Comparison of the Cortical Activity Index and Bispectral Index During General Anesthesia: A Prospective Observational...

Background and Objectives: No processed electroencephalographic index is a gold standard for anesthetic depth. We compared the Cortical Activity Index (CAI) with the bispectral index (BIS) as a pragmatic, widely studied clinical comparator, without assuming that BIS measures the true hypnotic state or that CAI is superior. Materials and Methods: This single-center prospective observational study enrolled 100 adult surgical patients; seven cases with acquisition-device data loss were excluded according to the final research-results report, leaving 93 participants for analysis. Concurrent CAI and BIS values were evaluated across 31 analyzable perioperative stages. The prespecified similarity criterion was a mean patient-level Pearson correlation coefficient of at least 0.85. Results: After protocol-specified last observation carried forward handling, 2876 paired CAI-BIS observations were available. The mean patient-level correlation was 0.778 (SD 0.200), with a median of 0.845 (IQR 0.716-0.913); therefore, the prespecified similarity criterion was not met. The pooled observation-level correlation was 0.750. Both indices showed broadly similar perioperative trajectories, but BIS values were generally higher, and intubation-period spikes occurred more often for BIS than CAI (73.1% vs. 19.4%). Conclusions: Comparison with BIS provided a clinically relevant benchmark for CAI behavior but did not establish the absolute accuracy or superiority of either index. CAI did not meet the prespecified similarity criterion, supporting further validation against independent EEG and clinical endpoints.