Automatic versus visual EEG sleep staging in preadolescent children

Ambulant sleep polygrams were obtained from 14 normal subjects (9 boys and 5 girls) and from 3 boys with attention deficit disorder. The children were aged 8-12 years. Two consecutive nights were recorded with an eight-channel electroencephalographic (EEG) tape cassette recorder. The results were analyzed automatically by the Oxford Medilog 9000 Sleep Stager and by visual scoring from the Medilog Display Unit. Twenty-seven nights were analyzed; 7 nights were excluded because of electrode problems or other technical failures. The main sleep stage shifts and the length of sleep cycles as measured from the hypnogram of the automatic printout agreed well with corresponding values from our visual scoring. In the automatic scoring, rapid eye movement (REM) sleep time was shorter and slow wave sleep was longer than in the visual rating. This can be explained partly by specific properties of the EEG in this age group. The Oxford Medilog 9000 sleep stager can be used to survey sleep quality, but the results must be carefully checked visually. In cases of EEG pathology or sleep abnormalities in childhood it is doubtful if any time or labor is saved by using automatic scoring.