Effects of hypocapnia on canine spinal, subcortical, and cortical somatosensory-evoked potentials during isoflurane anesthesia.

Although hyperventilation with hypocapnia is frequently used in the management of neurosurgical patients in whom sensory-evoked potentials may be monitored, the effects of hypocapnia on evoked potentials have not been described with precision. In the present experiment, the effects of randomized arterial carbon dioxide tensions of 20, 25, 30, and 35 mm Hg on spinal, subcortical, and cortical somatosensory-evoked potentials (SEPs) were measured in dogs anesthetized with 1.40% isoflurane. Other variables known to affect the SEP (temperature, blood pressure, and arterial oxygen tension) were stable throughout the experiment. Hypocapnia caused reductions in the latencies of the early peaks of the spinal and subcortical SEPs. These differences were small, consisting of a 2% shortening of latency at 20 mm Hg carbon dioxide tension when compared with 35 mm Hg. No changes were detected in the later subcortical and cortical latencies. SEP amplitudes were also unchanged. These results in a controlled animal study corroborate the direction and magnitude of changes due to hypocapnia observed by other investigators in surgical patients. The magnitude of the changes indicates that SEP monitoring sensitivity is not compromised by clinically useful levels of induced hypocapnia during isoflurane anesthesia. Because hypocapnia may produce small SEP changes, baseline recordings should be acquired prior to initiation of hyperventilation. It is not warranted, however, to impute a severe deterioration of the SEP to hypocapnia alone, and causes must be sought elsewhere in a patient's status and management.