Epidural needle insertion : A large registry analysis.

BACKGROUND: Dural puncture, paraesthesia and vascular puncture are the most common complications of epidural catheter insertion. Their association with variation in midline needle insertion depth is unknown. OBJECTIVE: This study evaluated the risk of dural and vascular punctures and the unwanted events paraesthesia and multiple skin punctures related to midline needle insertion depth. MATERIAL AND METHODS: A total of 14,503 epidural catheter insertions including lumbar (L1-L5; n = 5367), low thoracic (T7-T12, n = 8234) and upper thoracic (T1-T6, n = 902) insertions, were extracted from the German Network for Regional Anaesthesia registry between 2007 and 2015. The primary outcomes were compared with logistic regression and adjusted (adj) for confounders to determine the risk of complications/events. Results are presented as odds ratios (OR, [95% confidence interval]). MAIN RESULTS: Midline insertion depth depended on body mass index, sex, and spinal level. After adjusting for confounders increased puncture depth (cm) remained an independent risk factor for vascular puncture (adjOR 1.27 [1.09-1.47], p = 0.002) and multiple skin punctures (adjOR 1.25 [1.21-1.29], p < 0.001). In contrast, dural punctures occurred at significantly shallower depths (adjOR 0.73 [0.60-0.89], p = 0.002). Paraesthesia was unrelated to insertion depth. Body mass index and sex had no influence on paraesthesia, dural and vascular punctures. Thoracic epidural insertion was associated with a lower risk of vascular puncture than at lumbar sites (adjOR 0.39 [0.18-0.84], p = 0.02). CONCLUSION: Variation in midline insertion depth is an independent risk factor for epidural complications; however, variability precludes use of depth as a reliable guide to insertion in individual patients.

[Neuraxial anaesthesia and NOACs]. / Rückenmarknahe Anästhesie und NOAK.

BACKGROUND: Cardiovascular comorbidities in surgical patients are frequent and have a substantial impact on the postoperative outcome. Neuraxial blockades are able to reduce perioperative morbidity and mortality. The increasing use of new oral anticoagulants (NOAC) requires a high level of attention, especially in patients undergoing neuraxial blockades or requiring postoperative analgesia. OBJECTIVE: The goal of this article is to present the benefit of neuraxial anaesthesia and analgesia in patients with cardiovascular risks and perioperative management of NOAC in this setting. MATERIALS AND METHODS: Review of the respective literature in PubMed during the last 25 years as well as presentation of the S1 guideline "Neuraxial anaesthesia and thrombo-embolic prophylaxis/antithrombotic medication" of the German Society of Anaesthesiology and Intensive Care Medicine (DGAI). RESULTS: Thoracic epidural anaesthesia and analgesia contribute to an improved outcome in surgical patients with high cardiovascular risk. In order to avoid severe complications in patients on NOACs undergoing neuraxial blockades the S1 guideline of the DGAI must be respected and close interdisciplinary consultations between anaesthetist, cardiologist and surgeon are mandatory. CONCLUSION: In consideration of the respective guideline neuraxial blockades can be performed in cardiovascular risk patients on NOACs, since these techniques contribute to an improved postoperative outcome.

Tunnelling of thoracic epidural catheters is associated with fewer catheter-related infections: a retrospective registry analysis.

BACKGROUND: Catheter-related infections are a serious complication of continuous thoracic epidural analgesia. Tunnelling catheters subcutaneously may reduce infection risk. We thus tested the hypothesis that tunnelling of thoracic epidural catheters is associated with a lower risk of catheter-related infections. METHODS: Twenty-two thousand, four hundred and eleven surgical patients with continuous thoracic epidural analgesia included in the German Network for Regional Anaesthesia registry between 2007 and 2014 were grouped by whether their catheters were tunnelled (n=12 870) or not (n=9541). Catheter-related infections in each group were compared with Student's unpaired t and χ(2) tests. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated with logistic regression, adjusting for potential confounding factors, including age, ASA physical status score, use of catheter for ≥4 days, multiple skin puncture, hospital, and surgical department. RESULTS: There were fewer catheter-related infections in patients with tunnelled catheters (4.5 vs 5.5%, P<0.001). Mild infections were also less common (4.0 vs 4.6%, P=0.009), as were moderate infections (0.4 vs 0.8%, P<0.001). After adjustment for potential confounding factors, tunnelling remained an independent prevention for any grade of infection (adjusted OR 0.51, 95% CI 0.42-0.61, P<0.001) and for mild infections (adjusted OR 0.54, 95% CI 0.43-0.66, P<0.001) and moderate and severe infections (adjusted OR 0.44, 95% CI 0.28-0.70, P=0.001). CONCLUSION: Tunnelling was associated with a lower risk of thoracic epidural catheter-related infections.

[Anesthesia in obstetrics: Tried and trusted methods, current standards and new challenges]. / Anästhesie in der Geburtshilfe: Altbewährtes, gegenwärtige Standards und neue Herausforderungen.

Obstetric analgesia and anesthesia have some specific aspects, which in particular are directly related to pathophysiological alterations during pregnancy and also to the circumstance that two or even more individuals are always affected by complications or therapeutic measures. This review article deals with some evergreens and hot topics of obstetric anesthesia and essential new knowledge on these aspects is described. The article summarizes the talks given at the 16th symposium on obstetric anesthesia organized by the Scientific Committee for Regional Anaesthesia and Obstetric Anaesthesia within the German Society of Anaesthesiology. The topics are in particular, special features and pitfalls of informed consent in the delivery room, challenges in education and training in obstetric anesthesia, expedient inclusion of simulation-assisted training and further education on risk minimization, knowledge and recommendations on fasting for the delivery room and cesarean sections, monitoring in obstetric anesthesia by neuraxial and alternative procedures, the possibilities and limitations of using ultrasound for lumbal epidural catheter positioning in the delivery room, recommended approaches in preparing peridural catheters for cesarean section, basic principles of cardiotocography, postoperative analgesia after cesarean section, the practice of early bonding in the delivery room during cesarean section births and the management of postpartum hemorrhage.

Obesity in regional anesthesia--a risk factor for peripheral catheter-related infections.

BACKGROUND: Obesity is believed to increase the risk of surgical site infections and possibly increase the risk of catheter-related infections in regional anesthesia. We, therefore, analyzed the influence of obesity on catheter-related infections defined within a national registry for regional anesthesia. METHODS: The German Network for Regional Anesthesia database with 25 participating clinical centers was analyzed between 2007 and 2012. Exactly, 28,249 cases (13,239 peripheral nerve and 15,010 neuraxial blocks) of patients ≥ 14 years were grouped in I: underweight (BMI 13.2-18.49 kg/m(2) , n = 597), II: normal weight (BMI 18.5-24.9 kg/m(2) , n = 9272), III: overweight (BMI 25.0-29.9 kg/m(2) , n = 10,632), and IV: obese (BMI 30.0-70.3 kg/m(2) , n = 7,744). The analysis focused on peripheral and neuraxial catheter-related infections. Differences between the groups were tested with non-parametric ANOVA and chi-square (P < 0.05). Binary logistic regression was used to compare obese, overweight, or underweight patients with normal weight patients. Odds ratios (OR and 95% confidence interval) were calculated and adjusted for potential confounders. RESULTS: Confounders with significant influence on the risk for catheter-related infections were gender, age, ASA score, diabetes, preoperative infection, multiple skin puncture, and prolonged catheter use. The incidence (normal weight: 2.1%, obese: 3.6%; P < 0.001) and the risk of peripheral catheter-related infection was increased in obese compared to normal weight patients [adjusted OR: 1.69 (1.25-2.28); P < 0.001]. In neuraxial sites, the incidence of catheter-related infections differed significantly between normal weight and obese patients (normal weight: 3.2%, obese: 2.3%; P = 0.01), whereas the risk was comparable [adjusted OR: 0.95 (0.71-1.28); P = 0.92]. CONCLUSION: This retrospective cohort study suggests that obesity is an independent risk factor for peripheral, but not neuraxial, catheter-related infections.

Administration of bovine polymerized haemoglobin before and during coronary occlusion reduces infarct size in rabbits.

BACKGROUND: Haemoglobin-based oxygen carriers (HBOC) seem to increase the risk of mortality and myocardial infarction in clinical trials. Therefore, we designed this randomized placebo-controlled animal study to evaluate the effects of prophylactic and therapeutic administration of HBOC in a myocardial ischaemia-reperfusion model with respect to infarct size and areas of impaired perfusion (no reflow, NR). METHODS: Thirty-two anaesthetized, mechanically ventilated rabbits were randomized to one of the four groups. Group G1 received 0.4 g kg(-1) i.v. HBOC-200 25 min before coronary artery occlusion, G2 received the same dose i.v. 10 min after occlusion, and G3 and 4 received i.v. saline. G1, 2, and 3 were subjected to 30 min occlusion of left coronary artery followed by 240 min of reperfusion. G4 was treated without ischaemia-reperfusion. Measurement included assessment of the area at risk and infarct size using triphenyltetrazolium chloride stain and areas of NR using thioflavin stain. Ischaemia-reperfusion was confirmed by microspheres technique. RESULTS: Infarct size as a percentage of the area at risk was significantly reduced in G1 [25 (sd 13)%, P=0.026] and G2 [22 (20)%, P=0.009] compared with G3 [48 (17)%]. The areas of NR in percentage of the area at risk [G1, 26 (15)%; G2, 34 (22)%; G3, 36 (12)%; G4, 5 (3)%] did not differ between the groups of animals undergoing coronary occlusion and reperfusion. CONCLUSIONS: Prophylactic and therapeutic administration of HBOC-200 reduces infarct size in myocardial ischaemia and reperfusion in rabbits. This reduction of infarct size is not accompanied by an improvement of areas of NR.

Effects of a prophylactic or therapeutic application of perflubron emulsion on myocardial ischaemia-reperfusion injury in rats.

BACKGROUND AND OBJECTIVE: The efficacy of administering a perfluorochemical-based oxygen therapeutic such as perflubron emulsion (Oxygen) prior to ischaemia is currently unknown, although there is evidence for potential beneficial effects for the perioperative treatment in cardiac risk patients. This experimental study investigated the efficacy of perflubron emulsion in preventing reperfusion injury and myocardial infarction size after coronary ischaemia and reperfusion. The perflubron emulsion was given either in a prophylactic manner, prior to induction of myocardial ischaemia, or as a therapeutic agent given during ischaemia. METHODS: Thirty-two anaesthetized and mechanically ventilated rats were subjected to 25 min occlusion of the left coronary artery followed by 120 min reperfusion. Animals were randomized to one of four groups:Group 1 was treated with administration of 6 g kg (-1) intravenous perflubron emulsion 25 min before occlusion; Group 2 received the same dose 10 min after occlusion; and Groups 3 and 4 received no perflubron emulsion. Inspired O2 (FiO2) concentration was maintained at 1.0 in Groups 1, 2 and 3 and at 0.35 in Group 4. RESULTS: Neither prophylactic nor therapeutic perflubron emulsion treatment reduced infarct size measurements by triphenyltetrazolium-chloride staining or severity of cardiac arrhythmias in comparison to the hyperoxic control group. However, prophylactic application of perflubron emulsion reduced areas of impaired perfusion vs. Group 3 assessed by in vivo staining with Thioflavin-S while no significant effect was seen in Groups 2 and 4 vs. 3. Density of DNA single-strand breaks in the ventricle was increased in all groups ventilated with 100% oxygen. CONCLUSION: Although administration of perflubron emulsion did not reduce infarct size, areas of impaired perfusion were significantly mitigated when perflubron emulsion was administered prior to coronary occlusion. However, a high oxygen concentration may provoke DNA strand breaks during reperfusion after ischaemia. Further studies must clarify whether enhanced oxidative stress outweighs the advantage of improved areas of impaired perfusion following perflubron emulsion.

Levobupivacaine for epidural anaesthesia and postoperative analgesia in hip surgery: a multi-center efficacy and safety equivalence study with bupivacaine and ropivacaine.

BACKGROUND AND OBJECTIVES: The aim of this randomized, single blind phase IIIb study was to evaluate the efficacy of 0.5% levobupivacaine versus 0.5% bupivacaine and 0.75% ropivacaine administered as epidural anesthesia and 0.125% levobupivacaine versus 0.125% bupivacaine and 0.2% ropivacaine for postoperative analgesia. The study was designed to test the equivalence of the overall profile of levobupivacaine against bupivacaine and ropivacaine. In addition, parameters of clinical safety were assessed. METHODS: A total of 88 patients undergoing hip surgery at 12 German academic hospitals were randomly assigned to 3 different treatment groups. Criteria for drug evaluation were the required epidural volume and time until onset and offset of sensory and motor block, the quality of postoperative analgesia using a pain visual analogue scale and verbal rating scale, as well as the need for rescue medication based on statistical non-inferiority testing. RESULTS: With respect to onset and offset of sensory and motor blockade, 0.5% levobupivacaine, 0.5% bupivacaine and 0.75% ropivacaine showed clinically significant equivalent profiles for all primary study endpoints. However, the levobupivacaine group showed a higher demand for intraoperative anesthesia. Postoperative analgesia request and pain scales did not differ significantly between groups, but comparatively lower total drug volumes were required in the bupivacaine group. No relevant differences between the trial groups concerning safety parameters were observed. CONCLUSIONS: The efficacy of epidural levobupivacaine for hip surgery and postoperative analgesia is equivalent and shows a comparable clinical profile to bupivacaine and 50-60% higher concentrated ropivacaine. The results of this equivalence study confirm suggestions derived from previous comparative studies.

HES 130/0.4 (Voluven) or human albumin in children younger than 2 yr undergoing non-cardiac surgery. A prospective, randomized, open label, multicentre trial.

BACKGROUND AND OBJECTIVE: For perioperative volume therapy in infants and young children, human albumin has frequently been the colloid of choice. Recently, HES 130/0.4 (6% hydroxyethyl starch, Voluven; Fresenius Kabi, Bad Homburg, Germany) was developed, which demonstrated improved pharmacokinetics and a favourable safety profile in adults compared with hydroxyethyl starch products with a less rapid metabolization. METHODS: Our prospective, controlled, randomized, open, multicentre pilot study was designed to obtain data on the effects of HES 130/0.4 compared with human albumin 5% with regard to haemodynamics in children <2 yr scheduled for elective non-cardiac surgery. RESULTS: A total of 81 patients were treated. Comparable amounts of both study solutions (16.0 mL kg(-1) hydroxyethyl starch 130/0.4 vs. 16.9 mL kg(-1) human albumin 5%) as well as add-on crystalloids were used until 4-6 h postoperatively. No differences were detected between the two treatment groups regarding perioperative stabilization of haemodynamics, coagulation parameters, blood gas analyses or other laboratory values. Blood loss was 96 +/-143 mL for hydroxyethyl starch and 145+/- 290 mL for human albumin (P > 0.05). There were no relevant differences in the amount of red blood cells, fresh frozen plasma or platelet concentrates in both treatment groups. Median length of ICU stay was 3.5 days (range 1-57 days, mean +/- SD 7.6 +/- 11.5 days) in the hydroxyethyl starch group and 6.0 days (range 1-71 days; mean +/- SD 9.1 +/- 14.2 days) in the human albumin group. There was no difference for hospital stay (median: 12 days for both groups). CONCLUSIONS: Both HES 130/0.4 and human albumin 5% were effective for haemodynamic stabilization in non-cardiac surgery of young infants with no adverse impact on coagulation or other safety parameters in our study population.

Pre-incisional epidural ropivacaine, sufentanil, clonidine, and (S)+-ketamine does not provide pre-emptive analgesia in patients undergoing major pancreatic surgery.

BACKGROUND: The concept of pre-emptive analgesia remains controversial. This prospective, randomized, and double-blind study compared epidural administration of ropivacaine 2 mg ml(-1), sufentanil 0.5 microg ml(-1), clonidine 3 microg ml(-1), and S(+)-ketamine 0.25 mg ml(-1) (study solution) given before incision with the same combination started at the end of the operation. METHODS: After testing the stability of the solution using high performance liquid chromatography (HPLC) and examining 12 patients for possible side-effects in comparison with the epidural infusion of ropivacaine 2 mg ml(-1) and sufentanil 0.5 microg ml(-1), 30 patients undergoing major pancreatic surgery were recruited into the study. Before induction of anaesthesia, an epidural catheter was inserted (TH6-8). Patients in Group 1 received a bolus of 8 ml followed by a continuous infusion (8 ml h(-1)) of the study solution before induction of anaesthesia. In Group 2, patients received the same volume of saline before operation, the study solution was started at the end of surgery. After operation, the infusion was maintained for at least 96 h using a patient-controlled epidural analgesia (PCEA) pump in both groups. Patients were evaluated up to the seventh postoperative day for pain and side-effects. RESULTS: Visual analogue scale (VAS) values at rest were as follows: G1 vs G2: 24 h, 19 (sd 23) vs 6 (13); 48 h, 4 (10) vs 11 (21); and 72 h, 12 (22) vs 13 (21). VAS values during coughing and mobilization were also comparable. Total volume of epidural infusion was 904 (114) ml in G1 vs 892 (154) ml in G2. The incidence of side-effects (nausea, vomiting, and motor block) was low and not different between the groups. CONCLUSIONS: Pre-incisional epidural analgesic infusion did not provide pre-emptive analgesia compared with administration started at the end of surgery, but both groups had low pain scores.

[Removal of an epidural catheter under ongoing antithrombotic therapy]. / Entfernung eines Epiduralkatheters unter antithrombotischer Therapie.

A rare though extremely harmful complication in neuraxial anaesthesia is an epidural hematoma which can be associated with deleterious consequences for the patient, e. g. persistent paraplegia. The risk of epidural haematomas after neuraxial blockade is dependent on abnormal anatomy of the spine, difficult and multiple punctures and coagulation disorders. Especially when patients undergo therapy with anticoagulants like low molecular heparin or platelet inhibitors (tyclopidine) or a combination of them, the indication for neuraxial blockade must strictly outweigh risk of spinal bleeding. In this context, the precautions and contraindications are the same for spinal puncture and catheter insertion as for catheter removal. We describe the case of a patient who underwent emergency coronary angioplasty in combination with coronary stent implantation due to acute postoperative myocardial infarction following knee replacement in continuous epidural anaesthesia. Under the symptoms of a beginning local infection at the puncture site the epidural catheter had to be removed in spite of ongoing antithrombotic therapy. A possible management of such cases is discussed with regard to risk minimization.

Comparative study of different concentrations of prilocaine and ropivacaine for intraoperative axillary brachial plexus block.

BACKGROUND AND OBJECTIVE: To compare the anaesthetic characteristics in terms of onset and offset times of the sensory and motor blocks of prilocaine 1% and ropivacaine 0.75% alone and in different combinations when used for brachial plexus anaesthesia in axillary perivascular blocks. METHODS: After informed consent 96 ASA I-III patients undergoing forearm or hand surgery participated in this prospective, randomized, double-blind study. Patients received either prilocaine 1% 40 mL (G1), prilocaine 1% 30 mL and ropivacaine 0.75% 10 mL (G2), prilocaine 1% 20 mL and ropivacaine 0.75% 20 mL (G3) or ropivacaine 0.75% 40 mL (G4) for axillary perivascular brachial plexus anaesthesia. Onset and duration of sensory and motor blocks in the distribution of the musculocutaneous, radial, median and ulnar nerves were assessed. RESULTS: The onset time of the sensory and motor blocks of the whole brachial plexus differed only between patients in G4 with ropivacaine 0.75% 40 mL demonstrating a later motor onset in comparison to all other groups and a later sensory onset in comparison to G1 and G2 (P < 0.01). The addition of ropivacaine resulted in longer offset times of the sensory and motor blocks. The median offset time of the motor block was 179.5 min in G1, 262 min in G2, 389.5 min in G3 and 745 min in G4 (P < 0.01). The median offset time of the sensory block was 163.5 min in G1, 277 min in G2, 383.5 min in G3 and 784 min in G4 (P < 0.01). There was no difference in onset and offset times between sensory and motor blocks within the groups. CONCLUSIONS: For axillary perivascular brachial plexus block prilocaine 1% alone and in combination with ropivacaine 0.75% was similar in terms of onset of sensory and motor blocks but different in duration of sensory and motor blocks without a differential sensory and motor offset.

Pulmonary embolism caused by polymethylmethacrylate during percutaneous vertebroplasty in orthopaedic surgery.

Vertebroplasty consists of percutaneous injection of acrylic cement--polymethylmethacrylate (PMMA)--into a partially collapsed vertebral body in order to obtain pain relief and augment mechanical stability of the vertebral body. Although vertebroplasty is an efficient treatment it is not free of complications. Our present case report describes a woman with pulmonary polymethylmethacrylate embolism during percutaneous vertebroplasty who presented with hypotension, arrhythmia and hypocapnia.

[Continuous interscalenic nerve block in a 3 year old child after amputation of upper extremity]. / Kontinuierliche interscalenäre Plexusblockade bei einem dreijährigen Kind nach Armamputation -- Ein Fallbericht.

Amputations of extremities, especially in the childhood, impose high demands on the perioperative management. Apart from the intraoperative care of these children, the postoperative pain therapy has to do one's utmost in the avoidance of the development of phantom limb pain, which can, especially in the childhood, be associated with far reaching psychological consequences. We report the case of a 3-year old boy who had to undergo exarticualtion of his left arm due to an osteosarcoma of the humerus. The perioperative pain management was performed by a preoperatively placed interscalene catheter and infusion of 0.2 % ropivacaine. Within the first six days postoperatively complete pain relief could be ensured with this analgetic regimen.

[Requirements and expectations for optimal volume replacement]. / Forderungen und Erwartungen an einen optimalen Volumenersatz.

A volume replacement should compensate a reduction in the intravascular volume and counteract a hypovolemia so that hemodynamics and vital functions can be maintained. For this therapy, a physiologically-based solution comprising both osmotically and colloid osmotically active components should be administered. A consensus is proposed for this purpose which takes into consideration the following aspects: The optimum colloid, the questionable use of albumin, the physiological electrolyte pattern encompassing sodium, potassium, chloride and phosphate and their contributions to osmolality, an eventual addition of glucose, the physiological acid-base status with bicarbonate or alternately with metabolisable anions, and the importance of a clear declaration of all ingredients. The consensus distinguishes between compulsory requirements derived from evidence-based medicine and physiological data and the potential expectations of an optimal volume replacement, including well-grounded wishes and aspirations for the future.

Effects of prophylactic or therapeutic application of bovine haemoglobin HBOC-200 on ischaemia-reperfusion injury following acute coronary ligature in rats.

BACKGROUND: Haemoglobin-based oxygen carriers (HBOCs) are assessed as blood substitutes in patients with perioperative anaemia including patients at risk for perioperative cardiac ischaemia. There is controversy as to whether HBOCs are beneficial or deleterious during ischaemia-reperfusion (I-R). Therefore the effects of HBOC-200 on I-R injury were evaluated in a randomized placebo-controlled animal trial. METHODS: Animals were randomized to receive either placebo i.v. without I-R (sham group, n=9), placebo i.v. with I-R (control group, n=10), HBOC-200 0.4 g kg(-1) i.v. prior to I-R (prophylaxis group, n=12) or HBOC-200 0.4 g kg(-1) i.v. during I-R (therapy group, n=15). I-R consisted of 25 min of acute ligature of the left coronary artery followed by 120 min of reperfusion. Measurements included assessment of the area at risk and infarct size using triphenyl tetrazolium chloride (TTC) stain, DNA single-strand breaks (in situ nick translation with autoradiography/densitometry) and cardiac arrhythmias. RESULTS: Infarct size within the area at risk was 62 (sd 15)% (control), 46 (10)% (prophylaxis, P<0.025 vs control) and 61 (9)% (therapy, P<0.85 vs control). The frequency of DNA single-strand breaks was reduced vs control in the sham (P<0.01) and prophylaxis (P<0.04) groups and was almost the same in the therapy group (P<0.75). The severity of cardiac arrhythmias during ischaemia was lower compared with control in the sham (P<0.001) and prophylaxis (P<0.039) groups, but there was no difference in the therapy group. CONCLUSION: This study demonstrates that neither prophylactic nor therapeutic application of the cell-free haemoglobin solution HBOC-200 aggravates cardiac I-R injury. Furthermore, the prophylactic approach may offer a new opportunity for pretreatment of patients at risk for perioperative ischaemic cardiac events.

Effects of isovolaemic haemodilution on oxygenation of liver and skeletal muscle.

BACKGROUND AND OBJECTIVE: Hydroxyethyl starch is frequently used for volume substitution during surgical procedures and for isovolaemic haemodilution. Haemodilution has also been shown to improve tissue oxygen tension in skeletal muscle: However, effects of this volume substitute on tissue oxygen tension of the liver during haemodilution remains unknown. METHODS: Fourteen foxhounds were anaesthetized with fentanyl/midazolam and mechanically ventilated with 30% oxygen. Following splenectomy animals were randomly assigned to a control group without haemodilution but fluid substitution with Ringer's lactate (Group C) or underwent isovolaemic haemodilution to a haematocrit of 25% with hydroxyethyl starch 70/0.5 (Group H). Haemodynamic parameters and oxygen transport during 100 min following isovolaemic haemodilution were measured. Liver oxygen tension was recorded using a flexible polarographic electrode tonometer, whereas in the muscle a polarographic needle probe was used. RESULTS: Animal characteristics and baseline haematocrit were similar in both groups. At baseline the tissue oxygen tension of liver and skeletal muscle were not different between groups. Haemodilution with hydroxyethyl starch 70/0.5 provided augmentation of mean liver tissue oxygen tension (baseline: 46 +/- 13 mmHg; 20 min: 60.3 +/- 12 mmHg; 60 min: 60 +/- 16 mmHg; 100 min: 63 +/- 16 mmHg; P < 0.05 vs. baseline), while oxygen tensions in Group C remained unchanged (baseline: 48 +/- 16 mmHg; 20 min: 52 +/- 19 mmHg; 60 min: 49 +/- 12 mmHg; 100 min: 52 +/- 16 mmHg) and no differences could be detected between groups. Oxygen tension in skeletal muscle changed as follows: Group H - baseline: 24 +/- 32 mmHg; 20 min: 32 +/- 3 mmHg; 60 min: 33 +/- 7 mmHg; 100 min: 33 +/- 11 mmHg. Group C - baseline: 22 +/- 6 mmHg; 20 min: 21 +/- 3 mmHg; 60 min: 24 +/- 4 mmHg; 100 min: 18 +/- 4 mmHg (P < 0.05 vs. baseline, p < 0.05 vs. Group C). CONCLUSION: In this animal model, isovolaemic haemodilution with hydroxyethyl starch 70/0.5 increased tissue oxygen tension in liver and skeletal muscle in comparison with baseline values. However, when compared between groups haemodilution only resulted in an increase of tissue oxygen tension in the muscle but not in the liver.