Bilateral Bispectral Index Monitoring Performance in the Detection of Seizures in Nonanesthetized Epileptic Patients: An Observational Study.

Background The aim of this observational study was to determine whether bilateral bispectral index (BIS) monitoring can detect seizures in epileptic patients. Methods Four-channel frontal BIS monitoring and standard 40-channel electroencephalography monitoring were conducted in epileptic patients undergoing evaluation for epilepsy surgery. The BIS numerical value, signal quality index, electromyography, suppression ratio, and color density spectral array were continuously recorded. In patients with electroencephalography-confirmed seizures, the mean value and trend (slope of linear regression) of bilateral BIS monitor parameters were analyzed from 1 minute before to 1 minute after seizure onset. Results Of 48 patients included in the study, 21 (43.8%) had at least 1 seizure. BIS numerical value was not able to detect focal or focal to bilateral tonic-clonic seizures. Considering all seizures, the only significant differences between recordings 1 minute before and 1 minute after seizure onset were a decrease in the signal quality index slope from 1 hemisphere (0.039±0.297 vs. -0.085±0.321, respectively; P=0.029) and in the mean signal quality index recorded from both hemispheres (left hemisphere: 65.775±30.599 vs. 61.032±26.285; P=0.016 and right hemisphere: 63.244±31.985 vs. 59.837±27.360; 0.029); these differences were not maintained after Hochberg adjustment for multiple comparisons. In seizures occurring during sleep, there was a change in the electromyography slope of 1 hemisphere before and after seizure onset (-0.141±0.176 vs. 0.162±0.140, respectively; P=0.038). There were variable responses in BIS parameters in the 3 patients who developed focal nonconvulsive seizure clusters. Conclusion Bilateral BIS monitoring was not able to detect the occurrence of seizures in epileptic patients.