A European consensus statement on the use of four-factor prothrombin complex concentrate for cardiac and non-cardiac surgical patients.

Modern four-factor prothrombin complex concentrate was designed originally for rapid targeted replacement of the coagulation factors II, VII, IX and X. Dosing strategies for the approved indication of vitamin K antagonist-related bleeding vary greatly. They include INR and bodyweight-related protocols as well as fixed dose regimens. Particularly in the massively bleeding trauma and cardiac surgery patient, four-factor prothrombin complex concentrate is used increasingly for haemostatic resuscitation. Members of the Transfusion and Haemostasis Subcommittee of the European Association of Cardiothoracic Anaesthesiology performed a systematic literature review on four-factor prothrombin complex concentrate. The available evidence has been summarised for dosing, efficacy, drug safety and monitoring strategies in different scenarios. Whereas there is evidence for the efficacy of four-factor prothrombin concentrate for a variety of bleeding scenarios, convincing safety data are clearly missing. In the massively bleeding patient with coagulopathy, our group recommends the administration of an initial bolus of 25 IU.kg-1 . This applies for: the acute reversal of vitamin K antagonist therapy; haemostatic resuscitation, particularly in trauma; and the reversal of direct oral anticoagulants when no specific antidote is available. In patients with a high risk for thromboembolic complications, e.g. cardiac surgery, the administration of an initial half-dose bolus (12.5 IU.kg-1 ) should be considered. A second bolus may be indicated if coagulopathy and microvascular bleeding persists and other reasons for bleeding are largely ruled out. Tissue-factor-activated, factor VII-dependent and heparin insensitive point-of-care tests may be used for peri-operative monitoring and guiding of prothrombin complex concentrate therapy.

The role of fibrinogen and fibrinogen concentrate in cardiac surgery: an international consensus statement from the Haemostasis and Transfusion Scientific Subcommittee of the European Association of Cardiothoracic Anaesthesiology.

To date, data regarding the efficacy and safety of administering fibrinogen concentrate in cardiac surgery are limited. Studies are limited by their low sample size and large heterogeneity with regard to the patient population, by the timing of fibrinogen concentrate administration, and by the definition of transfusion trigger and target levels. Assessment of fibrinogen activity using viscoelastic point-of-care testing shortly before or after weaning from cardiopulmonary bypass in patients and procedures with a high risk of bleeding appears to be a rational strategy. In contrast, the use of Clauss fibrinogen test for determination of plasma fibrinogen level can no longer be recommended without restrictions due to its long turnaround time, high inter-assay variability and interference with high heparin levels and fibrin degradation products. Administration of fibrinogen concentrate for maintaining physiological fibrinogen activity in the case of microvascular post-cardiopulmonary bypass bleeding appears to be indicated. The available evidence does not suggest aiming for supranormal levels, however. Use of cryoprecipitate as an alternative to fibrinogen concentrate might be considered to increase plasma fibrinogen levels. Although conclusive evidence is lacking, fibrinogen concentrate does not seem to increase adverse outcomes (i.e., thromboembolic events). Large prospective multi-centre studies are needed to better define the optimal perioperative monitoring tool, transfusion trigger and target levels for fibrinogen replacement in cardiac surgery.

International consensus statement on the peri-operative management of direct oral anticoagulants in cardiac surgery.

Despite current recommendations on the management of severe peri-operative bleeding, there is no pragmatic guidance for the peri-operative monitoring and management of cardiac surgical patients taking direct oral anticoagulants. Members of the Transfusion and Haemostasis Subcommittee of the European Association of Cardiothoracic Anaesthesiology, of their own volition, performed an independent systematic review of peer-reviewed original research, review articles and case reports and developed the following consensus statement. This has been endorsed by the European Association of Cardiothoracic Anaesthesiology. In our opinion, most patients on direct oral anticoagulant therapy presenting for elective cardiac surgery can be safely managed in the peri-operative period if the following conditions are fulfilled: direct oral anticoagulants have been discontinued two days before cardiac surgery, corresponding to five elimination half-live periods; in patients with renal or hepatic impairment or a high risk of bleeding, a pre-operative plasma level of direct oral anticoagulants has been determined and found to be below 30 ng.ml-1 (currently only valid for dabigatran, rivaroxaban and apixaban). In cases where plasma level monitoring is not possible (e.g. assay was not available), discontinuation for 10 elimination half-live periods (four days) is required. For FXa inhibitors, a standard heparin-calibrated anti-Xa level of < 0.1 IU.ml-1 should be measured, indicating sufficient reduction in the anticoagulant effect. Finally, short-term bridging with heparin is not required in the pre-operative period.

[Hereditary heterozygous factor VII deficiency in patients undergoing surgery : Clinical relevance]. / Hereditärer heterozygoter Faktor VII-Mangel beim chirurgischen Patienten : Klinische Relevanz.

BACKGROUND: A hereditary deficiency in coagulation factor VII (FVII) may affect the international normalized ratio (INR) value. However, FVII deficiency is occasionally associated with a tendency to bleed spontaneously. We hypothesized that perioperative substitution with coagulation factor concentrates might not be indicated in most patients. METHODS: In this retrospective data analysis, we included all patients with hereditary heterozygous FVII deficiency who underwent surgical procedures at the University Hospital Basel between December 2010 and November 2015. In addition, by searching the literature, we identified publications reporting patients with FVII deficiency undergoing surgical procedures without perioperative substitution. RESULTS: We identified 22 patients undergoing 46 surgical procedures, resulting in a prevalence of 1:1500-2000. Coagulation factor concentrates were administered during the perioperative period in 15 procedures (33 %), whereas in the other 31 procedures (66 %), FVII deficiency was not substituted. No postoperative bleeding or thromboembolic events were reported. In addition, we found no differences in pre- and postoperative hemoglobin and coagulation parameters, with the exception of an improved postoperative INR value in the substituted group. In the literature review, we identified five publications, including 125 patients with FVII deficiency, undergoing 213 surgical procedures with no perioperative substitution. DISCUSSION: Preoperative substitution using coagulation factor concentrates does not seem to be mandatory in patients with an FVII level ≥15 %. For decision-making on preoperative substitution, patient history of an increased tendency to bleed may be more important than the FVII level or increased INR value.

Perioperative administration of fibrinogen does not increase adverse cardiac and thromboembolic events after cardiac surgery.

BACKGROUND: Although infusion of fibrinogen concentrate is increasingly used in bleeding patients after cardiac surgery, safety data are scarce. We aimed to evaluate the effect of perioperative administration of fibrinogen concentrate on postoperative morbidity and mortality in patients undergoing cardiac surgery. METHODS: During a 2 yr study period, 991 patients underwent cardiac surgery at a single university centre and were eligible for propensity score (PS) matching. We matched 190 patients with perioperative infusion of fibrinogen concentrate (median dose 2 g) with 190 controls without fibrinogen administration. After PS matching, crude outcome was analysed. Further, a multivariate logistic regression including additional risk factors for adverse outcome was performed. The primary endpoint was a composite of mortality and the occurrence of major cardiac and thromboembolic events within 1 yr. Secondary outcomes included mortality after 30 days and 1 yr and the composite of mortality and adverse events after 30 days. RESULTS: The administration of fibrinogen concentrate was not associated with an increased risk for mortality and thromboembolic or cardiac events within 1 yr after cardiac surgery [unadjusted hazard ratio (HR) 0.91; 95% confidence interval (CI) 0.55-1.49; P=0.697]. When using multivariate logistic regression model, the HR for adverse outcome in patients with administration of fibrinogen concentrate was 0.57 (95% CI 0.25-1.17; P=0.101). Similarly, the administration of fibrinogen concentrate did not adversely affect the secondary outcomes when applying unadjusted and multivariate regression analyses. CONCLUSIONS: Our study strongly suggests that the administration of fibrinogen concentrates at low dose is not associated with thromboembolic complications or adverse outcomes after cardiac surgery.

Intra-operative assessment of pulmonary artery pressure by transoesophageal echocardiography.

The clinical value of the estimation of systolic pulmonary artery pressure, based on Doppler assessment of peak tricuspid regurgitant velocity using transoesophageal echocardiography, is unclear. We studied 109 patients to evaluate the feasibility of obtaining adequate Doppler recordings, and compared Doppler estimates with values measured using a pulmonary artery catheter in a subset of 33 patients. Tricuspid regurgitation was evaluated at the mid-oesophageal level at 0-120° using Doppler echocardiography. A Doppler signal was defined as adequate if there was a ≤ 20° alignment and a full envelope. Doppler estimates of systolic pulmonary artery pressure within 10 mmHg and 15% of the value recorded with the pulmonary artery catheter were considered to be in sufficient agreement. Adequate Doppler signals were obtained in 64/109 (59%) patients before and 54/103 (52%) after surgery. Doppler estimates by transoesophageal echocardiography were within 10 mmHg and 15% of values recorded with the pulmonary artery catheter in 28/33 (75%) patients and 22/31 (55%) patients, respectively. In 7 (21%) patients, the echocardiographic Doppler measurement exceeded the measured systolic pulmonary artery pressure by more than 30%. Our study indicates that estimation of the systolic pulmonary artery pressure using transoesophageal Doppler echocardiography is not a reliable and clinically useful method in anaesthetised patients undergoing mechanical ventilation.

Volatile anaesthetics and positive pressure ventilation reduce left atrial performance: a transthoracic echocardiographic study in young healthy adults.

BACKGROUND: Animal and in vitro studies suggest that volatile anaesthetics affect left atrial (LA) performance. We hypothesized that human LA pump function and dimensions are altered by volatile anaesthetics in vivo. METHODS: We performed transthoracic echocardiographic (TTE) measurements in 59 healthy subjects (aged 18-48 yr) undergoing minor surgery under general anaesthesia. The unpremedicated patients were randomly assigned to anaesthesia with sevoflurane, desflurane, or isoflurane. TTE examinations were performed at baseline and after induction of anaesthesia and upon placement of a laryngeal mask during spontaneous breathing. After changing to intermittent positive pressure ventilation (IPPV), an additional TTE was performed. The study focused on the velocity-time integral of late peak transmitral inflow velocity (AVTI) and maximum LA volume. RESULTS: We found no evidence for relevant differences in the effects of the three volatile anaesthetics. AVTI decreased significantly from 4.1 (1.2) cm at baseline to 3.2 (1.1) cm during spontaneous breathing of 1 minimum alveolar concentration of volatile anaesthetics. AVTI decreased further to 2.8 (1.0) cm after changing to IPPV. The maximum LA volume was 45.4 (18.6) cm(3) at baseline and remained unchanged during spontaneous breathing but decreased to 34.5 (16.7) cm(3) during IPPV. Other parameters of LA pump function and dimensions decreased similarly. CONCLUSIONS: Volatile anaesthetics reduced active LA pump function in humans in vivo. Addition of IPPV decreased LA dimensions and further reduced LA pump function. Effects in vivo were less pronounced than previously found in in vitro and animal studies. Further studies are warranted to evaluate the clinical implications of these findings. CLINICAL TRIAL REGISTRATION: NCT0024451.

Perioperative factor concentrate therapy.

Transfusion of allogeneic plasma has been a life-saving measure for decades in patients with severe trauma or suffering from major surgical blood loss. The safety of allogeneic blood components has improved in terms of pathogen transmission, but haemostatic efficacy of plasma is hindered by the large volume and time required for thawing and infusion. Several plasma-derived and recombinant factor concentrates are clinically available and indicated for targeted replacement of missing coagulation elements in hereditary disorders of thrombosis and haemostasis. When used appropriately, factor concentrate therapy can rapidly restore deficient factor(s) without causing volume overload. The haemostatic defect in perioperative patients is often multifactorial, and therefore careful clinical judgement and timely coagulation testing must be exercised before the administration of factor concentrates. In this review, the rationale for including factor concentrates in perioperative haemostatic management will be discussed in conjunction with the limitations of plasma transfusion.

Evaluation of a novel flow chamber system to assess clot formation in factor VIII-deficient mouse and anti-factor IXa-treated human blood.

Blood flow properties play important roles in the regulation and formation of thrombus. To evaluate the influence of blood flow on thrombus formation in haemophilia, whole blood samples were obtained from FVIII-deficient (FVIII(-/-) ) and wild-type (FVIII(+/+) ) mice (n = 6 respectively), and from six human volunteers. Anti-FIXa aptamer was added to human blood to model acquired haemophilia B. Recalcified whole blood samples containing corn trypsin inhibitor and danaproid were perfused over the microchip coated with collagen and tissue thromboplastin at shear rates of 1100 and 110 s(-1) . Thrombus formation in the capillary was quantified by monitoring flow pressure changes. The intervals to 5 kPa (T(5) ) and 40 k Pa (T(40) ) reflect the onset and growth of thrombus formation respectively. Furthermore, fibrin and platelets in thrombi were quantified by immunostaining. T(5) at both shear rates were similar in FVIII(-/-) and FVIII(+/+) mice. T(40) of FVIII(-/-) mice (1569 ± 565 s) was significantly delayed compared with FVIII(+/+) mice (339 ± 78 s) at 110 s(-1) (P < 0.05), but not at 1100 s(-1) . The delay was normalized by adding human FVIII (2 IU mL(-1) ). Similarly, adding anti-FIXa aptamer to human blood prolonged T(40) at 110 s(-1) (P < 0.01), but not at 1100 s(-1) . Impaired production of fibrin due to anti-FIXa aptamer at 110 s(-1) was shown in the immunostained thrombus. Our perfusion experiments demonstrated that shear rates influence thrombus formation patterns in haemophilia, and that reduced activity of intrinsic tenase (FIXa-FVIIIa) becomes evident under venous shear rates.

Impact of loss of high-molecular-weight von Willebrand factor multimers on blood loss after aortic valve replacement.

BACKGROUND: Severe aortic stenosis is associated with loss of the largest von Willebrand factor (vWF) multimers, which could affect primary haemostasis. We hypothesized that the altered multimer structure with the loss of the largest multimers increases postoperative bleeding in patients undergoing aortic valve replacement. METHODS: We prospectively included 60 subjects with severe aortic stenosis. Before and after aortic valve replacement, vWF antigen, activity, and multimer structure were determined and platelet function was measured by impedance aggregometry. Blood loss from mediastinal drainage and the use of blood and haemostatic products were evaluated perioperatively. RESULTS: Before operation, the altered multimer structure was present in 48 subjects (80%). Baseline characteristics and laboratory data were similar in all subjects. The median blood loss after 6 h was 250 (105-400) and 145 (85-240) ml in the groups with the altered and normal multimer structures, respectively (P=0.182). After 24 h, the cumulative loss was 495 (270-650) and 375 (310-600) ml in the groups with the altered and normal multimer structures, respectively (P=0.713). Multivariable analysis revealed no significant influence of multimer structure and platelet function on bleeding volumes after 6 and 24 h. After 24 h, there was no obvious difference in vWF antigen, activity, and multimer structure in subjects with and without the altered multimer structure before operation or in subjects with and without perioperative plasma transfusion. CONCLUSIONS: The altered vWF multimer structure before operation was not associated with increased bleeding after aortic valve replacement. Our findings might be explained by perioperative release of vWF and rapid recovery of the largest vWF multimers.

The influence of pre-admission hypoglycaemic therapy on cardiac morbidity and mortality in type 2 diabetic patients undergoing major non-cardiac surgery: a prospective observational study.

It remains unclear whether type 2 diabetics treated with either insulin or oral hypoglycaemic agents have the same incidence of cardiac morbidity and mortality after major non-cardiac surgery. We prospectively studied 360 type 2 diabetic patients undergoing major non-cardiac surgery of which 105 were treated with insulin only, 171 were treated with oral hypoglycaemics only and 84 were treated with a combination of insulin and oral hypoglycaemics. All-cause mortality after 30 days and after 12 months was highest in the insulin (10% and 26%) and lowest in the oral hypoglycaemics group (2% and 13%; p = 0.02 and 0.007, respectively). Insulin treatment was independently associated with increased mortality after 30 days (hazard ratio 3.93; 95% CI 1.22-12.64; p = 0.022) and 12 months (hazard ratio 2.03; 95% CI 1.16-3.58; p = 0.014) after multivariate adjustment for age, sex and the revised cardiac risk index (insulin treatment excluded). The increased mortality in insulin-treated diabetic patients may be due to a more progressive disease state in these patients rather than the treatment modality itself.

Remifentanil does not impair left ventricular systolic and diastolic function in young healthy patients.

BACKGROUND: Experimental studies and investigations in patients with cardiac diseases suggest that opioids at clinical concentrations have no important direct effect on myocardial relaxation and contractility. In vivo data on the effect of remifentanil on myocardial function in humans are scarce. This study aimed to investigate the effects of remifentanil on left ventricular (LV) function in young healthy humans by transthoracic echocardiography (TTE). We hypothesized that remifentanil does not impair systolic, diastolic LV function, or both. METHODS: Twelve individuals (aged 18-48 yr) without any history or signs of cardiovascular disease and undergoing minor surgical procedures under general anaesthesia were studied. Echocardiographic examinations were performed in the spontaneously breathing subjects before (baseline) and during administration of remifentanil at a target effect-site concentration of 2 ng ml(-1) by target-controlled infusion. Analysis of systolic function focused on fractional area change (FAC). Analysis of diastolic function focused on peak early diastolic velocity of the mitral annulus (e') and on transmitral peak flow velocity (E). RESULTS: Remifentanil infusion at a target concentration of 2 ng ml(-1) did not affect heart rate or arterial pressure. There was no evidence of systolic or diastolic dysfunction during remifentanil infusion, as the echocardiographic measure of systolic function (FAC) was similar to baseline, and measures of diastolic function remained unchanged (e') or improved slightly (E). CONCLUSION: Continuous infusion of remifentanil in a clinically relevant concentration did not affect systolic and diastolic LV function in young healthy subjects during spontaneous breathing as indicated by TTE.

Thrombin generation and fibrinolysis in anti-factor IX treated blood and plasma spiked with factor VIII inhibitor bypassing activity or recombinant factor VIIa.

Activated prothrombin complex concentrates (aPCC) and recombinant activated factor VIIa (rFVIIa) are two important therapies in haemophilia patients with inhibitors and improve clot stability. We hypothesized that potential differences in procoagulant and fibrinolytic actions of aPCC and rFVIIa may lie in the clot stability against fibrinolytic activation. We used thrombin generation, fluorescence detection and thromboelastometry in anti-factor IXa (FIXa) aptamer-treated whole blood (WB) and plasma to evaluate: (i) generation of thrombin and activated factor X (FXa) and (ii) viscoelastic properties of blood clots in the presence of tissue plasminogen activator (tPA) after addition of aPCC (0.4 U mL(-1)) or rFVIIa (60 nm). Peak thrombin generation increased from 85 +/- 19 nm in aptamer-treated plasma to 276 +/- 83 nm and 119 +/- 22 nm after addition of aPCC and rFVIIa respectively (P < 0.001). FXa activity increased within 20 min by 87 +/- 6% and by 660 +/- 97% after addition of aPCC and rFVIIa respectively (P < 0.001). TPA-induced lysis time increased from 458 +/- 378 s in aptamer-treated WB to 1597 +/- 366 s (P = 0.001) and 1132 +/- 214 s (P = 0.075), after addition of aPCC and rFVIIa respectively. In this haemophilia model using the anti-FIXa aptamer, the larger amount of thrombin was generated with aPCC compared with rFVIIa, while FXa generation was more rapidly increased in the presence of rFVIIa. Furthermore, clot formation in anti-FIXa aptamer-treated WB was less susceptible to tPA-induced fibrinolysis after adding aPCC compared with rFVIIa.

Haemodilution-induced profibrinolytic state is mitigated by fresh-frozen plasma: implications for early haemostatic intervention in massive haemorrhage.

BACKGROUND: Fibrinolysis contributes to coagulopathy after major trauma and surgery. We hypothesized that progressive haemodilution is responsible, at least in part, for increased fibrinolytic tendency of blood clot. METHODS: The study was performed in two parts. First, whole blood (WB) samples collected from six healthy, consented volunteers were diluted in vitro with either saline or fresh-frozen plasma (FFP) to 40% and 15% of baseline. We quantified factor levels related to coagulation and fibrinolysis, and measured endogenous thrombin generation in undiluted control plasma samples and in samples diluted with saline or FFP. Additionally, thromboelastometry was used to assess susceptibility to fibrinolysis after adding tissue plasminogen activator in undiluted WB samples and in samples diluted with saline before and after substitution of fibrinogen or FFP. Secondly, as a model of in vivo haemodilution, we evaluated the same parameters before and after operation in nine consented patients undergoing off-pump coronary artery bypass surgery. RESULTS: The dilution with saline caused dose-dependent decreases in plasma levels of coagulation and antifibrinolytic factors, and in thrombin generation. In FFP-supplemented samples, factor levels and thrombin generation were maintained within normal ranges. Fibrinolytic tendency was significantly higher after haemodilution with saline independent of fibrinogen substitution compared with FFP. Similarly, increased tendency for fibrinolysis was also observed in the in vivo haemodilution. CONCLUSIONS: We demonstrated in vitro and in vivo that progressive haemodilution decreases endogenous antifibrinolytic proteins including alpha(2)-antiplasmin and thrombin-activatable fibrinolysis inhibitor, resulting in increased fibrinolytic tendency. Therefore, early fluid replacement therapy with FFP might be advantageous after massive haemorrhage.

Different effects of sevoflurane, desflurane, and isoflurane on early and late left ventricular diastolic function in young healthy adults.

BACKGROUND: Knowledge on the effects of volatile anaesthetics on left ventricular (LV) diastolic function in humans in vivo is limited. We tested the hypothesis that sevoflurane, desflurane, and isoflurane do not impair LV diastolic function in young healthy humans. METHODS: Sixty otherwise healthy subjects (aged 18-48 yr) undergoing minor procedures under general anaesthesia were studied. After randomization for the anaesthetic, transthoracic echocardiographic examinations were performed at baseline and under anaesthesia with 1 minimum alveolar concentration (MAC) of the volatile anaesthetics during spontaneous breathing and intermittent positive pressure ventilation (IPPV). Peak early (E') and late (A') diastolic velocities of the mitral annulus were studied as the main echocardiographic indicators of diastolic function. RESULTS: During anaesthesia with 1 MAC under spontaneous breathing, E' increased with desflurane (P<0.001), was not significantly different with isoflurane (P=0.030), and decreased with sevoflurane (P=0.006). During IPPV, E' was similar to baseline with desflurane (P=0.550), insignificantly decreased with isoflurane (P=0.029), and decreased with the sevoflurane group (P<0.001). In contrast, A' was similarly reduced in all groups during spontaneous breathing without further changes during IPPV. Haemodynamic changes were comparable in all study groups. CONCLUSIONS: The findings of this in vivo study indicate that desflurane and isoflurane, and most likely sevoflurane, have no relevant direct negative effect on early diastolic relaxation in young healthy humans. In contrast, all three volatile anaesthetics appear to impair late diastolic LV filling during atrial contraction.

Finding the optimal concentration range for fibrinogen replacement after severe haemodilution: an in vitro model.

BACKGROUND: Replacement of fibrinogen is presumably the key step in managing dilutional coagulopathy. We performed an in vitro study hypothesizing that there is a minimal fibrinogen concentration in diluted whole blood above which the rate of clot formation approaches normal. METHODS: Blood samples from six healthy volunteers were diluted 1:5 v/v with saline keeping haematocrit at 24% using red cell concentrates. We measured coagulation factors and thrombin generation in plasma at baseline and after dilution. Thromboelastometry was used to evaluate (i) speed and quality of clot formation in diluted samples supplemented with fibrinogen 50-300 mg dl(-1) and (ii) clot resistance to fibrinolysis. Diluted and undiluted samples with no added fibrinogen served as controls. RESULTS: Coagulation parameters and platelets were reduced by 74-85% after dilution. Peak thrombin generation was reduced by 56%. Adding fibrinogen led to a concentration-dependent improvement of all thromboelastometric parameters. The half maximal effective concentration (EC50) for fibrinogen replacement in haemodiluted blood was calculated to be 125 mg dl(-1). Adding tissue plasminogen activator, 0.15 microg ml(-1), led to a decrease of clot firmness and lysis time. CONCLUSIONS: The target plasma concentration for fibrinogen replacement was predicted by these in vitro results to be greater than 200 mg dl(-1) as only these concentrations optimized the rate of clot formation. This concentration is twice the level suggested by the current transfusion guidelines. Although improved, clots were prone to fibrinolysis indicating that the efficacy of fibrinogen therapy may be influenced by co-existing fibrinolytic tendency occurring during dilutional coagulopathy.

Randomized clinical trial of moxonidine in patients undergoing major vascular surgery.

BACKGROUND: Myocardial ischaemia is the leading cause of perioperative morbidity and mortality after surgery in patients with coronary artery disease. The aim of this study was to evaluate the effects of moxonidine, a centrally acting sympatholytic agent, on perioperative myocardial ischaemia and 1-year mortality in patients undergoing major vascular surgery. METHODS: In this double-blind, placebo-controlled two-centre trial, 141 patients were randomly assigned to receive moxonidine or placebo on the morning before surgery and on the following 4 days. Levels of cardiac troponin I (cTnI) were analysed before surgery and on days 1, 2, 3 and 7 thereafter. Holter electrocardiograms were recorded for 48 h starting before the administration of the study drug. Patients were followed daily during admission and by telephone interview 12 months after surgery. RESULTS: The incidence of raised perioperative cTnI levels or alteration in the ST segment in the Holter electrocardiogram or both was 40 per cent in the moxonidine group and 37 per cent in the placebo group (P = 0.694). All-cause mortality rates within 12 months were 10 per cent in the moxonidine group and 11 per cent in the placebo group (P = 0.870). CONCLUSION: Small oral doses of moxonidine did not reduce the incidence of perioperative myocardial ischaemia and had no effect on mortality in patients undergoing vascular surgery. REGISTRATION NUMBER: NCT00244504 (http://www.clinicaltrials.gov).