The efficacy of an integrated autotransfusion regimen, including pre-donation and perioperative salvage of autologous blood, was prospectively evaluated in 2884 patients undergoing total hip (n=2016) or knee arthroplasty (n = 480), and hip revision (n = 388) with either balanced general, regional, or integrated epidural/general anaesthesia. Allogenic concentrated red blood cells were transfused in the presence of symptomatic anaemia or when haemoglobin concentration was < 6gdL−1 (10gdL−1 in patients affected by cerebrovascular or coronary artery disease) after all salvaged and pre-donated autologous blood had been transfused. A total of 278 patients (9.6%) received allogenic blood. Risk factors for allogenic blood transfusion were: preoperative haemoglobin concentration < 10gdL−1 (after autologous blood pre-donations) (Odds ratio: 8.7; 95% CI: 6.5–16.8; P=0.004), hip revision versus hip or knee arthroplasty (Odds ratio: 5.8; 95% CI: 3.9–8.5; P=0.0001) and inability in obtaining the number of predonations required by the Maximum Surgery Blood Order on Schedule (Odds ratio: 3.4; 95% CI: 2.7–4.1; P=0.0001). The incidence of perioperative complications, including wound infection and haematoma, as well as myocardial ischaemia, respiratory failure and thromboembolic complications, was higher in those patients requiring allogenic blood transfusion (29.8%) than that observed in patients receiving only autologous blood (6.6%) (P=0.0005); while the mean time duration from surgical procedure to patient discharge from the orthopaedic ward was shorter in those patients not receiving allogenic blood transfusion (12 days; 25–75th percentiles: 8–14 days) than in those patients who required perioperative transfusion with allogenic blood (15 days; 25–75th percentiles: 10–17 days) (P=0.0005). In conclusion, this prospective study highlighted the clinical relevance of applying an extensive and integrated autotransfusion regimen in order to reduce allogenic blood transfusion and associated complications in patients undergoing major joint replacement.

Anaesthetic agents have been implicated in the development of postoperative pneumonia, but the direct effect of volatile anaesthetics on bacterial growth has given contradictory results. The effects of isoflurane on the growth of Staphylococcus aureus and Escherichia coli were investigated under conditions similar to those of clinical practice, using standardized microbiological methods. An open anaesthetic circuit was adjusted to the normal ventilatory settings of an adult patient. A spray, installed on the inspiratory side of the circuit, ensured the delivery of isoflurane at 1.5 minimal alveolar concentration. The bacterial strains studied were both wild-type and reference strains. Bacterial inoculums were prepared to obtain a bacterial exponential growth of 103 colony-forming units per mL in 10mL of nutritive broth. Each strain was studied with and without exposure to isoflurane and measured by the usual criteria of bacterial growth, and by bacterial regrowth after exposure to isoflurane. Under experimental conditions close to clinical practice, exposure to isoflurane did not alter bacterial growth of S aureus and E. coli, or their bacterial regrowth when isoflurane exposure is over.

To evaluate the effects of adding low-dose fentanyl to 0.75% ropivacaine during peripheral nerve blocks, 30 ASA physical status I–II patients undergoing hallux valgus repair under combined sciatic–femoral nerve block were randomly allocated in a double-blind fashion to receive nerve block placement with 30 mL of either 0.75% ropivacaine alone (group: ropivacaine, n=15) or 0.75% ropivacaine plus fentanyl 1 μg kg−1 (group: ropivacaine–fentanyl, n=15). A blinded observer recorded haemodynamic variables and sedation, as well as the time required to achieve surgical block and the first request for analgesia. Readiness to surgery required 10 min (5–20 min) with 0.75% ropivacaine and 10 min (3–20 min) with the ropivacaine–fentanyl mixture. No differences in the degree of sedation, peripheral oxygen saturation, and haemodynamic variables were observed between the two groups. The degree of pain measured at first analgesic request, and the consumption of postoperative analgesics, was similar in the two groups, while the mean time from block placement to the first request for pain medication was 13.7 h (25–75th percentiles: 11.8–14.5 h) in the ropivacaine group and 13.9 h (25–75th percentiles: 10.5–14.5 h) in the ropivacaine–fentanyl group (P=not significant). We conclude that adding fentanyl 1 μg kg−1 to 0.75% ropivacaine did not provide clinically relevant advantages in terms of onset time, quality and duration of combined sciatic–femoral nerve block in patients undergoing elective hallux valgus repair.

Recently, it was suggested that peripherally-mediated analgesia can be accomplished by the intra-articular delivery of the mu-opioid morphine or of the a2-agonist clonidine. This clinical study assesses the potential peripheral analgesic effect of the combination of morphine and clonidine after intra-articular administration. Sixty patients (American Society of Anesthesiologists status I or II) undergoing arthroscopic repair of the knee during general anaesthesia were randomized to receive after operation, in a double-blind manner, either 1 μg morphine intra-articularly (group 1); 150 μg clonidine intra-articularly (group 2); or 1 μg morphine + 150 μg clonidine intra-articularly (group 3); or normal saline intra-articularly (group 4) in a volume of 30 mL, respectively. Visual analogue pain scores (VAS), duration of analgesia as defined by first demand for supplemental analgesics, subsequent 24 h consumption of postoperative supplementary supplementary analgesics, and patient satisfaction were evaluated. Co-administration of morphine + clonidine (group 3) resulted in a significant VAS reduction at 2 h after injection compared with the other groups. There was a tendency towards a lower need for supplementary rescue analgesia and towards a more prolonged analgesia in group 3 (211 min ± 224 min SD) compared with group 1 (173 min ± 197 min SD) and group 4 (91 min ± 21 min SD). More patients were very satisfied with the postoperative analgesic regimen receiving the combination of morphine and clonidine (group 3) at 24 h postoperatively. Thus we conclude, that the peripheral co-delivery of an opioid and an a2-agonist will result in improved postoperative pain relief, when compared with each single agent given alone.

For supportive therapy in sepsis, an adequate volume loading is likely the first step, and possibly the most important step in the treatment of patients with septic shock. However, it is still a matter of debate what kind of fluids, endpoints for resuscitation and monitoring techniques should be used. Nevertheless, therapies which closely control volume loading and regional perfusion are becoming more utilized. An elevated global O2-supply (DO2) may be necessary and beneficial in most of these patients but the increase of DO2 should be guided by the measurement of parameters assessing global and regional oxygenation. Routine strategies for elevating DO2 by the use of very high dosages of catecholamines cannot be recommended. Vasopressors should be used to achieve an adequate perfusion pressure. With norepinephrine, no negative effects on regional perfusion have been demonstrated when the patient is adequately volume resuscitated and the DO2 is normal or even slightly elevated. In contrast, epinephrine should be avoided because it seems to redistribute blood flow away from the splanchnic region. There is controversy whether dopamine should still be used as a first-line drug in patients with septic shock, because there are some clinical and experimental data that indicate unfavourable effects on mucosal perfusion of the gut.

To date, there are no convincing data to support the routine use of low-dose dopamine or dopexamine in patients with sepsis. Neither low-dose dopamine nor dopexamine have been proven to prevent renal failure in septic patients. Furthermore, there is evidence that low-dose dopamine may reduce mucosal perfusion in the gut in some patients. Dopexamine has been suggested to improve splanchnic perfusion but because these effects remain somewhat controversial, there is as yet no reason for a general recommendation for dopexamine in septic patients.

These recommendations are currently limited by the lack of sufficient outcome studies and studies which evaluate regional perfusion. Until the various catecholamine regimes are more fully examined, recommendations for catecholamine support in sepsis must be considered ‘conditional’.

The aim of this study was to elucidate the influence of four neuromuscular blocking substances on coronary vascular tone using the model of isolated porcine coronary artery segments. We studied the effects of four muscle relaxants, atracurium, pancuronium, rocuronium, and vecuronium (0.1, 1, and 10 μg mL−1 each), on the contractile response to three vasoconstrictors: acetylcholine, histamine, and serotonin. None of the neuromuscular blocking agents under investigation exerted a significant influence on the vasoconstricting effects of these mediators except for pancuronium, which dose-dependently attenuated acetylcholine-mediated contractions (−10.8% attenuation for 10 μg mL−1 pancuronium, P < 0.05). There was no difference between vessels with intact endothelium and denuded preparations. It is concluded that high-dose pancuronium exerts an antimuscarinic effect in vascular smooth muscle. The other neuromuscular agents studied do not alter vascular reactivity of isolated porcine coronary arteries.

Environmental monitoring of anaesthetic gases was carried out in theatre areas in eight hospitals as well as in Delivery suites, radiology, radiotherapy and Dental suites. High staff exposures occurred in nontheatre areas although exposures in theatres were generally satisfactory. Environmental control measures are required where staff exposures exceed legal standards.

There is a risk of regurgitation if a laryngeal mask airway is used with intermittent positive pressure ventilation. The lower the pressure gradient between airway and stomach, the less will be the likelihood of stomach insufflation with gas and consequent regurgitation. A flow generator ventilator, such as the Penlon Nuffield 200 which achieves a peak pressure at the end of inspiration at any given tidal volume, is likely to produce a higher peak airway pressure than a pressure generator. Replacing the expiratory valve by a venturi gives a preferable flow pattern with an inspiratory pressure plateau. At equal tidal volumes, the venturi modification provided a 10% lower inflation pressure (standard deviation 6%, P < 0.01) and a 1.8% smaller leak (P < 0.05) past the laryngeal mask airway. The same comparison but using identical inflation pressures, gave 8% larger tidal volumes (standard deviation 5.1%, P < 0.01) and a 2% smaller leak (P < 0.01) past the laryngeal mask airway.

In anaesthetized dogs, stellate ganglion blockade led to a moderate disturbance in left ventricular diastolic function. We investigated the effect of a left-sided block, following injection of 10 mL bupivacaine 0.5%, on echocardiographic variables of ventricular function in eight otherwise healthy patients with sympathetically mediated pain syndromes. After the blockade, heart rate (control: 66±3 (mean± SEM), block: 64± 3 min−1) and mean arterial blood pressure (88.5±6.0 vs. 84.0±8.1 mmHg) were unchanged, but afterload decreased (end-systolic meridional wall stress; 69.6±9.9 vs. 59.8±7.1, P<0.05). Stroke volume increased from 71.2±8.1 to 79.6±7.4 mL, P <0.05. Variables of systolic function were unchanged, but relaxation was prolonged (isovolumic relaxation time; 71±5 vs. 81±4 ms, P<0.05). In patients who were ASA I, there was a small impairment in echocardiographic variables during ventricular relaxation after a left stellate ganglion blockade. This small effect did not compromise ventricular function, and the heart responded with a small stroke volume increase to the simultaneous afterload reduction.

Epidural analgesia is the most effective and innocuous technique for obstetrics. Pain relief is its main indication but maternal diseases that might be decompensated by labour and delivery are also accepted indications. Low doses of long-acting local anaesthetics alone or in combination with low doses of fentanyl or sufentanil provide good quality analgesia and are safe for mother and fetus. Test doses in parturients lack sufficient specificity and sensitivity for detecting inadvertent intravascular injection, and subarachnoid migration of the catheter is possible at any time during the procedure. Therefore, every injection must be considered as a test dose and only fractionated injections must be made. Epidural block to T10 is needed for labour and to level T4 for Caesarean section. Maintenance of the block with a continuous infusion, or patient-controlled epidural analgesia with a background continuous infusion, provides more stable analgesia than by intermittent injection. Technical difficulties, dural tap, bloody tap, hypotension and insufficient block are most frequent complications of epidural block in obstetrics. Excessive motor block prolongs the second stage of labour and increases the frequency for instrumental delivery and is therefore considered a complication

We conducted a prospective, randomized, double-blind clinical trial comparing droperidol 1.25 mg intravenously (i.v.) (group 1, n = 30), ondansetron 4 mg i.v. (group 2, n = 30), or both (group 3, n = 30) in the prevention of postoperative nausea and vomiting (PONV) in the first 24 h following major gynaecological procedures under combined general and epidural anaesthesia. PONV was analysed by a linear nausea/vomiting score, incidence of nausea and vomiting, and the need for antiemetic rescue. Our results showed a similar incidence of nausea and vomiting in all groups (G1 33%, G2 40%, G3 43%). However, when comparisons were made according to the time of assessment, combination therapy resulted in significantly lower PONV than droperidol in the first hour (0% vs. 13%, P < 0.05) and second hour (0% vs. 13%, P < 0.05), and than ondansetron on the first hour (0% vs. 13%, P < 0.05). A trend persisted up to the fourth hour but was not statistically significant in either group. In conclusion, droperidol and ondansetron are effective agents in the prevention of PONV, and their combination seems to provide slightly better results than either drug alone.

Respiratory support with high-frequency jet ventilation has been advocated during airway surgery or to improve pulmonary mechanics and gas exchange in patients with bronchopleural fistulae or pulmonary insufficiency. Despite a large body of published evidence describing its benefits as an alternative ventilatory approach in anaesthesia and intensive care medicine, its application has not gained widespread acceptance and is restricted to specialized centres. To review the literature on high-frequency jet ventilation in European and North American institutions, we performed a search in a computerized database (MEDLINE) covering the period from 1990 until the present time, describing the use of high-frequency jet ventilation in over 7000 patients. Various modes to apply high-frequency jet ventilation during airway surgery have been established, but its value in intensive care is controversial. We report our experience with combined high-frequency jet ventilation and provide guidelines for its safe application.

The elimination of procalcitonin and the course of plasma concentrations during continuous veno-venous haemodiafiltration were measured in patients with sepsis or multiple organ dysfunction syndrome, because these patients are a main target group for the measurement of procalcitonin and often require renal replacement therapy. Procalcitonin was measured in the prefilter plasma and the filtrate at 5 min, 15 min and 1, 2, 4, 6, 12, 24 h after set-up of continuous veno-venous haemodiafiltration. In a prospective study, 19 patients with plasma levels of procalcitonin >3ng mL−1 and acute oliguric renal failure treated with continuous veno-venous haemodiafiltration using a polysulphone membrane, were evaluated for the study of clearance. Twenty-one control patients (procalcitonin <2ng mL−1) were studied to determine whether filtration itself induced a procalcitonin response. No interventions were required. In patients with low procalcitonin concentrations (procalcitonin <2ng mL−1) continuous veno-venous haemodiafltration did not cause a rise in procalcitonin. In patients with increased procalcitonin plasma concentrations (>3ng mL−1), the protein was removed through the polysulphone membrane, with a final clearance of 4 mL min−1 after the initial adsorption period (clearance 0.4–0.9mL min−1 during the first hour of continuous veno-venous haemodiafiltration). Thus, on the average, approximately 10% of plasma concentrations were measurable in the filtrate ultimately. However, procalcitonin plasma levels were not significantly altered during continuous veno-venous haemodiafiltration (86% of the initial concentration after 24 h). Although procalcitonin is removed from the plasma during continuous veno-venous haemodiafiltration in measurable amounts plasma procalcitonin concentrations did not change significantly during haemodiafiltration. Procalcitonin thus can also be used as a diagnostic parameter in patients undergoing continuous veno-venous haemodiafiltration.

It is assumed that diabetic patients with uraemia have more complications at renal transplantation than those who are not diabetic. We compared the preoperative ECGs, and invasive perioperative haemodynamic and oxygenation parameters in 15 diabetic and 15 non-diabetic uraemic patients undergoing renal transplantation. The number of patients with increased QT dispersion in the ECG was higher in diabetic than in non-diabetic patients (P < 0.05). Before anaesthesia, heart rate and mean arterial pressure were higher (P <0.05) in the diabetic than in the non-diabetic group. After preanaesthetic volume loading all patients showed a hyperdynamic circulation, which subsided during anaesthesia. However, stroke volume index remained unchanged. Four patients in the diabetic group and six in the non-diabetic group needed additional oxygen therapy after surgery. No cardiac dysrhythmias were noted. However, the increased QT dispersion in diabetic patients calls for an adequate perioperative ECG monitoring for dysrhythmias. The diabetic and non-diabetic uraemic patients performed equally well at renal transplantation. In conclusion, renal transplantation for diabetics is justified.

It is hypothesized that tissue dysoxia and O2 debt are major factors in the development and the propagation of multiple organ failure in critically ill patients. Dysoxia is the result of an abnormal relationship between O2 supply (DO2) and O2 demand and translates into increased anaerobic metabolism and tissue and blood lactate concentration. First-line therapeutic strategies used to avoid the development of an O2 debt involve correction of cardiac output, haemoglobin, and O2 saturation in order to increase DO2 above its critical value. They are not sufficient, however, to ensure appropriate end-organ perfusion and oxygenation. The adequacy of cardiac output towards tissue metabolic requirements may be appreciated by venous-to-arterial and gut mucosal-to-arterial PCO2 differences. This review details these strategies and discusses their usefulness in current practice.

The effects of propofol and thiopental on three vasoconstrictors, acetylcholine, histamine, and serotonin were studied in isolated porcine and human coronary artery rings. Propofol and thiopental attenuated the contractile response to all mediators in a dose-dependent manner. This dilating effect was fairly weak using low concentrations (propofol 1 μg mL−1 and 10 μg mL−1, thiopental 5 μg mL−1 and 10 μg mL−1). In the presence of high concentrations (propofol 100 μg mL−1, thiopental 50 μg mL−1) marked relaxation was observed (propofol −32% up to −49%, P < 0,05; thiopental −23% up to −67%, P < 0.05). These dilating effects were seen both in intact and denuded rings, the differences were not significant. Human coronary artery segments were relaxed by thiopental (−22% to −76%) and propofol (−11% to −67%) to a similar extent. Our data indicate that propofol and thiopental relax isolated coronary segments in a dose-dependent manner, and that there is no evidence that these effects are dependent of endothelial factors.

The retrospective analysis of 1700 forensic autopsies over 17 years (1981–97) following unexpected sudden cardiac death revealed a group of 50 cases that could have been related to surgery and/or anaesthesia. Patients were young with no history of cardiac disease. Surgery was performed for uncomplicated disorders, all classified as ASA 1. Cardiac arrest took place at induction of anaesthesia in 16% of cases, during surgery in 64% and at the end of surgery in 20%. Investigation and expertise reports ordered by the public prosecutor revealed none of the typical causes of death usually associated with surgery or anaesthesia. Pathological examination showed cardiac lesions in 47 cases: arrhythmogenic right ventricular cardiomyopathy in 18 cases, coronary artery disease in 10 cases, cardiomyopathy in eight cases, structural abnormalities of the His bundle in nine cases, mitral valve prolapse in one case, and acute myocarditis in one case. Identification of the cause of death of patients at low risk may provide major relief to the family of the patient and the medical staff.

The concentrations of nitrous oxide, sevoflurane and oxygen in the circle system of a closed-circuit anaesthesia machine, the PhysioFlex™, were measured in seven patients. During anaesthesia, the settings for each gas were changed and their concentrations recorded. At the induction of anaesthesia, it took 80– 510s (median 190 s) for the end-tidal sevoflurane concentration to reach 2.0%, and 920–2640 s (median 1500 s) for the oxygen in the breathing circuit to reach 30%. At this time, the nitrous oxide concentration was 60±3% (mean±SD). During anaesthesia, it took 90–480 s (median 140 s) for the end-tidal sevoflurane concentration setting to decrease from 3.0 to 1.0%, and 90–400 s (median 110 s) to return from 1.0 to 3.0%. When the inspired oxygen was increased from 30 to 50%, circuit concentrations reached equilibrium in 40–60 s (median 40 s), and when decreased from 50% back to 30%, equilibrium took 310–470 s (median 450 s). During recovery from anaesthesia, inspiratory sevoflurane concentration took 40–70 s (median 50 s) to decrease to 0.2%. The PhysioFlex™ provided adequate control of sevoflurane and oxygen concentrations, but not of increasing nitrous oxide concentrations.

A small decrease in cerebral blood flow (approximately 10%) in response to 100% oxygen (O2) administration is well recognized. This observation was based on human volunteer studies, which employed a nitrous oxide washout method for the measurement of cerebral blood flow. Because this method is now appreciated to be subject to potential errors we have examined the cerebral blood flow response to 100% oxygen using a magnetic resonance imaging technique to quantify changes in carotid and basilar artery flow. The study, was performed in 12 normal male subjects aged 23–42 years. We report decreases in cerebral blood flow ranging from 9 to 31% with a mean value of over 20%. The decrease in cerebral blood flow was greater in seven young subjects (aged 23–26 years) with decreases in cerebral blood flow of 19.3–31.4% (mean 26.8%). In five older subjects (aged 32–42 years), decreases in CBF were smaller (mean 16.2%). The administration of 100% O2 was accompanied by a small decrease in end-tidal CO2 (3.7–7.1%), insufficient to explain the changes in cerebral blood flow. We conclude that the decrease in cerebral blood flow in response to O2 administration is greater than previously described and appears to be greater in young adults.