Is procedural sedation by trained sedation nurses safe? Does the use of EEG monitoring of depth of sedation improve safety compared to use of another tool, specifically the Ramsay Sedation Scale?
Dr. Tong J. Gan, Department of Anesthesiology, Duke University Medical Center, Durham, NC and colleagues endeavored to answer these questions by using data collected from patients who received sedation by trained sedation nurses under the supervision of a physician over a 15-month period. The results of their study weree published in the July 2014 issue of Anesthesia & Analgesia in the article titled “A Prospective Evaluation of the Incidence of Adverse Events in Nurse-Administered Moderate Sedation Guided by Sedation Scores or Bispectral Index.”
In the pretraining baseline phase, the authors used IV midazolam 1 to 2 mg and fentanyl 50 mcg or hydromorphone 0.2 mg, followed by additional doses of midazolam, fentanyl, or hydromorphone, to maintain a Ramsay Sedation Score of 2 to 3 (cooperative, oriented, and responds to verbal command). Propofol was not used. A BIS sensor was applied but nurses could not see the values. In the training period, the nurses received formalized and comprehensive education on the use of BIS to guide drug administration, pharmacology of the drugs that were commonly used, and methods to rescue patients from oversedation. The intent was to keep BIS values between 70 and 90. In the implementation phase, sedation was conducted as in the pretraining baseline phase, but BIS was used to guide administration. If BIS was more than 90 or sedation was inadequate, bolus doses of midazolam 1 mg were administered, and fentanyl 25 mcg or hydromorphone 0.2 mg was administered to provide analgesia.
The most common procedures undergone by patients in the study included colonoscopy, upper gastrointestinal endoscopy, examination under sedation, endoscopic retrograde cholangiopancreatography, and central venous catheter placement. Data were available on almost 2,000 patients. As expected, the Ramsay Sedation Scale was inversely correlated with BIS value. In both the pretraining and implementation phases, 7% of patients had BIS values of 45-60 during some point of the procedure, and 12 patients reached BIS values < 45. Overall, however, mean BIS values for the entire case were 80. Adverse event numbers were low and are listed in table 2 in the article.
When BIS monitoring was used, there were fewer adverse events, though the actual number of adverse events was low. The differences were statistically significant for desaturation, but not for other adverse events. When compared to the pretraining phase, the number of adverse events during the implementation phase was lower.
The study was small and this report was from a single institution. However, it suggests monitoring brain function may be useful to guide sedation.
Is there something specific to the BIS monitor or might the same be found with other EEG monitors, e.g., entropy? Might other sedation regimens be safer or less safe? Further research is needed.