Previous Abstract | Next Abstract
Printable Version
October 18, 2008
10:00 AM - 11:30 AM
Room Room 230D
78% of Patients Who Had Nurse Delivered Sedation Have BIS Levels Consistent with General Anesthesia
Tong J. Gan, M.B.B.S., FRCA, Meryt Hanna, M.P.H., MB.B.Ch, Anne Madden, MSN, Minyi Lu, M.D., Ph.D., Holly Muir, M.D.
Anesthesiology, Duke University Medical Center, Durham, NC

Moderate sedation is commonly used for patients undergoing interventional procedures outside the operating room environment. A previous smaller study found that patients undergoing colonoscopy under propofol sedation frequently were taken to levels of general anesthesia and were at risk for respiratory depression and airway obstruction1. This prospective quality improvement project investigates the depth of sedation and safety profile of patients undergoing procedural sedation in various off-site hospital locations.


Following IRB approval and informed patient consent, adult patients undergoing procedural sedation outside the OR were monitored with routine monitors including ECG, pulse oximetry and non-invasive blood pressure. Nurses were instructed to provide sedation according to the normal clinical practice, guided by Ramsay sedation scores. All patients also had bispectral index (BIS) sensor (Aspect Medical System, Norwood, MA) attached before the commencement of the sedation. In order to maintain the blind, the face of the BIS monitor was covered. Midazolam was used as the sedative and fentanyl was used as the analgesic. Additional drugs were administered when deemed clinically necessary. The drugs used, doses and time of administration were recorded on a computer scanable form. BIS data were downloaded. Descriptive statistics and correlation were used to analyze the data. A p value <0.05 was considered significant.


A total of 595 patients were enrolled to date. Data presented represent mean ± SD or percentages. The age was 58 ± 13 and M:F ratio was 46%:53%. Procedure time was 27 ± 35 min, comprising of colonoscopy (42%), upper GI endoscopy (15%), bronchoscopy (12%), examination under sedation (10%), radiology (10%) and others (11%). The dose of midazolam and fentanyl administered were 1.5 ± 3.6 mg and 43 ± 21 mcg, respectively. The average BIS values during the procedure were 49 ± 17. BIS values were ≤ 45 in 52% of patients and ≥ 45 and ≤ 60 in 23% of patients. In 78% of patient, BIS was < 60 for greater than 5 min and this finding is consistent across all sedation settings. Adverse events occurred in 6% of the patients which included episodes of oxygen desaturation, pain, hyper and hypotension, restlessness, difficulty to arouse and tachycardia, among the most common. Lower BIS values were associated with incidence of oxygen desaturation.


Nurses delivering sedation outside the OR frequently administer relatively deep sedation which could be associated with sedation related adverse events. It remains to be tested if the routine use of BIS monitoring could reduce sedation related adverse events and improve outcomes in these populations.


Ramsay M et al. Anesthesiology 2007;107:A1014.

Anesthesiology 2008; 109 A192