration, consulting and software.