Article Text
PDF
Statistics from Altmetric.com
- Conscious sedation
- emergency medicine
- dislocation
- hypnotics and sedatives
- propofol
- anaesthesia
- analgesia/pain control
- musculo-skeletal, fractures and dislocations
Introduction
Closed reduction in a dislocated shoulder is painful and requires both analgesia and a degree of muscle ‘relaxation.’ Non-anaesthetic physicians in the Emergency Department often combine an opiate and benzodiazepine to provide these effects. In our institution, we use morphine and midazolam. Although this combination usually gives satisfactory conditions, these agents act for a long time compared with more recently developed drugs such as remifentanil and propofol. Prolonged sedation can affect staff deployment in the emergency department.1
Remifentanil and propofol are commonly combined for total intravenous anaesthesia and are used as conscious sedation for painful procedures. They have a rapid onset of action and with small doses, rapid recovery. Remifentanil has an ester link that is quickly destroyed in blood and tissues, and has an elimination half-life of approximately 10 min.2 Propofol has an initial distribution half-life of 2–4 min.3
Morphine and midazolam are commonly used for procedural sedation. They have slow onset and offset times. Intravenous morphine has an onset time of 5.4–13 min4 and mean elimination half life of 1.4–3.4 h.5 Although intravenous midazolam has an onset time of under 2 min, the half life is 60–90 min.6
Propofol and remifentanil are synergistic pharmacodynamically7 and pharmacokinetically.8 We showed that a combination of remifentanil and propofol was effective in providing analgesia and sedation for relocation of a dislocated shoulder.9 We wished to compare the recovery time of these newer agents (R&P), with our current ED sedation regimen (M&M) to see if the use of remifentanil and propofol could allow quicker recovery, reduce the duration of nursing supervision required and allow early discharge. We considered that if the time to full recovery was reduced by 15 min, this would be clinically important.
Methods
Ethics
We obtained ethical permission from our local research ethics committee to compare recovery from sedation with remifentanil and propofol with recovery from our standard regimen of morphine and midazolam, using a randomised controlled trial. We also assessed operating conditions and patient comfort. Although the study was not designed to assess safety, we recorded all adverse events.
Recruitment
We recruited 40 patients with shoulder dislocation between the ages of 16 and 65 presenting to a large university hospital emergency department at times when suitably trained staff were present between 18 February 2004 and 26 May 2005. We considered patients with anterior glenohumeral dislocation including those with avulsion fracture of the greater tuberosity or of the glenoid labrum. Patients provided informed written consent before entering the study. We excluded patients with American Society of Anesthesiologists (ASA) physical status greater than II, those with more major fracture dislocations, other major injuries, posterior shoulder dislocations, a history of drug abuse, and those who were intoxicated.
Treatment allocation
Treatment was allocated in equal proportions (20 per group) to either morphine and midazolam or remifentanil and propofol by sequentially selecting sealed envelopes that had been shuffled and numbered.
Operators
Sedation was not by anaesthetists but by suitably trained emergency department doctors. Suitable training was at least 3 months' training in anaesthesia, and successful assessment for practical competency in the sedative technique by a consultant anaesthetist. Our standard procedural sedation protocol is for two doctors and one nurse to be present. One doctor (the operator) performs the procedure, one doctor administers the sedation, and the emergency department nurse provides assistance as necessary.
Dosing regime
We used a specific dosing regimen for each form of sedation. Remifentanil was prepared as a 10 μg/ml solution by reconstituting 1 mg of remifentanil in 100 ml of 0.9% w/v sodium chloride. Ten millilitres of this was drawn into a 10 ml syringe and labelled. Propofol was prepared as a 5 mg/ml solution by mixing 10 ml of propofol 1% with 2 ml of lidocaine 1% and 8 ml of 0.9% w/v sodium chloride to obtain a total syringe volume of 20 ml, and labelled. The initial propofol dose was 0.5 mg/kg (ie, 0.1 ml/kg of the syringe volume), and the initial remifentanil dose was 0.5 μg/kg (ie, 0.05 ml/kg of the syringe volume). The intravenous cannula was flushed with 0.9% w/v sodium chloride after each dose. If further sedation was required, subsequent propofol doses were 0.25 mg/kg, and subsequent remifentanil doses were 0.5 μg/kg. A maximum dose was not set for propofol and remifentanil. The operator and doctor giving the sedation waited until they felt the previous dose had ‘worn off’ before giving further doses.
Morphine and midazolam were each prepared at a concentration of 1 mg/ml (10 mg drug in 10 ml 0.9% w/v sodium chloride) and labelled. Morphine was given in 2.5 mg increments intravenously at 3 min intervals until either analgesia was achieved or a total dose of 0.15 mg/kg had been given. Midazolam was then given in 1 mg increments intravenously at 3 min intervals until sedation was adequate for the procedure, or either 10 mg or 0.15 mg/kg had been given. The intravenous cannula was flushed with 0.9% w/v sodium chloride after each dose.
Monitoring
The doctor administered the sedation and the nurse monitored the patient. We continuously measured the patient's pulse rate and Spo2 throughout the sedation period by pulse oximeter. Automated cuff blood pressure was assessed every 2 min (Propaq CS Model 242, Skaneateles Falls, New York).
Preoxygenation
The patients to be given propofol and remifentanil were first given oxygen from a tight-fitting mask, reservoir bag and spill valve (Mapleson C circuit, Intersurgical, Wokingham, UK) supplied with oxygen at 6 l/min for 3 min. This technique prevents hypocapnia, with the aim of avoiding apnoea. Departmental practice does not routinely require oxygen therapy for patients given morphine and midazolam sedation, as this hinders the detection of respiratory depression. If oxygen saturation decreased during monitoring, oxygen was given immediately to restore the SpO2.
Treatment failure was not specifically defined. If several attempts were not successful, the operator and doctor administering sedation jointly decided to stop further attempts at shoulder reduction.
Data collection and methods of assessment
We used a specifically designed form to record the study data. The operator assessed the operating conditions for manual reduction using a five-point ordinal scale of impossible, poor, adequate, good or excellent. We assessed sedation using the Observer Assessment of Alertness/Sedation Score (table 1)10 at 15 min intervals after the first dosing had finished.
Observer Assessment of Alertness/Sedation Score scoring system
After the procedure, the patient was asked to rate pain or discomfort during the procedure on a four-point scale of 0—nil, 1—mild, 2—moderate and 3—severe. Adverse events including oxygen desaturation, apnoea, hypotension or other symptoms such as nausea, fainting or dizziness were recorded.
Using expected ratios for recovery time of 0.5 for remifentanil and propofol and 1.0 for morphine and midazolam, allocation ratio of 1:1 and continuity correction for small sample size, we determined that a sample size of 18 per group was required for a power of 0.9 and α of 0.05 (double-sided analysis)
Discharge criteria
We followed our usual departmental practice to assess patients for discharge. They had to be alert and orientated, walk normally without assistance and eat and drink prior to discharge. They had to have accompanied transport home.
Analysis of results
Our primary outcome measure was time to full recovery. Since these times are often not normally distributed, we compared the median time to full recovery using the Mann–Whitney U test. Although the study was not primarily designed to compare operating conditions or patient discomfort, we compared these assessments with the Wilcoxon–Mann–Whitney test for ordered categories and present the data as distribution histograms.
Results
Recruitment of 40 suitable patients required 15 months. In that time, 407 patients were seen with shoulder dislocation, one of whom had already taken part in the study. The reasons for exclusion are shown in table 2.
Patient recruitment and exclusions
The characteristics of the patients included in the study are shown in table 3. There were no evident systematic differences between the patients randomised to receive the two treatments.
Patient characteristics
The drug doses administered are shown in table 4.
Drug dosages
Two patients allocated to receive remifentanil and propofol, and one patient allocated to receive morphine and midazolam had already been given morphine for pain control. Two patients (one M&M, one R&P) had been given nalbuphine by the ambulance service prior to admission.
Figure 1 represents the salient results of the study. All subjects given remifentanil and propofol were completely recovered after 30 min, in contrast to the patients given morphine and midazolam, where 17 out of 20 had recovered completely after 1 h. The median recovery times were 15 min (95% CI 15 to 20) and 45 min (95% CI 29 to 48), respectively.
Time to full recovery.
There were no significant differences in the reduction conditions (figure 2) or the scores for pain or distress, between the groups (figure 3). Reduction was not obtained in seven patients with the sedation provided by the study regimen and further medication was required (3 M&M, 4 R&P).
Reduction conditions.
Patient experience.
Adverse effects
Two patients given midazolam required flumazenil within 10 min of shoulder reduction to counter oversedation. One of these patients had been given additional morphine (total 20 mg), as shoulder reduction proved impossible within the morphine dose limit set by the trial. No other adverse events were recorded.
Discussion
We have shown that these two sedation techniques, used in relatively fit patients, provide equivalent treatment conditions. Remifentanil and propofol gave the additional advantage that all patients were ready for discharge after 30 min. At that time, only half of the patients receiving standard treatment were fit for discharge. It took a further 60 min for the remaining morphine and midazolam patients to recover fully. Our findings support and extend those of Taylor and colleagues, who used propofol alone to reduce shoulder dislocation.11
Remifentanil is now widely used in the UK for anaesthesia and sedation. The plasma breakdown confers a short half-life, and emergence is rapid even after prolonged infusion. Little has been published on remifentanil use in the Emergency Department. We considered that the features of remifentanil could provide profound analgesia and quick recovery in short but painful procedures such as relocation of a dislocated shoulder.
Is this an important positive outcome, and what are the possible dangers of the new technique? We were required, by our ethics committee and by our management, to conduct this study only if an additional appropriately trained doctor was present to administer the agents and monitor the patients. We used up-to-date monitors and methods that were designed to ensure safety. A primary safety feature was partial rebreathing preoxygenation carried out by a separate doctor who administered the sedation. This increases safety in three ways:
Even in fit young volunteers, sedatives and opioids can cause apnoea.12 Patients who are in pain and worried hyperventilate and become hypocapnic, so apnoea might occur when propofol is given.13 By administering oxygen through a tight-fitting face mask-reservoir bag system with the oxygen flow adjusted to allow some rebreathing, the effects of hyperventilation on carbon dioxide levels are prevented and the likelihood of apnoea diminished.14
By having a separate doctor to sedate and observe the patient, it was possible to detect apnoea early and remind the patient to breathe if necessary.
Giving supplemental oxygen ensures that the lung contains sufficient oxygen to sustain satisfactory oxygen saturation, despite apnoea, for longer than if only air were breathed.15
We consider that preoxygenation with partial rebreathing as described is an important safety feature of this technique.
Can this new technique using remifentanil and propofol be used to allow patients to leave the emergency department more promptly after shoulder reduction, freeing nursing staff for other tasks, who would otherwise have to monitor sedated patients?1 We measured sedation using a widely used and validated scale that is designed for clinical assessment of degrees of sedation.10 However, this scale was not designed to assess adequacy of recovery, so even when scoring a maximum on this scale, patients may still be impaired. This is more likely with midazolam, whose action persists far longer than propofol. We suggest, however, that the difference in time to recovery when using remifentanil and propofol is clinically significant.
The use of propofol for sedation continues to provoke discussion. Some support its use,16 while others emphasise the unpredictability of this agent and its capacity to induce anaesthesia when only sedation is intended.17 It is relevant that 21% of monitored anaesthesia care malpractice cases on the ASA closed claims database from 1990 to 2006 related to respiratory depression after giving sedative or opioid drugs.18 We calculated our doses on the basis of a study of patients of similar physical status undergoing endoscopy19 and our pilot study,9 and diluted the agent to reduce its rate of onset.20 Nevertheless, previous studies show that some patients may become unresponsive after even these doses, and recommendations from bodies in the UK21 and USA17 are that during such sedation, trained anaesthesia personnel should be present to ‘rescue patients whose level of sedation becomes deeper than initially intended.’ Other suggestions are even stronger: ‘(Propofol) should be administered only by persons trained in the administration of general anaesthesia, who are not simultaneously involved in these surgical or diagnostic procedures.’22 This restriction is concordant with specific language in the propofol package insert. We considered that the ED doctors who performed this technique were suitably trained and experienced in the administration of general anaesthesia to allow them to safely perform this sedation. We define suitably trained as, ‘emergency department doctors who have had at least a 3-month secondment for training in anaesthesia and have been assessed for practical competency in this technique by a consultant anaesthetist.’ This would normally occur in the UK during training for the award of CCT in Emergency Medicine.
Propofol is often used for sedation by non-anaesthetists23 and programmes have been developed to train non-anaesthesiologists in its use.24 We agree that anaesthesia support should be available when propofol infusions are used for sedation,24 but consider that the anaesthesia training provided for our ED doctors should allow safe sedation with remifentanil and propofol using the technique we describe. We suggest that this study describes an effective sedation technique for non-anaesthetists, provided they are suitably experienced and equipped.
The safety of the method described in this paper can only be measured in a much larger study. We intend to implement this, using the same method of partial rebreathing preoxygenation to reduce the risk of apnoea if consciousness is lost. Reduced dosage will be necessary for older people.19
In summary, this study shows that remifentanil and propofol for sedation provide earlier patient recovery, compared with morphine and midazolam, while maintaining equivalent treatment conditions. It also describes an effective preoxygenation and sedation technique for suitably experienced and equipped non-anaesthetic doctors to use remifentanil and propofol to allow reduction of anterior shoulder dislocation.
References
Footnotes
Competing interests None.
Patient consent Obtained.
Ethics approval Ethics approval was provided by the Lothian Research Ethics Committee.
Provenance and peer review Not commissioned; externally peer reviewed.