ACADEMIC EMERGENCY MEDICINE 2011; 18:800–806 © 2011 by the Society for Academic Emergency Medicine
Objectives: The authors performed a prospective, double-blinded, randomized trial with emergency department (ED) patients requiring procedural sedation and analgesia (PSA) for repair of deep traumatic lacerations and reduction of bone fractures, to compare the ketamine/propofol (ketofol) combination with the midazolam/fentanyl (MF) combination.
Methods: Sixty-two patients scheduled for PSA who presented between January 2009 and June 2009 were enrolled prospectively. Thirty-one were randomly assigned to the ketofol group, and 31 were assigned to the MF group.
Results: The median starting doses were 0.75 mg/kg of both ketamine and propofol (interquartile range [IQR] = 0.75 to 1.5 mg/kg), 0.04 mg/kg midazolam (IQR = 0.04 to 0.06 mg/kg), and 2 μg/kg fentanyl (IQR = 2 to 3 μg/kg). There were no significant differences in sedation time between the groups. There were no differences in physician satisfaction (p = 0.065). Perceived pain in the ketofol group, as measured by the Visual Analog Scale (VAS), was significantly lower than in the MF group (median ketofol = 0, IQR = 0–1 vs. median MF = 3, IQR = 1–6; p < 0.001). Only one patient in each group required bag–mask ventilation, and neither of them were intubated.
Conclusions: The ketamine/propofol combination provides adequate sedation and analgesia for painful procedures and appears to be a safe and useful technique in the ED.
Today, procedural sedation and analgesia (PSA) with a wide different spectrum of drugs has an increased role in the emergency department (ED). Propofol is one of these drugs in the category of nonopioid, nonbarbiturate, sedative–hypnotic substances with rapid onset and short recovery time.1 Propofol has some drawbacks, such as painful infusion, transient cognitive dysfunction, and a relatively high occurrence of cardiovascular and respiratory depression. Because it is not known as an analgesic, sometimes, especially in painful procedures, a sufficient level of analgesia cannot be obtained.2–4
Ketamine is a phencyclidine derivative with dissociative, sedative, analgesic, and amnestic properties that preserves muscle tone and protects airway reflexes and spontaneous respiration.5 Pretreatment with ketamine has proved effective in preventing propofol infusion pain6,7 and counteracts the hemodynamic depression of propofol.8 When prescribed alone, side effects such as emergence phenomena, postoperative dysphoria, vomiting, or laryngospasm can be seen.8–10 The combination of ketamine plus propofol (ketofol), with opposing hemodynamic and respiratory effects, leads to use of smaller doses of each one and results in diminished dose-dependent side effects.11
Administering a bolus form of ketofol has been shown to be efficacious in the operating room, in the ambulatory settings,11,12 in the ED as an induction regimen for rapid sequence intubation,13 and in performing painful procedures with PSA.14 The authors found ketofol to be both safe and efficacious, but it was a purely descriptive study. Thus, we carried out a prospective, randomized trial to determine if ketofol has advantages over the more commonly used midazolam/fentanyl (MF) combination in our ED.
This was a prospective, double-blinded, randomized trial comparing two drug combinations for efficacy, physician satisfaction, and adverse events. The trial registration number is IRCT201009204784N1. The study was approved by our institutional review board, and written informed consent was obtained from all participants or legal guardians.
Study Setting and Population
The study hospital is a referral (Level 3) hospital with about 56,000 ED patients per year. Our study population was the patients requiring PSA for repair of deep traumatic lacerations and reduction of bone fractures in the ED, who presented to our center between January 2009 and June 2009. Patients were considered for inclusion if they were aged 18 years or older. Exclusion criteria included American Society of Anesthesia (ASA) physical status ≥ 3; a positive history for adverse reaction to ketamine, propofol, midazolam, fentanyl, or egg; pregnancy; increased intracranial pressure; multiple trauma; and patients with a major psychiatric disease (e.g., psychosis) who were unable to complete the Visual Analog Scale (VAS).
A convenience sample of patients was enrolled (n = 62), with enrollment largely dependent on the availability of emergency physicians (EPs), especially during busy periods and overnight. Enrollment and the study procedure are further described in Figure 1. Patients were randomized to receive either ketofol or MF. Before commencement of the study, sedation packs were assembled. One pack was used for each patient and contained all data collection documents and a sealed envelope indicating the drug to which the patient had been randomized. An anesthesiologist who did not participate in the collection of data used Random Allocation Software15 to block-randomize subjects. No other person knew the nature of the drug to be used for any patient until the study packs were opened. The anesthesiologist prepared two sets of syringes containing the drug dosage required for each 10 kg body weight. Syringes were prepared as a 1:1 mixture of ketamine (7.5 mg) and propofol (7.5 mg) for the ketofol group and midazolam 0.4 mg + fentanyl 20 μg for the MF group. All syringes were covered with aluminum foil, to appear the same. All procedures were performed in our ED, which is equipped for pulse oximetry and cardiac monitoring, emergency resuscitation, and airway management. Supplementary oxygen was used only for pulse O2 saturations less than 90%. If necessary, head tilt and chin lift maneuvers, bag–mask ventilation, or endotracheal intubation was applied.
Three staff members attended each patient. An EP enrolled the patient; opened the study pack; administered the predetermined drug; and documented patient demographics, drug doses, times, intervention for hemodynamic changes, patient pain, physician satisfaction, and the major causes that negatively affect physician satisfaction negatively. The EP alone was responsible for all decisions relating to sedative drug administration. An ED resident (“operator”), blinded to the study drug, performed the procedure and documented adverse events. The data collection sheet contained a list of specific adverse events to be reported. An ED nurse monitored the patient throughout and recorded the times to first awakening, heart and respiratory rate, pulse oximetry, and blood pressure at the beginning and after administration of the drugs until specific discharge criteria were met. The study drugs were given slowly over 1 minute as an initial dosage. The clinical sedation endpoint was spontaneous eyelid closure, with a variable degree of sedation allowing safe execution of the procedure without any visible pain. A clinical sedation end point was chosen rather than a prescriptive dose (mg/kg) of sedative agent, because it was believed to better reflect common clinical sedation practice, account for variability in patient response, and attempt to standardize all patients to a clinically relevant level of sedation. If the patient exhibited grimacing and/or extremity movement requiring restraint, additional drug was then administered again in half of the last used dosage until little or no patient movement and/or grimacing occurred with the advancement of the procedure. Repeat small boluses were chosen rather than a continuous infusion because it has been observed that the duration of the procedures is short, and occasionally after induction no further boluses are needed. After the first dose was prescribed, the patients’ eyes were covered with sterile gauze to mask the nystagmus effect of ketamine. Because propofol (an opaque, milky liquid) is easily distinguishable from midazolam and fentanyl (clear liquids), true blinding to the nature of the study drugs was not possible. However, to effectively blind the operator, drugs were given by an intravenous (IV) catheter connected to a 3-way stopcock. The IV tubing was covered, and supplementary drug doses or alternative drugs were administered under a drape by the EP if required.
In this study, the primary outcome measures were sedation time (defined as drug prescription to first eye opening to verbal stimulus), patient’s pain severity, and practitioner’s satisfaction. Pain intensity was assessed by using a 10-cm VAS, with 0 = ”no pain” and 10 = ”worst possible pain.” Physician’s satisfaction ratings were recorded on a VAS, with 0 = ”very unsatisfied” and 10 = ”completely satisfied.” Both physician and patients were aware of the mapping from numbers into descriptive anchors at the time of the selection.
Secondary outcomes were sedation level, changes in oxygen saturation and vital signs, and adverse event rates for each study group. Sedation level was evaluated by Ramsay Sedation Scores (Table 1).16 At the completion of the procedure, all patients were monitored for a further minimum period of 30 minutes on site. Patients were discharged from the ED or transferred to other wards after meeting discharge criteria (Table 2); the total score must be equal to or greater than 7 before completion of the protocol.
|I: Nervous, agitated and/or restless|
|II: Cooperative, orientated, quite patient|
|III: Only obeying the orders|
|IV: Sleeping, responding to hitting the glabella, and high voice suddenly|
|V: Sleeping, responding to hitting the glabella, and high voice slowly|
|VI: No response to any of these stimulations|
|Activity||0 = Unable to lift head or move extremities voluntarily or on command|
|1 = Lifts head spontaneously and moves extremities voluntarily or on command|
|2 = Able to ambulate without assistance|
|Breathing||0 = Apneic|
|1 = Dyspnea or shallow, irregular breathing|
|2 = Able to breathe deeply and cough on command|
|Circulation||0 = Systolic blood pressure < 80 mm Hg|
|1 = Systolic blood pressure > 80 and < 100 mm Hg|
|2 = Systolic blood pressure within normal range for patient|
|Consciousness||0 = Not responding or responding only to painful stimuli|
|1 = Responds to verbal stimuli but falls asleep readily|
|2 = Awake, alert, and oriented to time, person, place (child oriented to parent)|
Data were analyzed using SPSS (SPSS Inc., Chicago, IL) and values were expressed as number (%) or mean ± standard deviation (SD). Because primary outcome variables were unknown, and power analysis was not available to determine sample size, a pilot study with a sample size of 10 patients (five in each group) was done, which indicated a mean (±SD) time to first awakening of 18 (±10.4) minutes for ketofol and 29.1 (±10.2) minutes for MF (difference mean = 11.1). Thereafter, a sample size calculation with a power of 80% and alpha level of 0.05 statistical significance estimated that 29 patients would be required in each arm. Unpaired Student’s t-test was employed for comparison of hemodynamic measurements and duration of interventions; chi-square and Fisher’s exact tests were employed to compare distribution of sexes and differences between side effects in the groups. Pain VAS and satisfaction VAS were analyzed using the Mann-Whitney rank sum test. Statistical significance was accepted at values less than 0.05.
Thirty-one of the 62 patients were randomly assigned to the ketofol group, and 31 patients were assigned to the MF group. The ages of the patients ranged between 20 and 37 years. There was no difference between two groups in age, sex, initial oxygen saturation, blood pressure, ASA classification, and performed procedures. Demographic data and median total doses of propofol, ketamine, fentanyl, and midazolam are presented in Table 3. Additional doses were needed for 23 cases in the ketofol group and 11 cases in the MF group (p = 0.05). The median administered dose of ketofol was 1.125 mg/kg for propofol and 1.125 mg/kg for ketamine (interquartile range [IQR] = 0.75 to 1.5 mg/kg). In the MF group, the median dose of midazolam was 0.04 mg/kg (IQR = 0.04 to 0.06 mg/kg) and of fentanyl was 2 μg/kg (IQR = 2 to 3 μg/kg). There were no significant differences in sedation time between the groups. The mean (±SD) total sedation time was 25.1 (±13.8) minutes (median = 20 minutes, IQR = 18 to 30 minutes) in the ketofol group and 26.1 (±12.6) minutes (median = 25, IQR = 15 to 33) in the MF group (p = 0.77). Most patients in the ketofol group underwent PSA with a Ramsay score between IV and VI (87.1%), while most of the patients in the MF group had a Ramsay score of III or less (58.1%).
|Ketofol (n = 31)||MF (n = 31)|
|Age, yr||25 (23–37)||25 (20–32)|
|Ketamine (mg/kg)||1.125 (0.75–1.5)||—|
|Propofol (mg/kg)||1.125 (0.75–1.5)||—|
|Fentanyl (μg/kg)||—||2 (2–3)|
|Midazolam (mg/kg)||—||0.04 (0.04–0.06)|
There were no differences in median physician satisfaction score (p = 0.065). Physician satisfaction for sedation in MF group was rated as ≥8 by all except two. The median satisfaction scores for physicians in the ketofol group was 8 (range = 2 to 10; IQR = 5 to 9), and the most common reason for reduced satisfaction was the intraoperative conditions such as clonic movements and emergent phenomena during the procedure.
Perceived pain, as measured by the VAS, was significantly lower in the ketofol group than the MF group (median ketofol = 0, IQR = 0 to 1 vs. median MF = 3, IQR = 1 to 6; p < 0.001). Twenty-two of the 31 (83.9%) patients in the ketofol group were satisfied with the painless procedure, while 26 of the 31 (93.3%) patients in the MF group experienced pain (p < 0.0001) (Table 4). No patient in the ketofol group experienced high-grade pain (all nine patients had score ≤ 6), while in the MF group only nine patients (29%) had a pain score ≤ 6.
|Ketofol Group (n = 31)||MF Group (n = 31)||p-value|
|Vital sign changes|
|Pulse rate, median, beats/min (range) IQR||−3 (−33 to 14) −9 to 4||−4 (−75 to 55) −15 to 5||0.943|
|Systolic blood pressure, median, mmHg (range) IQR||4 (−30 to 33) −4 to 10||0 (−40 to 10) −10 to 0||0.006|
|Respiratory rate, median, /min (range) IQR||0 (−9 to 2) −2 to 0||−1 (−10 to 3) −2 to 0||0.589|
|Median decrease in oxygen saturation from baseline, % (range) IQR||0 (−37 to 2) −1 to 1||−2 (−16 to 5) −5 to 0||0.499|
|Number of patients experiencing a decrease in oxygen saturation (SpaO2 < 90%).||1 (3.2%)||11 (35.5%)||0.003|
|BMV during anesthesia||1 (3.2%)||1 (3.2%)||1|
|Patients requiring additional drug||23 (74.2%)||11 (35.5%)||0.005|
|Duration of anesthesia (min), median change (IQR)||20 (18 to 30)||25 (15 to 33)||0.767|
|Pain VAS, median change (IQR)||0 (0 to 1)||3 (1 to 6)||<0.001|
|Doctor’s satisfaction scale, median change (IQR)||8 (5 to 9)||9 (8 to 10)||0.065|
|Ramsay sedation score, median change (IQR)||3 (2 to 4)||6 (5 to 6)||<0.001|
During PSA in the ketofol group, most patients experienced decreases in heart rate, respiratory rate, and blood pressure, except systolic arterial pressure. Vital sign changes after the intervention are summarized in Table 4. There was statistically (but likely not clinically) significant difference in the systolic arterial pressure change between subjects treated with ketofol and MF (medians = 4 and 0 mm Hg, respectively; p = 0.006). However, among the two study groups, reduction in the heart rate and respiratory rate values was not statistically significant compared with baseline values. There was no hypotension or evidence of poor perfusion in either group. One patient in the MF group experienced bradycardia that resolved spontaneously. There were no hypoxic patients before PSA. Six cases of oxygen desaturation occurred in the MF group and two occurred in the ketofol group, one of whom suffered apnea (p = 0.120). Only one patient in each group required bag–mask ventilation, but neither of them were intubated.
Side effects that occurred during or after intervention are shown on Table 5. Between the two groups, there was significant difference in the incidence of adverse events (p = 0.001). Most complications were dose-dependent and were observed in higher doses of anesthetic drugs. Nine patients (29%) in the ketofol group developed an unpleasant emergence reaction consisting of agitation, but only one of them was treated with midazolam 0.025 mg/kg IV, with prompt resolution of the event. Four patients in the ketofol group had involuntary movements, and one experienced restlessness. One patient (3.2%) in the ketofol group had a brief self-limited episode of cough immediately after prescribing PSA drugs.
|Side Effect||Ketofol (n = 31)||MF (n = 31)|
|None||13 (41.9)||25 (80.6)|
|Emergence reaction||9 (29.0)||0|
|Involuntary movements||4 (12.9)||0|
|Oxygen desaturation||2 (6.2)||5 (16.1)|
Nausea, vomiting, dizziness, confusion, or agitation during PSA were not observed. At discharge, no patient demonstrated any sequela from any adverse event, and all patients were discharged in a good condition.
In this prospective, randomized study, we compared the safety and efficacy of the ketofol mixture with the commonly used mixture for PSA in our ED, MF. There were no significant differences in sedation time between the groups. In a prior randomized, nonblinded, prospective clinical trial of adult patients undergoing procedural sedation for painful procedures in the ED, the median time to return to baseline mental status after the procedure was completed was 14 minutes (range = 2 to 47 minutes) for the ketamine group and 5 minutes (range = 1 to 32 minutes) for the propofol group (p < 0.001).17 Although propofol has been shown to have negative effects on hemodynamic and respiratory parameters; in our study there were no cases of hypotension, bradycardia, or hypoxia that required invasive management. In the MF group, the rate of desaturation (SpaO2 < 90%) was 16.1% and in the ketofol group it was 6.2% probably due to airway patency and muscle tone preservation by ketamine17 and using a smaller dose of propofol in the ketofol combination. In previous reports, hypoxia secondary to propofol alone occurred in about 5% of patients, and bag–mask ventilation assistance has been required in 0.8%.18,19 The rate of hypoxia in our study is comparable to that of reports in the literature for sedation with ketofol. Akin et al.20 showed that the addition of low-dose ketamine to propofol preserved mean arterial pressure and reduced the risk of respiratory depression and the need for repeat medication administration.21 Goh et al.22 found that ketofol provided equivalent laryngeal mask airway insertion conditions while optimizing hemodynamics and minimizing apnea. Green et al.23 found a significant incidence of mild transient laryngospasm (8.2%) in a group of children sedated for gastroscopy with ketamine as a single drug at a total dose of 1.3 mg/kg. Miner et al.17 detected a higher rate of subclinical respiratory depression in patients undergoing procedural sedation with ketamine for painful procedures in the ED than propofol, but there was no difference in the rate of clinical interventions related to respiratory depression between the groups. Risk factors that predict ketamine-associated airway and respiratory adverse events are high IV doses and the use of coadministered anticholinergics or benzodiazepines.24
Emergence delirium occurs more often in adults than children and is an adverse effect of ketamine.25,26 Nine patients (29%) in our study had emergence phenomena, and only one of them was severe enough to be treated with midazolam. In a study of 1,022 pediatric patients, Green et al.12 reported mild emergence in 17.6% and moderate to severe agitation in 1.6% of patients. Chudnofsky et al.27 described emergence phenomena in up to 50% of adults. According to the expected rate of emergence phenomena in adults described in the literature, our results suggest that ketofol may be associated with a lower rate of unpleasant emergence than ketamine alone.28 The effect is self-limited,29 and treatment remains symptomatic, relying on benzodiazepines and/or antipsychotics (e.g., haloperidol).30 It has been shown that coadministered midazolam reduces the incidence of recovery agitation after ketamine PSA in ED adults,31 but this is not a consistent finding among studies.31,32 Nagata et al.33 and Mortero et al.2 suggested that ketamine in sedative doses is associated with electroencephalographic activation. Furthermore, small doses of ketamine increase thalamic sensory output and arousal. Arousal effects of ketamine may partially antagonize sedative effects of propofol. This could be a dose-dependent interaction of the excitatory anesthetic ketamine with a pure central nervous system depressant, such as propofol.34
Ketamine can induce emesis;35 it is reported that lower doses of propofol had antiemetic effects via antagonizing dopamine D2 receptors, and it proved useful to treat refractory nausea and vomiting in patients receiving chemotherapy.36 Early adolescence is the peak age for ketamine-associated emesis, and its rate is higher with intramuscular administration and with unusually high IV doses.37 In our study, we did not observe any nausea or vomiting.
In a descriptive study conducted in the ED setting, 114 patients requiring PSA for mainly orthopedic procedures were given a 1:1 mixture of propofol and ketamine in 1- to 3-mL aliquots titrated at the discretion of the treating physician.4 In that study, transient hypoxia occurred in 2.6% of patients, out of whom one patient required bag–valve mask ventilation. Emergence reactions occurred in three patients, and one of them received midazolam. No patient had vomiting or aspiration; they had less pain; and the median physician, nurse, and patient satisfaction scores were 10 (on a 1 to 10 scale). In keeping with this earlier report, in our study the perceived pain, as measured by the VAS, was significantly lower in the ketofol group than in the MF group. Physician satisfaction was not different between groups.
The median total dose of ketofol administered was 1.125 mg/kg for both propofol and ketamine. Badrinath et al.38 compared multiple concentrations of ketamine in a ketamine-propofol combination during monitored anesthesia care and found an increase in postoperative nausea, vomiting, and psychotomimetic effects; prolonged recovery; and delay in discharge with an increasing dose of ketamine. Use of equivalent doses of ketamine and propofol in our study could not prevent adverse effects of ketamine. We think that a small dose of midazolam as pretreatment before administration of ketofol may preserve the sedative effect and increase physician’s satisfaction by reducing clonic movements and emergence phenomenon during and after the procedure.
The relatively small number of patients examined in our study precludes generalizations regarding safety and the incidence of rare adverse effects. Further randomized, prospective studies conducted in the ED with larger samples comparing ketofol to other common PSA agents could further document the safety, efficacy, and effectiveness of the ketamine and propofol combination for PSA in the ED setting.
Because propofol (an opaque, milky liquid) is easily distinguishable from midazolam and fentanyl (clear liquids), true blinding to the nature of the study drugs was not possible. The study was performed in a single ED, and confirming the results will require a multicenter study. Finally, our ED is dedicated to adults, so the results of our study cannot be used for young children.
Coadministration of propofol and ketamine provided adequate sedation and analgesia for painful ED procedures with less oxygen desaturation than the combination of midazolam and fentanyl and a deeper level of sedation and analgesia according to the Ramsay score. As with all procedural sedation and analgesia regimens, adverse effects are possible, and thus, appropriate monitoring and the ability to intervene with cardiorespiratory support remain essential. Further randomized, prospective studies conducted in the ED with larger samples comparing ketofol to other common procedural sedation and analgesia agents could further document the safety, efficacy, and effectiveness of the ketamine and propofol combination in the ED setting.