Etomidate for Procedural Sedation in the Emergency Department

Samuel M. Keim, MD, Brian L. Erstad, PharmD, FCCM, John C. Sakles, MD, Virgil Davis, MD


Pharmacotherapy. 2002;22(5) 

In This Article


Study Design

We conducted a 2-year retrospective analysis of all patients who received etomidate for sedation in the Arizona Health Sciences Center emergency department without concomitant intubation from May 1997 to May 1999. Because the study was retrospective, it was approved by the institutional review board with waiver of consent. Patients charged for etomidate were identified from a billing database. Initially, this database was compared with a separate computerized database to exclude patients who underwent planned rapid-sequence intubation. The remaining patients constituted the study population, and their medical records were reviewed. Data collection was accomplished using a standardized data collection form. With a timed flow sheet method, data were obtained primarily from standard procedural sedation and emergency department records of vital signs, level of consciousness, drugs administered, and adverse events. These data were recorded concurrent with the sedation by a registered nurse. Nurse and physician charts also were reviewed. The remainder of the medical records was reviewed to ascertain whether any adverse events led to significant morbidity. Patients were observed in the emergency department until recovery from sedation was complete.

Setting and Population

All patients were seen in the emergency department of an urban university teaching hospital with an annual census of approximately 55,000. Patients with conditions that required painful emergency procedures were considered potential candidates for sedation with etomidate. Selection criteria for administration of etomidate versus other agents were specified only as "for moderate to deep sedation needs." We did not attempt to differentiate which types of patients were selected. All sedation procedures were conducted in a monitored room with an available airway cart in the attendance of the treating physician and a nurse certified in advanced cardiac life support. Monitoring of cardiac and pulse oximetry was continuous during all sedation procedures.

An emergency physician who was competent in both airway management and resuscitation administered the etomidate. This physician was responsible solely for patient sedation and analgesia. The recommended protocol was to administer etomidate as a single intravenous bolus of 0.1 mg/kg over 1 minute to induce deep sedation. Administration of concomitant analgesics and other sedatives (e.g., midazolam) was neither recommended nor discouraged. The final choice and dosages of adjunctive drugs were at the discretion of the sedating physician.

Study Protocol

Data were collected regarding age, sex, initial and total etomidate doses, adjunctive drugs (those given within 15 minutes of etomidate), time from initial dose to recovery (defined as being responsive to verbal commands), emergency department observation time, and adverse events. Adverse events, which were classified as either present or absent, were apnea (> 30 sec), desaturation (< 90% oxygen saturation by pulse oximetry with or without oxygen supplementation with nasal cannula), bag valve mask ventilation or intubation, emesis, aspiration, hypotension, myoclonus, inadequate sedation, unsuccessful procedure, and cardiac arrest.