Appropriate airway management is an essential part of the anaethetist's role. Difficult intubation, which can now be quantified using the ‘Intubation Difficulty Scale’, should be anticipated whenever possible. A strategy needs to be developed in order to anticipate problems. The first part of this paper reviews the different factors that contribute to make intubation and/or ventilation difficult. Problems with intubation (or ventilation of the lungs) can be caused by abnormal laryngeal structures (e.g. tumour, stenosis), or by difficulty in seeing the glottis. The clinical history will usually help identify the former problem, while physical examination of the airway is required to reveal either disproportion between the various structures of the airway (e.g. tongue, larynx), and/or difficulties in aligning the oral, pharyngeal, and laryngeal axes. The different techniques used to diagnose these problems are described. The second part of this paper summarizes the algorithms used by the anaesthetist when management of the airway is found difficult. Three situations are considered: (a) anticipated difficult intubation, for which awake fibreoptic intubation would appear to be the technique of choice in the majority of cases, (b) unforeseen difficult intubation in a patient whose lungs can be ventilated; here, various techniques for control of the airway will be briefly described, and (c) both tracheal intubation and lung ventilation are impossible; this is a life-threatening emergency, for which three solutions are proposed. These include use of the laryngeal mask airway, the Combi-tube®, or transtracheal ventilation. These situations will be analysed with the aim of proposing management strategies that always guarantee the safety of the patient.

Background and aim We investigated the haemody-namic stability and emergence characteristics of isoflur-ane/nitrous oxide anaesthesia supplemented with remifentanil or fentanyl in patients undergoing carotid endarterectomy.

Methods Anaesthesia was induced with propofol (1–2mgkg−1) and either remifentanil (O.5μgkg−1) or fentanyl (1 μgkg−1), followed by an infusion of remifentanil (0.2 μg kg−1 min−1) or fentanyl (2 μg kg−1 h−1).

Results There were no significant differences between the groups in haemodynamic variables, postoperative pain, nausea or vomiting. After induction there was a significant decrease in mean arterial pressure for both groups (P<0.001) and a decrease in heart rate (P=0.001) in the remifentanil group. In both groups these haemodynamic changes continued during maintenance of anaesthesia (P<0.05). The time to eye opening after surgery was significantly shorter with remifentanil compared with fentanyl (6.62±3.89 vs. 18.0±15.18min, P=0.015).

Conclusion Remifentanil appears to be a comparable opioid to fentanyl when supplementing isoflurane/nitrous oxide anaesthesia for carotid endarterectomy.

Background and aim We evaluated the costs and benefits of total intravenous anaesthesia compared with a balanced anaesthesia regimen.

Methods One-hundred and twenty-four patients undergoing cataract surgery were randomized to either a pro-pofol/remifentanil or an isoflurane/fentanyl group. In the propofol/remifentanil group, both drugs were used for induction and maintenance of anaesthesia; in the isoflurane/fentanyl group, anaesthesia was induced with etomi-date and fentanyl and maintained with isoflurane and fentanyl. All patients received mivacurium for muscle relaxation and the lungs were ventilated mechanically. The use of propofol and remifentanil resulted in a faster emergence and an overall savings per case of € 12.25 due to a reduction in personnel costs which outweighs the higher drug acquisition costs.

Results In the propofol and remifentanil group, more patients were satisfied and would accept the same anaesthetic again.

Conclusion We conclude that propofol and remifentanil is more cost-effective than isoflurane/fentanyl due to its better recovery profile, reduced total direct costs and higher patient satisfaction.

Background and aim Infusion of ketamine and midazolam can maintain prolonged anaesthesia, but delayed recovery is a limitation. We aimed to develop an approximation regimen for the infusion of ketamine and midazolam to obtain steady-state anaesthesia with acceptable recovery.

Methods Thirty-one patients undergoing radical cystect-omy were studied. The initial regimen was calculated from drug pharmacokinetic variables and tailored in a pilot study (15 patients) to develop the approximation regimen dosage. Anaesthesia was induced with midazolam (150μgkg−1) and ketamine (2mgkg−1). Tracheal intubation and ventilation with oxygen enriched air (FiO2 = 0.35) were facilitated by muscle relaxants. Anaesthesia was maintained by the approximation regimen doses. Routine monitoring was used for all patients, but pulmonary artery catheters were inserted in 11 patients, to obtain haemo-dynamic and oxygenation variables.

Results Steady-state anaesthesia was obtained with minimal deviations in the regimen in some patients followed by reasonable recovery.

Conclusion It is concluded that infusion of ketamine and midazolam in the approximation regimen doses can be used to maintain anaesthesia for prolonged abdominal surgery.

Background and aim The effects of analgesic drugs on sleep are poorly understood. We investigated short- and medium-term effects of tramadol on sleep structure.

Methods Eight healthy volunteers received a placebo (predrug placebo-night), then, in a randomized, double-blind, cross-over fashion a single oral dose of tramadol 50 mg or 100mg (drug-night), and finally, again a placebo (postdrug placebo-night). Standardized polysomnography (electroencephalogram, electro-oculogram, submental electromyogram) was continuously recorded during placebo- and drug-nights

Results During drug-nights both doses of tramadol significantly increased the duration of stage 2 sleep, and significantly decreased the duration of slow-wave sleep (stage 4). Tramadol 100mg but not 50 mg significantly decreased the duration of paradoxical (rapid eye movement) sleep. In the placebo-night after tramadol 100mg (but not after 50 mg) duration of stage 2 sleep was significantly shorter, and duration of stage 4 sleep was significantly longer compared with the predrug placebo-night. Conclusion In healthy volunteers, a single dose of tramadol 50 mg disturbs sleep in the night of drug application. With 100mg, sleep is disturbed in both the night of drug application and in the subsequent night.

Background and aim Compound A generation and accumulation in sevoflurane anaesthesia is dependent on fresh gas flow. We investigated the extent of generation of compound A.

Methods After Institutional Review Board approval and informed consent, patients with normal renal function were randomized to receive either sevoflurane (n = 33) or isoflurane (n = 43) minimal flow anaesthesia (0.5 L min−1) for at least 2 h under standardized conditions. Compound A concentrations were quantified and blood and urine samples were taken to assess renal involvement. Both groups were comparable.

Results No significant differences concerning blood chemistry and urine measurements were found. The maximum mean compound A concentration was observed 90 min after flow reduction being 40±9p.p.m. at a corresponding mean sevoflurane concentration of 2.1 ±0.5 vol%. Mean inspiratory compound A exposure was 102±l33p.p.mh−1.

Conclusion Compound A concentrations using 0.5 L min−1 fresh gas flow and a heated absorber were higher than previously published values using an inflow of 1 Lmin−1. Compound A exposure was similar to other clinical studies which did not show changes in renal and hepatic function.

Background and aim This study assessed the accuracy of using capnography with a modified, hollow gum elastic bougie in predicting tracheal intubation, and its effectiveness as a method of apnoeic oxygenation.

Methods Patients were randomly allocated to having the gum elastic bougie inserted, under anaesthesia, in the trachea or the oesophagus. End-tidal carbon dioxide measurements were made at 10 and 20 s. The position of the gum elastic bougie was correctly predicted in 89.2% of patients. We tested the apnoeic oxygenation on an anaes-hetic simulator model, which is housed in the Scottish Anaesthesia Simulator Centre, Stirling, UK.

Results The time taken for the oxygen saturation to fall to 90% was significantly prolonged when the gum elastic bougie was used for apnoeic oxygenation.

Conclusion The modification of the gum elastic bougie allows a more objective assessment of correct placement than the previous tactile method. The current design of bougie is unsuitable but can be modified.

Background and aim We evaluated the effect of the injection technique on the onset time and efficacy of interscalene brachial plexus anaesthesia.

Methods With Ethical Committee approval and written consent, 30 patients undergoing elective shoulder acromioplasty or capsuloplasty were randomly allocated to receive interscalene brachial plexus block with 20 mL of ropivacaine 0.75% by using either a single injection (Single group, n = 15) or multiple injection (Multiple group, n = 15). Nerve blocks were placed with the aid of a nerve stimulator using short bevelled, Teflon® coated needles. The stimulation frequency was set at 2 Hz and the intensity of stimulating current, initially set at 1 mA, was gradually decreased to ≤0.5 mA after each muscular twitch was observed. In the Single group, the anaesthetic solution was slowly injected after the first muscular twitch had been observed. In the Multiple group, 6mL were injected at shoulder abduction, 6mL were njected at arm flexion, and 6ml_ at the extension of the arm.

Results Placing the block required 5min (4–8 min) in the Multiple group and 3 min (1–10 min) in the Single group (P=0.001); however, total preoperative time (from skin disinfection to complete loss of pinprick sensation from C4 to C7 with inability to elevate the limb from the operating table) was shorter in the Multiple group (15 min; range 10–28 min) than in the Single group (23 min; range 14–60 min) (P=0.03). Additional intravenous fentanyl supplementation was required in two patients of the Multiple group (13%) and eight patients of the Single group (53%) (P=0.05).

Conclusion We conclude that using a multiple injection technique shortened the preparation time and improved the quality of interscalene brachial plexus anaesthesia performed with small volumes of ropivacaine 0.75%.

Background and aim This study determines whether epidural fentanyl given before incision decreases the requirements for opioid analgesia postoperatively, compared with the same dose of epidural fentanyl given after the surgery.

Methods Forty patients scheduled to undergo elective abdominal surgery were randomly allocated between two groups according to the time of administered of fentanyl. None of the patients in either group received premedica-tion. Prior to induction of general anaesthesia an epidural catheter was inserted at the L2–3 interspace and flushed with 0.9% NaCI. Patients then received 100μg fentanyl in 10mL 0.9% NaCI through this catheter either 15min before awaking at the end of the operation (group I), or else the same dose given at an estimated time of 15 min before the start of surgery (group II). Postoperative analgesia consisted of patient-controlled intravenous fentanyl. The amount of fentanyl used by the patients was noted at 2, 4, 8, 12 and 24 h after surgery. Pain scores and sedation scores were assessed at 0, 2, 4, 8, 12 and 24 h postoperatively.

Results The consumption of fentanyl was similar in both groups in all studied periods postoperatively. The mean pain score was lower for patients in group I than group II immediately after operation. There were no statistically significant differences between the mean pain scores of groups at 2, 4, 8, 12 and 24 h after operation. Mean sedation scores were similar in both groups at all times postoperatively.

Conclusion This study showed that the dose of fentanyl administered epidurally prior to surgical incision did not produce any clinically useful pre-emptive analgesic effect.