Anesthesia depth monitoring using alternative placement of entropy sensors: a prospective study.

Abstract

Spectral entropy is based on analysis of variations in electroencephalography and frontal electromyography, and is a safe and reliable method for anesthesia depth monitoring. However, standard frontal positioning of entropy electrodes in patients undergoing cardiac surgery is sometimes challenging. The present study aimed to compare standard entropy sensor placement with an alternative (infraorbital) site. This prospective study included 20 patients who underwent cardiovascular surgery at the authors' center. Monitoring was performed with standard and alternative entropy electrode positions from patient admission to surgery to transfer to the intensive care unit. Data were recorded every 15s; all data were analyzed and compared using Bland-Altman, scatter plot with Pearson correlation coefficient, and sensitivity/specificity analyses. Overall, 20,784 pairs of response entropy (RE) and state entropy (SE) indexes were collected. Bland-Altman analysis revealed a mean difference in RE of 0.37 (95% LOA - 7.09, 7.88) and SE 0.69 (95% LOA - 5.95, 7.31); with 3.46% (720/20,784) RE and 3.40% (706/10,790) SE values lying outside of the limits of agreement. Correlation analysis revealed strong positive correlation in both cases: RE, r = 0.983, p < 0.05; SE, r = 0.984; p < 0.05. Sensitivity/specificity analysis revealed 98.1% sensitivity, 93.3% specificity and 97.1% test efficiency for RE, and 99.2%, 95.1% and 98.5% for SE, respectively. Infraorbital entropy sensor placement in patients undergoing cardiovascular surgery is reliable and effective. The strong positive correlation between the two methods of registration enables alternative entropy measurement when frontal placement is not possible.