Combined Ketamine/Propofol for Emergency Department Procedural Sedation

Introduction

Numerous drugs and combinations of drugs are used for procedural sedation and analgesia (PSA) in Emergency Departments, including propofol, ketamine, benzodiazepines, narcotics, barbiturates, and others, but propofol has gained popularity despite its potential to cause cardiac and respiratory depression. Obviously the optimal agent or combination of agents has not been identified. There are reasons to believe that a combination of ketamine and propofol may have advantages over other agents/combinations. These include better hemodynamic stability at equal depth of anesthesia with a combination of ketamine/propofol than with propofol alone, less respiratory depression with the combination in comparison to propofol alone, and preservation of respiratory drive with the combination. There is one study of ketamine/propofol in Emergency Department (ED) procedural sedation which demonstrated the safety and effectiveness of the combination, but did not compare it to any other agents or combinations. We designed a randomized, placebo controlled study to compare propofol to propofol and ketamine for adequacy of sedation and respiratory depression in Emergency Department procedural sedation and analgesia. Our hypothesis was that the combination of propofol/ketamine would produce better sedation and/or less respiratory depression than propofol alone.

Methods

Study design

We conducted a randomized, prospective, double-blinded study of all patients receiving procedural sedation. From May 2007 until March 2009 in the ED of a 274 bed university teaching hospital. The study was approved by the University of Missouri's Institutional Review Board and informed consent was obtained from all participants.

Study setting and population

All patients requiring PSA in the ED were viewed as potential subjects unless they were pregnant, less than 1 year of age, history of prior adverse reaction to anesthesia, underlying cardiac or pulmonary disease, hepatic dysfunction, porphyria, psychiatric illness, allergy to eggs/soybeans, increased intracranial or intraocular pressure, abnormal airway pathology or an American Society of Anesthesiologists (ASA) score of 3 or greater. The attending ED physician would then approach the patient to enroll them in the study. If the patient accepted they were randomized by the hospital pharmacy.

Study protocol

An ED attending physician was dedicated to PSA throughout the procedure. Patients had EKG, blood pressure, respiratory rate, pulse oximetry, and end-tidal carbon dioxide (PetCO2) monitored, had IV access obtained and were placed on nasal cannula oxygen supplementation. All patients received pre-procedure analgesia with 0.5 to 1.5 mcg/Kg of fentanyl and all patients had reflective sunglasses placed so as to obscure eye movements from the staff.

Subjects were randomized by the pharmacy in blocks of ten. Consecutively numbered pre-filled 3cc syringes were prepared by the pharmacy staff once they received a signed and dated study enrollment sheet from the ER staff with the patients weight in kilograms provided. All physicians, nurses, patients and study personnel were blinded to the contents of the syringes which were hand delivered by pharmacy personnel.

Patients were randomized to receive either 0.5 mg/Kg of ketamine or placebo (normal saline) delivered to the emergency room sedating physician in a 3 cc syringe containing a clear/colorless solution. This solution was delivered intravenously over a one minute infusion. On completion of this infusion all patients received propofol starting at 1 mg/Kg over 2 minutes and supplemented with repeated boluses of 0.5 mg/Kg to maintain adequate sedation. Patients were felt to be adequately sedated once they received a Colorado Behavioral Numerical Pain Scale (CBNPS) score of 0 to 113(table 1.) Patients were monitored after the procedure until a normal level of consciousness was observed.

The quantity of all drugs delivered were recorded. During the procedure all patients were monitored for 5 respiratory depression markers:

* PetCO2 rise of ³ 5 mm/Hg
* Respiratory rate \< 8 br/min
* arterial oxygen saturation (SaO2) \< 90%
* Apnea ³ 15 seconds
* Airway manipulation

Physicians were permitted to intervene and provide any supportive/resuscitative measures at there discretion despite the pre-specified respiratory depression markers.

All data were collected and recorded on standardized Hospital PSA forms. Data was collected for the entire time frame of the individual procedural sedations. Following the completion of the procedural sedation a second form was filled out by both the sedating physician and monitoring nurse recording specifically any respiratory events/rescue interventions and overall satisfaction with the procedural sedation. The overall quality of the PSA was evaluated by the physician/nurse performing the sedation as one of the following:

1. Not satisfied
2. Somewhat satisfied
3. Satisfied
4. Very satisfied
5. Excellent

All data was recorded on a secure computer in spreadsheet form (Microsoft Excel 2003, Microsoft Corporation, Redmond, WA) for later analysis.

Outcome Measures

Four endpoints were defined prior to study initiation:

Respiratory Depression. A difference in evidence of respiratory depression between the groups. Respiratory depression was defined as the occurrence of any of the 5 markers.

Satisfaction with PSA. A difference in the evaluation of the quality of the sedation by the providers.

Quality of PSA. Number of patients with a CBNPS of 0.

Propofol usage. Did the addition of ketamine significantly reduce the amount of propofol needed to produce adequate PSA.

Data analysis We plan to enroll 100 patients in each group. This is based on an estimate of a 40% incidence of respiratory depression with propofol alone, a reduction to 20% with the combination of ketamine/propofol, an a of 0.05 and a power of 0.8.14-16. An interim analysis was conducted at enrollment of 100 patients using a significance of 0.025 for difference in respiratory depression.

Respiratory depression and CBNPS were compared using a chi-squared test and Satisfaction with PSA and Quality of PSA were compared using a t-test. Tests were done with Primer of Biostatistics (Version 6.0, Stanton A. Glantz, 2005).