Dexmedetomidine is an alpha 2-adrenoreceptor agonist, which provides sedation, analgesia and anxiolysis in clinical practice (Cortinez et al., 2004,Hall et al., 2000). Three types of alpha 2-adrenergic receptor subtypes are found in the human body and they have been designated alpha 2A, alpha 2B and alpha 2C. The alpha 2A subtype is most likely responsible for the analgesic properties of dexmedetomidine in both peripheral and central sites (Kingery et al., 2000, Smith et al., 2001). Activation of central alpha 2-adrenoreceptors in the locus ceruleus (Correa-Sales et al., 1992) and the dorsal horn of the spinal cord (Gaumann et al., 1992b) are responsible for both analgesic and sedative effects. Dexmedetomidine has a very high alpha 2 to alpha 1 selectivity, 1620 to 1, or approximately 8 times that of clonidine. It is also 4 to 6 times more potent than clonidine by weight (Bhana et al., 2000).

Although dexmedetomidine produces dose dependent sedation upon intravenous administration, its the analgesic effect is of dexmedetomidine is more variable and controversial. In an ischaemic pain model in healthy volunteers, a single bolus of dexmedetomidine produced a 50% reduction in pain scores when compared to placebo (Jaakola et al., 1991). In another volunteer study using the cold pressor test, dexmedetomidine 1 µg/kg over 10 minutes followed by an infusion of 0.2 to 0.6 µg/kg/hour reduced pain by approximately 30% (Hall et al., 2000). However, when administered as a target controlled infusion at concentrations ranging from 0.09 to 1.23 ng/mL, dexmedetomidine had no analgesic effect in human volunteers subjected to heat and electrical pain, although sedation was produced (Memis et al., 2004).

Clonidine and dexmedetomidine are two common alpha 2 agonists used clinically. Although clonidine former has been used successfully in regional analgesia and anesthesia (Gabriel et al., 2001)., There are only very few studies evaluating the peripheral analgesic effects of dexmedetomidine. Since acute postoperative dental pain is a common analgesia model (Cooper, 1991; US Food and Drug Administration 1992), the investigators conducted this study, aiming to assess the postoperative analgesic efficacy of peripheral dexmedetomidine after bilateral third molar surgery under general anaesthesia. The analgesic effects were compared up to the 72nd hour postoperatively in order to evaluate any potential preventive analgesic effect.

The study protocol was approved by our local Institutional Review Board and written consent was obtained from all the participants. Eligibility for recruitment included American Society of Anesthesiologists (ASA) physical status I and II, age between 18 and 50 years of age with 4 bilateral impacted third molar teeth scheduled for extraction under general anaesthesia. Exclusion criteria included clinical history or electrocardiographic evidence of heart block, ischaemic heart disease, asthma, sleep apnoea syndrome, impaired liver or renal function, alcohol consumption in excess of 28 units per week, pregnancy, patient refusal, known psychiatric illness, chronic sedative or analgesic use, and regular use of or known allergy to dexmedetomidine, paracetamol or dextropropoxypheneopioids. Patients with preoperative inflammation at the site of surgery were also excluded.

After obtaining written, informed consent, patients were randomly allocated to receive 1)Preoperative IV dexmedetomidine 1mcg/kg and normal saline infiltrated to wound at the end of surgery; 2) Preoperative normal saline IV infusion and dexmedetomidine 1mcg/kg infiltrated to wound or 3) Preoperative IV normal saline infusion and normal saline infiltrated to wound at the end of surgery. A computer generated random sequence was used for drug allocation and this was prepared by a statistician who was unaware of the clinical nature of the study. The patients were assessed by the list anaesthetist the day before surgery. They were fasted for at least 6 hours and did not receive any premedication before arrival at the operation theatre. Patients were educated on the use of the Numerical Rating Scale (NRS) for pain assessment, where zero corresponds to no pain and 10 represents the worst pain imaginable.

A P-deletion test was performed in the ward preoperatively. The P-deletionThis test consists of 58 lines of closely typed text containing between 74 to 97 occurrences of the letter "p". The patients are instructed to cross out every letter "p" they can find as quickly as possible, reading from left to right down the page. The number deleted in 180 seconds and the number of errors counts. This test assesses speed, accuracy, concentration, mental activity, hand-eye coordination and vigilance. It was firstly described by Dixon and colleagues who found it to be an accurate test for psychomotor recovery from anesthesia and sedation (Dixon et al., 1973).

On arrival at the operating theatre, a 20-gauge intravenous cannula was inserted in the dorsum of each patient's left hand. Either dexmedetomidine 1 µg/kg (Group D) diluted to 4ml with normal saline or the same volume of 0.9% saline alone (Group P) was prepared by another anaesthetist who did not participate in patient management or data collection. Both preparations were clear solutions, therefore patients, medical, nursing staff, and data collectors were all blinded to the allocated drug. Heart rate, blood pressure, respiratory rate and oxygen saturation were recorded (S/5 Anesthesia Monitor, Datex-Ohmeda, WI, USA) before general anaesthesia and thereafter at five-minute intervals from the time of commencing surgery until discharge from the recovery room.

Induction and maintenance of anaesthesia were done according to a standard protocol. Anaesthesia was induced with propofol titration of 2 to 3 mg/kg, fentanyl 1 µg/kg and muscle relaxation was achieved with rocuronium 1 mg/kg. The trachea was intubated via the nasal route. General anaesthesia was maintained with a mixture of air, oxygen and sevoflurane at 1 to 1.5 MAC. Before the start of the surgical procedures, regional anaesthesia were performed by infiltrating 2.7 mL of 2% lignocaine with 1 in 80,000 adrenaline around the base of the gum of each third molar by the same team of dental surgeons. Sevoflurane was stopped when the last suture had been inserted. andVolume of 1 ml of either dexmedetomidine (Group D, totally dexmedetomidine 1 µg/kg for 4 surgical sites) or saline (Group P) was infiltrated to the submucosa of each surgical site. Neuromuscular block was antagonized with neostigmine 50 µg/kg and atropine 20 µg/kg, and followed by extubation after recovery of consciousness. The patients were then monitored in the recovery room for at least 30 minutes. Resting pain scores, blood pressure, heart rate, oxygen saturation and the Ramsay sedation score (RSS) (appendix 1) were assessed and recorded every 5 minutes. Intravenous ketorolac or morphine would be administered upon patients' request. Patients were transferred back to the general ward when fully consciousness with normal physiological parameters.

The P-deletion test was repeated hourly from the first postoperative hour in the general ward until the performance reached the same level as that before surgery. Blood pressure, heart rate, oxygen saturation, and RSS were monitored hourly for 4 hours. NRS pain scores were charted hourly for 6 hours then 4-hourly thereafter. Patients were prescribed two analgesic tablets, each containing paracetamol 320 mg and dextropropoxyphene 32.5 mg (Dolpocetmol®, Synco Limited, Hong Kong, China), on an as required basis to a maximum of four times daily. The patients were explained told that the pain medications prescribed could be taken if the postoperative NRS pain score was more than 3.

Final hospital discharge was at the discretion of the attending dental surgeon. The oral analgesic medication described above was prescribed for three days postoperatively and patients were given aA diary was given to our patients. diary to record NRS pain scores at rest and upon mouth opening, analgesic consumption and side effects were self recorded by each patient up to the 72nd hour postoperatively. Global pain satisfaction using NRS (zero being least satisfied and 10 being most satisfied) was recorded on postoperative day 3.

By considering a difference of 2 units in NRS pain scores to be clinically significant, and an estimated standard deviation of 2.5 units among patients, we calculated that a sample size of 30 per group would give an 80% power of the test at the 5% level of significance. In order to account for possible dropouts due to protocol violation or missing data, 35 patients were recruited in each group. Statistical analysis was performed using SPSS 14.0 for Windows (SPSS Inc., IL, USA). NRS pain scores, analgesic consumption, NRS global pain satisfaction and the difference in pre- and post-operative P-deletion scores were compared using the Mann-Whitney U test. Perioperative vital signs and NRS pain scores up to the 72nd postoperative 72nd hour were plotted graphically using GraphPad Prism 4.03 (GraphPad Software Inc., CA, USA) and the mean area under the curve AUC were compared between groups using the Mann Whitney U test. All categorical data were analyzed using χ2 test. Time to first analgesic use was compared using the log-rank test in Kaplan-Meier survival analysis. The incidence of side effects was analyzed by χ2 test or Fisher's exact test as appropriate.

Inclusion Criteria:

• American Society of Anesthesiologists (ASA) physical status I and II
• Age between 18 and 50 years of age with 4 bilateral impacted third molar teeth scheduled for extraction under general anaesthesia

Exclusion criteria:

• Clinical history or electrocardiographic evidence of heart block
• Ischaemic heart disease
• Asthma
• Sleep apnoea syndrome
• Impaired liver or renal function
• Alcohol consumption in excess of 28 units per week
• Pregnancy
• Patient refusal
• Known psychiatric illness
• Chronic sedative or analgesic use, and regular use of or known allergy to dexmedetomidine, paracetamol or dextropropoxypheneopioids
• Patients with preoperative inflammation at the site of surgery were also excluded

1 day after surgery