ABSTRACT
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 an 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-minute Diagnostic Interview for Confusion Assessment Method (3D-CAM) 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 3D-CAM. 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; 3D-CAM: 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 3D-CAM achieved 50%/78%/0.639 (P = 0.04). CONCLUSION: Bispectral index monitor + DSA was similar to 3D-CAM 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.
Subject(s)
Consciousness Monitors , Delirium , Delirium/diagnosis , Electroencephalography , Humans , Sensitivity and Specificity , SoftwareABSTRACT
BACKGROUND: The Accreditation Council for Graduate Medical Education (ACGME) has introduced competency-based assessments (milestones) for resident education. However, the existing milestones for Anesthesiology are not specific to Neuroanesthesiology. The Society for Neuroscience in Anesthesiology & Critical Care (SNACC) commissioned a task force to adapt the ACGME anesthesiology milestones for use in Neuroanesthesiology training, and to provide recommendations for implementing milestones. METHODS: A 7-member expert task force supported by an advisory committee developed the initial milestones by consensus. Written permission was given by the ACGME. The milestones were refined following 3-month pilot use in 14 departments across the United States and inputs from SNACC members. Final milestones were approved by the SNACC Board of Directors. RESULTS: Twelve Neuroanesthesiology-specific milestones in 5 major ACGME domains are recommended; these were identified as most pertinent to this subspecialty rotation. These pertain to patient care (7 milestones), medical knowledge (2 milestones), practice-based learning and improvement (1 milestone), and interpersonal and communication skills (2 milestones). Each milestone was described in detail, with clear outline of expectations at various levels of training. CONCLUSIONS: The SNACC Neuroanesthesiology milestones provide a framework for reviewing resident performance and are expected to facilitate improved use of ACGME milestones during Neuroanesthesiology subspecialty training. The task force recommends that the target should be to accomplish level 4 or higher milestones by the end of residency training. Individual programs should decide the implications of a resident not meeting the expected milestones.
