RESUMO
Visual evoked potentials and electroretinograms were elicited by light-emitting diode stimulation and recorded simultaneously, with cephalic and noncephalic references, in 30 normal subjects and in 30 brain-dead patients. A characteristic pattern was found in the group of patients: when a cephalic reference was used for both visual evoked potentials and the electroretinogram, the a- and b-waves of the electroretinogram were recognized in all patients, and visual evoked responses consisted of waves with inverse polarity, similar morphologic characteristics, the same latency, and less amplitude than those of the electroretinogram. When a noncephalic derivation was chosen for the electroretinogram and visual evoked potentials, electroretinogram did not change in either morphologic features or latency, whereas the visual evoked potential channel showed no response. Only in two cases was it possible to record waves in the visual evoked potential lead with a noncephalic reference, showing a spread of the electroretinogram to the occipital area, with a considerably reduced amplitude. These results suggest that, although contamination of visual evoked potential records by the spread of the electroretinogram to the occipital area could occur, it is easy to confirm the absence of a true cortical visual response in brain-dead patients by means of a noncephalic reference. This pattern clearly confirms that in the visual pathways of brain-dead patients, electrical activity is confined to the retina.