The following abstracts were presented at the Annual Meeting of the Group of Anaesthetists in Training (GAT) in Glasgow, June 2012


To see or not to see: radio-opacity of epidural catheters

S. Gibb, V. Bythell, C. Ripley and R. Gauntlett

Royal Victoria Infirmary, Newcastle upon Tyne, UK
Email:
sarahgibb@doctors.org.uk

An obstetric patient developed back pain following accidental shearing of an epidural catheter. Due to her symptoms, it was decided to attempt to localise the retained segment. The manufacturer’s advice stated that the epidural catheter which had been used was radio-opaque; however plain x-ray, ultrasound scan and magnetic resonance imaging (MRI) all failed to reveal the fragment. We aimed to further investigate the radio-opacity of a variety of epidural catheters.

Methods

We obtained samples of seven epidural catheters, six of which were described as radio-opaque. We used a loin of pig (lumbar spine, muscles, connective tissues and skin) to simulate the radiological appearances of the human lumbar spine. We performed an initial control x-ray of all seven catheters. We then inserted each epidural catheter into the porcine lumbar epidural space, leaving a radio-opaque marker at the site of skin insertion, and x-rayed the model in the lateral position with a horizontal beam (Siemens Aristos, Camberley, Surrey, UK). The films were reviewed by a consultant radiologist.

Results

All seven epidural catheters were visible on the initial control x-ray. Of the six radio-opaque catheters, only three were visible on plain x-ray after insertion in the porcine model.

Discussion

Although rare, shearing of an epidural catheter with retention of a fragment is distressing for both the patient and the clinician responsible. The available literature advises leaving the retained piece in-situ unless neurological symptoms develop [1]. We have demonstrated that 50% of the epidural catheters marketed as radio-opaque are not visible on plain x-ray when inserted into a porcine model. Localisation of those catheters that could be identified was aided by the presence of the full length of catheter and the radio-opaque insertion site marker. On radiology review it was felt that a short segment retained within the soft tissues would likely have been less readily visible. Ideally, we would like to repeat this investigation using other imaging modalities and shorter fragments.

Acknowledgments

We would like to acknowledge the assistance of the research radiographers, Clare Moody and Celia Miller.

Reference

Emergency use of airway equipment

T. Barge and P. Duggleby

Stoke Mandeville Hospital, Aylesbury, UK
Email:
tom.barge@doctors.org.uk

The Difficult Airway Society provides robust guidelines for the management of unanticipated difficult tracheal intubation [1]. The guidelines involve the use of a range of techniques and airway equipment. We investigated whether the frequency of use of airway equipment electively correlated with the anaesthetist’s confidence to use the equipment in an emergency, such as unanticipated difficult or failed tracheal intubation.

Methods

We surveyed anaesthetists in our hospital using a standardised written questionnaire. We asked them to rate how frequently they use various pieces of airway equipment electively, and how confident they would feel using the equipment in an emergency. The equipment surveyed (15 different types) was all available on our hospital’s difficult airway trolley [2]. We used contingency tables and the chi-squared test for statistical analysis.

Results

Twenty-two consultants, 16 middle grades (registrars, staff grades and associate specialists) and 12 junior trainees (CT 1-2) were surveyed. There was a significant correlation between the frequency a piece of airway equipment was used electively and how confident responders would feel using the equipment in an emergency (p < 0.001) (Table 1). The decline in confidence was significant once equipment was used on a yearly or less frequent basis (p < 0.001), but not if used monthly or more frequently (p = 0.096). The rate of decline in confidence was significantly greater the more junior the anaesthetist (p < 0.001).

Table 1.   Anaesthetists’ confidence to use airway equipment in an emergency relative how frequently they use the equipment electively. Values are number (proportion).
Frequency of useVery confidentPartially confidentNot confident
Weekly146 (97%)3 (2%)1 (1%)
Monthly101 (93%)8 (7%)0
Yearly62 (53%)46 (40%)8 (7%)
>Yearly83 (22%)159 (42%)133 (35%)

Discussion

Our data demonstrates that confidence to use a piece of airway equipment in an emergency correlates with how frequently it is used on an elective basis. Monthly usage appears to be necessary to maintain confidence, particularly for more junior anaesthetists. We suggest anaesthetists should incorporate equipment they may use in an emergency into their elective practice. For some equipment, such as cannula or surgical crycothyroidotomy sets, this would not be feasible. Regular use of such equipment in simulators may improve confidence to use such equipment in an emergency.

References

What becomes of a broken heart?

K. Morton

Lister Hospital, Stevenage, UK
Email:
katmorton@doctors.net.uk

This case report describes a rare but potentially fatal complication of atrial ablation treatment. The difficulty in diagnosis and perils of investigation are discussed.

Case report

A 46 year-old man presented to the Emergency Department having collapsed at home. On examination he had a Glasgow Coma Score of 8/15, was hypotensive and tachycardic. Past medical history included paroxysmal atrial fibrillation for which he had undergone atrial ablation three weeks ago. Initial investigations included a normal computerised tomography (CT) scan of the brain, however echocardiography revealed an echogenic, globular structure in the left atrial area. This led to a presumed diagnosis of endocarditis, and broad spectrum antibiotics were commenced. The patient then deteriorated with unstable cardiac dysrhythmias and an episode of haematemesis. Oesophagastroduodenoscopy (OGD) demonstrated clotted blood but nothing else of note. A repeat brain and chest CT scan revealed multiple cerebral haemorrhagic infarcts with pockets of air within the brain as well as a small collection of air in the left atrial area. The diagnosis of atrial-oesophageal fistula was made and the patient was transferred to a cardiothoracic centre. Unfortunately he suffered a pulmonary embolism and died before surgery.

Discussion

Atrial-oesophageal fistula is a rare complication of radiofrequency ablation, with an incidence of 0.01–0.2% [1]; mortality is high (up to 68%). This case report highlights the difficulty with diagnosis as the condition mimics more common pathology such as sepsis and endocarditis. In one study of nine cases of atrial-oesophageal fistula, all of which died, only four were correctly diagnosed before death [2]. As demonstrated in this case, a common presentation is haematemesis, and OGD may increase mortality by precipitating complications such as cerebral infarction due to air. Mortality may be improved if clinicians are more aware of the diagnosis and OGD is not performed.

References

Automatic referral of patients after thoracic surgery to the pain team

S. Soni, S. Maitre, S. Thomas and M. Koertzen

Hammersmith Hospital, London, UK
Email:
sanooj_soni@yahoo.co.uk

Chronic pain after thoracic surgery represents a significant clinical problem. Thoracotomy, along with limb amputation, is the procedure that elicits the highest risk of severe chronic postoperative pain [1]. Studies have shown that in this patient group, high levels of immediate postoperative pain are associated with an increase in the incidence of chronic pain [2]. The pain team play an important role in managing acute pain in patients undergoing thoracic surgery and therefore minimising the frequency of chronic pain. We examined whether post-thoracic surgical patients were being referred to and reviewed by the acute pain team.

Methods

Our local guidelines state anaesthetists should refer every patient undergoing thoracic surgery to the acute pain team, ensuring that pain specialists see all these patients postoperatively. We performed a two-month prospective audit examining how often thoracic patients were referred to the acute pain team. Following this, we developed a series of recommendations and then re-audited.

Results

Between October and November 2011, a total of 68 patients underwent thoracic surgery. Of these, only 14 (20%) were referred by the anaesthetist to the pain team for postoperative review. Seventeen patients (25%) had an epidural or paravertebral catheter and none of these were referred to or reviewed by the pain team. These results were deemed to be unacceptable, therefore a department-led teaching programme was undertaken, highlighting the pain team guidelines. Furthermore, we created a separate database of all thoracic surgical patients, which triggered an automatic referral to the acute pain team. Following these interventions, a re-audit of all thoracic patients from December 2011 to January 2012 demonstrated a dramatic improvement. Referrals by anaesthetists improved to 57/83 (68%). However, this was further improved by the automatic referral system to 75 patients (90%, p < 0.0001).

Discussion

Education and initiation of an automatic referral system resulted in a large improvement in referrals to the pain team. This resulted in daily review by the pain team, with the aim of improving the management acute postoperative pain and reducing the incidence of future chronic pain. We suggest that similar automatic referral systems should be instituted in other thoracic units across the country.

References

An unusual case of bilateral vocal cord palsy

S. Bratanow and M. Davidson

Royal Devon & Exeter Hospital, Exeter, UK
Email:
s.bratanow@nhs.net

A recent national survey showed that almost 30% of all adverse events associated with anaesthesia occur at tracheal extubation or during recovery [1]. We present a case of unexpected bilateral vocal cord palsy causing severe stridor after repeated attempts at extubating the trachea.

Case report

A 24-year old previously fit and well woman presented to our emergency department after a witnessed prolonged tonic-clonic seizure that followed exercise, which was complicated by vomiting. Her Glasgow Coma Score was three, and she required tracheal intubation and mechanical ventilation of the lungs. A computerised tomography scan of the brain was unremarkable. On day-two, magnetic resonance imaging of her brainstem, lumbar puncture and cerebro-spinal fluid analysis were also normal. The electroencephalogram (EEG) was consistent with global hypoxic injury. Failed tracheal extubation on day-three was thought to be secondary to a chest infection, although expiratory stridor was also noted. Neurological recovery was slow, but by day-six she was starting to obey commands. After a further attempt at tracheal extubation, there was marked stridor, unresponsive to corticosteroids or nebulised adrenaline. We therefore performed nasendoscopy under remifentanil sedation, which showed bilateral vocal cord palsy with no other abnormality to the larynx and its surrounding structures. The trachea was re-intubated and a percutaneous tracheostomy performed. Serial nasendoscopies confirmed recovering vocal cord function and five weeks after discharge from the intensive care the tracheostomy was removed and normal bulbar function demonstrated.

Discussion

Vocal cord palsy is most commonly caused by injury to the recurrent laryngeal nerve during head and neck procedures. Rarely, as in our case, it can also result from a hypoxic insult to the brainstem, although this is more frequently seen in children [2]. Flexible nasendoscopy / laryngoscopy allows the vocal cords to be visualised and movement patterns during phonation observed. It is a simple technique, which we believe can easily be learned by trainee anaesthetists and performed at the bedside. We found it very useful in guiding further management in our patient where we had not expected any difficulties with tracheal extubation. Recent guidelines from the Difficult Airway Society promote the concept of an extubation strategy for each individual patient [3]. It must not be underestimated that problems can occur even in the expected ‘low-risk airway’ and a backup strategy must always be in place.

References

An audit of the peri-operative care of high-risk surgical patients

D. Palmer, J. Oates and S. Mercer

Aintree University Hospitals NHS Foundation Trust, Liverpool, UK
Email:
davidpalmer@nhs.net

Of patients undergoing surgery in the UK, 1.6% die within 30 days; the great majority (80%) come from a ‘high-risk’ group of patients [1]. A recent UK audit of high-risk patients highlighted deficiencies in peri-operative care [1]. We performed an audit to examine the care of high-risk patients at our institution.

Methods

A prospective audit was carried out over seven days; patients were included if their predicted mortality exceeded 5% using the P-POSSUM score [2], or if their anaesthetist identified them as high-risk. The seven audit criteria were adapted from national guidance [3]: elective high-risk patients should be seen in a pre-operative clinic; estimated mortality should be recorded on the consent form; patients should have active consultant input in their care; predicted mortality greater than 10% requires direct supervision by consultants in surgery and anaesthesia; an ‘end-of-surgery bundle’ should determine optimal location for immediate post-operative care; high-risk patients should be referred to critical care; and patients with an estimated mortality greater than 10% should be admitted to critical care.

Results

A total of 205 patients were underwent surgery during the audit period, and 160 (78%) were screened to determine risk. Of these, 35 (22%) were identified as high-risk by their anaesthetist and 20 (13%) had a P-POSSUM predicted mortality ≥5%; overall 40 (25%) were high-risk by either measure. Thirty-four (86%) high-risk elective cases were seen in a pre-operative clinic. Only 18% of all high-risk cases had ‘death’ mentioned on the consent form. Overall, 90% and 85% of cases were conducted by consultants in surgery and anaesthesia, respectively. An end-of-surgery bundle was not implemented in any case. Critical care referral took place in 37% of unplanned cases, and a further 38% of elective cases were referred to critical care, although this improved to 62% if day-case urology patients were excluded (risk often mitigated by use of subarachnoid block). Of those patients with a P-POSSUM predicted mortality ≥10%, seven (58%) went to critical care, two (17%) were considered ‘too stable’ and were not referred and two (17%) were not referred because of medical co-morbidity. An end-of-surgery bundle comprising arterial blood gas, serum lactate and P-POSSUM score was introduced in the emergency theatre. A sticker to record the bundle elements was appended to blank anaesthetic charts. A further seven day period of data collection identified six high-risk cases from 36 patients, all of whom were referred to critical care. The bundle was used in three cases.

Discussion

National guidelines recommend the end-of-surgery bundle to determine location for immediate post-operative care [3]. This is likely to have greatest utility in unplanned cases. We illustrated how a novel bundle using commonly measured parameters can facilitate critical care referral in unplanned surgery.

References

Delayed diagnosis of non-convulsive status epilepticus

R. Parmar, R. Dolan, Y. W. Li, R. Singh and M. S. Youssef

Walsall Healthcare NHS Trust, Walsall, UK
Email:
rinesh.parmar@doctors.org.uk

Presentation of non- convulsive status epilepticus may be subtle and a high index of suspicion is required. Potentially reversible delayed diagnosis can result in preventable morbidity and mortality. We present a recent case which cause diagnostic difficulty.

Case report

A 58-year old man with a history of alcoholic liver and chronic kidney disease presented with a normal Glasgow Coma Score, a flail chest and bilateral pneumothoraces following a road traffic collision. Chest drains were inserted and the patient was monitored in critical care, where he deteriorated requiring respiratory followed by multi-organ support. During a long admission complicated by sepsis, hepato-renal syndrome and severe thrombocytopenia, he had a single tonic-clonic seizure. Imaging revealed no acute pathology. No further seizures were observed, although in the following days it was noted there was some periodic twitching of his right arm. However, despite propofol and alfentanil sedation being discontinued for a week, treatment for hepatic encephalopathy, and biochemical abnormalities being at pre-seizure levels, the patient failed to rouse appropriately. An electroencephalogram (EEG) was performed, which demonstrated repetitive periodic epileptiform complexes, confirming that the twitching was not correlated with EEG activity. A diagnosis of non-convulsive status epilepticus was made and phenytoin was administered, following which the patient became rousable and was eventually discharged home.

Discussion

Non- convulsive status epilepticus is characterised by impaired consciousness, minimal motor features in the form of facial or limb twitching and an EEG which displays general, lateralised or regional periodic patterns [1]. Estimated incidence is 5.6–18.3/100 000 and it may be found in 8–30% of patients with altered mental state. Associated morbidity and mortality is reported to be 39% and 18% [2]. Risk factors include previous epilepsy, extremes of age, critical illness, female gender, genetic predisposition and acquired brain insults. Management involves preventing seizures and treatment of the underlying cause. Diagnosis in critical care may be difficult as we found, and a high index of suspicion is required; EEG should be considered early.

Acknowledgments

Published with the written consent of the patient.

References

Emergency laparotomy: comparison with national standards.

P. Howells and V. Poongavanam

George Eliot Hospital, Nuneaton, UK
Email:
vivekananthan.poongavanam@geh.nhs.uk

Emergency laparotomy is common, with the potential for substantial morbidity. We wanted to compare local performance to a recent national audit [1].

Methods

A pre-existing audit template was taken from the Emergency Laparotomy Network website [1]. All laparotomies undertaken in the year up to 30th November 2011 were identified and available notes were retrospectively analysed.

Results

A total of 102 laparotomies were performed between November 2010 and November 2011, and fifty case notes were analysed. The majority of patients were admitted under general surgery, but 12 (24%) were initially admitted under medical specialities. Thirty-five (70%) laparotomies were performed by consultant surgeons, whilst just 20 (40%) of the patients had a consultant anaesthetist in theatre. None of our patients were managed using cardiac output monitoring. Eight patients (16%) were admitted to high dependency or intensive care. All 50 patients survived their in-patient stay and were alive at 30 days.

Discussion

A substantial proportion of our patients are admitted under the care of the physicians; this has been shown to be associated with worse outcome [2]. A higher than expected proportion of procedures is carried out by consultants. There is currently under-utilisation of cardiac output monitoring and high dependency/intensive care facilities for the patient population we are treating. Overall, our audit has shown multiple areas for improvement and we await with interest the update on the national audit to benchmark our performance.

References

Post-cardiac arrest hyperoxia

J. Harkins, P. Jefferson and D. R. Ball

Dumfries and Galloway Royal Infirmary, Dumfries, UK
Email:
jacquelineharkins@nhs.net

There are approximately 50 000 cardiac arrests in the UK every year, of which approximately 6000 are admitted to intensive care [1]. Recent guidelines have advised patients’ inspired oxygen concentration should be adjusted to achieve arterial oxygenation between 94 and 98% [1].

Methods

We looked at all cases of out-of-hospital cardiac arrest admitted to the intensive care unit (ICU) at Dumfries and Galloway Royal Infirmary from January 2008. We excluded cases without a primary cardiac cause and those referred to the Procurator Fiscal. The data from 17 cases was analysed. We accessed the results of the first three arterial blood gases (ABG) using the laboratory browser: the first ABG from each patient in the emergency department; the first ABG after admission to ICU; and a subsequent ABG in ICU. The standard for the audit was that no patients should have a PaO2 > 40 kPa, and all patients should have a PaO2 between 10 and 14.9 kPa.

Results

Eight patients (62%) had PaO2 > 40 kPa in the emergency department, and only two (15%) had PaO2 between 10 and 14.9 kPa. Looking at the first ABG taken in ICU, eight (47%) had PaO2 > 40 kPa and only one (6%) had PaO2 between 10 and 14.9 kPa. None of the patients had PaO2 > 40 kPa subsequently, and only 4 (27%) patients had PaO2 between 10 and 14.9 kPa.

Discussion

Hyperoxia was common in our patient cohort. There may be an association between arterial hyperoxia following resuscitation from cardiac arrest and in-hospital mortality [2], and a significant linear trend between increasing PaO2 after cardiac arrest and in-hospital mortality [3]. We suggest updating our local policies in the emergency department and the ICU, and perhaps using oxygen saturation levels to guide titration of inspired oxygen concentration rather than ABG results.

References

Adrenaline resistance in anaphylaxis and the use of angiotensin converting-enzyme inhibitors

S. Hodge

Dorset County Hospital, Dorchester, UK
Email:
sarahehodge@doctors.org.uk

I present a case of cardiac arrest secondary to anaphylaxis that was refractory to intravenous adrenaline but responsive to metaraminol.

Case report

A 59-year old gentleman attended for elective arthroscopic subacromial decompression of the shoulder. Regular medication included lisinopril. He had no known drug allergies and had undergone previous uneventful general anaesthesia. Due to his hiatus hernia and obesity, a modified rapid sequence induction technique was performed, using fentanyl 100 mcg, propofol 180 mg and suxamethonium 100 mg. Within one minute of tracheal intubation the patient suffered a cardiac arrest (pulseless electrical activity). There were no stereotypical signs of an anaphylactic reaction. There was no cardiovascular response to 4 mg intravenous adrenaline, but following 5 mg metaraminol and a fifth dose of 1 mg adrenaline he developed ventricular fibrillation. We defibrillated the patient twice and administered a further 5 mg bolus of meteraminol, at which point spontaneous circulation was restored after a total of 35 minutes of cardiac arrest. The patient’s lungs were mechanically ventilated overnight and the next day there were no signs of neurological injury. On subsequent testing, he was shown to be allergic to suxamethonium and chlorhexidine (to which he had not been exposed).

Discussion

Anaphylaxis leading to cardio-respiratory arrest is relatively uncommon, but as this case demonstrates the typical physical signs of anaphylaxis may not present. Anaphylaxis may be resistant to adrenaline, in which case the use of vasoconstrictors is well described [1]. Angiotensin converting enzyme (ACE) inhibitors may play a role in anaphylaxis. Blockade of ACE inhibits the normal compensatory vasoconstrictive response to hypotension, thereby confounding cardiovascular collapse. Furthermore, the use of ACE-inhibitors is proven to be an independent risk factor for developing severe allergies to food [2] and insect venom [3] by increasing plasma bradykinin levels, so it may be postulated that their use may also exacerbate allergic reactions to drugs, such as in this case.

References

Evaluation of new guidelines for checking the anaesthetic machine

H. Reddy, L. Bowen, C. Bailey, P. Clyburn and L. Gemmell

Wrexham Maelor Hospital Wrexham, and University Hospital of Wales, Cardiff, UK
Email:
lowribowen@gmail.com

This evaluation was undertaken to benchmark test the new Association of Anaesthetists of Great Britain and Ireland (AAGBI) guidelines for checking the anaesthetic machine against current practice [1]. We engaged a range of anaesthetists to test and obtain feedback on the proposed format.

Methods

Participants were randomised into two equal-sized groups via sealed envelopes. Each group was assigned two different pre-determined faults. Whilst carrying out their usual machine checks, participants were timed and scored on the completeness of their check compared to the new guidelines and on their identification of a fault. The process was then repeated whilst following a copy of the new guidelines, and a different fault used. Only one fault was placed in each check depending on what envelope the participants had chosen. Participants were asked if they routinely carried out any functional checks between cases (which is a new six point additional check in the guidelines) and to comment on the guidelines’ format.

Results

We recruited 52 non-consultant anaesthetists: 16 CT1-2; 29 ST3-7; and seven non-consultant career-grade doctors. The usual machine check scored a median (IQR [range]) 14 (13–16 [4–20]) and took a 239 (180–351 [5–695]) seconds to complete. The time to finding a fault was 161 (123–257 [11–611]) seconds. Using the new guidelines, completion time was 290 (232–381 [92–714]) seconds, and time to finding a fault was 158 (100–281 [36–500]) seconds. Four anaesthetists failed to identify a fault during their routine check compared with two when using the guidelines. Participants commented on the format of the guideline using a visual analogue score of 0 (awful)–10 (fantastic). Layout was awarded 8 (8–9 [3–10]), legibility 9 (8–9 [5–10]), and safety implications 8 (8–9 [2–10]).

Discussion

Although the new guideline check took longer to perform, we anticipate that times will improve with repeated practice and exposure. The time to fault finding was recorded but is inevitably dependant on what the fault is and the order of the checking. The important safety point is whether the fault was identified, not how long it took. The functional check was identified as emphasising what should be checked between cases as this wasn’t addressed in the previous guidelines. Most anaesthetists were positive about the legibility, layout and safety implications. Qualitative comments included the ease and logical progression of the guidelines. Several people suggested the check should either be recorded in the notes or become part of the World Health Organisation safety checklist documentation

Reference

Measurement of high sensitivity troponin-T in critical care patients with acute brain injury

G. Campbell, R. Milton, P. Bunting, J. Baldwin and N. Doherty

Royal Preston Hospital, Preston, UK
Email:
gecampbell34@doctors.org.uk

Release of cardiac biomarkers is well documented following subarachnoid haemorrhage (SAH) [1], though little is written about traumatic brain injury or on the early release of troponins. High sensitivity troponin-T is a cardiac-specific structural protein that is released after myocyte necrosis; it is more sensitive than other assays [2]. The primary aim of this study is to monitor changes in serum high-sensititivty troponin-T after SAH and traumatic brain injury

Methods

Following ethical approval, adult patients with severe traumatic brain injury or SAH requiring mechanical ventilation of the lungs were enrolled in the study. Time-zero was set as the time the ambulance arrived at the scene, and arterial blood samples were taken after four, eight, 12, 16, 20, 24 and 48 hours. These patients were later followed up after one month, and Glasgow Outcome Scores (GOS) [3] were recorded: GOS-1 death; GOS-2 persistive vegetative state; GOS-3 severe disability; GOS-4 moderate disability; and GOS-5 good recovery.

Results

Thirteen (52%) patients had SAH and 12 (48%) traumatic brain injury. Eighteen (72%) patients had a rise in plasma high-sensitivity troponin-T of any magnitude at any time. Higher values were seen after SAH, median (range) 275, (22–696), compared to traumatic brain injury 31 (19–171). Values peaked before 24 hours, with three patients returning to normal within 48 hours. At one month, 10 (40%) patients were GOS-1; four (16%) GOS-2; four (16%) GOS-3; four (16%) GOS-4; and three (12%) GOS-5. There was no association between GOS and peak high-sensitivity troponin-T in patients with SAH, but in patients with traumatic brain injury higher values were associated with GOS-1,2 and 3.

Discussion

The use of a high-sensitivity troponin-T assay increased the detection of abnormal levels of cardiac troponin after both SAH and traumatic brain injury. These findings suggest that neurogenic myocardial injury may be more prevalent following acute brain injury than previously thought [3].

References

1. Priebe HJ. Aneurysmal subarachnoid haemorrhage and the anaesthetist. British Journal of Anaesthesia2007; 99: 102– 18.

2. Clerico A, Giannoni A, Prontera C, Giovanni S. High-sensitivity troponin: a new tool for pathophsiological investigation and clinical practice. Advances in Clinical Chemistry2009; 49: 1– 30.

3. Lim HB, Smith M. Systemic complications after head injury: a clinical review. Anaesthesia2007; 62: 474– 82.

Ancillary