Blood transfusion after cardiac surgery: is it the patient or the transfusion that carries the risk?

BACKGROUND: The transfusion of red blood cells (RBCs) after cardiac surgery has been associated with increased long-term mortality. This study reexamines this hypothesis by including pre-operative hemoglobin (Hb) levels and renal function in the analysis. METHODS: A retrospective single-center study was performed including 5261 coronary artery bypass grafting (CABG) patients in a Cox proportional hazard survival analysis. Patients with more than eight RBC transfusions, early death (7 days), and emergent cases were excluded. Patients were followed for 7.5 years. Previously known risk factors were entered into the analysis together with pre-operative Hb and estimated glomerular filtration rate (eGFR). In addition, subgroups were formed based on the patients' pre-operative renal function and Hb levels. RESULTS: When classical risk factors were entered into the analysis, transfusion of RBCs was associated with reduced long-term survival. When pre-operative eGFR and Hb was entered into the analysis, however, transfusion of RBCs did not affect survival significantly. In the subgroups, transfusion of RBCs did not have any effect on long-term survival. CONCLUSIONS: When pre-operative Hb levels and renal function are taken into account, moderate transfusions of RBC after CABG surgery do not seem to be associated with reduced long-term survival.

Levosimendan cardioprotection in acutely beta-1 adrenergic receptor blocked open chest pigs.

BACKGROUND: Levosimendan and volatile anesthetics have myocardial pre-conditioning effects. beta-1 adrenergic receptor antagonists may inhibit the protective effect of volatile anesthetics. No information exists as to whether this also applies to the pre-conditioning effect of levosimendan. We therefore investigated whether levosimendan added to metoprolol would demonstrate a cardioprotective effect. METHODS: Three groups of anesthetized open chest pigs underwent 30 min of myocardial ischemia and 90 min of reperfusion by temporary occlusion of the largest side branch from the circumflex artery or the left anterior descending artery. One group (CTRL) served as a control, in another group (BETA), a metoprolol-loading dose was intravenously injected 30 min before ischemia, and in a third group (BETA+L), a levosimendan infusion was added to metoprolol. Myocardial tissue concentrations of glucose, glycerol, and lactate/pyruvate ratio as the primary end-points were investigated with microdialysis in ischemic and non-ischemic tissues. RESULTS: At the end of the ischemic period, statistically significant differences were only found between CTRL and BETA+L in the ischemic myocardium, with a lower lactate/pyruvate ratio, lower glycerol, and higher glucose concentrations in BETA+L as compared with CTRL. There were no differences in non-ischemic myocardium. From 10 to 90 min of reperfusion, no more differences were found between groups. CONCLUSION: The cardioprotective effect of levosimendan on ischemic metabolism with a reduction in the myocardial lactate/pyruvate ratio, less glycerol accumulation, and better preserved glucose concentration does not seem to be prevented by beta-1 adrenergic receptor antagonism with metoprolol.

Preoperative and perioperative use of levosimendan in cardiac surgery: European expert opinion.

In cardiac surgery, postoperative low cardiac output has been shown to correlate with increased rates of organ failure and mortality. Catecholamines have been the standard therapy for many years, although they carry substantial risk for adverse cardiac and systemic effects, and have been reported to be associated with increased mortality. On the other hand, the calcium sensitiser and potassium channel opener levosimendan has been shown to improve cardiac function with no imbalance in oxygen consumption, and to have protective effects in other organs. Numerous clinical trials have indicated favourable cardiac and non-cardiac effects of preoperative and perioperative administration of levosimendan. A panel of 27 experts from 18 countries has now reviewed the literature on the use of levosimendan in on-pump and off-pump coronary artery bypass grafting and in heart valve surgery. This panel discussed the published evidence in these various settings, and agreed to vote on a set of questions related to the cardioprotective effects of levosimendan when administered preoperatively, with the purpose of reaching a consensus on which patients could benefit from the preoperative use of levosimendan and in which kind of procedures, and at which doses and timing should levosimendan be administered. Here, we present a systematic review of the literature to report on the completed and ongoing studies on levosimendan, including the newly commenced LEVO-CTS phase III study (NCT02025621), and on the consensus reached on the recommendations proposed for the use of preoperative levosimendan.

Cardiopulmonary perfusion and cerebral blood flow in bilateral carotid artery disease.

The fear of cerebral complications after cardiopulmonary bypass in patients with heart disease and severe carotid artery disease has led many authors to suggest combined approaches in these patients. The pathogenetic mechanism for stroke is based partly on the stenotic narrowing of the carotid artery. A diameter reduction of 75% is frequently considered hemodynamically significant and indicative of an increased risk for neurological morbidity. We studied the cerebral blood flow in 7 patients undergoing coronary artery bypass grafting who also had severe bilateral carotid disease. The results were compared with the results in 17 patients without carotid disease who had bypass grafting. The cerebral blood flow was measured by xenon 133 washout technique before, during, and after cardiopulmonary bypass with moderate hypothermia. Acid-base regulation was according to the alpha-stat theory, and blood pressure was kept greater than 50 mm Hg. The cerebral blood flow levels (mL.100g-1.min-1) before, during, and after cardiopulmonary bypass in the study group (30 +/- 11, 31 +/- 8, 47 +/- 20) (mean +/- standard deviation) were almost identical to those in the control group (30 +/- 11, 28 +/- 8, 47 +/- 12). The cerebral blood flow levels for the left and right hemispheres in the group with carotid disease were comparable and within normal ranges. In 2 patients, slight differences were noted between hemispheres, and this finding may indicate an increased risk for ischemia. These patients, however, did not show any signs of postoperative deficit. The flow limitations of critical carotid stenoses do not seem to imply a risk for cerebral hypoperfusion if cardiopulmonary perfusion is performed in a controlled manner.(ABSTRACT TRUNCATED AT 250 WORDS)

Prediction and prevention of delayed ischemic dysfunction after aneurysmal subarachnoid hemorrhage and early operation.

Mean hemispheric cerebral blood flow (CBF) was studied after the intravenous administration of xenon-133 in 20 anesthetized patients with aneurysmal subarachnoid hemorrhage. Before early aneurysm operation, repeated CBF measurements were made to evaluate the cerebral vascular reactivity to controlled hyperventilation. Thirteen individuals received intravenous treatment with the calcium channel blocker nimodipine, whereas the other seven patients did not receive such specific antiischemic treatment. Five of the latter patients had an impaired CO2 response, and three showed delayed ischemic deterioration (DID), whereas none of the seven nimodipine-treated patients with impaired CO2 response showed DID. One nimodipine-treated patient with a preserved CO2 response, in whom operation was complicated, developed DID. The observed findings indicate that DID after uncomplicated early aneurysm operation may be associated with an early disturbance of cerebral vasoreactivity. Treatment with nimodipine may counteract the development of DID in patients with an impaired CBF CO2 response.

Cerebral hemodynamics in patients with acute severe head trauma.

Mean hemispheric cerebral blood flow (CBF) was studied in 11 comatose brain-injured patients following intravenous administration of xenon-133. Repeated measurements were performed in order to evaluate cerebral vasoreactivity following a decrease in PaCO2. In addition, the effect of induced barbiturate coma was evaluated in patients with intracranial hypertension. The cerebral vasoreactivity and the CBF response following induction of barbiturate coma varied. In patients with normal CO2 reactivity, barbiturate treatment was accompanied by a considerable decrease in CBF as compared to patients with decreased or abolished CO2 response. During barbiturate treatment the intracranial pressure (ICP) became normal in three of four patients with preserved CO2 response, but reached normal levels in only one of five patients with impaired CO2 reactivity. Patients whose ICP became normal recovered. The data suggest a positive correlation between CO2 reactivity and the effect of barbiturate treatment. Furthermore, preserved cerebral vasoreactivity after severe head injury may be of prognostic value.

Cerebral hemodynamic changes and electroencephalography during carotid endarterectomy.

Some patients undergoing endarterectomy for occlusive carotid artery disease run a risk of brain ischemia during cross-clamping of the artery. The present study of 15 patients was undertaken to evaluate changes in cerebral blood flow (CBF), as measured with an intravenous (IV) tracer (133Xenon) technique, and to relate CBF changes to changes in the electroencephalogram (EEG). CBF was measured before and after induction of anesthesia, during cross-clamping of the carotid artery, after release of the clamps, and at 24 hours after the operation. All the patients were anesthetized with methohexitone, fentanyl, and nitrous oxide and oxygen. EEG was continuously recorded during the operation. Carotid artery shunts were not used. In 8 patients, cross-clamping of the carotid artery did not influence the EEG. In this group of patients, induction of anesthesia caused a 38% decrease in CBF, which presumably reflects the normal reaction to the anesthetic agent given. There were no further changes in CBF during cross-clamping. In 7 patients, the EEG showed signs of deterioration during the intraoperative vascular occlusion. In these patients, anesthesia did not cause any CBF change, whereas cross-clamping the artery induced a 33% decrease in CBF. In individual patients, the severity of EEG changes correlated with the decrease in CBF. The absence of a change in CBF by anesthesia and a decrease due to cross-clamping of the carotid artery may be explained by the presence of a more advanced cerebrovascular disease and an insufficiency to maintain CBF during cross-clamping.(ABSTRACT TRUNCATED AT 250 WORDS)

Acupuncture for pain relief during childbirth.

BACKGROUND: There are few studies on acupuncture in childbirth despite the generally established analgesic effect of acupuncture treatment. METHODS: The analgesic effect of acupuncture during childbirth was assessed by comparing the need for other pain treatments (epidural analgesia using bupivacaine, pudendal nerve block, intramuscular meperidine, nitrous oxide/oxygen, intracutaneous sterile water injections) in 90 women given acupuncture (acupuncture group) with that in 90 women not given acupuncture (control group). RESULTS: 52 women (58%) in the acupuncture group and 13 (14%) in the control group managed their deliveries without further pain treatment (p < 0.001). The groups were similar with respect to age, pariety, duration of delivery, use of oxytocine and incidence of Caesarean section. Acupuncture treatment was found to have no major side effects, and 85 women (94%) given acupuncture reported that they would reconsider acupuncture in future deliveries. CONCLUSION: Acupuncture reduces the need for other methods of analgesia in childbirth.

[Immediate relief in malignant respiratory tract neoplasms by endoscopic neodymium-YAG laser treatment]. / Omedelbar lindring vid maligna tumörer i luftvägarna med endoskopisk neodym-YAG-laserbehandling.

Primary and secondary malignant tumours in the central airways can cause obstruction or chronic haemoptyses. Conventional forms of treatment are insufficient. Bronchoscopic Neodymium-YAG laser treatment, hitherto unused in Sweden, can give immediate palliation, Indications and contraindications are given, and the first fourteen patients treated are presented.

Levosimendan cardioprotection reduces the metabolic response during temporary regional coronary occlusion in an open chest pig model.

BACKGROUND: Inotropic and myocardial anti-ischemic effects have been demonstrated with levosimendan. The comparison of levosimendan started before an ischemia-reperfusion event as compared with levosimendan started during ischemia has not been studied. METHODS: In anesthetized pigs, a major branch of the circumflex artery was completely occluded for 30 min and then reperfused. The metabolism in the ischemic myocardium and in non-ischemic control myocardium was studied with microdialysis concomitantly with monitoring of global hemodynamics and coronary artery flow in the chosen artery. In the protection group (PRO, n= 6), a levosimendan infusion was started 30 min before coronary artery occlusion, and in the treatment group (TRE, n= 6), a levosimendan infusion was started 10 min after the coronary artery occlusion with a loading dose of 13.3 microg/kg followed by an infusion of 0.67 microg/kg/min. A two-way repeated measures ANOVA completed with Bonferroni's multiple comparison procedure was applied to the data. A P < 0.05 was considered significant. RESULTS: During the ischemic period, the cardiac output and contractility (dp/dt(max)) were higher in the PRO as compared with the TRE and the systemic vascular resistance was lower. The myocardial microdialysate glucose concentration in the ischemic area during ischemia was higher in the PRO as compared with the TRE, and the lactate/pyruvate ratio and the lactate concentration were lower. The differences in the metabolites persisted into the first 10 min of reperfusion. No differences were found for the non-ischemic areas. CONCLUSION: Levosimendan used throughout myocardial ischemia-reperfusion might have a cardioprotective affect on the response to myocardial ischemia as compared with levosimendan started during the ischemia.

Myocardial glycerol release, arrhythmias and hemodynamic instability during regional ischemia-reperfusion in an open chest pig model.

BACKGROUND: To develop cardioprotection against peri-operative myocardial ischemia-reperfusion injury, we need animal models where the local metabolism and blood flow are studied concomitantly with the global circulatory consequences during regional coronary occlusion. METHODS: In six anesthetized domestic pigs, the largest branch of the circumflex artery was occluded for 30 min. Microdialysate was sampled from the ischemic and non-ischemic myocardium along with continuous measurements of local coronary artery flow, global hemodynamics and registration of arrhythmias, from baseline through to 30 min of ischemia and 180 min of reperfusion. RESULTS: During ischemia, the microdialysate glucose concentration decreased, the glycerol concentration increased and the lactate/pyruvate ratio increased significantly. For glycerol, there was a further increase at reperfusion. During ischemia, cardiac output was unchanged; however, during reperfusion there was a significant drop lasting for several minutes, longer than the period in which an increased number of arrhythmias were registered. CONCLUSION: The present study demonstrates deranged circulation and arrhythmias corresponding to ischemic metabolism after regional myocardial ischemia and reperfusion. Reperfusion induced more pronounced circulatory changes than the actual ischemia. A substantial increase in myocardial glycerol release seems to be a marker of ischemic metabolism and may prove to be an indicator of reperfusion injury.

Surgery for primary hyperparathyroidism performed under local anaesthesia.

Patients with primary hyperparathyroidism are often elderly with cardiovascular disease and in some an operation might be hazardous owing to anaesthetic complications. A technique for operation for primary hyperparathyroidism under local anaesthesia is described. The method uses a unilateral approach. Seventeen consecutive patients operated on under local anaesthesia were compared with a group of 15 patients undergoing surgery under general anaesthesia. Normocalcaemia was achieved in 14 patients in each group. There was no difference in the extent of pain or the overall well-being between the two groups as determined by a visual analogue scale. Patients receiving local anaesthesia, however, experienced significantly less nausea after operation (P < 0.01). There was more fluctuation in blood pressure and heart rate in the general anaesthesia group compared with the other group. Surgery for primary hyperparathyroidism can be performed safely under local anaesthesia, and could be offered to patients if general anaesthesia were not suitable or involved an increased perioperative risk. It should not be recommended for routine use in patients who are fit for general anaesthesia.

Anaesthesia for resection of tumours in the trachea and central bronchi using the Nd-YAG-laser technique.

The use of lasers in upper airway surgery is now common practice. The introduction of the Nd-YAG laser technique makes it possible to perform endoscopic resection of tumours located in the trachea and central bronchi. Usually these patients require general anaesthesia. Our experience in 13 patients using total intravenous anaesthesia and jet ventilation with air is reported. Oxygen saturation was maintained at a higher level than when the patients were breathing 100% oxygen before anaesthesia. Occasional reductions in saturation were due to airway obstruction and were easily corrected by a short interruption of the procedure. All patients tolerated the anaesthesia and surgery well. No complications related to the anaesthetic method or the use of the Nd-YAG-laser occurred. A review of the possible hazards in these procedures is given together with advice on safety precautions needed.

Cerebral blood flow and oxygen consumption during isoflurane and halothane anesthesia in man.

In 13 patients, the effects on cerebral blood flow (CBF) and cerebral metabolic rate of oxygen (CMRO2) of isoflurane and halothane administered in a clinically relevant situation were studied. Measurements were performed during fentanyl/nitrous oxide (65%) anesthesia together with moderate hyperventilation (PaCO2 approx 4.5 kPa), and repeated after addition of 0.65 MAC of isoflurane (n = 6) or halothane (n = 7). CBF was measured after intravenous administration of 133xenon and CMRO2 was calculated from the arterial venous differences of oxygen content (AVDO2) determined in arterial and jugular venous bulb blood. CBF and CMRO2 (means +/- s.e. mean) determined prior to administration of volatile agents were 28 +/- 5 ml x 100(-1) x min-1 and 2.0 +/- 0.3 ml x 100 g-1 x min-1, respectively, in the isoflurane group. In the halothane group, CBF was 25 +/- 0.4 ml x 100 g-1 x min-1 and CMRO2 was 2.0 +/- 0.4 ml x 100 g-1 x ml-1. There were no significant intergroup differences. Isoflurane did not change CBF, whereas halothane produced an increase of 36% (P less than 0.05) compared to values obtained during fentanyl/N2O anesthesia. In addition, isoflurane caused a further decrease in CMRO2 of 12% (P less than 0.01) as compared to a 20% increase (P less than 0.05) with halothane. The cerebral metabolic depression caused by the short-acting anesthetic induction agents would be expected to decrease with time, and could partly explain the observed increase in CMRO2 produced by halothane. The study suggests that the cerebrovascular and metabolic properties of isoflurane differ from those of halothane, also in man.

Effects of nitrous oxide on cerebral haemodynamics and metabolism during isoflurane anaesthesia in man.

Seven normoventilated and five hyperventilated healthy adults undergoing cholecystectomy and anaesthetized with methohexitone, fentanyl and pancuronium were studied with measurement of cerebral blood flow (CBF), cerebral metabolic rate of oxygen (CMRO2), and quantified electroencephalography (EEG) under two sets of conditions: 1) 1.7% end-tidal concentration of isoflurane in air/oxygen; 2) 0.85% end-tidal concentration of isoflurane in nitrous oxide (N2O)/oxygen. The object was to study the effects of N2O during isoflurane anaesthesia on cerebral circulation, metabolism and neuroelectric activity. N2O in the anaesthetic gas mixture caused a 43% (P less than 0.05) increase in CBF during normocarbic conditions but no significant change during hypocapnia. CMRO2 was not significantly altered by N2O. EEG demonstrated an activated pattern with decreased low frequency activity and increased high frequency activity. The results confirm that N2O is a potent cerebral vasodilator in man, although the mechanisms underlying the effects on CBF are still unclear.

Comparison of ethanol absorption during continuous and intermittent flow irrigation in transurethral resection.

Transurethral resection of the prostate was performed using intermittent-flow bladder irrigation (n = 50), or by continuous-flow irrigation, using a suprapubic trocar (n = 50). The irrigant solution contained 1.5% glycine +1% ethanol and fluid absorption was measured from the ethanol content of the expired breath. Fluid absorption was significantly lower in patients receiving continuous-flow irrigation (p less than 0.007) although major absorption occurred in one of these patients. The immediate detection of absorption with the ethanol method allowed us to stop one of the operations performed with intermittent bladder irrigation, at which 2 l of fluid had been absorbed in 20 min. With correction for the amount of removed prostatic tissue, there were no differences in operation time or blood loss between the two types of irrigation.

Cerebral blood flow during carotid endarterectomy determined by three dimensional SPECT measurement; relation to preoperative risk assessment.

Cross-clamping of the carotid artery during carotid endarterectomy implies a risk of developing an ischaemic insult. To evaluate the effects of carotid artery occlusion on cerebral blood flow (CBF), both hemispheric and regional CBF (rCBF) were investigated using intravenously (i.v.) administered 133Xenon with 3 min clearance recording time for two-dimensionally (hemispheric CBF) and 99m-technetium-hexamethylpropylene amine oxime (99mTC-HMPAO) for three-dimensionally single photon emission computed tomography (SPECT) measurements (rCBF). Thirteen patients scheduled to undergo carotid endarterectomy anaesthetised with fentanyl/isoflurane participated in the study. Preoperative evaluation included investigation of rCBF with SPECT in all participants. Two intraoperative 133Xe CBF measurements were performed in each patient, before and after occlusion of the carotid artery. The preoperative rCBF measurement constituted the reference, for technical reasons, for the intraoperative investigations of rCBF during cross-clamping, which was completed immediately after the hemispheric measurements. The increase in preoperative risk evaluation as described by Sundt et al. and modified by Cho et al. corresponded excellently to a decrease in hemispheric CBF due to cross-clamping. A significant decrease in rCBF (p < 0.005) was present between patients with high and low preoperative risk score for the region of the middle cerebral artery. In this region, a correlation between decrease in rCBF and corresponding decrease in hemispheric CBF was also present. The present study demonstrates that the vascular regions of the ipsilateral middle cerebral artery are the most vulnerable vascular area during cross-clamping in individuals with high preoperative risk score.(ABSTRACT TRUNCATED AT 250 WORDS)

Effects of nitrous oxide on human regional cerebral blood flow and isolated pial arteries.

BACKGROUND: Results from previous studies on the effect of nitrous oxide (N2O) on the cerebral circulation are conflicting. Early reports claim N2O to have no effect whereas recent findings demonstrate a cerebral cortical vasodilatation during N2O inhalation, but the regional cerebral blood flow (CBF) in the subcortical structures is unknown. METHODS: Regional CBF was measured three-dimensionally with single photon emission computer-aided tomography after injection of xenon 133 in 8 spontaneously breathing men (mean age 29.6 yr) during normocapnia and hypocapnia with and without inhalation of 50% N2O. 8 isolated human pial arterial segments were mounted in organ baths. The segments were contracted with prostaglandin F2 alpha and subjected to 30% oxygen and 5.6% carbon dioxide in nitrogen or N2O. RESULTS: Normocapnic young men had a global CBF of 55 +/- 4 ml.100 g-1.min-1. Decreasing end-tidal CO2 tension by 1.3 kPa (9.3 mmHg) reduced CBF uniformly, with a decrease in global CBF to 45 +/- 2 ml.100 g-1.min-1 (P < 0.0001). During normocapnia, inhalation of 50% N2O increased mean CBF to 67 +/- 7 ml.100 g-1.min-1 (P < 0.0001). Inhalation of 50% N2O during hypocapnia increased mean CBF to 63 +/- 5 ml.100 g-1.min-1 (P < 0.0001). During N2O inhalation there was no significant difference in mean CBF between normo- and hypocapnia. However, during hypocapnia, but not during normocapnia, N2O inhalation significantly changed the distribution of regional CBF (P < 0.0001). Compared with hypocapnia without N2O, flow increased through the frontal (143%), parietal (140%) and temporal (133%) regions as well as through insula (151%), basal ganglia (145%) and thalamus (133%). In isolated human pial arteries, addition of N2O changed neither basal tension, nor the contraction elicited by prostaglandin F2 alpha. CONCLUSIONS: Inhalation of 50% N2O increased global CBF mainly by augmenting flow in frontal brain structures. In contrast, changes in carbon dioxide without N2O affected CBF uniformly in the brain. The uneven change in distribution of the CBF when N2O was added during hypocapnia, the reduced carbon dioxide response, and the lack of effect of N2O on isolated human pial arteries suggest that N2O may increase metabolism in selected brain areas.

Transplantation of lungs from a non-heart-beating donor.

BACKGROUND: In animals, we have previously done successful lung transplantations using organs from non-heart-beating donors. We have also developed an ex-vivo system of assessing the function of such organs before transplantation. The next stage was to try the technique in human beings. Bearing in mind the sensitive ethical issues involved, our first aim was to find out what procedures would be acceptable, and to use the results to guide a clinical lung transplantation from a non-heart-beating donor. METHODS: The ethical acceptability of the study was gauged from the results of a broad information programme directed at the general public in Sweden, and from discussions with professionals including doctors, nurses, hospital chaplains, and judges. The donor was a patient dying of acute myocardial infarction in a cardiac intensive-care unit after failed cardiopulmonary resuscitation. The next of kin gave permission to cool the lungs within the intact body, and intrapleural cooling was started 65 min after death. Blood samples were sent for virological testing and cross matching. The next of kin then had time to be alone with the deceased. After 3 h, the body was transported to the operating theatre and the heart-lung block removed. The lungs were assessed ex vivo, and the body was transported to the pathology department for necropsy. RESULTS: No contraindications to transplantation were found, and the right lung was transplanted successfully into a 54-year-old woman with chronic obstructive pulmonary disease. The donor lung showed excellent function only 5 min after reperfusion and ventilation, and during the first 5 months of follow-up, the function of the transplanted lung has been good. INTERPRETATION: About half the deaths in Sweden are caused by cardiac and cerebrovascular disease. This group could be a potential source of lung donors. When all hospitals and ambulance personnel in Sweden have received training in non-heart-beating lung donation, we hope that there will be enough donor lungs of good quality for all patients needing a lung transplant.

Simultaneous transcranial Doppler sonography and cerebral blood flow measurements of cerebrovascular CO2-reactivity in patients with aneurysmal subarachnoid haemorrhage.

Transcranial Doppler sonography (TCD) flow velocities and cerebral blood flow (CBF) measurements were evaluated in 14 patients who had suffered a major aneurysmal subarachnoid hemorrhage (SAH). Cerebrovascular reactivity to hypocapnia was evaluated simultaneously by the two methods. The measurements were performed under general anaesthesia preoperatively, within 72 hours after the bleed, during normocapnia and hypocapnia. There was poor correlation between absolute values of hemispheric CBF and corresponding TCD mean flow velocity. Controlled hyperventilation was associated with a significant decrease in CBF as well as TCD flow velocity (p less than 0.001). In terms of reactivity indices the correlation between the two methods was poor and not significant (r = 0.33, p = 0.09). The principal differences between the methods are discussed as well as the application of TCD in the evaluation of cerebrovascular reactivity.