Systemic complications
Injured brains cause impairment of systemic organs systems and, because non-neurological organ dysfunction and failure are independent contributors to morbidity and mortality after brain injury, they represent potentially modifiable risk factors. However, their management presents significant challenges because the optimum treatment for the failing systemic organ system can have potentially adverse effects on the injured brain and vice versa.Brain-injured patients cared for by specialist neurointensive care teams suffer fewer significant non-neurological complications compared with those managed in general ICUs.
Protocol-guided management
Protocol-guided treatment improves clinical outcome in all areas of medicine and is effective in reducing mortality and improving outcome after brain injury.
In a UK study published in 2002, the establishment of an evidence-based management protocol aimed at control of ICP and CPP after TBI resulted in a significant reduction in mortality compared with historic controls, from 59.6 to 40.4%. Furthermore, 66% of patients with raised ICP in the absence of a mass lesion, and 60% of those who required complex interventions to optimize ICP and CPP, had a favourable outcome. A US study that included all categories of head injury admitted to an ICU compared outcomes after the introduction of a standardized treatment protocol with two previous time periods – one before the availability of a neurointensive care unit and the other after the establishment of a basic neurointensive care unit without protocolized treatment strategies. here was a decrease in mortality from 40 to 27 to 2.8% in the three time periods, respectively, in association with an increase in the incidence of good functional outcome in survivors from 40 to 68 to 84%. In another US study, management in a level I centre was associated with better outcome after severe head injury than outcome in a level II centre, but it was significantly improved in both following the introduction of standardized treatment protocols.
Inevitably, the majority of studies examining the introduction of protocolized management strategies have used historic control groups, and these limit interpretation to some extent because of the unknown impact of the temporal differences in outcome that might have resulted from unrelated changes in technology, patient management, personnel and organization. Both practical and ethical issues are likely to prevent randomized controlled trials in this area, but a UK study published in 2004 attempted to deal with this issue. In this study, the introduction of an evidence-based management protocol significantly reduced ICU mortality from around 20 to 13.5% and overall hospital mortality from 24.5 to 20.8% in patients with severe TBI. hese improvements occurred despite an increase in the median age and APACHE II score of the patient population after implementation of the protocol. Although historic controls were also used in this study, the mortality of patients with non-neurological disease admitted to the same (mixed) ICU did not change significantly over the same period. his strongly suggests, although it does not prove, that the benefits to the head-injured patients was related to the introduction of the protocol-driven management paradigm, rather than because of other changes.
Despite the available evidence and guidance, there is considerable variation in the implementation of established neurointensive care management strategies. Several studies have confirmed that units that aggressively monitor, and therefore presumably aggressively manage, ICP after severe head injury have better outcomes than those that do not. However, such studies do not confirm whether aggressive management per se improves outcome or whether the implementation of complex monitoring and management strategies are proxy markers for units that provide an integrated approach to management and higher standards of care overall. Although the application of protocol-guided monitoring and management after head injury increases resource usage, the improvements in outcome are likely to justify the increased cost of the treatment episode. For example, a study from the Centers for Disease Control and Prevention in the USA demonstrated that a significantly greater proportion (66 vs. 35%) ofpatients with severe TBI had a good outcome when their management was based on Brain Trauma Foundation guidelines. Extrapolation of these findings suggests that widespread implementation of guideline-based management would result in cost savings of US$250 million per year in the USA alone.
Notwithstanding the multiple evidence that aggressive ICP- and CPP-targeted treatment after head injury is beneficial, there is some suggestion that it is associated with increased levels of therapy intensity but not necessarily with improved outcome. A Dutch study published in 2005 compared 333 patients with severe TBI managed in two head-injury centres. One provided supportive intensive care (mean arterial pressure >90 mmHg and other therapeutic interventions directed by clinical observations and CT findings) without ICP monitoring, whereas the other provided protocol-driven intensive care, guided by ICP monitoring, to maintain ICP <20 mmHg and CPP >70 mmHg, according to Brain Trauma Foundation guidelines. Hospital mortality was similar in the two centres (34 vs. 33%) and the odds ratio for a more favourable outcome following ICP- and CPP-targeted therapy was 0.95 (95% CI 0.62-1.44). However, intensity of treatment, assessed by the use of sedatives, vasoactive drugs, mannitol and barbiturates, was greater in the centre providing ICP-guided care, and the median time on mechanical ventilation was also significantly longer (12 days vs. 5 days). In this study, statistical uncertainty allowed for the possible benefit of ICP monitoring and management, but the authors calculated that this potential benefit would be rather small, with a number needed to treat of 16. In another European head-injury study published in 2008, patients who received ICP monitoring and management had a tendency to lower raw and risk-adjusted mortality rates than those who were not monitored, but this difference was not statistically significant. As in the Dutch study, the use of ICP monitoring and management was associated with increased treatment intensity, including higher vaso-pressor use, and did not show a significant association with neurological outcome at discharge from hospital.
Although the Dutch study has been criticized, it is of some interest that ICP-guided treatment in this study failed to control ICP within its target range (<20 mmHg) in more than one-quarter ofpatients. his raises the important question whether continuation of ICP- and CPP-guided treatment, with its associated side effects, is warranted if it is clearly not achieving its aim. Shifting the CPP target downwards is likely to be more beneficial than continuing to increase vasopres-sor support in patients without evidence of brain tissue ischaemia/hypoxia or metabolic disturbance. he concept of individualized CPP management, targeting a patient-specific CPP rather than a generic target, is gaining acceptance and is likely to reduce the incidence of treatment-related complications.
Benefits of neurointensive care units
A consensus is emerging that treatment in dedicated neurointensive care units is beneficial for patients with neurological disease generally but particularly for those with acute brain inj ury. Several studies have shown that neurointensive care units, staffed by dedicated neu-rointensivists, not only save lives and improve outcome but that they are also associated with better resource utilization compared with management of critically ill neurological patients in a general ICU. he potential benefits of neurointensive care are likely to be multifactorial and are summarized in Table 18.2. he fact that protocol-guided brain resuscitation within the context of excellent general critical care reduces mortality and improves outcome after brain injury is clearly a major contributor to these benefits, but the advantages of specialized neurointensive care must surely extend beyond this single issue.
Table 18.2 Aspects of neurointensive care that contribute to improved outcome
|
Delivery of individualized, protocol-guided care |
|
Multimodal brain monitoring-guided treatment strategies |
|
Dedicated, specialist multidisciplinary team including specialist neuroscience critical care nurses and therapists |
|
Supervision of management by dedicated neurointensivists |
|
Rapid access to neurosurgical services |
|
Increased expertise from higher caseload |
|
Awareness of the interplay between the injured brain and systemic organ systems: |
|
• Improved control of systemic physiology |
|
• Greater understanding of the causes and treatment of non-neurological organ system dysfunction and failure |
Caseload
Specialization attracts a greater caseload and this is likely to increase expertise. In an Austrian study of 1856 patients with severe TBI, those admitted to units treating more than 30 cases per year had lower mortality compared with those admitted to medium (10-30 cases per year) and small (<10 cases per year) centres. Compared with large units, the odds ratio for hospital mortality was 1.85 (95% CI 1.42-2.40) for patients managed in medium-sized units and 1.91 (95% CI 1.24-2.93) for those managed in small units. Analysis of data collected as part of the National Acute Brain Injury Study: Hypothermia also examined inter-centre differences in physiological and treatment variables during the intensive care management of head-injured patients included in the parent study. Tere were no significant differences in the incidence or severity of intracranial hypertension between units, but there were significant differences in the maintenance of arterial blood pressure and CPP targets, and in the use of vasopressors and mannitol. Specialist units admitting larger numbers of head-injured patients performed better. Two US studies also showed that mortality after SAH is significantly reduced in high-volume centres that provide access to specialized multidisciplinary neurocritical care. Finally, among 4674 post-cardiac arrest patients admitted to ICUs in 39 hospitals in the USA, age- and severity of illness-adjusted institutional mortality ranged from 46 to 68%, with the lowest mortality rates occurring in ICUs treating a higher volume of post-cardiac arrest survivors.
The neurointensive care team
It is well recognized that the organization of ICU teams can directly affect patient outcome and resource utilization. In particular, staffing units with critical care physicians is associated with positive outcome benefits. In a systematic review of 26 observational studies of staffing models in general medical and surgical ICUs, high-intensity staffing (full-time presence of an intensivist or closed ICU) was associated with a reduced mortality in 94% of studies compared with low-intensity staffing (no full-time intensivist). Tere was also reduced hospital length of stay in all the studies that controlled for case mix. Similar findings have recently been confirmed in neurointensive care. In one study, the appointment of a full-time neurointensivist was associated with a 51% reduction in neurointensive care unit mortality, a 12% shorter hospital length of stay and 57% greater odds of being discharged to home or a rehabilitation unit rather than to a long-term care facility. Te presence of a neurocritical care team is also an independent predictor of decreased hospital mortality and reduced costs of care.
It is difficult to pinpoint exactly why neurointensive care teams influence patient outcome, but a major contributor is likely to be that a dedicated team provides standardized diagnosis and management of the most common problems, and explicit goal setting related to the prevention of secondary brain insults by control of ICP, CPP, arterial blood pressure, blood glucose and temperature. Round-the-clock provision of dedicated and experienced neurointensive care teams not only facilitates the application of individualized treatment strategies but also ensures that such therapies are applied in a timely and consistent fashion. In a US study, a significantly higher proportion of specialist centres had treatment guidelines in place than non-specialist centres (78.4 vs. 53.7%) and adherence to such guidelines was also much higher in specialist units.
A crucial component of the neurocritical care team is its nursing staff. Neuroscience critical care nurses not only need the skills held by their general ICU counterparts but must also become proficient at the neurological examination to a much greater degree of sophistication and precision. Despite advances in neu-romonitoring techniques, the neurocritical care nurse remains the most important neurological monitor and is able to detect subtle changes in neurological status and identify deterioration early. he bedside nurse is also in a unique position to make sure that everyone is aware of local management protocols and that they are followed. Acute rehabilitation plays a major role in securing improved long-term outcome after neurological illness, and intervention from neurophysi-otherapists is likely to occur earlier and more reliably in a specialist than in a general unit.
Management of systemic physiological variables
Neurointensivists and their teams focus on the interplay between the brain and other organ systems and integrate all aspects of neurological and medical management into a single care plan. he neurocriti-cal care team is familiar with the unique aspects of the primary disease processes and also with the effects of interventions on the injured brain. For example, blood pressure control is more aggressive in neuroin-tensive care units compared with general units, resulting in a lower incidence of systemic (often iatrogenic) hypotension. Other physiological derangements, such as fever, hyperglycaemia, anaemia, sodium disturbances and delirium, have specific consequences in the context of acute brain injury and require different management strategies than in general intensive care. Neurointensive care clinicians have a clear understanding of the pathophysiological basis of the systemic complications of brain injury, and balance treatment strategies to minimize adverse effects on the underlying brain injury.
Therapeutic nihilism
Because of the availability of less invasive and more effective treatments, patients with severe brain injury who would previously have been considered unsal-vageable are increasingly being offered treatment. here is considerable evidence that early aggressive intervention on the neurointensive care unit can result in excellent outcomes in substantial numbers of such patients, and recent attention has focused on the major role that therapeutic nihilism can have on determining outcome when severely brain-injured patients are cared for by non-specialist teams.
Despite maximal intervention, some patients will have a poor outcome, and it is essential that aggressive early treatment is linked to compassionate end-of-life care if a satisfactory degree of clinical improvement does not occur within an appropriate timescale. he confidence to withdraw treatment after a failed trial of early maximal intervention means that the usual justification for withholding treatment in the acute phase (i.e. survival with a devastating neurological injury) becomes irrelevant. his ensures that patients have access to care that might allow them to recover beyond initial expectations but also that decisions to withhold or withdraw support are made by those with sufficient expertise.
Training
Neurointensivists must be trained and experienced in general intensive care and, conversely, neurointensive care is a vital part of the training of general intensiv-ists. In October 2005, neurointensive care gained formal recognition by the United Council of Neurological Subspecialties (UCNS) in the USA. Subsequently, this has led to accreditation of neurointensive care training programmes and certification of neurointensive care physicians in the USA. his has been a major milestone in the recognition of neurointensive care as a subspe-cialty of intensive care medicine and, although there is no similar recognition in other countries, there are clearly opportunities to extend UCNS neurointensive care training standards outside the USA. Importantly, expertise in neurointensive care involves procedural skills, proficiency with standard (systemic) monitoring and management, as well as specialized neuro-monitoring techniques and interventions. Training programmes should therefore be split between neurological diseases and conditions, and medical diseases and conditions that commonly complicate acute neurological illness. Further information is available at http://www.ucns.org/ go/subspecialty/neurocritical/ certification.
Research
Although improved outcomes can be achieved by applying consistently what we have learned through research, there are limited data to guide most interventions on the neurointensive care unit. Over the years, numerous drugs with promising neuroprotective effects in the laboratory have been evaluated in large clinical studies and failed to offer outcome benefits to patients. In contrast, the effectiveness of many of the basic treatment algorithms and physiological interventions that are routinely applied during the critical care management ofbrain injury have not been evaluated in large studies. While there is always reluctance to subject long-standing clinical practices to rigorous investigation, there is no doubt that well-conducted studies are required to determine optimal strategies for many neurointensive care interventions.
There are many challenges to conducting research in critically ill neurological patients, particularly in the heterogeneous brain injury population. In order to be successful, future trials will need to be targeted to subgroups of patients with a specific pathology and an intermediate prognosis; the outcome of those who will inevitably fair well or badly is unlikely to be modified by the intervention being tested. In addition, blinding of treatment interventions, prevention of protocol violations and treatment cross-overs remain problematic. Clinical equipoise is also an issue, as the firmly held biases of some clinicians are likely to prevent randomization to reference or investigational therapy. For example, it is unlikely that experts in neurotrauma would support the notion that ICP monitoring is no better than no monitoring, and a randomized controlled trial of ICP-guided therapy, coordinated through the Traumatic Coma Data Bank, has previously been proposed but not funded. he ethical issues of such a study have in any case been questioned, and the prospect of embarking on a study of a technique that is considered by many experts to be indispensable is limited, not least because accepted treatments and decision-making in TBI are driven by protocols guided by ICP monitoring.
Despite these issues, neurointensivists and their teams are establishing excellent track records for the coordination of complex, multicentre randomized controlled studies. Ongoing studies in 2010 include those investigating brain tissue oxygenation-guided therapy and decompressive craniectomy after TBI, therapeutic hypothermia for control of ICP and the indications for surgery after intracranial haemorrhage. he UK-based Risk Adjustment in Neurocritical Care (RAIN) study is a prospective validation of risk prediction models for adult patients with acute TBI designed to evaluate the optimum location for management of TBI and the comparative costs of neurointensive care. It is anticipated that these and other studies will provide an evidence base for many of the interventions that are currently provided on an empirical basis during neurointensive care.
Summary
The management of critically ill neurological patients is complex and requires a coordinated and stepwise approach that includes clinical assessment, monitoring and individualized multifaceted management strategies to minimize secondary neurological injury. Improved understanding of the pathophysiology of the injured brain has allowed new diagnostic, prognostic and treatment modalities to be incorporated into routine management strategies. he complex treatment modalities applied in brain-injured patients call for interdisciplinary collaboration between neurointensivists, neuro-surgeons, neurologists, specialist nurses and therapists, and the neurointensive care unit serves as the focal point for these efforts. Although there is substantial evidence that patients managed in specialist neurointensive care units have better outcomes than those managed in general units, it is important that we continue to strive to determine exactly why this is the case.
