989
ACADEMIC EMERGENCY MEDICINE • October 1999, Volume 6, Number 10
CLINICAL INVESTIGATIONS A Clinical Trial of Propofol vs Midazolam for Procedural Sedation in a Pediatric Emergency Department CHARLES J. HAVEL JR., MD, RICHARD T. STRAIT, MD, HALIM HENNES, MD, MS
Abstract. Objective: To compare the effectiveness, recovery time from sedation, and complication rate of propofol with those of midazolam when used for procedural sedation in the pediatric emergency department (PED). Methods: A prospective, blinded, randomized, clinical trial comparing propofol and midazolam was conducted in the PED of a tertiary pediatric center. Eligible patients were aged 2–18 years with isolated extremity injuries necessitating closed reduction. All patients received morphine for pain, then were randomized to receive propofol or midazolam for sedation. Vital signs, pulse oximetry, and sedation scores were recorded prior to sedation and every 5 minutes thereafter until recovery. Recovery time, time from cast completion to discharge, and other time intervals during the PED course and all sedation-related complications were also recorded. Results: Between August 1996 and October 1997, 91 patients were enrolled. Demographic data, morphine doses, and sedation scores were similar between the
P
EDIATRIC patients may experience a pronounced stress response to conditions that either are painful or necessitate prolonged immobility. Accordingly, the American College of Emergency Physicians, the American Academy of Pediatrics, and the American Society of Anesthesiologists (ASA) have all issued statements supporting the safe and appropriate use of sedative From the Section of Emergency Medicine, Department of Pediatrics (CJH, RTS, HH), and the Department of Emergency Medicine (CJH), Medical College of Wisconsin, Milwaukee, WI. Received December 2, 1998; revision received May 9, 1999; accepted May 26, 1999. Presented at the Ambulatory Pediatric Association annual meeting, Washington, DC, May 1997; the 10th International Trauma Anesthesia and Critical Care Symposium, Baltimore, MD, May 1997; and the SAEM annual meeting, Washington, DC, May 1997. Supported in part by a grant from the Children’s Hospital of Wisconsin Foundation, #3307.914. Address for correspondence and reprints: Charles J. Havel Jr., MD, Department of Emergency Medicine, 9200 West Wisconsin Avenue, Milwaukee, WI 53226. Fax: 414-257-8040; e-mail:
[email protected]
propofol and midazolam groups. Mean ⫾ SD recovery time for the propofol group was 14.9 ⫾ 11.1 minutes, compared with 76.4 ⫾ 47.5 minutes for the midazolam group, p < 0.001. Mild transient hypoxemia was the most significant complication, occurring in 5 of 43 (11.6%) patients given propofol and 5 of 46 (10.9%) patients given midazolam (odds ratio 1.08, 95% CI = 0.24 to 4.76). Conclusion: In this study, propofol induced sedation as effectively as midazolam but with a shorter recovery time. Complication rates for propofol and midazolam were comparable, though the small study population limits the power of this comparison. Propofol may be an appropriate agent for sedation in the PED; however, further study is necessary before routine use can be recommended. Key words: propofol; procedural sedation; closed orthopedic reduction; pediatric emergency department. ACADEMIC EMERGENCY MEDICINE 1999; 6: 989–997
medications to ameliorate this response.1–3 Midazolam is commonly used for procedural sedation in the pediatric emergency department (PED).4 However, reports of procedural sedation with propofol in other outpatient situations5–15 suggest that it may be a desirable alternative for PED patient sedation. This study compared propofol and midazolam when used for procedural sedation during closed orthopedic reductions in pediatric patients. The specific study hypotheses were: 1) propofol and midazolam would produce equally effective sedations; 2) patients receiving propofol would recover more rapidly from sedation than patients receiving midazolam; and 3) no difference in complication rates between groups would be observed.
METHODS Study Design. This study was a prospective, randomized, blinded trial of procedural sedation
990
PROPOFOL
Havel et al. • CLINICAL TRIAL OF PROPOFOL
TABLE 1. Ramsay Sedation Scale17 Awake 1 = anxious and agitated or restless 2 = cooperative, oriented, and tranquil 3 = responds to commands only
Asleep 4 = brisk response to light glabellar tap or loud auditory stimulus 5 = sluggish response to light glabellar tap or loud auditory stimulus 6 = no response to light glabellar tap or loud auditory stimulus
induced with propofol compared with that induced with midazolam. The institutional human rights review board approved the study protocol and written informed consent was obtained prior to patient enrollment. Study Setting and Population. Patients 2–18 years of age presenting to the ED of a tertiary care children’s hospital with an isolated extremity injury necessitating procedural sedation for closed reduction were eligible for enrollment. Exclusion criteria included: 1) hemodynamic compromise; 2) a history of cardiac disease; 3) a known allergy to any study medication, eggs, or soybeans; and 4) inability to obtain informed consent from a parent or guardian. A study investigator (CJH or RTS) was required to administer study medications and be present at the bedside throughout the course of sedation. Patients enrolled were judged by the consulting orthopedic physician to require urgent reduction of their injuries and were therefore not candidates for elective sedation and reduction. Sedation was instead performed with all patients considered to have ‘‘full stomachs’’ with close attention to the level of sedation induced. No specific data were collected regarding the fasting status of the study patients. Study Protocol. Intravenous opioid administration was the only means of providing analgesia in this study. Enrolled patients received morphine sulfate 0.05–0.1 mg/kg (maximum single dose 5 mg) intravenously (IV) prior to administration of any sedation medication. Additional 0.05–0.1-mg/ kg doses of morphine were repeated as required to ensure adequate analgesia. Patients were randomized in blocks of ten using the Moses-Oakford algorithm16 to then receive either midazolam or propofol for sedation. The depth of sedation for all patients was scored numerically using the Ramsay sedation scale17–20 (Table 1). The predetermined goal of sedation for all patients was a score of 4 or 5. Patients in the midazolam group received an initial dose of 0.1 mg/kg IV (maximum single dose 5 mg) over 1 to 2 minutes. Additional doses were administered in increments of 0.05–0.1 mg/kg as required to achieve adequate sedation. Patients in the propofol group were given preservative-free lidocaine 2% in a dose of 0.5 mg/kg IV push over 1 to 2 minutes to alleviate pain at the injection site.
Propofol was then administered by syringe pump as a 1-mg/kg bolus over 2 minutes followed by a propofol infusion at 67–100 g/kg/min continued until cast completion. Additional propofol boluses, up to 1 mg/kg each, were given and/or the infusion rate adjusted as required to maintain adequate sedation. Upon cast completion, no further analgesia or sedation was administered and the patient was allowed to recover without using any reversal agents. Blinding during sedation was achieved by shielding medications, infusion tubing, and IV site from the view of everyone but the study investigator. Patients in the propofol group were given normal saline by IV injection to simulate midazolam administration while patients in the midazolam group were given normal saline boluses and infusions to simulate propofol administration. A single syringe pump (Baxter Model AS40A infusion pump, Baxter IV Systems Division, Deerfield, IL) was used to administer all propofol and saline boluses and infusions. Measurements. Demographic data, medication times and doses, and admission and discharge times were obtained from the patients’ records. An ASA risk classification21 was retrospectively assigned by review of the patient PED record. All data recorded by the PED nurse and orthopedist were obtained prospectively in blinded fashion. Sedation scoring was accomplished using the Ramsay sedation scale. PED nurse sedation scores were recorded beginning immediately presedation and then every 5 minutes until prereduction. Sedation was then scored immediately before reduction was begun and then every 5 minutes until casting was completed. Immediately at cast completion, a sedation score was recorded and then sedation scores were obtained every 5 minutes until the patient was recovered from sedation as determined by the PED nurse. Patients with a sedation score of 6 had their responses to a painful stimulus assessed. Recovery time was defined as the period from when the last dose of sedation medication was administered (either discontinuation of propofol infusion or last dose of midazolam) to when the patient was determined by the nurse to meet standard postsedation discharge criteria.1,3,4 These criteria were applied uniformly to each study patient by the PED nurse according to institutional and study protocol. Study investigators recorded
991
ACADEMIC EMERGENCY MEDICINE • October 1999, Volume 6, Number 10
unblinded Ramsay sedation scores immediately presedation, prereduction, at completion of casting, and 5 minutes after cast completion. Investigator scores were used to calculate interrater reliability and validate the blinded nursing sedation scores. The PED nurse was responsible for monitoring each patient and accompanied the patient at all times throughout the procedural sedation. Such monitoring during sedation is included in routine PED nursing responsibilities and therefore no specific additional training was required other than review of the use of the Ramsay sedation scale. Pulse oximetry was monitored continuously and values were recorded every 5 minutes beginning immediately prior to sedative administration to the time at which postsedation discharge criteria were met. Supplemental oxygen was given only for hypoxemia, defined as a pulse oximetry reading of
