Treatment of severe sepsis and septic shock often focuses on resolving immediate life-threatening problems related to infection (source control, antibiotics) and providing circulatory, ventilatory and other organ support. Neurologic complications, such as sepsis-associated encephalopathy, frequently occur in septic patients and are associated with higher mortality and long-term complications. As case fatalities and overall mortality continue to decline, long-term cognitive problems are becoming more common among survivors. Although the aetiology of septic encephalopathy has not been clearly established, systemic inflammation appears to play a key role in altering both the blood–brain barrier permeability and amplifying the inflammatory response. Several new therapies are now aimed at reducing systemic inflammation. These may eventually play a role in reducing neurologic complications related to the acute pathophysiology of sepsis and may be able to reduce early cerebral dysfunction with the goal of reducing long-term neurologic complications. Coupled plasma filtration adsorption is an extracorporeal therapy aimed at the non-specific removal of cytokines and mediators involved in systemic inflammation and immune suppression by the use of plasma filtration coupled to an adsorbent resin cartridge with high affinity for many cytokines and mediators. Several cytokines that are removed by coupled plasma filtration adsorption have also been implicated in blood–brain barrier permeability, leucocyte recruitment and amplification of the inflammatory response. Current studies are ongoing to determine whether treatments such as coupled plasma filtration adsorption may also be beneficial in reducing long-term neurologic complications.

Current understanding of oral anticoagulant treatment and related intracerebral haemorrhage remains rather limited as compared to that of spontaneous intracerebral haemorrhage. Although intracerebral haemorrhage is the most serious complication of oral anticoagulant treatment, standardized treatment guidelines are still lacking. The currently employed treatments are aimed at normalization of the iatrogenic coagulation impairment, and are not based on randomized controlled trials evidence. Since most patients with oral anticoagulant treatment-intracerebral haemorrhage are at high risk of cardio-embolism and often myocardial infarction, it is uncertain whether the use of procoagulant treatments for oral anticoagulant treatment-intracerebral haemorrhage may increase their risk of thrombotic complications. Patients who receive chronic oral anticoagulant treatment urgently require effective treatments for acute oral anticoagulant treatment-intracerebral haemorrhage, and therefore controlled clinical trials are needed.

Drugs that inhibit platelet function are widely used to decrease the risk of occlusive arterial events in patients with atherosclerosis. There are three families of anti-platelet agents with proven clinical efficacy: (1) cyclo-oxygenase inhibitors, such as aspirin; (2) adenosine diphosphate receptor antagonists, such as the thienopyridine compounds ticlopidine and clopidogrel; and (3) glycoprotein IIb/IIIa antagonists. All these drugs are used during coronary interventions and in the medical management of acute coronary syndromes, while only aspirin and thienopyridine compounds are used in the long-term prevention of cardiovascular and cerebrovascular events in patients at risk. Despite the good risk-to-benefit ratio of anti-platelet agents, the risk of severe bleeding complications, including cerebral haemorrhage, is slightly increased, albeit to a much lesser extent than that associated with the use of other antithrombotic drugs, such as anticoagulants or thromobolytic agents. In addition, it must be noted that the increased incidence of haemorrhagic stroke is usually outweighed by a significant decrease in the incidence of ischaemic strokes. The combination of aspirin and vitamin K antagonists may be associated with the heightened risk of cerebral haemorrhage, compared to treatment with either drug alone.

Intracerebral haemorrhage accounts for 10–15% of strokes and is associated with high mortality and severe disability in survivors. Despite its seriousness, the treatment options for intracerebral haemorrhage are limited. Measures aimed at decreasing elevated intracranial pressure are of limited effectiveness. This has stimulated an interest in attempting to improve the prognosis of intracerebral haemorrhage by addressing the haematoma directly, either removing it by surgical means or limiting its early spontaneous growth. The international Surgical Trial in Intracerebral Haemorrhage (STICH), which randomized subjects with intracerebral haemorrhage within 72 h of symptom onset to medical management vs. surgery, failed to document the superiority of one treatment over the other, when compared with regard to mortality and functional outcome at 90 days. The subgroup of patients with lobar haematomas located at a depth of 1 cm or less from the cortical surface fared better with surgery than with medical management. A similar comparison trial is planned for this subgroup of patients. The neutral results of The international Surgical Trial in Intracerebral Haemorrhage (STICH) prompted the assessment of haemostatic therapies, based on the observation that haematomas often enlarge substantially in the hours that follow the onset of symptoms. Recombinant activated factor VII has been shown in a phase IIb, dose-finding trial to result in a significant reduction of haematoma growth, and both mortality and functional scales trended in favour of recombinant activated factor VIIa. The main complication of this therapy was arterial thromboembolic events (myocardial infarction and ischaemic stroke). A phase III randomized trial has recently been completed.

Multi-centred studies in patients who remain comatose after cardiac arrest and also in newborn babies with perinatal asphyxia have clearly demonstrated that mild hypothermia (32–34°C) can improve neurological outcome after post-anoxic injury. This represents a highly promising development in the field of neurocritical care. This review discusses the place of mild therapeutic hypothermia in the overall therapeutic strategy for cardiac arrest patients. Cooling should not be viewed in isolation but in the context of a ‘treatment bundle,’ which together can significantly improve outcome after cardiac arrest. Favourable outcomes of 50–60% are now routinely achieved in many centres in patients with witnessed arrest and an initial rhythm of ventricular fibrillation or ventricular tachycardia. These results have been achieved by combining a number of therapeutic strategies, including early and effective resuscitation with greater emphasis on continuing chest compressions throughout various procedures (including resumption of compressions immediately after defibrillation even if rhythm has been restored) as well as prevention of hypoxia and hypotension in all stages following restoration of spontaneous circulation. Regarding the use of hypothermia, early induction and proper management of side-effects are the key elements of successful implementation. Treatment should include the rapid infusion of 1500–3000 mL of cold fluids to induce hypothermia and prevent hypovolaemia and hypotension. Educational activities to increase awareness and acceptance of new therapeutic options and European Resuscitation Council guidelines are urgently required.

In industrial countries it is estimated that the incidence of out-of-hospital sudden cardiac arrest lies between 36 and 128 per 100 000 inhabitants per year. Almost 80% of patients who initially survive a cardiac arrest present with coma lasting more than 1 h. Current therapy during cardiac arrest concentrates on the external support of circulation and respiration with additional drug and electrical therapy. Therapeutic hypothermia provides a new and very effective therapy for neuroprotection in patients after cardiac arrest. It is critical that mild hypothermia has to be applied very early after the ischaemic insult to be effective, otherwise the beneficial effects would be diminished or even abrogated. There are numerous methods available for cooling patients after ischaemic states. Surface cooling devices are non-invasive and range from simple ice packs to sophisticated machines with automatic feedback control. Other non-invasive methods include drugs and cold liquid ventilation. The newer devices have cooling rates comparable to invasive catheter techniques. Invasive cooling methods include the administration of ice-cold fluids intravenously, the use of intravascular cooling catheters, body cavity lavage, extra-corporeal circuits and selective brain cooling. Most of these methods are quite invasive and are still in an experimental stage. The optimal timing and technique for the induction of hypothermia after cardiac arrest have not yet been defined, and it is currently a major topic of ongoing research. The induction of hypothermia after cardiac arrest needs to be an integral component of the initial evaluation and stabilization of the patient.

Postanoxic coma is a state of unconsciousness caused by global anoxia of the brain, most commonly due to cardiac arrest. Outcome after postanoxic coma lasting more than several hours is generally, but not invariably, poor. Recovery of consciousness reported in the literature varies from 8% to 72% of patients, but is mostly thought to be around 20–30% in patients surviving in coma for at least 24 h. Research is directed at defining factors that reliably predict poor outcome in these patients. Favourable outcome proves impossible to predict. Studies on outcome prediction have focussed mostly on neurological examination, clinical neurophysiological tests and biochemical parameters. The most recent and extensive study in this respect was the PROPAC study in The Netherlands (407 patients). This study confirmed earlier findings that bilaterally absent early cortical response after median nerve somatosensory potentials (absent somatosensory evoked potentials) is the most reliable predictor of poor outcome (no recovery of consciousness). A serum neuron-specific-enolase level >33 μg L−1 seemed equally reliable. In 2006, the American Practice Parameter on anoxic-ischaemic coma was published, summarizing the findings from the different studies. Poor outcome was defined as death, coma or severe disability after 6 months. The following factors were found to reliably predict this outcome: myoclonic status epilepticus within the first 24 h, absent pupillary responses after 24 h, absent corneal reflexes after 48 h, motor response to pain absent or extensor after 72 h and absent somatosensory evoked potentials (as defined above) after 1–3 days. Results for biochemical parameters (such as neuron-specific enolase) and neuroimaging are inconclusive.

Decision-making at the end-of-life in the United States has evolved over the last 50 yr, beginning with the development of the concept of brain death as a criterion for permitting patients who are in a state of ‘irreversible coma’ to be considered as ‘dead’ for purposes of ventilator withdrawal and organ transplantation. Since then, a firm consensus has emerged in American law and ethics that ‘Patients have a virtually unlimited right to refuse any unwanted medical treatment, even if necessary for life itself.’ With regard to patients who are unable to make decisions for themselves, both Europe and the United States are converging toward a view that respects a role for surrogates in decision-making while recognizing the need to limit their authority. Beyond decisions to withdraw and withhold treatments, both the United States and Europe are experimenting with active hastening of the dying process through euthanasia and physician-assisted suicide. In the author’s opinion, the next big question to be addressed in end-of-life decision-making is ‘Just how bad does the neurological condition and prognosis need to be before it is acceptable to allow a decision to withdraw life support’? Although the practices described here have wide acceptance throughout the United States and Europe, the worldwide emergence of religious fundamentalism and the associated vitalistic view about the sanctity of life may result in significant changes over the next few decades.

Managing end-of-life care can be difficult because of the particular nature of intensive care support, which can separate the biological and the biographical aspects of life. Artificial organ support can temporarily delay death but, at the same time, may fail to restore a quality of life that the patient judges acceptable. For this reason, two concepts must be considered: that the mission of the healthcare system should be to care for patients according to their interests and wishes and that quality of care is related above all to the careful commitment of healthcare workers to the patient’s best interests. Keeping these concepts in mind, the rule of the five Cs (competence, collegiality, communication, continuity of care and compassion) might be helpful in the management of end-of-life care. Unfortunately, neither the rule of the five Cs nor the careful use of moral principles in order to promote the patients’ dignity can assure a universally acceptable decision. A reasonable level of ‘moral certainty’, however, might be achieved using a deliberative approach, which provides for the inclusion of all the different subjects involved in the decision-making process (patient, family, doctors, nurses and other carers), in order to reach the best possible decision in a specific situation.

The interruption of life support poses different problems for he who interrogates himself regarding the possible juridical role of omissible behaviour or activities by part of the physician when dealing with end-of-life interventions within the boundary of life and death. The present contribution proposes to trace the coordinates necessary to answer the main query regarding the obligations which may be incumbent on the physician. For this reason, the necessity to interpret the legal sanctions in a technical key is highlighted. This is performed in sight of a progressive and inevitable adaptation to problems which are the result of a social evolution, and to the conception of values which consitute an object responsibility, as renewed by the constitution. The laws that discipline crimes against life and individual integrity must be interpreted while keeping in mind that the objective of maintaining the patient in life must be integrated with the control of suffering and the guarantee of a dignified death. When identifying the principles which have to inspire the decisions during ‘borderline or boundary situations’, it is highlighted the way the physician has to resort to a just equilibrium between benefit, which can be reasonably expected, and sacrifice, which should be imposed, taking into consideration the criteria of good clinical practice, among which attention to the patient’s will must be taken into consideration.

In 1892 Osler described ‘rapid loss of flesh’ in prolonged sepsis. Thereafter, for years, limb weakness was attributed to cachectic myopathy, and difficulty weaning from mechanical ventilation was attributed to diaphragmatic fatigue. In 1961 Mertens described ‘coma-polyneuropathies’, and in 1971 Henderson and colleagues described polyneuropathy in patients with burns. In 1984 Bolton and colleagues, in a series of reports, defined the clinical, electrophysiological and morphological features of septic encephalopathy and critical illness polyneuropathy. Evidence suggested that polyneuropathy was due to the ‘toxic’ effects of sepsis. Polyneuropathy was a common cause of difficulty in weaning when lung and cardiac cause had been excluded. Since 1984, cases of critical illness polyneuropathy have been reported from several countries. Moreover, a number of investigators reported instances of critical illness myopathy. Comprehensive studies by Latronico and colleagues indicated that polyneuropathy and myopathy often occurred together in the same patient. With successful treatment of sepsis, improvement often occurred in encephalopathy, polyneuropathy and myopathy, except in very severe cases.

Routine motor nerve conduction studies measure latencies, conduction velocities and amplitudes of compound action potentials. These measurements can be very useful in defining the pathology, while they provide little insight into the underlying disease mechanisms. Increasingly, the technique of ‘threshold tracking’ is being used in research and clinical studies on large myelinated axons. Nerve excitability testing is a non-invasive approach in investigating the pathophysiology of peripheral nerve disorders, which determines the electrical properties of the nerve membrane at the site of stimulation. We have found evidence that in patients with critical illness polyneuropathy peripheral nerves are depolarized. The correlations with serum factors suggest that this membrane depolarization is related to endoneurial hyperkalemia and/or hypoxia. While other mechanisms of depolarization may well be involved, the degree to which potential-sensitive nerve excitability indices are related to serum potassium and bicarbonate suggests that other factors, independent of potassium and acid–base balance, are likely to be of relatively minor significance.

With better survival of critically ill patients, ‘de novo’ arising neuromuscular complications like critical illness myopathy or polyneuropathy have been increasingly observed. Prolonged hospitalization not only imposes risks like pneumonia or thrombosis on patients but also represents a real budget threat to modern intensive-care medicine. Clinical symptoms like muscle weakness and weaning failure are common to critical illness myopathy and critical illness polyneuropathy and do not allow for distinction. Specific therapies are not yet available, and the quest for the pathomechanisms has proved more complicated than anticipated. Especially for critical illness myopathy, multiple sites of disturbances to the excitation–contraction coupling cascade are possible causes of muscle weakness. The present review summarizes the epidemiological, clinical and diagnostic features of critical illness myopathy and then focuses on current concepts of the presumed pathomechanisms of critical illness myopathy. Sepsis was shown to be a major cause of critical illness myopathy and special emphasis will be placed on how sepsis and inflammatory mediators influence (i) the membrane excitability at the level of voltage-gated ion channels and (ii) the intracellular protein signalling that results in selective loss of myosin protein content and muscle wasting. For (i), critical illness myopathy represents a new type of acquired channelopathy affecting the inactivation properties of Na+ channels. For (ii), both protein proteolysis and protein build up at the transcriptional level seem to be involved. Findings from different studies are put into a common context to propose a model for cytokine-mediated failure of muscle in severe sepsis. This can open a series of new possible trials to test specific therapeutic strategies in the future.

In spite of the many tools available for monitoring the central nervous system, there are no clinical trials which prove that continuous monitoring of any single variable in the intensive care unit has had any significant impact on the outcome of patients. Even in the absence of robust evidence proving the efficacy of neuromonitoring tools, we believe it is time to re-examine the basic objectives of neuromonitoring. The main reasons for monitoring neurocritical patients could be summarized as follows: (1) to detect early neurological worsening before irreversible brain damage occurs; (2) to individualize patient care decisions; (3) to guide patient management; (4) to monitor therapeutic response of some interventions and to avoid any consequent adverse effects; (5) to allow clinicians to be able to understand the pathophysiology of complex disorders; (6) to design and implement management protocols; and (7) to improve neurological outcome and quality of life in survivors of severe brain injuries. To reach these goals, there is a need to overcome some obstacles, such as the learning curve needed for any monitor and establishing consensus among experts on how to interpret monitor readings. In this review, the obstacles confronted in running randomized clinical trials in this field are discussed. The lack of equipoise and the ethical concerns in conducting such trials are discussed. In addition, the reasons for failure to improve outcome through the use of some monitoring devices are discussed and potential solutions proposed.

Raised intracranial pressure and low cerebral perfusion pressure are associated with ischaemia and poor outcome after brain injury. Therefore, many management protocols target these variables. However, there are no randomized controlled trials that have demonstrated the effectiveness of intracranial pressure-guided care in severely head-injured patients. Observational studies of such therapy have yielded inconsistent results, ranging from decreased mortality to no effect or increased morbidity or mortality. A recent cohort study supports the notion that the possible benefits of intracranial pressure monitoring after traumatic brain injury are small – if present – and would exceed a number needed for the treatment of 16. Furthermore, intracranial pressure monitoring and aggressive management of intracranial pressure and cerebral perfusion pressure have been associated with increased lengths of stay in the neurocritical care unit, conceivable costs and possibly an increased rate of complications. Against this background, there is sufficient clinical equipoise to warrant an adequately powered randomized controlled trial to compare intracranial pressure-guided care with supportive critical care without intracranial pressure monitoring in patients with severe traumatic brain injury. However, the realization of such a trial is likely to be problematic for a number of reasons, not least of which the firmly held biases of many clinicians.

This article will discuss the challenges related to the demonstration of the efficacy of monitoring in an intensive care environment. It will address interpretation, therapeutic intervention, sample size and compare efficacy to effectiveness in the context of the developing field of neurocritical care.

Data on the cerebrovascular effects of catecholamines after head injury are difficult both to interpret and to compare. Diverse parameters with regard to brain trauma animal models, methods of determining the effects on the cerebral blood flow and metabolism and choice of end-points have been used. Many studies investigate the cerebrovascular effects of catecholamines over a range of cerebral perfusion pressures above the range recommended by current guidelines. The relationship between patient outcome and the use of a specific substance to improve cerebral perfusion has not been investigated. Dopamine, norepinephrine and phenylephrine all seem to increase cerebral blood flow in various animal models and in patients. The data suggest that norepinephrine may be the most predictable. It is associated with an improved restoration of global and regional oxygenation when compared to dopamine. Dopamine has been associated with an increase in brain oedema. There is further evidence that dopamine has many disadvantages in critically ill patients due to its ability to suppress circulating concentrations of most anterior pituitary-dependent hormones. Both aspects would further discourage its use. Data on phenylephrine are scarce. It has been associated with increased intracranial pressure and a failure to improve cerebral oxygenation despite markedly improved cerebral perfusion pressure. For all other catecholamines and related substances there are insufficient data on the cerebrovascular effects after head injury. This suggests that norepinephrine may be the catecholamine that is the most suitable substance to maintain or restore adequate cerebral perfusion. The data, however, are insufficient to formulate a guideline.

Animal and human studies suggest that hypertonic saline is a potential therapeutic agent to assist with the medical treatment of patients with traumatic brain injury. It may have a place as osmotherapy to decrease brain size, predominately of uninjured brain and has several potential advantages over mannitol. Hypertonic saline has clinically desirable physiological effects on cerebral blood flow, intracranial pressure and inflammatory responses in models of neurotrauma. Animal studies support its use, but definitive human trials using mortality end-points in brain trauma are lacking. Hypertonic saline may be considered a therapeutic adjunct to the medical management of traumatic brain injury, awaiting definitive evidence to support routine use.

An uncontrolled rise in intracranial pressure is probably the most common cause of death in traumatic brain-injured patients. The intracranial pressure rise is often due to cerebral oedema. Diffusion-weighted imaging has been extensively used to study cerebral oedema formation after trauma in experimental studies. Nevertheless, this technology is difficult to perform at the acute phase, especially in unstable head trauma patients. For these reasons, a safe examination allowing us to better understand the pathophysiology of cerebral oedema formation in such patients would be of great interest. Radiological attenuation is linearly correlated with estimated specific gravity in human tissue. This property gives the opportunity to measure in vivo the volume, weight and specific gravity of any tissue by computed tomography. We recently developed a software package (BrainView) for Windows workstations, providing semi-automatic tools for brain analysis from DICOM images obtained from cerebral computed tomography. In this review, we will discuss the results of the in vivo analysis of brain weight, volume and specific gravity and consider the use of this software as a new technology to improve our knowledge of cerebral oedema formation after trauma and to evaluate the severity of traumatic brain-injured patients.

Follow-up studies on individuals who suffer severe head injury give a picture of poor cognitive and psychosocial outcome. However, recent evidence suggests that with the passage of time, many individuals make adjustments that compensate for early disability, leading to a reduction in social handicap with a corresponding improvement in life quality and personal satisfaction. This article will attempt to briefly review the main sources of evidence contributing to long-term outcome following serious head trauma.