Symposium: The Management of Head Injuries
The meeting commenced at 2pm with an introduction and welcome from Dr John Chambers and Dr Mike Trask, Presidents of the Liverpool Society of Anaesthetists and Section of Anaesthesia, Manchester Medical Society respectively.
1) Trends in Head Injury Outcome
Mr Andrew King, Consultant Neurosurgeon, Hope Hospital, Manchester
Mr King started by looking at the results of the trauma audit research network database for patients presenting with head injuries between 1989 and 2003. He commented that with medical management in the 1980’s, the general consensus was that a third of trauma deaths were avoidable. Survival improved between 1989 and 1994 with the introduction of paramedics and acute trauma life support courses. However, since 1994, there had been little change in outcome from head injuries. Of all trauma patients, 12% were associated with significant head injuries. They had a mortality of 25% and tended to be younger and more severely injured patients. Mr King then looked at the paper he had produced in association with H C Patel (Lancet 2005; 366: 1538-44), to investigate if there was any difference in outcome in patients treated in a neurosurgical, compared to a non-neurosurgical centre. The raw data showed that patients in a neurosurgical centre had a mortality of 36% and those in a non-neurosurgical centre had a mortality of 59% indicating that there was 2.4 times greater likelihood of a patient dying from a head injury in a non neuro-surgical centre. To some extent this could be put down to the fact that patients admitted to a neuro-surgical centre were selected. Those with diffuse head injuries were unlikely to be transferred, whereas those with a specific lesion, usually extra-dural or sub-dural haematomas that were subsequently evacuated, were always transferred. However, even without this difference for diffuse head injuries there was still an 18% difference in mortality.
Mr King then went on to talk about work from the intensive care unit at Hope Hospital, looking at patients who had been referred but were not transferred because of lack of intensive care facilities. The findings showed a significant number of patients may have benefited from neurosurgical intensive care, which equated to about a patient a week within the Region. There was consequently a small group of patients who should have been transferred but couldn’t due to lack of resources. He then went onto transfer times in patients with extra- or sub-dural haematomas. Ideally all patients with extra-dural haematomas should go to surgery within two hours of the onset of symptoms, or within four hours of an injury for a sub-dural haematoma. Surgery performed within these time constraints improves outcome. In a 30 month prospective study at Hope Hospital in 39 patients with extra-dural haematomas and 42 patients with sub-dural haematomas, the average time to surgical evacuation for extra-durals was over 6 hours with a similar time for sub-durals. He commented that there was no single cause for these delays, and they occurred at every stage. There were consequently major multi-factorial challenges ahead to greatly improve the service.
Mr King commented upon the NICE guidelines that were published this month, which recommended that patients with isolated head injuries should be taken directly to neurosurgical centres. He then went on to ask the question why outcome was better at neurosurgical centres when essentially the treatment was fairly straightforward. There appeared to be no simple explanation why outcome should be better in a neurosurgical centre for this group of patients.
Mr King concluded that outcome for severely head injured patients had not improved in the last 13 years but they appeared to be better managed in a dedicated neurosurgical unit. There seemed to be avoidable delays in patient transfer and these delays were not improving, and finally if a referring hospital had not been able to transfer a patient to a neurosurgical intensive care unit when they thought the patient required a bed, they should keep trying.
2) The Role of Hypothermia in the Management of Head Injuries
Dr Richard Protheroe, Consultant Anaesthetist, Hope Hospital, Manchester
Dr Protheroe started with a history of hypothermia in head injury patients, which had apparently first been cited by Hypocrates. Most people knew about the detrimental effects of hypothermia on the cardiovascular system with a fall in cardiac output, dysrhythmias and acute coronary syndromes, but there were many other complications of hypothermia, including renal tubular dysfunction, ileus, hepatic dysfunction, hypoglycaemia, and lactic acidosis. Coagulopathy and white cell dysfunction could also occur. Hypothermia was common after trauma with 50% of patients being admitted with hypothermia. This was worse with the severity of injury and was associated with a poorer outcome. He then asked the question “Was pyrexia a problem?” In the animal model there appeared to be more neuronal damage with a rise in temperature, and this appeared to be confirmed in humans with strokes. However, in trauma patients, the detrimental effects of pyrexia had not yet been proven.
Looking at therapeutic hypothermia in patients with head injuries, in the 1980s and 1990s, animal work showed there was significant benefit. However, the animals were rapidly cooled immediately after the injury and this treatment is not easily replicated in humans. There is some evidence in humans that with hypothermia, there is a reduction in intracranial pressure which may lead to better outcome but this has not been shown to be statistically significant and numbers are small. A paper by Marion (NEJM 336(8):540-6, 1997) showed that patients who were cooled to a temperature of 33oC for 24 hours compared to patients who were not cooled, had a better outcome at 3 and 6 months, but there was no difference at 12 months. Jiang (J Neurosurg. 93(4):546-9, 2000) showed that patients at 12 months had better outcomes, but if patients had hypothermia for more than 3 days they had an increased incidence of infections. Signorini et al went through the Cochrane database and did a metanalysis of 8 trials which showed good outcomes with hypothermia. However, Clifton, (N Engl J Med. 344(8):556-63, 2001) showed no benefits in survival or outcome and had more days of complications in the hypothermia group although it did show that hypothermia was better in younger patients. Dr Protheroe went through the papers of Clifton (Journal of Neurotrauma. 19(3):293-301, 2002 Mar)andZhi (Surg Neurol. 59(5):381-5, 2003 May.), which both indicated that outcome and mortality were better with hypothermia. However, there was little statistical significance. He then quoted the paper of Polderman (Intensive Care Medicine. 28(11):1563-73, 2002), which had actually changed his practice. Patients with a Glasgow coma score of less than 9 received intracranial pressure monitoring, and if their ICP went above 20mmHg, neuromuscular blockers were given. If this was ineffective, a barbiturate coma was introduced, and if there was still no improvement in ICP, hypothermia was instituted. This led to a better than expected mortality and morbidity outcome. He admitted that the literature was contradictory and a current US study maybe biased towards better outcome in hypothermic patients. The best temperature to cool patients to appeared to be 35oC because white cell dysfunction starts when temperatures below 34oC are reached, with an increase in infections.
In conclusion, Dr Protheroe stated there was excellent animal work to prove the case for hypothermia in head injury but this was not translated into human practice. He thought there was a requirement for more robust trials with the current lack of good evidence.
3) Final Whistle or Extra-Time?
Dr Steve Kerr, Consultant in Paediatric Intensive Care, Alder Hey Hospital, Liverpool
Dr Kerr started his talk with a history around the requirement for the diagnosis of brain stem death. This had been precipitated by the first heart transplant performed by Christian Barnard in 1967 and a need for the diagnosis of death to enable organs to be removed. A group based at Harvard published some guidance in JAMA in 1968. In 1976 the Medical Royal Colleges of the UK published their guidance for the diagnosis of brain death in the British Medical Journal. These guidelines have stood the test of time. The British Paediatric Association in 1991 also published some guidelines and agreed that the diagnosis of brainstem death could be used in children down to the age of two months. Below that age the criteria were inapplicable. There were 70 different guidelines in 80 countries, and there did not appear to be unanimity in the diagnosis of brain death. He then went onto clarify the reasons for the diagnosis of brainstem death were important. Firstly, for organ donation, secondly, to identify patients for whom further treatment was futile, and thirdly the cost and resource implications of continuing to treat patients in whom there was no likelihood of any recovery. He then went onto recommend the forthcoming document soon to be published on the diagnosis and certification of death.
4) Debate: 'This House believes that Head Injuries are best treated in a Neurosurgical Centre'
For: Dr Sean Tighe, Consultant Anaesthetist, Chester
Against: Mr Paul Eldridge, Consultant Neurosurgeon, Walton Centre, Liverpool.
For the motion, Dr Sean Tighe, Consultant Anaesthetist, Chester
Dr Tighe started the debate and stated that he would confine his arguments to major head injuries, and he defined that as anyone with a Glasgow Coma Score of less than 9.
He also stated that his argument was not that all major head injuries should be transferred, but that major head injuries are best treated in neurosurgical centres. In his favour he stated a number of documents including, recommendations from the Royal College of Surgeons, The European Brain Injury Consortium, the Brain Trauma Foundation 2000, the Neuro-Anaesthesia Society, the Trauma Audit Research Network, and most recently NICE. He also quoted the paper from Patel (Lancet 2005; 366: 1538-44), confirming that the latest figures show a better survival and outcome in a neurosurgical centre, and recent editorials in the British Journal of Anaesthesia (Smith M. Neurocritical care. Has it come of age? BJA 2004; 93: 753-5), and The Lancet (Rosenfeld JV, Cooper DJ. Management of severe head injury: can we do better? Lancet 2005; 366: 1509-10.) re-inforced this view.
Dr Tighe then asked the questions “What can neurosurgical centres offer which district general hospital can’t?” Dr Tighe thought that apart from certain surgery, there were fewer errors, there was greater experience from higher volumes, intracranial and cerebral perfusion pressure monitoring and more aggressive treatment. There were also opportunities for new therapies and research, and patients were surrounded by expert teams with the greater experience. There was also more rapid withdrawal from patients with very poor prognoses. He continued with the question “What was the problem with district general hospitals?” Firstly, he thought that volumes were clearly much smaller, medical cover out of hours in many intensive care units was by non-intensivists, and consequently there was significant inexperience in the team as a whole. Complex protocols for head injured patients including ICP monitoring were performed less well, and when patients were discharged to the general ward, their standard of care was poorer. They also had delayed rehabilitation, which was less specific to head injured patients. He made the point that in pure economic terms, head injuries were commonest in young people, particularly those under 16 years of age with consequent high costs of chronic care. There was also no financial incentive for a neurosurgical centre to take a patient, and there appeared to be a great difference in transfer rates to neurosurgical centres in different regions. He thought there was a significant apathy from neurosurgons to head injuries as they ‘got in the way’ of more interesting surgery. Dr Tighe then went on to give his opinion on how to improve the service by getting NICE more involved particularly after the latest recommendations, informing the public and the press to exert political pressure, locally, regionally and nationally, build business cases based on payment by results, reduced length of stays and better outcomes and the transfer of funding from DGHs.
In conclusion, Dr Tighe felt that all the guidelines confirmed his view, and that was indeed the opinion of experts, and publications proved the case. District general hospitals, in his opinion, were not good enough and there were major economic arguments.
Against the motion, Mr Paul Eldridge, Consultant Neurosurgeon, Walton Centre, University Hospital Aintree, Liverpool
In reply, Mr Eldridge started his argument looking at current practice. He said that patients with severe head injuries were currently treated in district general hospitals and therefore that evidence must justify the change in any current practice. He described the current arrangements of admission, CT scan and onward referral. He then asked the question “Does the risk of transfer affect the outcome?” He stated that there were still major problems with the transfer of patients suffering hypoxia, hypotension, unsecured airways and inadequate evaluation of the cervical spine. He said that the benefit of neurosurgical care must be greater than the risk of transfer in order to justify the transfer. He thought if the district general hospital was good at transfer, they would therefore be good at general care within the intensive care unit. Mr Eldridge then asked the question “What was the difference between neurosurgical critical care, and general critical care?” as intracranial pressure monitoring can be done in a district general hospital and cerebral perfusion pressure/intracranial pressure driven care can be done effectively in the district general hospital. He then looked at the TARN document more carefully and said that case matching doesn’t prove that critical care was an independent variable in outcome. He said that the primary injury was the most important determinant in influencing outcome. Therefore, protocol driven treatment was effective and not the location. He thought that the competency of the average district general hospital in intensive care units should be high enough for good quality care to be delivered.
In conclusion, Mr Eldridge felt the way forward was to improve the core skills of district general hospital intensive care units to allow intracranial pressure monitoring. He thought there was a lack of motivation in district general hospitals with a wish to withdraw from the treatment of severely head injured patients. There was also a resource issue at neurosurgical centres, which can’t cope with all the potential transfers, and payment by results was not yet applicable to intensive care unit patients, and therefore the proposition that severely head injured patients should be treated in a neurosurgical centre was still unproven.
The vote prior to the debate showed there were only 4 delegates against the motion. Following the two presentations, although there was still overwhelming support for the motion, several people changed their vote against the motion.
Following the vote, Mr Eldridge then stated that although he had presented the case, these were not actually his views, and that Walton as a whole, wanted to increase its critical care beds and increase rehabilitation for these patients. However, resources had not yet been found, and the centre was at major risk if financial support could not be provided.
The meeting closed at 5.15 p.m. with a vote of thanks from Dr Mike Trask, President of the Section of Anaesthesia, Manchester Medical Society.