You are on page 1of 9

clinical investigations in critical care

Propofol vs Midazolam for ICU Sedation*
A Canadian Multicenter Randomized Trial
Richard I. Hall, MD, FCCP; Dean Sandham, MD, FCCP; Pierre Cardinal, MD; Martin Tweeddale, MD; David Moher, MSc; Xiaohua Wang, MSc STAT; and Aslam H. Anis, PhD; for the Study Investigators†

Study objectives: To determine whether sedation with propofol would lead to shorter times to tracheal extubation and ICU length of stay than sedation with midazolam. Design: Multicenter, randomized, open label. Setting: Four academic tertiary-care ICUs in Canada. Patients: Critically ill patients requiring continuous sedation while receiving mechanical ventilation. Interventions: Random allocation by predicted requirement for mechanical ventilation (short sedation stratum, < 24 h; medium sedation stratum, > 24 and < 72 h; and long sedation stratum, > 72 h) to sedation regimens utilizing propofol or midazolam. Measurements and results: Using an intention-to-treat analysis, patients randomized to receive propofol in the short sedation stratum (propofol, 21 patients; midazolam, 26 patients) and the long sedation stratum (propofol, 4 patients; midazolam, 10 patients) were extubated earlier (short sedation stratum: propofol, 5.6 h; midazolam, 11.9 h; long sedation stratum: propofol, 8.4 h; midazolam, 46.8 h; p < 0.05). Pooled results showed that patients treated with propofol (n 46) were extubated earlier than those treated with midazolam (n 53) (6.7 vs 24.7 h, respectively; p < 0.05) following discontinuation of the sedation but were not discharged from ICU earlier (94.0 vs 63.7 h, respectively; p 0.26). Propofol-treated patients spent a larger percentage of time at the target Ramsay sedation level than midazolam-treated patients (60.2% vs 44.0%, respectively; p < 0.05). Using a treatment-received analysis, propofol sedation either did not differ from midazolam sedation in time to tracheal extubation or ICU discharge (sedation duration, < 24 h) or was associated with earlier tracheal extubation but longer time to ICU discharge (sedation duration, > 24 h, < 72 h, or > 72 h). Conclusions: The use of propofol sedation allowed for more rapid tracheal extubation than when midazolam sedation was employed. This did not result in earlier ICU discharge. (CHEST 2001; 119:1151–1159)
Key words: ICU; mechanical ventilation; midazolam; multicenter; propofol; randomized clinical trial; sedation; Abbreviations: APACHE acute physiology and chronic health evaluation; CI confidence interval

*From the Queen Elizabeth II Health Sciences Centre (Dr. Hall), Halifax, Nova Scotia, Canada; Foothills Hospital (Dr. Sandham), University of Calgary, Calgary, Alberta, Canada; Ottawa General Hospital (Dr. Cardinal), Ottawa, Ontario, Canada; Vancouver General Hospital (Dr. Tweeddale), Vancouver, British Columbia, Canada; Ottawa Civic Hospital (Mr. Moher), Ottawa, Ontario, Canada; St. Paul’s Hospital (Mrs. Wang), Vancouver, British Columbia, Canada; and the Department of Health Care and Epidemiology (Dr. Anis), University of British Columbia, Vancouver, British Columbia, Canada. †A list of additional study investigators is located in Appendix 1. Presented in part to the American College of Chest Physicians Annual Meeting, New Orleans, LA, October 26 –30, 1997.

This research was supported by Zeneca Pharma Inc. Canada. Dr. Hall has received consultation fees from Zeneca Pharma Inc. and Hoffman-LaRoche Limited. He has no equity interest in either company. Drs. Sandham, Cardinal, Tweeddale, and Anis, and Mr. Moher and Mrs. Wang have no financial relationship with either company. Manuscript received March 11, 1999; revision accepted September 7, 2000. Correspondence to: Richard I. Hall, MD, FCCP, Department of Anesthesia, Queen Elizabeth II Health Sciences Centre, 1796 Summer St, Halifax, Nova Scotia, Canada B3H 3A7; e-mail: rihall@is.dal.ca

CHEST / 119 / 4 / APRIL, 2001

1151

center 2. long-term mechanical ventilation problems. sealed. a surgical and trauma ICU.3–5 Both are rapid-acting drugs and do not suppress adrenal function when administered to the critically ill patient. Infusion Regimens Once randomized.024 mg/kg/h adjusted to achieve the target Ramsay sedation score. While a variety of agents have been employed for the provision of sedation. medium term ( 24 h but 72 h). given that midazolam is a longer-acting4 agent than propofol.18. Eligibility Criteria Inclusion Criteria: Patients were of either gender. and were entered into the trial. We also reasoned that there are sufficient differences in the onset of drug effect that knowledgeable caregivers would likely recognize treatments on this basis alone. and required immediate sedation so as to permit the initiation and tolerance of mechanical ventilation.15. Groups of four opaque. medium-term. The physical appearance of propofol (formulated in a white lipid emulsion) is different from midazolam (clear liquid). Physicians made an assessment as to whether patients would require sedation for short-term ( 24 h). and long term ( 72 h). frestimuli including pain after protocol and consent process. two currently popular ones are midazolam and propofol. center 3. this difference in pharmacokinetic properties would lead to significant differences in time to tracheal extubation and ICU discharge. 18 years of age. No restrictions were placed on the type or amount of analgesia required to achieve patient comfort.2 Sedation is frequently required as a component of compassionate care in these patients. surrogate consent was obtained from next-of-kin at the earliest possible time following ICU admission. knowledge of the treatment groups would prevent potentially hazardous hemodynamic changes through appropriate care with drug infusion rates. a mixed population of medical-surgical patients and the regional trauma unit. Canada) random block design.8 –17 Data regarding the safety and efficacy of propofol when used for prolonged sedation (ie.012 to 0. were randomized. As for the propofol group. open-label trial conducted in four ICUs in academic medical centers across Canada. or status epilepticus. The research review board of each institution approved the 1152 We elected not to mask the treatment groups. randomized.6. a secondary question was in which segment of the ICU population (ie. Because the benefits of a shorter-acting sedative might be offset by the duration of infusion21 (ie.12. or long-term ( 72 h) mechanical ventilation on admission to the ICU. Ontario.6 mg/kg/h initially. medium-term ( 24 and 72 h). Furthermore. which was subsequently titrated to achieve a target Ramsay sedation score22 (see Appendix 2) that was specified for each patient at least daily and was adjusted based on the patient’s response to therapy. In situations in which patients could not give consent.16. Masking Materials and Methods Trial Design This study was a multicenter.20 This multicenter study conducted in four Canadian ICUs sought to examine whether. to treat acute agitation). at all hours of the day or night was limited. the infusion rate could be increased so as to provide a bolus dose prior to adjustment of the standard infusion regimen. for example. coma due to a cerebrovascular accident or unknown etiology. The practicality of masking infusions by wrapping bags and tubing. 24 h). and center 4.1. patients requiring prolonged infusions during mechanical ventilation receive more sedation and have prolonged recovery times). when necessary to administer bolus infusions for acute agitation. Patients then were stratified by predicted sedation time while receiving mechanical ventilation. comparing propofol to midazolam for the sedation of ICU patients suggest that propofol is as good as midazolam for this purpose and that its pharmacokinetic profile may permit more rapid dissipation of effects.19 and it is not currently recommended for long-term sedation due to lack of such information.11.7 Randomized studies. thus allowing rapid weaning from mechanical ventilation and tracheal extubation. and discomfort from the presence of an endotracheal tube. or long-term mechanical ventilation) did this benefit apply? Our null hypothesis was that there would be no difference in time to tracheal extubation or ICU length of stay when propofol was compared to midazolam as the primary sedative agent in patients requiring mechanical ventilation. Indeed. largely from European centers. those requiring sedation for short-term. Patients randomized to the midazolam group received an infusion of 0. sequentially numbered envelopes with the group allocation coded for each center were provided for each of the three strata based on predicted sedation time while the patient was receiving mechanical ventilation as follows: short term (expected duration.3 to 0. 72 h) are limited. a mixed medical-surgical unit that also admitted postoperative cardiac surgical cases.quent venipuncture. cranial trauma or neurosurgical intervention.8. and any leakage of solution would unmask the study. Clinical Investigations in Critical Care . These centers were characterized as follows: center 1. Mississauga. Exclusion Criteria: Exclusion criteria included a known or suspected allergy or intolerance to propofol or midazolam. a mixed medical-surgical unit that was a referral center for patients with difficult. atients a variety of P noxious in an ICU are exposed to surgery. suspected pregnancy. in situations in which the rapid control of sedation was desired. All centers had 24-h coverage by staff intensivists and in-house coverage by resident house staff. patients allocated to the propofol group received an infusion of 0. When necessary to achieve a very rapid induction of sedation (eg. an infusion bolus could be administered. Randomization and Stratification Patients were individually assigned to treatment group by a random allocation process using a computer-generated (Zeneca Pharma Inc.

25 Patients were judged to be awake when the Glasgow coma score was 12. In situations in which patients required multiple independent periods of sedation or reintubation due to alterations in their disease process. whichever came first. CHEST / 119 / 4 / APRIL. all investigators agreed in advance that patient crossover between sedative regimens was prohibited unless a clear treatment failure could be demonstrated (eg. Patients who died or crossed over were eliminated from the study. time to ICU discharge. After 72 h. Time to tracheal extubation. Patients receiving muscle relaxants and sedation were given a Ramsay sedation score of 6.Figure 1.23 Once patients were entered into the study. then every 6 h for 48 h following extubation or until ICU discharge. Admission therapeutic intervention scoring system score26 and APACHE II score27 were recorded. increasing paradoxical excitement in response to benzodiazepine administration or hyperlipidemia in response to propofol administration). Primary Outcome Measures: The time from withdrawal of sedation until tracheal extubation and ICU discharge for each stratum was taken as the primary outcome measures. and requirements for reintubation were recorded. Decisions as to when a patient was ready for a trial of extubation or for discharge from the ICU were left to the attending intensivists. A record of vital signs was maintained every 20 min for 40 min. postoperative emergence delirium. Measurement Scales The Ramsay sedation score22 was utilized to quantitate the desired degree of sedation specified at regular intervals and adjusted as the patient’s condition (ie. Measurements The Ramsay score (target and actual) was recorded hourly for the first 72 h or up to the time of discharge from ICU if this occurred prior to 72 h. Recovery of consciousness was determined by measurement of the acute physiology and chronic health evaluation (APACHE) III modification of the Glasgow coma score to allow for scoring of the verbal component in intubated patients. recovery or deterioration) dictated. and the data were censured. The purpose of our study was to inquire whether real-world use of the agents in a diverse group of Canadian ICUs would lead to measurable differences in outcomes. Trial flow diagram. it was recorded as the patient’s condition or infusion rate was altered. 2001 1153 . eg. when sedation was required to rapidly achieve patient control of.24. the first period of sedation accompanied by tracheal extubation was utilized for data collection surrounding this event. the time factor required to prepare and mask solutions was thought to be such as to be detrimental to patient care.

22 patients per group to detect a difference of 12. The trial was terminated prematurely for fiscal reasons. We assumed an interest in detecting a 20% relative reduction in recovery time using propofol ( 0. A second analysis was performed post hoc to determine differences between groups based on the time from sedation reduction to tracheal extubation and ICU discharge using the data for patients as actually sedated (ie. those patients sedated for 24 h.6 0.5%) in the propofol group (p 0. There were no differences Table 1—Distribution of Study Population by Center.37). Lack of response to sedation occurred in four midazolam-treated patients (one of whom died) and in seven propofol-treated patients (four of whom died) (p 0. and 54. were examined by a blinded safety review board that determined whether the event was attributable to study drug treatment or not.95 h. and Diagnosis* Center 1 Diagnosis Surgical Elective postoperative Vascular surgery Cardiac surgery Trauma Liver transplant Intra-abdominal sepsis Necrotizing fasciitis Mediastinitis Burn Aspiration pneumonia Medical Pulmonary edema (CHF) COPD Acute MI Pneumonia Guillain-Barre ´ Pancreatitis Hemorrhage (GI bleed) Diagnosis unknown Total Center 2 Center 3 Center 4 Mid Prop Mid Prop Mid Prop Mid Prop 2 1 0 1 0 4 0 1 0 2 4 3 0 2 1 1 0 0 0 2 1 3 0 4 0 3 0 0 0 1 2 3 0 1 0 4 1 0 0 0 5 2 13 0 0 0 0 0 0 0 1 1 14 0 0 3 1 0 1 0 1 0 0 1 0 1 1 0 1 0 0 1 0 0 0 0 0 0 0 1 0 0 1 1 0 0 0 0 Results The trial flow diagram28 is presented in Figure 1.3 h in the long-term sedation group (ie. A two-way analysis of variance was used to assess differences in tracheal extubation time by treatment group and sedation stratum. A similar analysis was used to compare differences in the time of ICU length of stay. Lack of response to sedation was attributed to paradoxical excitement and agitation in three midazolam-treated patients and in one propofoltreated patient. The statistical power was calculated for each stratum separately (short term [ 24 h]. Secondary Outcome Measures: The time when the patient was deemed ready for extubation (as assessed by attending physicians clinically and by use of weaning parameters in most cases) was recorded as well as the interval between when the patient was deemed ready for discharge and when discharge actually occurred.Data were collected for the duration of the patient’s ICU stay.05.9%) and 15 deaths (19. Clinical Investigations in Critical Care . as identified by site investigators.1. Prop propofol.8 h in the medium-term sedation group (ie. Mid midazolam. ICU length of stay was recorded as the time from admission to ICU until the patient was discharged to the floor.11 In that study. Hypertriglyceridemia developed in one propofol-treated patient who subsequently died. Statistical Analysis The primary analysis was by intention-to-treat and was based on differences in time to tracheal extubation and time to ICU discharge from the withdrawal of sedation using data from patients as assigned to the original stratum.97 h).0 5.8 h in the short-term sedation group (ie. 22 per group to detect a difference of 0.7 12. and propofol administration was thought to have possibly contributed to that death by the safety review board. Hypotension and ventricular tachycardia 1154 0 0 0 0 2 0 0 13 0 0 1 0 0 1 0 15 2 2 4 0 0 1 0 22 2 0 4 0 0 0 0 18 0 1 1 0 1 0 1 24 0 0 1 1 0 1 0 24 0 0 0 1 0 0 0 6 0 0 0 0 0 0 0 2 *Excluding patients who died or were sedation failures. and long term [ 72 h].30). Ninety-five percent confidence intervals (CIs) were constructed for all data. Probability values 0.95 h. 31 patients per group to detect a difference of 4. Group. medium term [ 24 and 72 h]. Demographic information is provided in Table 2. or 72 h regardless of initial randomization assignment). Sample Size Calculation Statistical power was estimated using reduction in recovery time as the primary outcome. There were more women in the midazolam group with a tendency to a lower weight distribution. leading to death developed in one patient receiving propofol. 7 days). the mean ( SD) recovery time for midazolam was 3. 24 h) 21. The recovery time for patients assigned to midazolam treatment was taken from the data of Carrasco et al.05 were considered to be statistically significant. 7 days). The incidence of adverse events and mortality was compared using a contingency table analysis. 24 to 72 h. 0. 2-sided). There were 11 deaths in the midazolam group (13. MI myocardial infarction. CHF congestive heart failure. Diagnostic categories for the remaining 124 patients who survived and were discharged from the ICU are presented in Table 1. Safety Review Board All reports of deaths and serious adverse events. Unacceptable hypotension developed in two propofol-treated patients. Initial treatment assignment and stratification were equally distributed among centers (data not shown).

8 (55.9]).7) 6 72.1 46. 24.4) 72. 2001 1155 . mm Hg Repeated tracheal intubation.4) 67 (62. Data for the post hoc secondary analysis using treatment received stratification are given in Table 5.50 Propofol 70.† beats/min Respiratory rate. respiratory rate. Using this analysis.4–316. top) and vs time from end of sedation to ICU discharge (Fig 2. SBP systolic BP. Therefore. Data for the primary outcome variables for the remaining 99 patients by stratum.9 117.9].05). h Midazolam (n 65) 59. 60. 52.23 0.727 0.3 (55.94 *Excluding data from patients who died or did not respond to sedation.2–18. 44. One of these patients required repeated tracheal intubation.0].56 40.67 158. M F TISS score† APACHE II score† SBP.4) 21 (18.26 72. the mean time from reduction of sedation to tracheal extubation was shorter for propofol-treated patients than for midazolam-treated patients (midazolam.2 to 9.5 207.28 159.0% [range.9) 34 29 40 (35.36 0.0].1–16.7 46.897 0.07 0. bottom).45 83.4 94. Propofol-treated patients had lower diastolic BP.0) 115 (100.3–133.2) 98 (92.74 0. Of the four patients whose time from end of sedation to tracheal extubation exceeded 80 h (Fig 2. are listed in Table 4.† mm Hg Paco2. No.4) 39 (35.62 0. yr Weight.33 p Value 0.6–22.7 207.7–48.2) 42 (36.16 0. no differences were detected.8–73. midazolam. all were treated with midazolam. data were pooled (Table 3). 94.8 p Value 0.5 to 35.06 0.0–155. the use of propofol was associ- ated with earlier tracheal extubation and longer times to ICU discharge. propofol.94 89. Individual data are given for duration of sedation vs time from end of sedation to tracheal extubation (Fig 2. and respiratory rate than did midazolam-treated patients at the time of tracheal extubation (data not shown). 6. 12 patients. p 0. three were sedated with propofol Table 3—Overall Length of Stay by Stratum* Sedation Strata 24 h Mean 95% CI Median 24–72 h Mean 95% CI Median 72 h Mean 95% CI Median Midazolam 59. or arterial blood gas levels. were noted.6) 126 (119.0 to 143.2% [range. Median length of stay. 35.3–64.6) 15 (13.67 219.2–82. Respiratory rate continued to be lower over the next 12 h.0 to 52. Otherwise.0) 92 (86. Patients treated with propofol spent more time at the target mean Ramsay sedation score level (propofol.4–130. 4.7 Propofol (n 59) 60. 63.7 h [95% CI.82 0. propofol. preventing the ascertainment of extubation time.8–278. Values given as mean (95% CI). between groups for baseline hemodynamic parameters of BP.0) 78.4–71.6–42.Table 2—Demographic Information* Variables Age.6) 68 (62.2) 16 (13.9–111. top). kg Gender.† mm Hg Heart rate.† beats/min Pao2. Significant differences between groups for the short-term and long-term strata for time to extubation. no differences in either extubation time or time to ICU discharge were detected between groups for the short sedation stratum. In the other two strata. Of the patients who had a time from end of sedation to ICU discharge of 300 h (Fig 2. When the data were pooled.17 202. No.25 151.0) 22 (20.9–78.15 0. 13 patients).05 0. 14. DBP diastolic BP.7–45.9].7 h [44.2–75. †At admission.2–66. Center analysis showed that the average length of stay was shorter for center 3 but that there was no difference in length of stay between treatment groups within each center.50 0.1]) but not the time to ICU discharge (midazolam.3 to 83.0) 124 (116.4–98.4–24.606 0.3–131.1 (68. as determined using an intention-to-treat analysis.7 h [95% CI.3) 134 (113. bottom).0–47.45 0.0 (73.† mm Hg DBP. CHEST / 119 / 4 / APRIL. propofol.1) 12 69.0 h [44.990 *Excluding data from patients who died or were sedation failures. Tracheal extubation occurred while continuous sedation was ongoing in 25 patients (midazolam.6–243. TISS therapeutic intervention scoring system.800 0.6 to 67.82 29. unless otherwise indicated. heart rate.8) 40 16 43 (37. but not for time to ICU discharge.940 0. heart rate.5–219.1–104.

9 (7.7–57.7 (24.3–43.0 (47.5) (n 4) 8. thus necessitating retention of patients in the ICU for longer periods of time. and one with midazolam.7 (63. Perhaps the ongoing release of propofol from fat stores4 contributed to residual sedation leading to an impairment of cough.7–41.7) 33. propofol-treated patients.008 72 h Extubation ICUdischarge 0 0.004 *Positive values only were used. Alternatively.16.4 (3. However.3 (1. Although not statistically significant.2 (44.15 with the use of propofol for ICU sedation.3 (5.Table 4 —Time From Sedation Reduction to Tracheal Extubation and ICU Discharge by Strata* Sedation Stratum 24 h End Point Extubation ICUdischarge Midazolam (n 26) 11.7–123.5 (0–335.1 (5. the use of propofol compared to midazolam for sedation of patients in the ICU was associated with a reduced time to tracheal extubation in evaluable patients.8) (n 11) 5.0 (6. atelectasis. but there was either no difference or a prolonged time to ICU discharge.20 *Positive values only were used with an intention-to-treat analysis.22 24–72 h Extubation ICUdischarge 0. which have demonstrated more rapid times to awakening9.2) 43. We can only speculate as to why ICU discharge was delayed.9–55.0) 72.158 72 h Extubation ICUdischarge 0.5 (10.7) 76.068 0. and they required reintubation more frequently than midazolam-treated patients. The degree to which this may have contributed to the results is unknown.17 and reduced times to tracheal extubation11.6) 49. Propofol-treated patients had lower respiratory rates following extubation.394 24–72 h Extubation ICUdischarge 0.2–88.3–18.4 (0–25.1) p Value 0. Values given as mean (95% CI). All of these patients required repeated tracheal intubation. there were more deaths among the propofol-treated patients.3) 129.30 and midazolam31 have been demonstrated to reduce neutrophil function in vitro.6) p Value 0.1–11.3) Propofol (n 28) 6.7) (n 10) 46.17 0.5) Propofol (n 21) 5. Both propofol29. we did not capture the reason for reintubation in our data set. for example.7) (n 15) 52.7 (0–316.3 (3. although they were extubated earlier (and perhaps prematurely).7 (25.5–12. lack of floor beds available to receive these patients.0) 47. Unfortunately.8) 121. Values given as mean (95% CI). Higgins Table 5—Time From Sedation Reduction to Tracheal Extubation and ICU Discharge by Stratum as Determined by Post Hoc Analysis* Sedation Stratum 24 h End Point Extubation ICUdischarge Midazolam (n 30) 9. Discussion In this multicenter randomized trial. It could perhaps be explained by problems in the systematic handling of patients within these institutions.5–210.2) 154.8) 32.3–55. unless otherwise indicated.7 (23.6–79.0) (n 8) 31.4–9.9 (1.11–14.7 (29.029 0.7) (n 17) 31.0–16.9) (n 7) 9. any pharmacokinetic advantage to earlier tracheal extubation associated with propofol use will be lost if there are routine difficulties associated with discharging patients from the ICU due to.9–249.6–8.5) 156.8 (14. and pulmonary dysfunction.5–103.03 0.6 (2. For example.9) (n 21) 7. Our trial confirms the findings of the majority of previous randomized studies.4–79.4 (30.6 (41.014 0.4–9.6 (0–108. 1156 Clinical Investigations in Critical Care . may have required more ongoing care in the ICU than patients treated with midazolam.1–180.7) 147.

18. drugs that reduce the time that a patient receives mechanical ventilation should lead to reductions in such adverse events. The strengths of our study include its randomized nature.8. Bottom: time from the end of sedation to ICU discharge vs duration of sedation. 21 days) when patients were sedated for 24 h.19 In the current study. that to capitalize on the more rapid awakening and tracheal extubation produced by the use of propofol sedation. No differences between groups were identified.11. 24 days. However. showed an increased ICU length of stay for propofol-treated patients (propofol-treated patients. The Society of Critical Care Medicine has published guidelines for sedation in the ICU. no differences in adverse events were detected between the two drug regimens. Both studies were conducted in Spain and may reflect differences in ICU practice and patient mix in that country vs Canada. Top: time from the end of sedation to tracheal extubation vs duration of sedation. Each point represents results for an individual patient. trauma patients]) or those in which there are systematic difficulties in arranging the timely transfer of patients once they are extubated out of the ICU (as exists in the Canadian health-care system as a consequence of bed closures).16 A study conducted by Weinbroum et al17 in Israel. BarrientosVega and colleagues15 also were able to show cost savings due to earlier ICU discharge in patients sedated with propofol for approximately 150 h. with cost savings in patients having ICU stays of 1 day. propofol-treated patients. prompting a change in the sedation regimen. ICU costs were five times higher in propofoltreated patients in that study. Note the increased variability for patients who received midazolam for sedation.12.20 They recommend propofol or midazolam for sedation intervals of 24 h and lorazepam for sedation intervals of 24 h. and significant hypotension developed in three others (one of whom died). consequently.34 Intuitively. those in which the patient’s disease process will require further ICU stay after extubation [eg. multicenter design. and differences in CHEST / 119 / 4 / APRIL.16. Carrasco et al11 demonstrated that the use of propofol was associated with earlier ICU discharge and. hypertriglyceridemia developed in one patient receiving propofol who subsequently died. size. Our findings were that the use of propofol for sedation did not shorten ICU length of stay. Our results suggest that. In this study. In situations in which this is not possible (ie. another Spanish trial. A recent meta-analysis32 suggested a need for more randomized trials of ICU sedation. Our study adds to a growing body of knowledge concerning the safety of propofol for ICU sedation in adults. It would appear. therefore. but the study was not designed to detect measurable differences. while propofol may be a satisfactory agent for ICU sedation.15. the ability to rapidly discharge the patient once tracheal extubation occurs is essential. 2001 1157 . their recommendations are predicated on a lack of information concerning the safety of propofol for long-term sedation. conducted in a trauma population. attention should be given to the monitoring of hemodynamics and serum triglyceride levels on a regular basis as a proportion of patients will develop hemodynamic and metabolic complications. Comparative studies reporting the role of propofol sedation to accelerate ICU discharge have produced mixed results. 31 days. et al10 did not find a difference in time to tracheal extubation when comparing propofol to midazolam for sedation in a cardiac surgical patient population. Prolonged tracheal intubation may be associated with adverse clinical events.Figure 2. Each point represents results for an individual patient. In contrast. In part. midazolam-treated patients. 18 days). the use of propofol may not be cost-effective. did not measure a significant difference in ICU discharge time (midazolam-treated patients. while able to show a difference in time to awakening. including development of nosocomial pneumonia33 and barotrauma.

et al. 15 16 17 Appendix 2: Ramsay Sedation Scale 1. Costa J. Propofol for long-term sedation in the intensive care unit: a comparison with papaveretum and midazolam. Propofol infusion for sedation in intensive care. 22:1415–1423 Carrasco G. Canada. Draper EA. Can J Anaesth 1993. Crit Care Med 1994. 25:33– 40 Sanchez-Izquierdo-Riera JA. Mississauga. For budgetary reasons. but. Commercial Products Manager.case mix. The APACHE III prognostic system: risk prediction of hospital mortality for critically ill hospitalized adults. It permitted earlier tracheal extubation compared to midazolam but did not permit earlier ICU discharge. et al. de Latorre FJ. 277: 63– 64 Knaus WA. et al. Ontario. Propofol vs midazolam in short-. Crit Care Med 1995. et al. et al. Canada. Hugh Devitt. et al. Montero A. Propofol versus midazolam for intensive care unit sedation after coronary artery bypass grafting. Pottecher T. Ottawa Civic Hospital. 3. Ottawa. Practice parameters for intravenous analgesia and sedation for adult patients in the intensive care unit: an executive summary. 23:1596 –1600 Hughes MA. Safety Review Committee. Drugs 1995. Drummond M. Crit Care Med 1997. Grounds RM. 23:286 –293 Kress JP. Pharmacology of drugs frequently used in ICUs: midazolam and flumazenil. McDonald JC. 76:334 –341 Ramsay MA. Disadvantages of prolonged propofol sedation in the critical care unit. 17(suppl):S1–S10 4 Fulton B. eds. 45:366 – 372 Boyd O. MD. A weakness is the lack of masking. Yared JP. Canada. Anderson GL. Edmonton. Sorkin EM. Crit Care Med 1994. et al. By chance alone. Crit Care Med 1996. and tranquil Responding to commands only Brisk response to glabellar tap Sluggish response to glabellar tap No response to light glabellar tap 18 19 20 References 1 Aitkenhead AR. Should clinical trials with concurrent economic analyses be blinded? JAMA 1997. MD. Vancouver. Murray AM. Propofol as used for sedation in the ICU. MD. Chest 1995. Halpern P. PA: WB Saunders. Hetzel W. PerezVela JL. et al. randomized. Simpson BR. Morales-Garcia C. we did not complete the recruitment of sufficient numbers of patients in the medium-term sedation stratum (24 patients) and the long-term sedation stratum (40 patients). 153:1012–1018 Barrientos-Vega R. Prolonged sedation with propofol in ICU patients: recovery and blood concentration changes during periodic interruptions in infusion. Crit Care Med 1995. 40:1142–1147 Higgins TL. Wallace PG. Ontario. or restless. Warren J. 64:394 –398 Newman LH. Sunnybrook Health Sciences Centre. 63:196 –206 2 Park GR. Prolonged sedation of critically ill patients with midazolam or propofol: impact on weaning and costs. Branson RD. 2. 50:636 – 657 Mirenda J. Midazolam infusion for sedation in the intensive care unit: effect on adrenal function.. 1–19 3 Amrein R. Comparative study of propofol versus midazolam in the sedation of critically ill patients: results of a prospective. 24:932–939 Ronan KP. Egol AB. 12 13 14 Appendix 1: Additional Study Investigators David Stewart. Gempeler F. Anesthesiology 1986. 42:929 –937 Harris CE. Paul’s Hospital. Intensive Care Med 1997. 23:1258 –1263 Beller JP. Anxious and agitated. et al. et al. BSc STAT. and Hector Leon. or both22 Cooperative. Craig Guenther. Zeneca Pharma Inc. O’Connor MF. Propofol: an overview of its pharma1158 21 22 23 24 cology and a review of its clinical efficacy in intensive care sedation. 2:656 – 659 Freemantle N. which may have affected our ability to detect differences within the medium-term sedation stratum and suggests that caution should be exercised in the interpretation of the data from the long-term stratum. Unknown gender differences in pharmacokinetics may have played a role in the results. et al. 4. Study Monitor. et al. Anesthesia 1987. 5 6 7 8 9 10 11 Conclusion Propofol was a satisfactory agent for ICU sedation in this randomized multicenter trial. The need for sedation and analgesia. et al. Midazolam versus propofol for long-term sedation in the ICU: a randomized prospective comparison. Analgesia and sedation in intensive care. more women were randomized to the midazolam treatment group. Br J Anaesth 1989. oriented. for reasons already expounded. White PF. Sedation and analgesia. 1993. 5. et al. Data Entry and Statistical Analysis. Propofol or midazolam for short-term alterations in sedation. Jacobs JR. 103:557–564 Chamorro C. we did not think that we could mask the study treatments in any practical fashion. 86:1219 –1224 Weinbroum AA. BMJ 1974. Chest 1991. Anesthesia 1990. In: Park GR. Wagner DP. Context-sensitive half-time in multicompartment pharmacokinetic models for intravenous anesthetic drugs. Rudick V. 22:710 –712 Shapiro BA. British Columbia. Glass PS. Rushmer F. St. Estafanous FG. 108:539 –548 Shapiro JM. et al. Controlled sedation with alphaxalone-alphadolone. University of Alberta Hospitals. Mackay CJ. Alberta. Safety Review Committee. Ontario. MSc. Mar S. Broyles G. Intensive Care Med 1991. Molina R. Sedation of critically ill patients during mechanical ventilation: a comparison of propofol and midazolam. Toronto. Comparison of propofol and midazolam for sedation in intensive care unit patients. Society of Critical Care Medicine. Am J Respir Crit Care Med 1996. Anesthesiology 1992. Savege TM. 6. et al. Caballero-Cubedo RE. Anesth Analg 1998. 100:1619 –1636 Clinical Investigations in Critical Care . Propofol versus midazolam: safety and efficacy for sedating the severe trauma patient. Pohlman AS. et al. George B. Chest 1993. Canada. multicenter trial. Westphal LM. Mauricio Calero. medium-. 61:583–588 Valente JF. Br J Anaesth 1988. Canada. and long-term sedation of critically ill patients: a cost-benefit analysis. Lugnier A. Gallagher TJ. Philadelphia. Gempeler F.

et al. et al. Cullen DJ. Crit Care Med 1999. Crit Care Med 1983. et al. Anesth Analg 1998. et al. Mechanisms of ventilator-induced lung injury. et al. The inhibitory effects of thiopental. Sedation in the intensive care unit: a systematic review. JAMA 2000. 27:1843–1847 Nishina K. Propofol inhibits human neutrophil functions. Draper EA. ed. Risk Factors for nosocomial pneumonia. Wagner DP. Yearbook of intensive care medicine 1993. 86:159 –165 Ostermann ME. Nishina K. 87:695–700 30 Hofbauer R. 11:1–3 27 Knaus WA. et al. Mikawa K. NY: Springer-Verlag. 21:131–143 CHEST / 119 / 4 / APRIL. APACHE II: a severity of disease classification system. 273–280 Parker JC. Therapeutic intervention scoring system: update 1983. Akamatsu H. Anesth Analg 1998. New York. Hernandez LA. 276:637– 639 29 Mikawa K. Crit Care Med 1993. Lancet 1974. Cho M. Propofol reduces the 31 32 33 34 migration of human leukocytes through endothelial cell monolayers. Improving the quality of reporting of randomized controlled trials: the CONSORT statement. 13:818 – 829 28 Begg C. 2001 1159 . Seiferling RA. Jennett B. 283:1451–1459 Brun-Buisson C. 2:81– 84 26 Keene AR. Eastwood S. and ketamine on human neutrophil functions. Assessment of coma and impaired consciousness: a practical scale. Akamatsu H. Keenan SP. In: Vincent J-L.25 Teasdale G. 1993. midazolam. Frass M. Peevy KJ. Crit Care Med 1985. JAMA 1996. Salfinger H.