BACKGROUND Delirium in hospitalized patients often goes undetected. Cerebral state monitors, which measure limited-channel electroencephalography, have shown potential for improving delirium detection. OBJECTIVE The aim of this study was to compare a FDA-approved cerebral state monitor, bispectral index monitoring with density spectral array (DSA), for delirium identification with clinical screening methods. METHODS Hospitalized patients receiving psychiatric consultation were assessed for delirium using the 3-dimensional-Confusion Assessment Method and underwent bispectral index monitor + DSA monitoring. Visual inspection of frequency band power of the DSA was performed by 2 trained independent raters. Average hue values were calculated for each frequency band using image analysis software as the device did not allow for extraction of raw electroencephalography data. Delirious versus nondelirious group averages, sensitivity, specificity, and area under the curve were calculated for significant DSA variables and the 3-dimensional-Confusion Assessment Method. RESULTS In an initial cohort of 43 patients, visual ratings of the DSA were not associated with delirium (P > 0.1). In a larger cohort of 123 subjects, multiple band hue ratios were associated with delirium, although none survived correction for multiple comparisons. In a subgroup of 74 non-neurological patients, low theta/low delta ratio was significantly associated with delirium (P = 0.001) (sensitivity/specificity/area under the curve: 83%/70%/0.757; 3-dimensional-Confusion Assessment Method: 67%/77%/0.717; paired-sample area under the curve difference: -0.040, P = 0.68). In 21 patients with dementia, low theta power demonstrated significantly greater sensitivity/specificity/area under the curve of 83%/78%/0.824, whereas 3-dimensional-Confusion Assessment Method achieved 50%/78%/0.639 (P = 0.04). CONCLUSION Bispectral index monitor + DSA was similar to 3-dimensional-Confusion Assessment Method for detecting delirium in hospitalized patients with and without neurological disorders, and was significantly more accurate in patients with dementia. More studies are needed to validate the use of cerebral state monitors for quantitative delirium detection.

中文翻译：

---

用于谵妄检测的密度谱阵列双谱指数监测

背景 住院患者的谵妄常常未被发现。测量有限通道脑电图的脑状态监测器已显示出改善谵妄检测的潜力。目的 本研究的目的是比较 FDA 批准的脑状态监测器、双谱指数监测与密度谱阵列 (DSA)，用于谵妄识别与临床筛查方法。方法 接受精神科会诊的住院患者采用3维混乱评估法评估谵妄，并接受双频指数监测+DSA监测。DSA 频带功率的目视检查由 2 名训练有素的独立评估员进行。使用图像分析软件计算每个频带的平均色调值，因为该设备不允许提取原始脑电图数据。计算显着 DSA 变量和 3 维混淆评估方法的谵妄与非谵妄组平均值、敏感性、特异性和曲线下面积。结果 在 43 名患者的初始队列中，DSA 的视觉评级与谵妄无关（P > 0.1）。在 123 名受试者的更大队列中，多带色调比率与谵妄相关，尽管没有人在多重比较校正后幸存下来。在 74 名非神经系统患者的亚组中，低 theta/低 delta 比率与谵妄显着相关（P = 0.001）（敏感性/特异性/曲线下面积：83%/70%/0.757；3维混淆评估法：67%/77%/0.717；曲线差下的配对样本面积：-0.040，P = 0.68）。在 21 名痴呆患者中，低 theta 功效显示出显着更高的敏感性/特异性/曲线下面积 83%/78%/0.824，而 3 维混淆评估方法达到 50%/78%/0.639 (P = 0.04) . 结论双频指数监测器+DSA与3维混乱评估法在有和无神经系统疾病住院患者中检测谵妄相似，对痴呆患者更准确。需要更多的研究来验证使用脑状态监测器进行定量谵妄检测。低 theta 功效显示出显着更高的敏感性/特异性/曲线下面积 83%/78%/0.824，而 3 维混淆评估方法达到 50%/78%/0.639 (P = 0.04)。结论双频指数监测器+DSA与3维混乱评估法在有和无神经系统疾病住院患者中检测谵妄相似，对痴呆患者更准确。需要更多的研究来验证使用脑状态监测器进行定量谵妄检测。低 theta 功效显示出显着更高的敏感性/特异性/曲线下面积 83%/78%/0.824，而 3 维混淆评估方法达到 50%/78%/0.639 (P = 0.04)。结论双频指数监测器+DSA与3维混乱评估法在有和无神经系统疾病住院患者中检测谵妄相似，对痴呆患者更准确。需要更多的研究来验证使用脑状态监测器进行定量谵妄检测。结论双频指数监测器+DSA与3维混乱评估法在有和无神经系统疾病住院患者中检测谵妄相似，对痴呆患者更准确。需要更多的研究来验证使用脑状态监测器进行定量谵妄检测。结论双频指数监测器+DSA与3维混乱评估法在有和无神经系统疾病住院患者中检测谵妄相似，对痴呆患者更准确。需要更多的研究来验证使用脑状态监测器进行定量谵妄检测。