Seizure Detection Algorithms in Critically Ill Children: A Comparative Evaluation.

OBJECTIVES: To evaluate the performance of commercially available seizure detection algorithms in critically ill children. DESIGN: Diagnostic accuracy comparison between commercially available seizure detection algorithms referenced to electroencephalography experts using quantitative electroencephalography trends. SETTING: Multispecialty quaternary children's hospital in Canada. SUBJECTS: Critically ill children undergoing electroencephalography monitoring. INTERVENTIONS: Continuous raw electroencephalography recordings (n = 19) were analyzed by a neurophysiologist to identify seizures. Those recordings were then converted to quantitative electroencephalography displays (amplitude-integrated electroencephalography and color density spectral array) and evaluated by six independent electroencephalography experts to determine the sensitivity and specificity of the amplitude-integrated electroencephalography and color density spectral array displays for seizure identification in comparison to expert interpretation of raw electroencephalography data. Those evaluations were then compared with four commercial seizure detection algorithms: ICTA-S (Stellate Harmonie Version 7; Natus Medical, San Carlos, CA), NB (Stellate Harmonie Version 7; Natus Medical), Persyst 11 (Persyst Development, Prescott, AZ), and Persyst 13 (Persyst Development) to determine sensitivity and specificity in comparison to amplitude-integrated electroencephalography and color density spectral array. MEASUREMENTS AND MAIN RESULTS: Of the 379 seizures identified on raw electroencephalography, ICTA-S detected 36.9%, NB detected 92.3%, Persyst 11 detected 75.9%, and Persyst 13 detected 74.4%, whereas electroencephalography experts identified 76.5% of seizures using color density spectral array and 73.7% using amplitude-integrated electroencephalography. Daily false-positive rates averaged across all recordings were 4.7 with ICTA-S, 126.3 with NB, 5.1 with Persyst 11, 15.5 with Persyst 13, 1.7 with color density spectral array, and 1.5 with amplitude-integrated electroencephalography. Both Persyst 11 and Persyst 13 had sensitivity comparable to that of electroencephalography experts using amplitude-integrated electroencephalography and color density spectral array. Although Persyst 13 displayed the highest sensitivity for seizure count and seizure burden detected, Persyst 11 exhibited the best trade-off between sensitivity and false-positive rate among all seizure detection algorithms. CONCLUSIONS: Some commercially available seizure detection algorithms demonstrate performance for seizure detection that is comparable to that of electroencephalography experts using quantitative electroencephalography displays. These algorithms may have utility as early warning systems that prompt review of quantitative electroencephalography or raw electroencephalography tracings, potentially leading to more timely seizure identification in critically ill patients.

Seizure Identification by Critical Care Providers Using Quantitative Electroencephalography.

OBJECTIVES: To compare the performance of critical care providers with that of electroencephalography experts in identifying seizures using quantitative electroencephalography display tools. DESIGN: Diagnostic accuracy comparison among healthcare provider groups. SETTING: Multispecialty quaternary children's hospital in Canada. SUBJECTS: ICU fellows, ICU nurses, neurophysiologists, and electroencephalography technologists. INTERVENTION: Two-hour standardized one-on-one training, followed by a supervised individual review of 27 continuous electroencephalography recordings with the task of identifying individual seizures on eight-channel amplitude-integrated electroencephalography and color density spectral array displays. MEASUREMENTS AND MAIN RESULTS: Each participant reviewed 27 continuous electroencephalograms comprising 487 hours of recording containing a total of 553 seizures. Performance for seizure identification was compared among groups using a nested model analysis with adjustment for interparticipant variability within groups and collinearity among recordings. Using amplitude-integrated electroencephalography, sensitivity for seizure identification was comparable among ICU fellows (83.8%), ICU nurses (73.1%), and neurophysiologists (81.5%) but lower among electroencephalographic technologists (66.7%) (p = 0.003). Using color density spectral array, sensitivity was comparable among ICU fellows (82.4%), ICU nurses (88.2%), neurophysiologists (83.3%), and electroencephalographic technologists (73.3%) (p = 0.09). Daily false-positive rates were also comparable among ICU fellows (2.8 for amplitude-integrated electroencephalography, 7.7 for color density spectral array), ICU nurses (4.2, 7.1), neurophysiologists (1.2, 1.5), and electroencephalographic technologists (0, 0) (p = 0.41 for amplitude-integrated electroencephalography; p = 0.13 for color density spectral array). However, performance varied greatly across individual electroencephalogram recordings. Professional background generally played a greater role in determining performance than individual skill or electroencephalogram recording characteristics. CONCLUSIONS: Following standardized training, critical care providers and electroencephalography experts displayed similar performance for identifying individual seizures using both amplitude-integrated electroencephalography and color density spectral array displays. Although these quantitative electroencephalographic trends show promise as a tool for bedside seizure screening by critical care providers, these findings require confirmation in a real-world ICU environment and in daily clinical use.

Early somatosensory evoked potential grades in comatose traumatic brain injury patients predict cognitive and functional outcome.

OBJECTIVES: To relate early somatosensory evoked potential grades from comatose traumatic brain injury patients to neuropsychological and functional outcome 1 yr later; to determine the day (within the first week after traumatic brain injury) that somatosensory evoked potential grade best correlates with outcome; to determine whether somatosensory evoked potential grade improvement in the first week after traumatic brain injury is associated with improved outcome. DESIGN: Prospective cohort study. SETTING: Critical care unit at a university hospital. PATIENTS: Median nerve somatosensory evoked potentials were obtained from 81 comatose patients with traumatic brain injury. Somatosensory evoked potential grades were calculated from results obtained on days 1, 3, and 7 after traumatic brain injury. Glasgow Outcome Scale, Barthel Index, Rivermead Head Injury Follow-up Questionnaire, General Health Questionnaire, Stroop Color-Word Test, Paced Auditory Serial Addition Task, and Symbol-Digit Modalities Test scores were obtained 1 yr after injury. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Somatosensory evoked potential grade on days 1, 3, and 7 related significantly with Glasgow Outcome Scale and Barthel scores (day 3 better than day 1) but did not relate with Rivermead Head Injury Follow-up Questionnaire or General Health Questionnaire scores. Day 3 and day 7 somatosensory evoked potential grades related significantly with Stroop scores. Day 3 somatosensory evoked potential grades related significantly with Symbol-Digit Modalities Test scores. Patients with bilaterally present but abnormal somatosensory evoked potentials, whose somatosensory evoked potential grade improved between days 1 and 3, had marginally better functional outcome than those without somatosensory evoked potential grade improvement. CONCLUSIONS: Day 3 somatosensory evoked potential grade related to information-processing speed, working memory, and the ability to attend to tasks 1 yr after traumatic brain injury. Day 3 somatosensory evoked potential grade had the strongest relationship with functional outcome. Somatosensory evoked potential grades were not related to emotional well-being.

Optimal intraoperative somatosensory evoked potential stimulus intensity can be determined by nerve action potential amplitude.

PURPOSE: Muscle twitch threshold has been used to determine optimal stimulus intensity for somatosensory evoked potentials but neuromuscular blockade precludes the use of muscle twitch during surgery. Accordingly, nerve action potential (NAP) amplitude was investigated as a surrogate to muscle twitch. METHODS: The ulnar and tibial nerves were stimulated at the wrist and ankle, respectively, in 27 patients undergoing spine and brain surgery. After neuromuscular blockade was gone, the stimulus intensity for just maximal NAP amplitude recorded from Erb's point and the popliteal fossa was compared with the stimulus intensity for hypothenar and plantar foot muscle twitch threshold (times two), respectively (Wilcoxon matched pairs test). RESULTS: There was no significant difference between stimulus intensity for just maximal Erb's point and popliteal fossa NAP amplitude when compared with stimulus intensity for hypothenar and plantar foot twitch threshold (times two), respectively. Eight patients required more than twitch intensity (times two) to obtain maximum NAP. CONCLUSIONS: The NAP amplitude may be used to determine optimal somatosensory evoked potential stimulus intensity when muscle twitch is not visible. This method should improve the success of intraoperative somatosensory evoked potential monitoring and decrease erroneous interpretation.