Approximately 200 000 patients per year will require mechanical ventilation secondary to neurological injury or disease. The associated mortality, morbidity, and costs are significant. The neurological patient presents a unique set of challenges to airway management, mechanical ventilation, and defining extubation readiness. Neurological injury and disease can directly or indirectly involve the process involved with respiration or airway control. This article will review the basics of airway management and mechanical ventilation in the neurological patient. The current state of the literature evaluating extubation criteria in the neurological patient will also be reviewed.

I read with interest the recent article concerning the evaluation of an expired fraction carbon dioxide monitor (Ratnasabapathy et al. Anaesthesia 2002; 57: 900–4). Figures 5 and 6 both illustrate the difference between two variables (arterial PaCO2– portable ) plotted against the one of the same variables (arterial PaCO2). If the arterial PaCO2 and portable are considered to be two different methods of estimating a single clinical measurement (legitimised by plotting the difference between these two variables on the y-axis of Figs 5 and 6) then the correct method would be to plot the arithmetic mean of the arterial PaCO2 and portable on the x-axis. This error may arise with the use of correlation to study the relationship between an initial measurement and a change in that measurement over time, and for any two quantities X and Y, X will be correlated with X–Y; indeed, even if X and Y are samples of random numbers then one would expect the correlation between X and X–Y to be 0.7 [1]. The same problem appears to occur when measuring the agreement between two methods of clinical measurement: Ratnasabapathy et al. cite a reference by Bland and Altman which reports a statistical method for assessing agreement between two methods of clinical measurement [2]. Bland and Altman point out that it is a mistake to plot the difference between two variables against either value separately because the difference will be directly related to each; this statistical artefact is further illustrated in an earlier article by Gill et al. [3]. The correct method, as indicated above, is to plot the difference between two variables against the arithmetic mean of the two variables. Figure 3, from the article by Ratnasabapathy et al. appears to use the correct method in illustrating the relationship between fixed and portable . The general effect of using the incorrect method appears to be to increase the value of the correlation coefficient [3]. I would be pleased to hear if Ratnasabapathy et al. agree with my interpretation of the above methods. I cannot comment upon the effect of any new analysis of the data by Ratnasabapathy et al. We are grateful firstly to Dr Lewis for his interest in our paper, and for cogently illustrating the statistical argument, and secondly for the opportunity to reply to his points. He is absolutely correct in terms of statistical accuracy − there are errors inherent in charting the difference between a standard and test measurement against only the standard, in terms of the correlation produced. For interest's sake, we include the revised graphs of our data using the ‘statistically correct’ methodology (1, 2). They indeed show a change in the slope of correlation in Fig. 2, whilst Fig. 1 is largely unchanged. There remains, however, a measurement difference of 0.4 kPa at an average of 5 kPa pCO2 in Fig. 6. Herein lies part of the problem facing any researcher in depicting this type of data. What does charting the average of pCO2 and FÉCO2 tell the clinical user? We would argue that this was a study of clinical accuracy, and the graph was structured to depict the deviation between what we accept as our clinically important variable (i.e. pCO2) and the tested variable (FÉCO2). Our error here lay in including the correlation lines − we should have refrained from including these in the graphs. While Dr Lewis is entirely correct to say that the correlation between the measurements is more accurately related to the average of the measurements, the charting methodology does not change the difference between the two measurement techniques, which we believe to be significant. We also believe the reader needs to appreciate the difference over the clinical range of pCO2, so that they may then relate that to the clinical changes that appear with increasing pCO2. An average of the test and standard do not tell the reader that at a pCO2 of 5.6 kPa, the portable capnometer under-read by 1.5 kPa. The figure of 5.6 has more immediate clinical relevance, in terms of cerebral blood flow and intracranial pressure, than the average of 4.85. This is perhaps an argument between the practically useful graphical depiction of data and the accurate statistical interpretation of graphed data. Our essential point remains, however, that this type of monitor needs to be tested against the clinically relevant variable, i.e. pCO2, and in a variety of circumstances, so as to gain insight into its performance in the various environments it may be used. M. J. Souter Harborview Medical Center, Seattle WA 98104–2499, USA E-mail: [email protected]

Respiratory monitoring is an important aspect of critical care, especially in neurosurgery and neuro-intensive care. Fixed capnographs are too cumbersome to allow monitoring during patient transport. Recently several portable expired fraction carbon dioxide devices have been developed, but no evaluation of their clinical peformance has been reported. We compared the Criticare POET LT Handheld expired fraction carbon dioxide monitor, in three different settings, to fixed capnographs and arterial blood gas analysis. A methodology for systematic appraisal of end-tidal capnographs is proposed.

The apolipoprotein E (apoE) gene (APOE) is polymorphic with three alleles, e2, e3, and e4, which give rise to three isoforms, E2, E3 and E4. Many reports have now described a strong association between the e4 allele and risk of developing late onset Alzheimer's disease as the result of the E4 isoform binding to β-amyloid protein and accelerating the deposition of amyloid, which is the main constituent of senile plaques.1 The APOE e4 allele also appears to be associated with deposition of β-amyloid after traumatic brain injury, which is also accompanied by increased APOE expression in the central and peripheral nervous systems. Neurological and cognitive decrements are well documented complications of coronary artery bypass grafting (CABG) surgery. Given that APOE e4 is associated with deposition of β-amyloid after traumatic brain injury, and poor neurological outcome after subarachnoid haemorrhage and stroke, it may also adversely influence neurocognitive outcome after CABG surgery. In a preliminary report, Tardiff and colleagues2 found that the APOE e4 allele was associated with greater risk of cognitive impairment, especially in those patients with lower educational levels. More recently, Steed and colleagues3 were unable to replicate the findings in …

Systemic inflammatory response syndrome is a common clinical entity often presenting as a complication of neurological disease or trauma. Clear diagnostic criteria exist. Induced pathological mechanisms include both immunological and endothelial dysfunction, and coagulopathy, which may lead to multiple organ failure and significant morbidity. Possible therapeutic mechanisms are discussed, but this complex syndrome is poorly understood and successful treatment may depend on further research into control mechanisms.

Perioperative levels of jugular bulb oxyhaemoglobin saturation (Sj(O(2))) and lactate concentration (Lj), and postoperative duration of Sj(O(2))<50% were compared between patients undergoing coronary artery bypass grafting (CABG) (n=86), heart valve (n=14) and abdominal aortic (n=16) surgery. Radial artery and jugular bulb blood samples were aspirated after induction of anaesthesia, during re-warming on cardiopulmonary bypass (CPB) (36 degrees C), on arrival in the intensive care unit (ICU) and, subsequently, at 1, 2 and 6 h after ICU admission. Most patients having heart surgery were hypocapnic at 36 degrees C on CPB. Following CABG and heart valve surgery, many patients were hypocapnic whereas after abdominal aortic surgery, most were hypercapnic. During CPB and postoperatively, Sj(O(2)) and Lj were significantly correlated to Pa(CO(2)) and the arterial concentration of lactate (La) respectively (P<0.05). After correction for arterial carbon dioxide tension (Pa(CO(2))) and La, there were no significant changes in Sj(O(2)) or Lj on CPB. Postoperatively, having corrected for Pa(CO(2)), there were significant effects on Sj(O(2)) over all groups as a result of time from surgery (P<0.001) and its interaction with operation type (P<0.001). Following correction for La, there were no postoperative effects on Lj. No significant differences (P=0.2) in duration of Sj(O(2))<50% existed between patients undergoing CABG (1054 (82) min), abdominal aortic (893 (113) min) and heart valve (1073 (91) min) surgery. The lack of significant reciprocal effects on Lj combined with the frequency of hypocapnia and strong influence of Pa(CO(2))()on Sj(O(2)), suggest that Sj(O(2))<50% during CPB and after cardiac surgery represents hypoperfusion as a consequence of hypocapnia rather than cerebral ischaemia.

Systemic inflammatory response syndrome is a common clinical entity often presenting as a complication of neurological disease or trauma. Clear diagnostic criteria exist. Induced pathological mechanisms include both immunological and endothelial dysfunction, and coagulopathy, which may lead to multiple organ failure and significant morbidity. Possible therapeutic mechanisms are discussed, but this complex syndrome is poorly understood and successful treatment may depend on further research into control mechanisms.

Greater levels of S100β after coronary artery bypass grafting (CABG) surgery are thought to indicate cerebral injury [1]. Lipid peroxidation arising from oxidative stress occurs during cardiopulmonary bypass (CPB), and is indicated by increased malondialdehyde concentration [2]. We report the relationships between malondialdehyde production, S100β, and neurological and cognitive scoring 3 months after CABG.

During the early postoperative period after coronary artery bypass grafting (CABG) surgery, many patients experience jugular bulb oxyhemoglobin desaturation (Sjo(2) < 50%). We sought to determine whether Sjo(2) during cardiopulmonary bypass and the early postoperative period influenced long-term cognitive performance after CABG surgery. One hundred two patients completed a battery of cognitive tests the day before and 3 mo after CABG surgery. A General Cognitive Score was generated from these tests as an overall measure of cognitive function. Intraoperatively, Sjo(2) was determined by intermittent blood sampling, and postoperatively, the area under the curve of Sjo(2) < 50% and time was calculated from continuous reflectance oximetry. No significant correlations between cognitive performance and either intra- or postoperative Sjo(2) were found. Preoperative cognitive performance was the main determinant of cognition at 3 mo (r(2) = 0.83, P ( 0.001), and palpable atheroma of the ascending aorta made a small, but significant, contribution to a decline in cognition (r(2) = 0.018, P = 0.001).

Objective: To ascertain if norepinephrine can be used as part of the cerebral perfusion pressure (CPP) management to increase arterial blood pressure (MAP) without causing cerebral hyperemia after severe head injury (HI).¶Design: Prospective, interventional study.¶Setting: Intensive care unit in a university hospital.¶Patients: Twelve severely HI patients; median Glasgow Coma Scale was 6 (range 3–8).¶Interventions: CPP management ( = 70 mmHg). Pressure autoregulation (assessed by norepinephrine infusion) was defined intact if %CPP/%CVR ≤ 2.¶Results: Cerebral blood flow (CBF: Xe133 inhalation technique), jugular bulb oxygen saturation (SjO2) and transcranial Doppler (TCD) were recorded during the test. Norepinephrine increased CPP by 33 % ( ± 4). Autoregulation was found to be intact in ten patients and defective in two. In the ten patients with preserved autoregulation, CBF decreased from 31 ± 3 to 28 ± 3 ml/100 g/min; in the two patients with impaired autoregulation CBF increased respectively from 16 to 35 and from 21 to 70 ml/100 g/min. SjO2 did not change significantly from baseline. TCD remained within the normal range.¶Conclusions: During CPP management norepinephrine can be used to increase MAP without potentiating hyperemia if pressure autoregulation is preserved. The assessment of pressure autoregulation should be considered as a guide for arterial pressure-oriented therapy after HI.

Objective: To investigate the role of the endothelin system in pressure autoregulation of cerebral blood flow (CBF) in rats.Design: We tested pressure autoregulation by increasing cerebral perfusion pressure (CPP; mean arterial pressure–intracranial pressure) with norepinephrine (0.08 μg · kg−1· min−1 for 30 min) twice in ten anesthetized normocapnic rats. The first test was performed without (control test) and the second test after administration of the combined endothelin ETA/B receptor antagonist, bosentan, i. v. (30 mg/kg; drug test). CBF was measured by the hydrogen clearance technique.Results: During the control test, norepinephrine infusion increased CPP by 21 ± 2 (23 ± 2 %) mmHg (mean ± SEM; p < 0.001) and CBF by 3.6 ± 3.1 (6 ± 8 %) ml/100 g/min (p = 0.5, Fig. 1); during the drug test, norepinephrine infusion increased CPP by 18 ± 1 (20 ± 2 %) mmHg (p < 0.001) and CBF by 15.8 ± 4.1 (46 ± 13 %) ml/100 g/min (p = 0.004). Mean arterial pressure was not affected by bosentan infusion (p = 0.2). PaC02 levels were stable during the tests (40.2 ± 1.4 mmHg).Conclusions: The endothelin system is involved in cerebral pressure autoregulation in a rodent model in vivo. The role of this system under pathophysiologic conditions such as subarachnoid hemorrhage, where basal vascular tone and its regulation may be altered, remains to be defined.

Introduction: Jugular bulb oxyhaemoglobin desaturation (SjO 2 ≤ 50%) during the rewarming phase of cardiopulmonary bypass (CPB) is associated with postoperative cognitive deficits [1].Isoflurane effects cerebral blood flow and cerebral metabolic rate, both of which affect SjO 2 [2].We report the effect of isoflurane on the incidence of desaturation during rewarming from CPB for coronary artery surgery.

We studied 15 patients undergoing cardiac surgery involving hypothermic cardiopulmonary bypass (CPB). Cerebral arteriovenous difference in oxygen content (AVDO2) was significantly less during CPB and for up to 18 h after operation compared with pre-CPB values (P < 0.05). There were no significant changes in mean jugular bulb oxyhaemoglobin saturation (SjvO2), cerebral arteriovenous difference in lactate content or lactate-oxygen index (LOI). SjVO2 and arterial carbon dioxide tension (PaCO2) (P = 0.005) were positively correlated as were AVDO2 and haemoglobin concentration (P = 0.012). AVDO2 and PaCO2 (P = 0.007) were negatively correlated as were LOI and arterial oxyhaemoglobin saturation (P = 0.037). There were no significant correlations between mean arterial pressure and any of the variables. SjVO2 and AVDO2 may require correction for changes in PaCO2 and haemoglobin concentration before relating these variables to cerebral outcome.

Fibreoptic jugular bulb oximetry has been validated for use in the care of severely head-injured patients. We compared bench and fibreoptic methods of measuring jugular bulb oxyhaemoglobin saturation (SjO2) in 33 patients undergoing cardiac surgery both during cardiopulmonary bypass (CPB) and in the early postoperative period. After insertion of a fibreoptic reflectance oximetry catheter into the jugular bulb, it was calibrated against a bench oximeter. Comparisons were made while on CPB (n = 60) and in the postoperative period for up to 18 h (n = 215). There was negligible bias throughout. There were wide limits of agreements (mean difference +/- 2SD) between the two methods during operation (-20.29% to 18.05%), whereas after operation the limits of agreement were far narrower (-6.39% and 7.45%). Measurement of SjO2 by the fibreoptic method compared poorly with bench oximetry during CPB but there was good agreement between the two methods in the early postoperative period.

Technical innovations in neuroimaging have improved diagnosis and prognosis, whereas developments in interventional neuroradiology have extended the range of therapy to different patient populations. These changes in service demand the identification of those clinical and technical factors distinguishing feasibility from futility, in order to increase population efficiency and reduce the harm associated with inappropriate therapy.

Reductions in cerebral venous oxyhemoglobin saturation (SjO2) occur during the rewarming phase of hypothermic cardiopulmonary bypass (CPB).We prospectively investigated the effects of propofol on these reductions in SjO2 (SjO2 <50%). Fiberoptic jugular bulb catheters were inserted in 30 patients undergoing coronary artery bypass grafting. Patients were randomly allocated to a test or control group. Test group patients (n = 15) received a propofol IV infusion titrated to electroencephalographic burst suppression during CPB. No significant differences in SjO2 <50% were found between the groups either by blood sampling and bench oximetry or fiberoptic oximetry. The arteriovenous difference in lactate concentration became negative in 59 of 120 samples. Propofol was associated with an increased incidence of hypotension (mean arterial pressure <50 mm Hg) (P = 0.023), an increased requirement for vasoconstrictor therapy (P = 0.025), and increases in the lactate oxygen index (P < 0.01). Propofol, when administered in doses that produce electroencephalographic burst suppression, does not attenuate the frequency or extent of reductions of SjO2 below 50% during rewarming from hypothermic CPB. However, it is associated with arterial hypotension and an increase in cerebral anaerobic metabolism. Implications: Reductions in cerebral venous oxyhemoglobin saturation during the rewarming phase of cardiopulmonary bypass may be related to brain injury. When administered in doses sufficient to produce electroencephalographic burst suppression, propofol did not attenuate the frequency or extent of such reductions in cerebral venous oxyhemoglobin saturation. (Anesth Analg 1998;86:926-31)

Diaspirin cross-linked haemoglobin (DCLHb) is a new oxygen carrying blood substitute with vasoactive properties. Vasoactive properties may be mediated via high affinity binding of nitric oxide by the haem moiety. Using a rodent model of head injury combined with ischaemia, we studied the effects of DCLHb on cerebral blood flow (CBF) and intracranial pressure (ICP). Twenty anaesthetized rats were allocated randomly to receive treatment with DCLHb 400 mg kg-1 i.v. or placebo (oncotically matched plasma protein substitute 4.5% i.v.). To produce diffusely increased ICP, after a severe weight drop injury, all animals underwent a 30-min period of bilateral carotid ligation combined with a period of induced hypotension. After reperfusion, DCLHb or placebo was infused and the animals instrumented for measurement of intraventricular ICP and CBF in the region of the sensorimotor cortex using the hydrogen clearance technique. Mean arterial pressure (MAP), ICP, cerebral perfusion pressure (CPP) (CPP = MAP - ICP) and CBF were measured 4 h after injury in all animals. DCLHb significantly reduced ICP from mean 13 (SEM 2) to 3 (1) mm Hg (P < 0.001), increased CPP from 52 (8) to 95 (6) mm Hg (P < 0.001) and increased CBF from 21 (2) to 29 (2) ml 100 g-1 min-1 (P = 0.032). We conclude that DCLHb improved CPP without a reduction in CBF in a rodent model of post-traumatic brain swelling.