Bispectral Index Monitoring in Terminally Ill Patients: A Validation Study

Context If regular therapies cannot relieve symptoms sufficiently in the last days of life, continuous palliative sedation may serve to reduce consciousness. Sedation level can be measured with EEG monitoring with the bispectral index (BIS) monitor. Objectives To determine the feasibility and validity of BIS monitoring in terminally ill patients. Methods In this prospective study, BIS registrations were performed in unconscious end-of-life patients admitted to a palliative care center. Validated scores were used to measure level of sedation (Ramsay score), pain (Numeric Rating Scale or Rotterdam Elderly Pain Observations Scale), delirium (Delirium Observation Screening score), and overall comfort (Numeric Rating Scale). Validity and sensitivity to change of BIS values were considered, and the effects of medication and the time till death on BIS values were evaluated in a linear mixed model analysis. Results Fifty-eight patients were included for analysis. BIS monitoring was acceptable to patients, relatives, and medical staff. BIS values were moderately correlated with Ramsay scores (0.46) but were highly variable for deeply sedated patients. BIS values changed significantly before and after a midazolam dose (P < 0.001). Midazolam treatment resulted on average in a statistically significant reduction of the BIS values (−4.5, 95% CI −7.0 to −2.0), whereas morphine and haloperidol did not. Conclusion This is one of the first validation studies in which BIS monitoring in end-of-life patients is described. BIS monitoring is feasible in unconscious terminally ill patients. However, based on our results, the wide range of BIS values in deeply sedated and comfortable patients seems to hamper its use in daily clinical practice.

• View at Publisher
• View at Scopus