Mortality prediction by bispectral electroencephalography among 502 patients: its role in dementia

Abstract

Complications of delirium and dementia increase mortality; however, it is difficult to diagnose delirium accurately, especially among dementia patients. The bispectral electroencephalography (BSEEG) score can detect delirium and predict mortality in elderly patients. We aimed to develop an efficient and reliable BSEEG device for high-throughput screening. We also hypothesized that BSEEG score can predict mortality among dementia patients. A prospective cohort study was conducted between January 2016 to December 2018 to measure BSEEG from elderly patients and correlate with outcomes. A total of 502 elderly (55 years old or older) patients with and without dementia were enrolled. For a replication of the utility of BSEEG, mortalities between BSEEG-positive and BSEEG-negative group were compared. In addition, patients with and without dementia status was added to examine the utility of BSEEG among dementia patients. The mortality within 180 days in the BSEEG-positive group was higher than that of the BSEEG-negative group in both the replication and the total cohorts. Mortality of those in the BSEEG-positive group showed a dose-dependent increase in both cohorts. When the dementia patients showed BSEEG positive, their mortality was significantly higher than those with dementia but who were BSEEG-negative. Mortality within 30 days in the BSEEG-positive group was significantly higher than that of the BSEEG-negative group. The utility of the BSEEG to predict mortality among large sample of 502 elderly patients was shown. The BSEEG score can predict mortality among elderly patients in general, and even among dementia patients, as soon as 30 days.

Evaluation of point-of-care thumb-size bispectral electroencephalography device to quantify delirium severity and predict mortality

“…Those data on detection of delirium, proportionality to delirium severity and association with poor outcomes based on EEG slowing are consistent with our data based on the BSEEG method. 23 Because their results corroborate the results of our previous and present studies, 12–14 their data from traditional multichannel EEG supports the validity of our BSEEG method with the limited electrodes (just two leads in the present report) using our BSEEG algorithm. Moreover, the user-friendly BSEEG method has additional advantages.…”

“…Our novel EEG algorithm was found to detect delirium and reported promising data in general medicine settings, the emergency room, in electroconvulsive therapy patients and in lipopolysaccharide-induced delirium in rodents. [12][13][14][15][16] Although our previously published data using the original BSEEG device has shown its effectiveness in detecting delirium, our previous data did not report the level of correlation between delirium severity and the BSEEG score. Also, the device used had some limitations.…”

“…4(e)). 13,14 Remarkably, subgrouping survival analysis using the s.d.-filtered BSEEG scores indicated that long-term mortality was predicted almost exclusively by BSEEG scores, regardless of the clinical status of delirium (Fig. 4(f)).…”

“…Clinical data, including the Confusion Assessment Method for the Intensive Care Unit (CAM-ICU), 17 the Delirium Rating Scale-Revised-98 (DRS) 18 and the Delirium Observation Screening Score (DOSS), 19 were collected as described in previous papers. [12][13][14] We defined delirium status by a CAM-ICU-positive score, DRS score ≥18, DOSS score ≥3, or clinical documentation of delirium. Additional details are described in the supplementary material.…”

“…All-cause mortality data were obtained from each patient's hospital records and obituary records as previously reported. 13,14,21 Hospital LOS data were collected by reviewing hospital records.…”