- Email: [email protected]
The Effect of the Assignment of a Pre-Sedation Target Level on Procedural Sedation Using Propofol
The Journal of Emergency Medicine, Vol. 32, No. 3, pp. 249 –255, 2007 Copyright © 2007 Elsevier Inc. Printed in the USA. All rights reserved 0736-4679/07 $–see front matter
doi:10.1016/j.jemermed.2006.07.023
Original Contributions
THE EFFECT OF THE ASSIGNMENT OF A PRE-SEDATION TARGET LEVEL ON PROCEDURAL SEDATION USING PROPOFOL James R. Miner,
MD,
Darren Huber,
MD,
Scott Nichols,
MD,
and Michelle Biros,
MS, MD
Department of Emergency Medicine, Hennepin County Medical Center, Minneapolis, Minnesota Reprint Address: James R. Miner, MD, Department of Emergency Medicine, Hennepin County Medical Center, 701 Park Ave. S, Minneapolis, MN 55415
e Abstract—The goal of this study was to determine if there is a difference in the achieved depth of sedation, the rate of respiratory depression, procedural difficulty, or patient perceived pain or recall between patients randomized to a pre-procedural target sedation level of moderate or deep procedural sedation using propofol during the reduction of fractures and dislocations in the Emergency Department (ED). This was a randomized, prospective study of adults undergoing procedural sedation (PS) with propofol for fracture or dislocation reduction in the ED between July 2003 and March 2004. Patients were randomized to a target sedation level of moderate or deep, using American Society of Anesthesiologists’ definitions. Doses, vital signs, nasal end tidal CO2 (ETCO2), pulse oximetry, and bispectral EEG analysis (BIS) scores were recorded. Respiratory depression was defined as a change in ETCO2 >10, an oxygen saturation of <90% at any time, or an absent ETCO2 waveform at any time. After the procedure, patients were asked if they perceived any pain or had any recall of the procedure. Physicians were asked to rate the difficulty of completing the reduction using a 100-mm visual analog scale (VAS). Respiratory depression rates were compared with chi-square tests, BIS and VAS scores were compared with t tests. Seventy-five patients were enrolled, 39 randomized to the target of moderate PS and 36 to the target of deep PS. No significant complications were noted. There were 25/36 (69%) of the patients assigned to the deep sedation target group who actually achieved a deep level of
sedation and 21/39 (54%) of the patients assigned to the moderate sedation target group who actually achieved a moderate level of sedation (p ⴝ 0.40). Respiratory depression was seen in 19/39 (49%) patients with the moderate PS target and 18/36 (50%) with the deep PS target (p ⴝ 0.91). The mean minimum recorded BIS score was 67.7 (95% confidence interval [CI] 62.2–73.3) for the moderate PS target group and 59.2 (95% CI 55.1– 64.2) for the deep PS target group (p ⴝ 0.03). There were 12/39 (31%) in the moderate PS target group and 4/36 (11%) in the deep PS target group who reported pain with or recall of the procedure (p ⴝ 0.04). The mean physician VAS for procedural difficulty was 34.0 (95% CI 23.7– 44.3) for the moderate PS group and 28.8 (95% CI 18.4 –39.2) for the deep PS group (p ⴝ 0.46). In this study, the assignment of a pre-procedural target sedation level of moderate or deep PS did not influence the level of sedation achieved, the rate of respiratory depression, the occurrence of complications, the time to return of baseline mental status, or the success of the procedure. It does not seem that the assignment of a preprocedural target sedation level is an effective means of changing the outcome of ED PS. © 2007 Elsevier Inc.
This work was presented at the Society for Academic Emergency Medicine Annual Meeting, Orlando, FL, May 2004.
The goal of Emergency Department (ED) procedural sedation is to allow the patient to tolerate a painful
RECEIVED: 20 May 2005; FINAL ACCEPTED: 5 July 2006
SUBMISSION RECEIVED:
e Keywords— emergency medicine; procedural sedation; monitoring/sedation; arousal/drug effects; propofol; end tidal CO2; bispectral analysis; emergency department
INTRODUCTION
13 December 2005; 249
250
J. R. Miner et al.
Figure 1. Definition of general anesthesia and levels of sedation/analgesia. *Monitored Anesthesia Care does not describe the continuum of depth of sedation; rather it describes “a specific anesthesia service in which an anesthesiologist has been requested to participate in the care of a patient undergoing a diagnostic or therapeutic procedure.” **Reflex withdrawal from a painful stimulus is NOT considered a purposeful response. Reproduced with permission from The American Society of Anesthesiologists. “Continuum of Depth of Sedation Definition of General Anesthesia and Levels of Sedation Analgesia”, ASA Standards, Guidelines & Statements. Accessed from asahq.com in May 2005.
procedure and to decrease the perception and recall of the pain. The necessary level of sedation to achieve tolerance of a given painful procedure, while at the same time producing a lack of recall of the procedure, is unknown, as is the optimal dose of sedatives necessary to achieve these results. Several levels of sedation have been described to exist in a continuum (1). These include minimal, moderate (conscious), and deep sedation (Figure 1). The criteria used to differentiate these levels have not been linked to a patient’s lack of recall or tolerance of a procedure, however, they have been linked to differing levels of respiratory depression (2,3). Although the frequency of reported complications is rare (4,5), the administration of sedatives used for moderate or deep sedation frequently results in some degree of respiratory depression. Although it is recognized that any patient sedated to a given level can inadvertently be over-sedated to a deeper level or undersedated to a lighter level, there is little information available that describes which level of sedation is necessary for any specific procedure or the comparable risks between each level of sedation. It is intuitive that patients more deeply sedated with propofol will experience more respiratory depression (6), episodes of aspiration, and other complications with the procedures when compared to less deeply sedated patients. It is uncertain, however, if choosing a level of sedation before the procedure will change the risk of these complications or affect the outcome of the procedure. It is also uncertain if such pre-assigned levels can consistently be achieved in the clinical setting. We recently observed that patients who were less deeply sedated displayed a lower frequency of respiratory depression with a similar rate of procedural recall and pain as more deeply sedated patients (6). With this in mind, it seems possible that by avoiding deeper levels of sedation, respiratory depression can be decreased with no increase in pain or procedural recall. We were unable to
determine, however, if the depth of sedation was due to the variable effects of propofol, the procedure for which the patient was being sedated, or the pre-procedural goal for the depth of sedation. In this study, we sought to determine whether giving the physician a target level of either moderate or deep sedation before the procedure was begun would affect the level of sedation achieved, the rate of respiratory depression, the rate of the patient’s perceived pain or procedural recall, or the physician’s perception of the difficulty of completing the procedure for which the patient was sedated. A secondary goal was to determine if there is a difference in the quantity of sedatives used when physicians aimed to deliver a specific level of procedural sedation.
METHODS Study Design This was a prospective, randomized clinical trial of adult ED patients undergoing procedural sedation using propofol in the ED for the reduction of fractures and dislocations. The study was conducted between July 2003 and March 2004. Patients were randomized to have the treating physician instructed to deliver the target sedation goal of either moderate or deep sedation. The Institutional Review Board (IRB) of Hennepin County Medical Center approved the study. Patients provided prospective, informed consent before enrollment.
Study Setting and Population This study was performed at Hennepin County Medical Center in Minneapolis, MN, an urban county medical
Procedural Sedation
center with approximately 93,000 patient visits per year. All adult (age ⱖ18 years) ED patients who were going to receive procedural sedation (PS) using propofol for the reduction of a fracture or dislocation were eligible for study enrollment. Patients were excluded if they were unable to give consent, had a known hypersensitivity to propofol, were pregnant, or had clinical evidence of intoxication before the start of the procedure. In our ED, PS is performed at the discretion of the treating emergency physician (EP). It is standard management in our ED that patients with fractures and dislocations receive intravenous morphine (0.1 mg/kg i.v. followed by 0.05 mg/kg i.v. q15 min as needed/tolerated) for pain control as soon as possible upon their arrival.
Study Protocol Before the procedure, informed consent was obtained. Patients were placed on cardiac, blood pressure, pulse oximeter and nasal sample end tidal CO2 (ETCO2) monitors. Supplemental oxygen was applied according to the standard guidelines for PS in our ED. Additionally, the patient was placed on a Bispectral Index monitor (BIS). Recent studies BIS to measure awareness during procedural sedation in the ED (7–9). The EEG analog of BIS has been shown to be more predictive of loss of awareness than the amount of sedative drug given. Patients were then randomized to a target sedation level of either moderate or deep sedation. The treating physician received a sealed envelope that contained the target sedation level before beginning the procedure and the definitions of the target levels (Figure 1). The randomization order was determined by a computer-generated random number. The patient was blinded to the sedation level they had been randomized to. Data was collected by trained research assistants. The sedation was performed by a separate physician from the one who performed the procedure. Propofol doses were at the discretion of the sedating physician. Before the start of the procedure, baseline values were recorded. During the procedure, pulse oximetry, heart rate, blood pressure, respiratory rate, ETCO2 and BIS scores were monitored continuously. The lowest value during every 1-min period was recorded. Any loss of ETCO2 waveform or the use of airway adjuncts, such as bag-valve-mask assisted respirations or oral airway placement, was also noted. Respiratory depression was defined as: an oxygen saturation of ⬍90%, a change from baseline ETCO2 of ⬎10 mm Hg, or airway obstruction with cessation of gas exchange at any time (noted by an absent ETCO2 waveform). After the procedure, the physician noted any complications, including, but not limited to: vomiting or aspiration, intubation, transfer to
251
a higher level of care after the procedure, hypotension (defined as a drop in systolic blood pressure ⬎20%), dysrhythmias, and whether or not the patient required assisted ventilations (by bag-valve mask [BVM]) due to decreased protective airway reflexes during the procedure. Physicians noted the level of sedation achieved during the procedure (minimal, moderate, deep, or general anesthesia). The end point of sedation was considered the time at which the patient had regained their baseline mental status. After the patient returned to their baseline mental status, patients were asked two “yes or no” questions: “Did you experience pain during the procedure?” and “Do you remember any of the procedure?” The physician who performed the procedure was then asked to complete a 100-mm VAS assessing perceived difficulty of the procedure. This VAS consisted of the question “How difficult was the procedure to perform?” and consisted of a 100-mm line with the word “easy” at one end of the line and “difficult” at the other end. The physician was asked to determine the level of sedation achieved as defined in Figure 1.
Data Analysis Data were collected by a designated research assistant during the procedure and were then entered into an EXCEL (Microsoft Corp., Redmond, WA) database for further analysis. All analysis and interpretation of data was done using STATA (STATA Corp., College Station, TX) statistical software. Descriptive statistics were used when appropriate. The rates of respiratory depression between patients in the moderate sedation and deep sedation group were compared with chi-squared tests. The mean BIS nadir between the moderate and deep sedation groups and the time to return of baseline mental status were compared with two-tailed t-tests. The presence or absence of reported pain or recall was compared between the two groups using Fisher’s exact tests. To detect an effect size of 0.44 in the achieved level of sedation between the groups using chi-squared tests, with an alpha of 0.05 and a beta of 0.2 (80% power), power analysis indicated that 28 patients per group were required.
RESULTS There were 77/135 eligible patients enrolled in the study. Two patients were excluded due to protocol violations (the procedure for which the patient was sedated was not a fracture or dislocation reduction). A total of 39 patients were randomized to the target of moderate sedation and 36
252
J. R. Miner et al.
Table 1. Results Summarized by Randomization Group
Initial BIS (95% CI) Initial systolic BP (mm Hg) (95% CI) Initial heart rate (95% CI) Initial ETCO2 (95% CI) Initial propofol dose (95% CI) Total number of doses given (95% CI) Total propofol dose Initial BIS (95% CI) BIS nadir (95% CI) Decrease in systolic blood pressure (95% CI) Time to baseline mental status Proc diff VAS Target sedation level achieved Respiratory depression
Moderate sedation (n ⫽ 39)
Deep sedation (n ⫽ 36)
95.3 (92.9–97.7) 144.8 (137.1–152.6) 87.0 (81.2–92.8) 37.5 (34.1–40.8) 0.88 mg/kg (0.83–0.93) 2.7 (2.1–3.3) 1.69 mg/kg (1.38–2.00) 95.3 (92.9–97.7) 67.8 (62.2–73.3) 11.4% (7.5–15.3)
96.3 (94.4–98.1) 140.6 (132.8–148.4) 85.0 (80.4–89.6) 42.4 (39.2–5.2) 0.98 mg/kg (0.88–1.09) 2.6 (2.1–3.1) 1.82 mg/kg (1.52–2.11) 96.3 (94.4–98.1) 59.9 (55.1–64.2) 9.3% (6.0–12.6)
8.0 min (5.9–10.5) 31.6 mm (19.3–43.9) 21/39 19/39
8.2 min (6.1–9.9) 30.6 mm (18.3–92.8) 25/36 18/36
Overall (n ⫽ 75) 95.8 (94.3–97.2) 142.8 (137.4–148.2) 86.0 (82.4–89.7) 39.8 (37.6–42.0) 0.93 (0.87–0.99) 2.6 (2.1–3.3) 1.75 (1.38–2.00) 95.8 (94.3–97.2) 64.0 (60.3–67.8) 10.4% (7.9–12.9) 8.1 min (6.6–9.5) 31.1 mm (22.7–39.5) 46/77 ( p ⫽ 0.40) 37/75 ( p ⫽ 0.92)
BIS ⫽ Bispectral index; BP ⫽ blood pressure; CI ⫽ confidence interval; ETCO2 ⫽ end tidal CO2; VAS ⫽ visual analog scale.
to the target of deep sedation. The procedures done included the reductions of 33 fractures and 42 dislocations. No significant adverse events were noted. Seventy-four of 75 (99%) procedures were successful; the one unsuccessful procedure was a reduction of a forearm fracture in a patient who had randomized to the goal of deep sedation. The baseline characteristics and results by the two randomized groups are summarized in Table 1. The mean initial dose of propofol was 0.88 mg/kg (95% confidence interval [CI] 0.83– 0.93) for the moderate sedation group and 0.98 mg/kg (95% CI 0.87–1.14) for the deep sedation group (p ⫽ 0.08). The mean total dose of propofol given over the entire procedure was 1.70 mg/kg (95% CI 1.38 – 2.00) for the moderate sedation group and 1.82 mg/kg (95% CI 1.53–2.12) for the deep sedation group (p ⫽ 0.54). By our pre-established criteria, respiratory depression was seen in 37/75 (49%) patients; 19/39 (49%) in patients assigned to receive moderate sedation and 18/36 (50%) in patients randomized to receive deep sedation (p ⫽ 0.92). All of the patients who met criteria for respiratory depression had a change from baseline ETCO2 ⬎10 mm Hg except for 1/39 (3%) in the moderate sedation assigned group and 1/36 (3%) in the deep sedation assigned group. These two patients had an absent ETCO2 waveform at some time during the procedure, without a change from baseline ETCO2 ⬎10 mm Hg noted. Among patients with a change in their ETCO2 ⬎10 mm Hg from baseline, the mean change was 21.1 mm Hg (95% CI 18.3–23.9). Five patients were noted to have an oxygen saturation ⬍90% at some point during the procedure; 2/39 (5%) in the moderate sedation assigned group and 3/36 (8%) in the deep sedation assigned group. The mean lowest oxygen saturation measured in these patients was 80.0 (95% CI 73.4 – 86.6). No patient had an absent ETCO2 waveform or an oxygen saturation ⬍90% for ⬎1 min.
No cardiac rhythm abnormalities were noted. The mean decrease in the systolic blood pressure was 9.0% in the deep sedation group (95% CI 6.0%–13.0%) and 11.0% (95% CI 7.0%–15.0%) in the moderate sedation group. Three patients required BVM respirations during the procedure: 2/39 (5%) in the moderate sedation assigned group and 1/36 (3%) in the deep sedation assigned group. None of these patients required BVM for more than 1 min. The patient in the deep sedation group had received a single bolus of 1.2 mg/kg before requiring BVM. The first patient in the moderate group received a single dose of propofol at 0.70 mg/kg before requiring BVM, the other had received an initial dose of 0.83 mg/kg followed by a repeat dose of the same size 2 min later before requiring BVM. The initial BIS score for the deep sedation assigned group was 96.3 (95% CI 94.4 –98.1) and for the moderate sedation assigned group was 95.3 (95% CI 92.9 –97.7) (p ⫽ 0.50). The mean minimum recorded BIS score was 67.7 (95% CI 62.2–73.3) for the moderate sedation assigned group and 59.9 (95% CI 55.1– 64.8) for the deep sedation assigned group (p ⫽ 0.03). Pain or recall was seen in 12/39 (31%) in the moderate sedation assigned group and 4/36 (11%) in the deep sedation assigned group (p ⫽ 0.04). The mean physician VAS for procedural difficulty was 34.0 mm (95% CI 23.7– 44.3) for the moderate sedation assigned group and 28.8 mm (95% CI 18.3–39.2) for the deep sedation assigned group (p ⫽ 0.72). The mean time to return of baseline mental status from the completion of the procedure was 7.97 min (95% CI 6.05–9.89) for the moderate sedation assigned group and 8.19 min (95% CI 5.87– 10.52) for the deep sedation assigned group (p ⫽ 0.88). All of the patients achieved either a moderate or deep level of sedation, as defined in Figure 1. However, only 25/36 (69%) of the patients randomized to a target of
Procedural Sedation
253
Table 2. Results Summarized by Level of Sedation Achieved
Initial propofol dose (95% CI) Total number of doses given (95% CI) Total propofol dose BIS nadir (95% CI) Change in systolic blood pressure (95% CI) Time to baseline mental status Proc diff VAS Respiratory depression
Patients achieving moderate sedation (n ⫽ 32)
Patients achieving deep sedation (n ⫽ 43)
Overall (n ⫽ 75)
0.92 mg/kg (0.81–1.04) 2.5 (1.9–3.3) 1.53 mg/kg (1.29–1.76) 78.4 (76.4–80.5) 10.6% (6.1–15.1) 7.4 min (5.8–9.1) 28.4 mm (22.9–44.6) 15/32 (46.9%)
0.93 mg/kg (0.88–0.99) 2.8 (2.3–3.3) 1.92 mg/kg (1.60–2.24) 53.7 (49.7–57.6) 10.2% (7.2–13.2) 8.6 min (6.3–11.9) 33.7 mm (19.3–37.4) 22/43 (51.2%)
0.93 (0.87–0.99) 2.6 (2.3–3.0) 1.75 (1.59–1.96) 64.0 (60.3–67.8) 10.4% (7.9–12.9) 8.1 min (6.7–9.6) 31.5 mm (24.3–38.6) 37/75 ( p ⫽ 0.71)
CI ⫽ confidence interval; BIS ⫽ Bispectral index; VAS ⫽ visual analog scale.
deep sedation achieved a deep level of sedation, and 21/39 (54%) of the patients randomized to a target of moderate sedation group actually achieved a moderate level of sedation (p ⫽ 0.40). The results of patients by achieved sedation level are summarized in Table 2. There were very few differences in the outcome of patients based on the level of sedation achieved. Respiratory depression was seen in 22/43 (51%) patients who achieved a deep level of sedation and 15/32 (47%) patients who achieved a moderate level of sedation (p ⫽ 0.71). There was no difference in the amount or number of doses of propofol given, the time to baseline mental status, the systolic blood pressure, or the perceived difficulty of the procedure. There was a large difference in the lowest BIS score achieved between the groups (p ⫽ 0.00). Six of 43 (14%) patients who achieved deep sedation and 10/32 (31%) patients who achieved a moderate level of sedation reported pain with or recall of the procedure for which they were sedated (p ⫽ 0.07).
DISCUSSION The assignment of a target level of moderate or deep procedural sedation did not seem to affect the actual level of sedation achieved, the rate of respiratory depression, or the amount of propofol given after the initial bolus. There was an increased rate of pain and recall in the moderate sedation goal group. Overall, the 75 procedures were performed without significant complications and, with the exception of one, all were successful. The patient with an unsuccessful procedure in the ED eventually underwent open reduction and internal fixation in the OR by orthopedics; the EP gave the procedure a VAS difficulty of 18 mm. This seems to indicate that the reduction itself was technically difficult rather than limited by the sedation target used. Evaluation of the dosing of the propofol shows a difference in the initial dose given to the patients who randomized to the moderate vs. the deep sedation group.
However, when the total amount of propofol given is compared, there is no difference between patients in either group. No specific dosing recommendations were given to EPs for use in this study, just a target sedation level. The doses of sedative given are similar to those described in similar ED procedural sedation studies (7,10 –13). Because individual patients respond differently to the same dose of medication (14), medications used for sedation are generally titrated until the patient seems adequately sedated. There was no difference in the rate of respiratory depression detected between the two groups. We have used these criteria for detecting respiratory depression in previous studies (6 – 8). They are not intended to detect clinically significant respiratory depression, but rather any changes, including sub-clinical changes, that indicate a decrease in respiratory effort. Changes in ETCO2 can detect airway obstruction (which is indicated by a decrease in ETCO2 as the exhaled sample mixes with ambient air), hypoventilation (which is indicated by rising ETCO2 as pCO2 rises), or apnea (which is indicated by a loss of the waveform). In fact, in this study, the oxygen saturation monitor did not detect any respiratory depression that was not detected by the ETCO2 monitor, but the ETCO2 monitor suggested respiratory depression that was not predicted by the change in oxygen saturation. Three patients in our study required BVM respirations. All three had received doses of propofol near the middle of the range for this study. We did not have specific criteria for the use of BVM, and its use was at the discretion of the treating physician, so it is possible that these results are due to the specific practices of individual EPs rather than actual loss of protective airway reflexes in these patients. The fact that these patients all received typical doses of propofol illustrates the need for close airway monitoring in sedated patients, whether they are being moderately or deeply sedated, and no matter what the target level of sedation is before the procedure is started or the dose of sedative given. The decrease in systolic blood pressure noted in our study has been previously described for the propofol
254
doses given (9,13). There was no difference in blood pressure changes between the groups noted, which is not surprising because the total dose of propofol given in the two groups was not different. The time to return to baseline mental status was not significantly different between the groups. This is not surprising because both groups had the same mean number of propofol doses and the same total dose of medication given, and is likely related to the half life of propofol. This does, however, raise the question of why there is a difference in the BIS nadir and the frequency of recall between the groups. Only 69% of the patients in the deep sedation target group and 54% of the patients in the moderate sedation target group were sedated to their targeted level of sedation. Overall, 43 patients achieved deep sedation and 32 patients achieved moderate sedation. The target level of sedation a patient was randomized to was not associated with the level achieved. This may be due to a number of factors. First, the EPs may have been unable to predictably and consistently achieve a specific level of sedation. Alternatively, it may be possible that a patient randomized to the deep sedation group achieved sedation adequate to tolerate the procedure before actually reaching deep sedation. The physician may then have decided to not give further medications, and the patient with the target of deep sedation was therefore only moderately sedated. By the same logic, a patient who randomized to the moderate sedation group but was not sufficiently sedated to undergo the procedure may have been given an additional bolus of propofol, which brought them to the level of deep sedation. Because the initial dose of medicine given is lower in patients who randomized to the moderate sedation group, we assume that the physician’s behavior was altered by entering patients into one target group over another. When patients were compared by the level of sedation actually achieved, there were no differences among them except in the lowest BIS score recorded and a trend toward a difference in the rate of pain or recall with the procedure. Interestingly, there was no difference in the amount of propofol, either in the initial bolus or the total amount given, between patients who achieved either moderate or deep level of sedation. This indicates that the distinction of moderate or deep procedural sedation may not be associated with the outcome parameters we measured in this study and which are routinely used in clinical practice. It is possible that by defining sedation criteria by markers of respiratory depression instead of markers more specific to the goals of procedural sedation (i.e., decreasing the patient’s awareness or the ability to form memories), we are unnecessarily approaching sedation levels associated with respiratory depression.
J. R. Miner et al.
The sedation level achieved (moderate or deep), the number of doses of medicine given and total dose of propofol, the rate of respiratory depression, and the time to return of baseline mental status do not seem to be influenced by the assignment of the patient to a pre-procedural target sedation level of moderate or deep procedural sedation. The risk of complications, such as respiratory depression or the inability to perform the procedure, was unchanged by the assignment of a pre-procedure target sedation level. Assigning a pre-procedural target sedation level of either moderate or deep does not seem to be an effective way of influencing the outcome of ED procedural sedation, except for patient recall.
LIMITATIONS The independent variable of this study was the treating physician’s behavior, which is difficult to control. We did not recommend doses but simply asked that they achieve a target sedation level. We chose this method because this is equivalent to any external control on a sedation level. Because we have no way to measure subtle changes in physician behavior, we cannot determine with this design what aspects of this control affected the outcomes of the sedation. We believe, however, that this design simulates clinical practice, where Emergency Physicians are sometimes limited to moderate rather than deep sedation, and the results are applicable to the clinical application of procedural sedation. The number of patients in this study accounts for the detection of only a rather large difference in the rates of respiratory depression (the study is powered to detect an effect size by chi-square of 0.44). A much larger study is necessary to detect subtle differences in the rates of respiratory depression, and it is likely one exists. However, due to the nature of our current respiratory depression detection method, we feel that the interpretation of a small detected difference between these groups may be difficult.
CONCLUSIONS In this study, the assignment of a pre-procedural target sedation level of moderate or deep PS did not influence the level of sedation achieved, the rate of respiratory depression, the occurrence of complications, the time to return of baseline mental status, or the success of the procedure. It does not seem that the assignment of a pre-procedural target sedation level is an effective means of changing most outcomes of ED PS.
Procedural Sedation
255
REFERENCES 1. Practice guidelines for sedation and analgesia by nonanesthesiologists. An updated report by the American Society of Anesthesiologists Task Force on Sedation and Analgesia by Nonanesthesiologists. Anesthesiology 2002;96:1004 –17. 2. Avramov MN, White PF. Use of alfentanil and propofol for outpatient monitored anesthesia care: determining the optimal dosing regimen. Anesth Analg 1997;85:566 –72. 3. Pratila MG, Fischer ME, Alagesan R, et al. Propofol versus midazolam for monitored sedation: a comparison of intraoperative and recovery parameters. J Clin Anesth 1993;5:268 –74. 4. Pitetti RD, Singh S, Pierce MC. Safe and efficacious use of procedural sedation and analgesia by nonanesthesiologists in a pediatric emergency department. Arch Pediatr Adolesc Med 2003; 157:1090 – 6. 5. Pena BM, Krauss B. Adverse events of procedural sedation and analgesia in a pediatric emergency department. Ann Emerg Med 1999;34:483–91. 6. Miner JR, Biros MH, Heegaard W, Plummer D. Bispectral electroencephalographic analysis of patients undergoing procedural sedation in the emergency department. Acad Emerg Med 2003;10: 638 – 43. 7. Miner JR, Biros M, Krieg S, Johnson C, Heegaard W, Plummer D. Randomized clinical trial of propofol versus methohexital for
8. 9.
10. 11. 12. 13.
14.
procedural sedation during fracture and dislocation reduction in the emergency department. Acad Emerg Med 2003;10:931–7. Miner JR, Heegaard W, Plummer D. End-tidal carbon dioxide monitoring during procedural sedation. Acad Emerg Med 2002;9: 275– 80. Hasan RA, Shayevitz JR, Patel V. Deep sedation with propofol for children undergoing ambulatory magnetic resonance imaging of the brain: experience from a pediatric intensive care unit. Pediatr Crit Care Med 2003;4:454 – 8. Coll-Vinent B, Sala X, Fernandez C, et al. Sedation for cardioversion in the emergency department: analysis of effectiveness in four protocols. Ann Emerg Med 2003;42:767–72. Bassett KE, Anderson JL, Pribble CG, Guenther E. Propofol for procedural sedation in children in the emergency department. Ann Emerg Med 2003;42:773– 82. Havel CJ Jr, Strait RT, Hennes H. A clinical trial of propofol vs midazolam for procedural sedation in a pediatric emergency department. Acad Emerg Med 1999;6:989 –97. Guenther E, Pribble CG, Junkins EP Jr, Kadish HA, Bassett KE, Nelson DS. Propofol sedation by emergency physicians for elective pediatric outpatient procedures. Ann Emerg Med 2003;42: 783–91. Glass PS, Bloom M, Kearse L, Rosow C, Sebel P, Manberg P. Bispectral analysis measures sedation and memory effects of propofol, midazolam, isoflurane, and alfentanil in healthy volunteers. Anesthesiology 1997;86:836 – 47.
doi:10.1016/j.jemermed.2006.07.023
Original Contributions
THE EFFECT OF THE ASSIGNMENT OF A PRE-SEDATION TARGET LEVEL ON PROCEDURAL SEDATION USING PROPOFOL James R. Miner,
MD,
Darren Huber,
MD,
Scott Nichols,
MD,
and Michelle Biros,
MS, MD
Department of Emergency Medicine, Hennepin County Medical Center, Minneapolis, Minnesota Reprint Address: James R. Miner, MD, Department of Emergency Medicine, Hennepin County Medical Center, 701 Park Ave. S, Minneapolis, MN 55415
e Abstract—The goal of this study was to determine if there is a difference in the achieved depth of sedation, the rate of respiratory depression, procedural difficulty, or patient perceived pain or recall between patients randomized to a pre-procedural target sedation level of moderate or deep procedural sedation using propofol during the reduction of fractures and dislocations in the Emergency Department (ED). This was a randomized, prospective study of adults undergoing procedural sedation (PS) with propofol for fracture or dislocation reduction in the ED between July 2003 and March 2004. Patients were randomized to a target sedation level of moderate or deep, using American Society of Anesthesiologists’ definitions. Doses, vital signs, nasal end tidal CO2 (ETCO2), pulse oximetry, and bispectral EEG analysis (BIS) scores were recorded. Respiratory depression was defined as a change in ETCO2 >10, an oxygen saturation of <90% at any time, or an absent ETCO2 waveform at any time. After the procedure, patients were asked if they perceived any pain or had any recall of the procedure. Physicians were asked to rate the difficulty of completing the reduction using a 100-mm visual analog scale (VAS). Respiratory depression rates were compared with chi-square tests, BIS and VAS scores were compared with t tests. Seventy-five patients were enrolled, 39 randomized to the target of moderate PS and 36 to the target of deep PS. No significant complications were noted. There were 25/36 (69%) of the patients assigned to the deep sedation target group who actually achieved a deep level of
sedation and 21/39 (54%) of the patients assigned to the moderate sedation target group who actually achieved a moderate level of sedation (p ⴝ 0.40). Respiratory depression was seen in 19/39 (49%) patients with the moderate PS target and 18/36 (50%) with the deep PS target (p ⴝ 0.91). The mean minimum recorded BIS score was 67.7 (95% confidence interval [CI] 62.2–73.3) for the moderate PS target group and 59.2 (95% CI 55.1– 64.2) for the deep PS target group (p ⴝ 0.03). There were 12/39 (31%) in the moderate PS target group and 4/36 (11%) in the deep PS target group who reported pain with or recall of the procedure (p ⴝ 0.04). The mean physician VAS for procedural difficulty was 34.0 (95% CI 23.7– 44.3) for the moderate PS group and 28.8 (95% CI 18.4 –39.2) for the deep PS group (p ⴝ 0.46). In this study, the assignment of a pre-procedural target sedation level of moderate or deep PS did not influence the level of sedation achieved, the rate of respiratory depression, the occurrence of complications, the time to return of baseline mental status, or the success of the procedure. It does not seem that the assignment of a preprocedural target sedation level is an effective means of changing the outcome of ED PS. © 2007 Elsevier Inc.
This work was presented at the Society for Academic Emergency Medicine Annual Meeting, Orlando, FL, May 2004.
The goal of Emergency Department (ED) procedural sedation is to allow the patient to tolerate a painful
RECEIVED: 20 May 2005; FINAL ACCEPTED: 5 July 2006
SUBMISSION RECEIVED:
e Keywords— emergency medicine; procedural sedation; monitoring/sedation; arousal/drug effects; propofol; end tidal CO2; bispectral analysis; emergency department
INTRODUCTION
13 December 2005; 249
250
J. R. Miner et al.
Figure 1. Definition of general anesthesia and levels of sedation/analgesia. *Monitored Anesthesia Care does not describe the continuum of depth of sedation; rather it describes “a specific anesthesia service in which an anesthesiologist has been requested to participate in the care of a patient undergoing a diagnostic or therapeutic procedure.” **Reflex withdrawal from a painful stimulus is NOT considered a purposeful response. Reproduced with permission from The American Society of Anesthesiologists. “Continuum of Depth of Sedation Definition of General Anesthesia and Levels of Sedation Analgesia”, ASA Standards, Guidelines & Statements. Accessed from asahq.com in May 2005.
procedure and to decrease the perception and recall of the pain. The necessary level of sedation to achieve tolerance of a given painful procedure, while at the same time producing a lack of recall of the procedure, is unknown, as is the optimal dose of sedatives necessary to achieve these results. Several levels of sedation have been described to exist in a continuum (1). These include minimal, moderate (conscious), and deep sedation (Figure 1). The criteria used to differentiate these levels have not been linked to a patient’s lack of recall or tolerance of a procedure, however, they have been linked to differing levels of respiratory depression (2,3). Although the frequency of reported complications is rare (4,5), the administration of sedatives used for moderate or deep sedation frequently results in some degree of respiratory depression. Although it is recognized that any patient sedated to a given level can inadvertently be over-sedated to a deeper level or undersedated to a lighter level, there is little information available that describes which level of sedation is necessary for any specific procedure or the comparable risks between each level of sedation. It is intuitive that patients more deeply sedated with propofol will experience more respiratory depression (6), episodes of aspiration, and other complications with the procedures when compared to less deeply sedated patients. It is uncertain, however, if choosing a level of sedation before the procedure will change the risk of these complications or affect the outcome of the procedure. It is also uncertain if such pre-assigned levels can consistently be achieved in the clinical setting. We recently observed that patients who were less deeply sedated displayed a lower frequency of respiratory depression with a similar rate of procedural recall and pain as more deeply sedated patients (6). With this in mind, it seems possible that by avoiding deeper levels of sedation, respiratory depression can be decreased with no increase in pain or procedural recall. We were unable to
determine, however, if the depth of sedation was due to the variable effects of propofol, the procedure for which the patient was being sedated, or the pre-procedural goal for the depth of sedation. In this study, we sought to determine whether giving the physician a target level of either moderate or deep sedation before the procedure was begun would affect the level of sedation achieved, the rate of respiratory depression, the rate of the patient’s perceived pain or procedural recall, or the physician’s perception of the difficulty of completing the procedure for which the patient was sedated. A secondary goal was to determine if there is a difference in the quantity of sedatives used when physicians aimed to deliver a specific level of procedural sedation.
METHODS Study Design This was a prospective, randomized clinical trial of adult ED patients undergoing procedural sedation using propofol in the ED for the reduction of fractures and dislocations. The study was conducted between July 2003 and March 2004. Patients were randomized to have the treating physician instructed to deliver the target sedation goal of either moderate or deep sedation. The Institutional Review Board (IRB) of Hennepin County Medical Center approved the study. Patients provided prospective, informed consent before enrollment.
Study Setting and Population This study was performed at Hennepin County Medical Center in Minneapolis, MN, an urban county medical
Procedural Sedation
center with approximately 93,000 patient visits per year. All adult (age ⱖ18 years) ED patients who were going to receive procedural sedation (PS) using propofol for the reduction of a fracture or dislocation were eligible for study enrollment. Patients were excluded if they were unable to give consent, had a known hypersensitivity to propofol, were pregnant, or had clinical evidence of intoxication before the start of the procedure. In our ED, PS is performed at the discretion of the treating emergency physician (EP). It is standard management in our ED that patients with fractures and dislocations receive intravenous morphine (0.1 mg/kg i.v. followed by 0.05 mg/kg i.v. q15 min as needed/tolerated) for pain control as soon as possible upon their arrival.
Study Protocol Before the procedure, informed consent was obtained. Patients were placed on cardiac, blood pressure, pulse oximeter and nasal sample end tidal CO2 (ETCO2) monitors. Supplemental oxygen was applied according to the standard guidelines for PS in our ED. Additionally, the patient was placed on a Bispectral Index monitor (BIS). Recent studies BIS to measure awareness during procedural sedation in the ED (7–9). The EEG analog of BIS has been shown to be more predictive of loss of awareness than the amount of sedative drug given. Patients were then randomized to a target sedation level of either moderate or deep sedation. The treating physician received a sealed envelope that contained the target sedation level before beginning the procedure and the definitions of the target levels (Figure 1). The randomization order was determined by a computer-generated random number. The patient was blinded to the sedation level they had been randomized to. Data was collected by trained research assistants. The sedation was performed by a separate physician from the one who performed the procedure. Propofol doses were at the discretion of the sedating physician. Before the start of the procedure, baseline values were recorded. During the procedure, pulse oximetry, heart rate, blood pressure, respiratory rate, ETCO2 and BIS scores were monitored continuously. The lowest value during every 1-min period was recorded. Any loss of ETCO2 waveform or the use of airway adjuncts, such as bag-valve-mask assisted respirations or oral airway placement, was also noted. Respiratory depression was defined as: an oxygen saturation of ⬍90%, a change from baseline ETCO2 of ⬎10 mm Hg, or airway obstruction with cessation of gas exchange at any time (noted by an absent ETCO2 waveform). After the procedure, the physician noted any complications, including, but not limited to: vomiting or aspiration, intubation, transfer to
251
a higher level of care after the procedure, hypotension (defined as a drop in systolic blood pressure ⬎20%), dysrhythmias, and whether or not the patient required assisted ventilations (by bag-valve mask [BVM]) due to decreased protective airway reflexes during the procedure. Physicians noted the level of sedation achieved during the procedure (minimal, moderate, deep, or general anesthesia). The end point of sedation was considered the time at which the patient had regained their baseline mental status. After the patient returned to their baseline mental status, patients were asked two “yes or no” questions: “Did you experience pain during the procedure?” and “Do you remember any of the procedure?” The physician who performed the procedure was then asked to complete a 100-mm VAS assessing perceived difficulty of the procedure. This VAS consisted of the question “How difficult was the procedure to perform?” and consisted of a 100-mm line with the word “easy” at one end of the line and “difficult” at the other end. The physician was asked to determine the level of sedation achieved as defined in Figure 1.
Data Analysis Data were collected by a designated research assistant during the procedure and were then entered into an EXCEL (Microsoft Corp., Redmond, WA) database for further analysis. All analysis and interpretation of data was done using STATA (STATA Corp., College Station, TX) statistical software. Descriptive statistics were used when appropriate. The rates of respiratory depression between patients in the moderate sedation and deep sedation group were compared with chi-squared tests. The mean BIS nadir between the moderate and deep sedation groups and the time to return of baseline mental status were compared with two-tailed t-tests. The presence or absence of reported pain or recall was compared between the two groups using Fisher’s exact tests. To detect an effect size of 0.44 in the achieved level of sedation between the groups using chi-squared tests, with an alpha of 0.05 and a beta of 0.2 (80% power), power analysis indicated that 28 patients per group were required.
RESULTS There were 77/135 eligible patients enrolled in the study. Two patients were excluded due to protocol violations (the procedure for which the patient was sedated was not a fracture or dislocation reduction). A total of 39 patients were randomized to the target of moderate sedation and 36
252
J. R. Miner et al.
Table 1. Results Summarized by Randomization Group
Initial BIS (95% CI) Initial systolic BP (mm Hg) (95% CI) Initial heart rate (95% CI) Initial ETCO2 (95% CI) Initial propofol dose (95% CI) Total number of doses given (95% CI) Total propofol dose Initial BIS (95% CI) BIS nadir (95% CI) Decrease in systolic blood pressure (95% CI) Time to baseline mental status Proc diff VAS Target sedation level achieved Respiratory depression
Moderate sedation (n ⫽ 39)
Deep sedation (n ⫽ 36)
95.3 (92.9–97.7) 144.8 (137.1–152.6) 87.0 (81.2–92.8) 37.5 (34.1–40.8) 0.88 mg/kg (0.83–0.93) 2.7 (2.1–3.3) 1.69 mg/kg (1.38–2.00) 95.3 (92.9–97.7) 67.8 (62.2–73.3) 11.4% (7.5–15.3)
96.3 (94.4–98.1) 140.6 (132.8–148.4) 85.0 (80.4–89.6) 42.4 (39.2–5.2) 0.98 mg/kg (0.88–1.09) 2.6 (2.1–3.1) 1.82 mg/kg (1.52–2.11) 96.3 (94.4–98.1) 59.9 (55.1–64.2) 9.3% (6.0–12.6)
8.0 min (5.9–10.5) 31.6 mm (19.3–43.9) 21/39 19/39
8.2 min (6.1–9.9) 30.6 mm (18.3–92.8) 25/36 18/36
Overall (n ⫽ 75) 95.8 (94.3–97.2) 142.8 (137.4–148.2) 86.0 (82.4–89.7) 39.8 (37.6–42.0) 0.93 (0.87–0.99) 2.6 (2.1–3.3) 1.75 (1.38–2.00) 95.8 (94.3–97.2) 64.0 (60.3–67.8) 10.4% (7.9–12.9) 8.1 min (6.6–9.5) 31.1 mm (22.7–39.5) 46/77 ( p ⫽ 0.40) 37/75 ( p ⫽ 0.92)
BIS ⫽ Bispectral index; BP ⫽ blood pressure; CI ⫽ confidence interval; ETCO2 ⫽ end tidal CO2; VAS ⫽ visual analog scale.
to the target of deep sedation. The procedures done included the reductions of 33 fractures and 42 dislocations. No significant adverse events were noted. Seventy-four of 75 (99%) procedures were successful; the one unsuccessful procedure was a reduction of a forearm fracture in a patient who had randomized to the goal of deep sedation. The baseline characteristics and results by the two randomized groups are summarized in Table 1. The mean initial dose of propofol was 0.88 mg/kg (95% confidence interval [CI] 0.83– 0.93) for the moderate sedation group and 0.98 mg/kg (95% CI 0.87–1.14) for the deep sedation group (p ⫽ 0.08). The mean total dose of propofol given over the entire procedure was 1.70 mg/kg (95% CI 1.38 – 2.00) for the moderate sedation group and 1.82 mg/kg (95% CI 1.53–2.12) for the deep sedation group (p ⫽ 0.54). By our pre-established criteria, respiratory depression was seen in 37/75 (49%) patients; 19/39 (49%) in patients assigned to receive moderate sedation and 18/36 (50%) in patients randomized to receive deep sedation (p ⫽ 0.92). All of the patients who met criteria for respiratory depression had a change from baseline ETCO2 ⬎10 mm Hg except for 1/39 (3%) in the moderate sedation assigned group and 1/36 (3%) in the deep sedation assigned group. These two patients had an absent ETCO2 waveform at some time during the procedure, without a change from baseline ETCO2 ⬎10 mm Hg noted. Among patients with a change in their ETCO2 ⬎10 mm Hg from baseline, the mean change was 21.1 mm Hg (95% CI 18.3–23.9). Five patients were noted to have an oxygen saturation ⬍90% at some point during the procedure; 2/39 (5%) in the moderate sedation assigned group and 3/36 (8%) in the deep sedation assigned group. The mean lowest oxygen saturation measured in these patients was 80.0 (95% CI 73.4 – 86.6). No patient had an absent ETCO2 waveform or an oxygen saturation ⬍90% for ⬎1 min.
No cardiac rhythm abnormalities were noted. The mean decrease in the systolic blood pressure was 9.0% in the deep sedation group (95% CI 6.0%–13.0%) and 11.0% (95% CI 7.0%–15.0%) in the moderate sedation group. Three patients required BVM respirations during the procedure: 2/39 (5%) in the moderate sedation assigned group and 1/36 (3%) in the deep sedation assigned group. None of these patients required BVM for more than 1 min. The patient in the deep sedation group had received a single bolus of 1.2 mg/kg before requiring BVM. The first patient in the moderate group received a single dose of propofol at 0.70 mg/kg before requiring BVM, the other had received an initial dose of 0.83 mg/kg followed by a repeat dose of the same size 2 min later before requiring BVM. The initial BIS score for the deep sedation assigned group was 96.3 (95% CI 94.4 –98.1) and for the moderate sedation assigned group was 95.3 (95% CI 92.9 –97.7) (p ⫽ 0.50). The mean minimum recorded BIS score was 67.7 (95% CI 62.2–73.3) for the moderate sedation assigned group and 59.9 (95% CI 55.1– 64.8) for the deep sedation assigned group (p ⫽ 0.03). Pain or recall was seen in 12/39 (31%) in the moderate sedation assigned group and 4/36 (11%) in the deep sedation assigned group (p ⫽ 0.04). The mean physician VAS for procedural difficulty was 34.0 mm (95% CI 23.7– 44.3) for the moderate sedation assigned group and 28.8 mm (95% CI 18.3–39.2) for the deep sedation assigned group (p ⫽ 0.72). The mean time to return of baseline mental status from the completion of the procedure was 7.97 min (95% CI 6.05–9.89) for the moderate sedation assigned group and 8.19 min (95% CI 5.87– 10.52) for the deep sedation assigned group (p ⫽ 0.88). All of the patients achieved either a moderate or deep level of sedation, as defined in Figure 1. However, only 25/36 (69%) of the patients randomized to a target of
Procedural Sedation
253
Table 2. Results Summarized by Level of Sedation Achieved
Initial propofol dose (95% CI) Total number of doses given (95% CI) Total propofol dose BIS nadir (95% CI) Change in systolic blood pressure (95% CI) Time to baseline mental status Proc diff VAS Respiratory depression
Patients achieving moderate sedation (n ⫽ 32)
Patients achieving deep sedation (n ⫽ 43)
Overall (n ⫽ 75)
0.92 mg/kg (0.81–1.04) 2.5 (1.9–3.3) 1.53 mg/kg (1.29–1.76) 78.4 (76.4–80.5) 10.6% (6.1–15.1) 7.4 min (5.8–9.1) 28.4 mm (22.9–44.6) 15/32 (46.9%)
0.93 mg/kg (0.88–0.99) 2.8 (2.3–3.3) 1.92 mg/kg (1.60–2.24) 53.7 (49.7–57.6) 10.2% (7.2–13.2) 8.6 min (6.3–11.9) 33.7 mm (19.3–37.4) 22/43 (51.2%)
0.93 (0.87–0.99) 2.6 (2.3–3.0) 1.75 (1.59–1.96) 64.0 (60.3–67.8) 10.4% (7.9–12.9) 8.1 min (6.7–9.6) 31.5 mm (24.3–38.6) 37/75 ( p ⫽ 0.71)
CI ⫽ confidence interval; BIS ⫽ Bispectral index; VAS ⫽ visual analog scale.
deep sedation achieved a deep level of sedation, and 21/39 (54%) of the patients randomized to a target of moderate sedation group actually achieved a moderate level of sedation (p ⫽ 0.40). The results of patients by achieved sedation level are summarized in Table 2. There were very few differences in the outcome of patients based on the level of sedation achieved. Respiratory depression was seen in 22/43 (51%) patients who achieved a deep level of sedation and 15/32 (47%) patients who achieved a moderate level of sedation (p ⫽ 0.71). There was no difference in the amount or number of doses of propofol given, the time to baseline mental status, the systolic blood pressure, or the perceived difficulty of the procedure. There was a large difference in the lowest BIS score achieved between the groups (p ⫽ 0.00). Six of 43 (14%) patients who achieved deep sedation and 10/32 (31%) patients who achieved a moderate level of sedation reported pain with or recall of the procedure for which they were sedated (p ⫽ 0.07).
DISCUSSION The assignment of a target level of moderate or deep procedural sedation did not seem to affect the actual level of sedation achieved, the rate of respiratory depression, or the amount of propofol given after the initial bolus. There was an increased rate of pain and recall in the moderate sedation goal group. Overall, the 75 procedures were performed without significant complications and, with the exception of one, all were successful. The patient with an unsuccessful procedure in the ED eventually underwent open reduction and internal fixation in the OR by orthopedics; the EP gave the procedure a VAS difficulty of 18 mm. This seems to indicate that the reduction itself was technically difficult rather than limited by the sedation target used. Evaluation of the dosing of the propofol shows a difference in the initial dose given to the patients who randomized to the moderate vs. the deep sedation group.
However, when the total amount of propofol given is compared, there is no difference between patients in either group. No specific dosing recommendations were given to EPs for use in this study, just a target sedation level. The doses of sedative given are similar to those described in similar ED procedural sedation studies (7,10 –13). Because individual patients respond differently to the same dose of medication (14), medications used for sedation are generally titrated until the patient seems adequately sedated. There was no difference in the rate of respiratory depression detected between the two groups. We have used these criteria for detecting respiratory depression in previous studies (6 – 8). They are not intended to detect clinically significant respiratory depression, but rather any changes, including sub-clinical changes, that indicate a decrease in respiratory effort. Changes in ETCO2 can detect airway obstruction (which is indicated by a decrease in ETCO2 as the exhaled sample mixes with ambient air), hypoventilation (which is indicated by rising ETCO2 as pCO2 rises), or apnea (which is indicated by a loss of the waveform). In fact, in this study, the oxygen saturation monitor did not detect any respiratory depression that was not detected by the ETCO2 monitor, but the ETCO2 monitor suggested respiratory depression that was not predicted by the change in oxygen saturation. Three patients in our study required BVM respirations. All three had received doses of propofol near the middle of the range for this study. We did not have specific criteria for the use of BVM, and its use was at the discretion of the treating physician, so it is possible that these results are due to the specific practices of individual EPs rather than actual loss of protective airway reflexes in these patients. The fact that these patients all received typical doses of propofol illustrates the need for close airway monitoring in sedated patients, whether they are being moderately or deeply sedated, and no matter what the target level of sedation is before the procedure is started or the dose of sedative given. The decrease in systolic blood pressure noted in our study has been previously described for the propofol
254
doses given (9,13). There was no difference in blood pressure changes between the groups noted, which is not surprising because the total dose of propofol given in the two groups was not different. The time to return to baseline mental status was not significantly different between the groups. This is not surprising because both groups had the same mean number of propofol doses and the same total dose of medication given, and is likely related to the half life of propofol. This does, however, raise the question of why there is a difference in the BIS nadir and the frequency of recall between the groups. Only 69% of the patients in the deep sedation target group and 54% of the patients in the moderate sedation target group were sedated to their targeted level of sedation. Overall, 43 patients achieved deep sedation and 32 patients achieved moderate sedation. The target level of sedation a patient was randomized to was not associated with the level achieved. This may be due to a number of factors. First, the EPs may have been unable to predictably and consistently achieve a specific level of sedation. Alternatively, it may be possible that a patient randomized to the deep sedation group achieved sedation adequate to tolerate the procedure before actually reaching deep sedation. The physician may then have decided to not give further medications, and the patient with the target of deep sedation was therefore only moderately sedated. By the same logic, a patient who randomized to the moderate sedation group but was not sufficiently sedated to undergo the procedure may have been given an additional bolus of propofol, which brought them to the level of deep sedation. Because the initial dose of medicine given is lower in patients who randomized to the moderate sedation group, we assume that the physician’s behavior was altered by entering patients into one target group over another. When patients were compared by the level of sedation actually achieved, there were no differences among them except in the lowest BIS score recorded and a trend toward a difference in the rate of pain or recall with the procedure. Interestingly, there was no difference in the amount of propofol, either in the initial bolus or the total amount given, between patients who achieved either moderate or deep level of sedation. This indicates that the distinction of moderate or deep procedural sedation may not be associated with the outcome parameters we measured in this study and which are routinely used in clinical practice. It is possible that by defining sedation criteria by markers of respiratory depression instead of markers more specific to the goals of procedural sedation (i.e., decreasing the patient’s awareness or the ability to form memories), we are unnecessarily approaching sedation levels associated with respiratory depression.
J. R. Miner et al.
The sedation level achieved (moderate or deep), the number of doses of medicine given and total dose of propofol, the rate of respiratory depression, and the time to return of baseline mental status do not seem to be influenced by the assignment of the patient to a pre-procedural target sedation level of moderate or deep procedural sedation. The risk of complications, such as respiratory depression or the inability to perform the procedure, was unchanged by the assignment of a pre-procedure target sedation level. Assigning a pre-procedural target sedation level of either moderate or deep does not seem to be an effective way of influencing the outcome of ED procedural sedation, except for patient recall.
LIMITATIONS The independent variable of this study was the treating physician’s behavior, which is difficult to control. We did not recommend doses but simply asked that they achieve a target sedation level. We chose this method because this is equivalent to any external control on a sedation level. Because we have no way to measure subtle changes in physician behavior, we cannot determine with this design what aspects of this control affected the outcomes of the sedation. We believe, however, that this design simulates clinical practice, where Emergency Physicians are sometimes limited to moderate rather than deep sedation, and the results are applicable to the clinical application of procedural sedation. The number of patients in this study accounts for the detection of only a rather large difference in the rates of respiratory depression (the study is powered to detect an effect size by chi-square of 0.44). A much larger study is necessary to detect subtle differences in the rates of respiratory depression, and it is likely one exists. However, due to the nature of our current respiratory depression detection method, we feel that the interpretation of a small detected difference between these groups may be difficult.
CONCLUSIONS In this study, the assignment of a pre-procedural target sedation level of moderate or deep PS did not influence the level of sedation achieved, the rate of respiratory depression, the occurrence of complications, the time to return of baseline mental status, or the success of the procedure. It does not seem that the assignment of a pre-procedural target sedation level is an effective means of changing most outcomes of ED PS.
Procedural Sedation
255
REFERENCES 1. Practice guidelines for sedation and analgesia by nonanesthesiologists. An updated report by the American Society of Anesthesiologists Task Force on Sedation and Analgesia by Nonanesthesiologists. Anesthesiology 2002;96:1004 –17. 2. Avramov MN, White PF. Use of alfentanil and propofol for outpatient monitored anesthesia care: determining the optimal dosing regimen. Anesth Analg 1997;85:566 –72. 3. Pratila MG, Fischer ME, Alagesan R, et al. Propofol versus midazolam for monitored sedation: a comparison of intraoperative and recovery parameters. J Clin Anesth 1993;5:268 –74. 4. Pitetti RD, Singh S, Pierce MC. Safe and efficacious use of procedural sedation and analgesia by nonanesthesiologists in a pediatric emergency department. Arch Pediatr Adolesc Med 2003; 157:1090 – 6. 5. Pena BM, Krauss B. Adverse events of procedural sedation and analgesia in a pediatric emergency department. Ann Emerg Med 1999;34:483–91. 6. Miner JR, Biros MH, Heegaard W, Plummer D. Bispectral electroencephalographic analysis of patients undergoing procedural sedation in the emergency department. Acad Emerg Med 2003;10: 638 – 43. 7. Miner JR, Biros M, Krieg S, Johnson C, Heegaard W, Plummer D. Randomized clinical trial of propofol versus methohexital for
8. 9.
10. 11. 12. 13.
14.
procedural sedation during fracture and dislocation reduction in the emergency department. Acad Emerg Med 2003;10:931–7. Miner JR, Heegaard W, Plummer D. End-tidal carbon dioxide monitoring during procedural sedation. Acad Emerg Med 2002;9: 275– 80. Hasan RA, Shayevitz JR, Patel V. Deep sedation with propofol for children undergoing ambulatory magnetic resonance imaging of the brain: experience from a pediatric intensive care unit. Pediatr Crit Care Med 2003;4:454 – 8. Coll-Vinent B, Sala X, Fernandez C, et al. Sedation for cardioversion in the emergency department: analysis of effectiveness in four protocols. Ann Emerg Med 2003;42:767–72. Bassett KE, Anderson JL, Pribble CG, Guenther E. Propofol for procedural sedation in children in the emergency department. Ann Emerg Med 2003;42:773– 82. Havel CJ Jr, Strait RT, Hennes H. A clinical trial of propofol vs midazolam for procedural sedation in a pediatric emergency department. Acad Emerg Med 1999;6:989 –97. Guenther E, Pribble CG, Junkins EP Jr, Kadish HA, Bassett KE, Nelson DS. Propofol sedation by emergency physicians for elective pediatric outpatient procedures. Ann Emerg Med 2003;42: 783–91. Glass PS, Bloom M, Kearse L, Rosow C, Sebel P, Manberg P. Bispectral analysis measures sedation and memory effects of propofol, midazolam, isoflurane, and alfentanil in healthy volunteers. Anesthesiology 1997;86:836 – 47.