Management of traumatic brain injury: a narrative review of current evidence
Globally, approximately 70 million people sustain traumatic brain injury each year and this can have significant physical, psychosocial and economic consequences for patients, their families and society. The aim of this review is to provide clinicians with a summary of recent studies of direct relevance to the management of traumatic brain injury in order to promote best clinical practice. The use of tranexamic acid in the management of traumatic brain injury has been the focus of several studies, with one large randomised controlled trial suggesting a reduction in all-cause mortality within 24 h of injury. The use of therapeutic hypothermia does not improve neurological outcomes and maintenance of normothermia remains the optimal management strategy. For seizure management, levetiracetam appears to be as effective as phenytoin, but the optimal dose remains unclear. There has been a lack of clear outcome benefit for any individual osmotherapy agent, with no difference in mortality or neurological recovery. Early tracheostomy (< 7 days from injury) for patients with traumatic brain injury is associated with a reduction in the incidence of ventilator-associated pneumonia and duration of mechanical ventilation, critical care and hospital stay. Further research is needed in order to determine the optimal package of care and interventions. There is a need for research studies to focus on patient-centred outcome measures such as long-term neurological recovery and quality of life.
Traumatic brain injury (TBI) is defined as “an alteration in brain function, or other evidence of brain pathology, caused by an external force” . Globally, it is estimated that around 70 million people sustain TBI each year , which can have significant physical, psychosocial and economic consequences for patients, their families and society as a whole. In 2016, TBI was estimated to have resulted in 8.1 million years lived with disability  and this is associated with a significant financial burden to healthcare systems and national economies due to the combination of high healthcare costs and loss of employment productivity .
The epidemiology of TBI in high-income countries is changing in line with ageing populations. Falls are now the leading cause of TBI and the median age of patients affected has doubled over the past 40 years . As a result, more patients with TBI have comorbid medical conditions which are associated with an increase in all-cause mortality rates . These changing patient characteristics have resulted in challenges for clinicians. Research investigating different therapies and interventions has often excluded older patients with pre-existing medical conditions, which makes extrapolation of results difficult.
Several organisations have produced guidelines for the management of TBI [7–9]. Key elements of the most commonly used are summarised in Table 1. However, adherence to guidelines is generally suboptimal, especially for older patients with TBI . The reasons for this are multifactorial, and one barrier is the low-quality evidence underpinning some recommendations. However, following certain aspects of a protocol-based management strategy for TBI, such as intracranial pressure (ICP) monitoring, has been shown to reduce mortality . The aim of this review is to provide clinicians with a summary of recent studies of direct relevance to the management of TBI in order to promote best clinical practice and facilitate the delivery of individualised patient care strategies. The key findings of the review are summarised in Fig. 1.
Table 1. Summary of key recommendations from guidelines on the management of traumatic brain injury.
National Institute for Health and Care Excellence: head injury: assessment and early management 
Brain Trauma Foundation: guidelines for the management of severe traumatic brain injury, 4th edn 
Association of Anaesthetists: guidelines for safe transfer of the brain-injured patient: trauma and stroke, 2019 
MAP ≥ 80 mmHg
SBP > 100 mmHg (age 50–69 y)
SBP ≥ 110 mmHg (ages 15–49 and >70 y)
SBP 110–150 mmHg
MAP > 90 mmHg
PaCO2 4.5–5.0 kPa
PaO2 > 13 kPa
Avoid PaCO2 < 3.33 kPa
PaCO2 4.5–5.0 kPa
PaO2 ≥ 13 kPa
No recommendation made
≤ 22 mmHg
No recommendation made
Cerebral perfusion pressure
No recommendation made
No recommendation made
No recommendation made
No recommendation made
No recommendation made
No recommendation made
Mannitol or hypertonic saline if impending uncal herniation
No recommendation made
Prophylactic hypothermia not recommended
Maintain normothermia (36–37°C)
SBP, systolic blood pressure; MAP, mean arterial pressure
Summary of recent evidence-based recommendations for the management of traumatic brain injury (TBI).
MEDLINE, PubMed, Embase and Google Scholar were searched for available evidence on the management of TBI in adults (age > 16 y). An example of the search strategy used is available as online Supporting Information Table S1. The search included clinical trials, meta-analyses, randomised controlled trials and systematic reviews. The search was restricted to literature published in English between 01 January 2018 and 01 August 2021. The main studies discussed in this narrative review are those that, in the opinion of the author, will impact on current clinical practice and future guidelines. Where possible, the included studies have been organised into clinically relevant themes.
Transfusion and coagulation
The publication of the third clinical randomisation of an antifibrinolytic in significant haemorrhage (CRASH-3) trial resulted in an upsurge in interest regarding the use of tranexamic acid (TXA) in the management of TBI. A summary of recent studies on TXA is shown in Table 2. The CRASH-3 trial allocated patients with TBI (Glasgow Coma Scale (GCS) ≤12 or any intracranial bleeding on CT scan) and no major extracranial bleeding at random to either a bolus plus infusion of TXA or placebo . Initially, the therapy had to be started within 8 h of injury, but this was subsequently revised to 3 h considering data from the earlier CRASH-2 study. In the 9127 patients with outcome data, there was no significant difference in head injury-related deaths in hospital within 28 days (18.5% vs. 19.8% for TXA and placebo, respectively (RR 0.94 (95%CI 0.86–1.02)). When patients with bilateral fixed dilated pupils and/or GCS 3 at baseline were excluded, the outcomes were similar (RR 0.89 (95%CI 0.80–1.00)). Further sub-group analyses suggested that patients with mild-moderate TBI (admission GCS 9–15) may have had a greater mortality benefit with TXA administration (RR 0.78 (95%CI (0.64–0.95)); the trial, however, was not powered for this outcome. In addition, there was no difference in neurological outcome between treatment groups.
Table 2. Summary of recent randomised controlled trials investigating tranexamic acid administration for patients with traumatic brain injury.
In-hospital 1 g TXA bolus then 1 g TXA 8-h infusion vs. placebo
Adults with TBI within 3 h of injury
GCS < 12
Any intracranial bleeding on CT scan
No major extracranial bleeding
Head injury-related death in hospital within 28 days in patients randomly allocated within 3 h of injury
Head injury-related mortality 18.5% with TXA vs. 19.8% for placebo (RR 0.94 (95%CI 0.86–1.02)). When patients with a GCS 3 or bilateral unreactive pupils were not studied, mortality 12.5% with TXA vs. 14.0% for placebo (RR 0.89 (95%CI 0.80–1.00)).
There have been several publications from subsequent sub-group analyses of the data from the CRASH-3 study. One criticism of the original study was that no details were provided as to the type of brain injury seen on CT. An analysis of 1767 patients showed that 61% of patients had more than one type of bleed, with subdural and intraparenchymal haemorrhage most common . However, this study was not able to confirm that TXA administration reduced intracranial bleeding. The mechanism of action for the beneficial effect of TXA, therefore, remains uncertain, with antifibrinolytic and anti-inflammatory mechanisms hypothesised. Patients with moderate/severe TBI who received TXA showed a lower degree of fibrinolysis as assessed by reductions in D-dimer and plasmin-antiplasmin levels ; of note, the antifibrinolytic effect was not demonstrated on thromboelastogram (TEG) analysis, which may reflect the insensitivity of this tool for the assessment of fibrinolysis. The other issue relating to the CRASH-3 study was the definition of what constituted a “head injury-related death.” This was addressed in a further analysis of the data that focused on the effect of TXA administration on early deaths (within 24 h and within 28 days) from any cause in patients with a baseline GCS ≥4 and reactive pupils . This showed that TXA administration within 3 h of injury reduced all-cause 24-h mortality (RR 0.74 (95%CI 0.58–0.94)) but had no effect on 28-day all-cause mortality (RR 0.93 (95%CI 0.83–1.03)). These findings were not affected by country income and vascular occlusive events were similar in both groups.
There is still uncertainty as to the role of TXA in the management of TBI, with recent meta-analyses showing that TXA has no effect on mortality or neurological recovery [18, 19]. However, more than 85% of participants in these analyses were from the CRASH-3 study. Given that a randomised controlled trial (RCT) of the size of CRASH-3 is unlikely to be repeated soon, it appears likely that it is the findings of this study which will influence clinical practice. As TXA is a low-cost drug  with minimal adverse effects, many centres will continue to administer this as an early treatment for all patients with isolated TBI despite clear evidence that this will not improve neurological outcome.
Red blood cell transfusion and erythropoietin
The optimal transfusion trigger for patients with TBI has been a matter of debate for some time, with guidelines generally recommending a target haemoglobin of 70–90 g.l−1. One systematic review identified four studies that compared transfusion triggers of < 70 g.l−1 and < 90 g.l−1 in patients with moderate to severe TBI . This showed a reduced risk of poor neurological outcome (defined as Glasgow Outcome Scale 4–5) with a transfusion trigger of < 70 g.l−1 (OR 0.64 (95%CI 0.42–0.97); p = 0.03), with no difference in mortality rates between the two targets. However, the studies included were judged to be at moderate to high risk of bias and were moderate to low in quality. The age of transfused red blood cells does not appear to be an important factor, with the use of fresh red cells (aged ≤7 days) not being shown to improve neurological outcome . Prospective RCTs are needed in order to determine the optimal transfusion trigger, which should ideally include the effect of transfusion on direct cerebral measures such as brain tissue oxygenation.
The hypothesised neuroprotective and neuroregenerative effects of erythropoietin have led to this being investigated as a therapy for TBI. Meta-analyses show a small improvement in mortality but no effect on neurological outcomes . Given the high cost of this therapy, further work is needed to clarify its precise role and indications before introduction into clinical practice.
With the increase in the age of the general population, a greater number of patients suffering TBI are now on antiplatelet therapy, such as aspirin and clopidogrel, at the time of their injury. Unsurprisingly, patients on antiplatelet therapy have a higher risk of intracranial haemorrhage after minor TBI (RR 1.51 (95%CI 1.21–1.88); p = 0.002)  and have a higher mortality rate ; the optimal management strategy for this patient population remains unclear. A meta-analysis examined the use of platelet transfusion to reverse the effects of antiplatelet drugs in patients with TBI and showed that while haematoma expansion was reduced, there were no improvements in mortality or neurological outcome . Further work is needed in this area, with the use of platelet function assays showing promise in helping determine precisely how and when platelet transfusion should be administered.
Like antiplatelet drugs, many patients who suffer TBI are now on anticoagulants, with an increasing proportion on direct oral anticoagulants (DOAC) rather than vitamin K antagonists such as warfarin. One systematic review showed that pre-injury anticoagulation was associated with an increased mortality risk (OR 2.39 (95%CI 1.63–3.50); p < 0.00001); however, only a small minority of patients were taking a DOAC (5%) with the remainder taking vitamin K antagonists . A small cohort study focusing on patients with minor head injury taking a DOAC showed an increase in the incidence of adverse outcomes, defined as a composite of requirement for neurosurgery, intracranial haemorrhage or death (3.4% (95%CI 1.4–8.0%)) , compared with patients who were not anticoagulated. However, a meta-analysis of patients with traumatic intracerebral haemorrhage reported that DOAC use was not associated with an increased risk of in-hospital mortality compared with vitamin K antagonists .
Patients with TBI can also develop a coagulopathy as a direct result of the initial trauma. Gruen et al. investigated the pre-hospital administration of plasma to patients who had suffered TBI (abbreviated injury score for the head > 2) in a post-hoc analysis of a RCT that had investigated the management of haemorrhagic shock . In total, 166 patients were included (74 received plasma and 92 standard care) of whom 86 sustained a subdural haematoma/haemorrhage and 11 sustained an extradural haematoma/haemorrhage. Compared with standard care, patients with TBI who received plasma had fewer crystalloid and packed red cell transfusions and a lower 30-day adjusted mortality (hazard ratio 0.55 (95%CI 0.33–0.94); p = 0.03). As this was a secondary analysis, these results should be seen as hypothesis generating, but appear to support the notion that the early treatment of trauma-induced coagulopathy may be of clinical benefit; similar studies examining the use of fibrinogen administration are ongoing (e.g. CRYOSTAT-2 https://www.isrctn.com/ISRCTN14998314).
Baksaas-Aasen et al. investigated the utility of viscoelastic testing as part of major trauma haemorrhage protocols compared with standard empiric protocols . There was no difference between groups for the primary outcome measure of 24-h mortality. However, in a pre-specified sub-group analysis of patients with TBI (n = 74), there was an improvement in 24-h survival with the use of viscoelastic testing (OR 2.12 (95%CI 0.84–5.34). However, these data should again only be viewed as hypothesis generating and further research is needed.
A large multicentre RCT has investigated the effect of early prophylactic hypothermia (33–35°C) for at least 72 h . In contrast to the earlier study by Andrews et al. , this intervention was started in the absence of elevations in ICP. The incidence of favourable neurological outcome was similar in both groups (unadjusted RR with hypothermia, 0.99 (95%CI 0.82–1.19)). Accidental hypothermia on hospital admission, which is common in patients who have suffered traumatic injuries, should also be actively managed as this is associated with an increased mortality risk after TBI ((OR 2.38 (95%CI 1.53–3.69)) . These studies add to the evidence suggesting that normothermia is key in the management of many types of brain injury  and hypo- and hyperthermia should both be managed aggressively at hospital admission. Normothermia after TBI has been defined variably in different studies, but in clinical practice a typical target is 36.5–37.5°C.
A multicentre RCT investigating the use of dexamethasone in patients with chronic subdural haematoma was identified during the review process . The management of this condition is covered in a separate review article in this supplement  and therefore the study by Hutchinson et al.  is discussed only briefly. This compared a 2-week tapering course of dexamethasone (initial dose 16 mg.day−1) with placebo in patients with chronic subdural haematoma, of whom 94% underwent surgical evacuation. Dexamethasone use was associated with an increased risk of an unfavourable outcome at 6 months (modified Rankin scale 4–6) and a greater incidence of adverse events. As such, prolonged steroid therapy should not be used as part of the clinical management of chronic subdural haematoma.
Seizures are common following TBI and may worsen secondary brain injury. Anti-epileptic medications are used commonly in the first 7 days following injury in order to reduce the incidence and severity of post-traumatic seizures. Phenytoin has been the drug of choice for many years, but has several adverse effects, including hypotension, cardiac arrhythmias and liver enzyme induction. Levetiracetam is a novel anti-epileptic drug that has become increasingly popular due to its low incidence of adverse effects. A meta-analysis of 10 studies that compared levetiracetam with phenytoin use in patients who had suffered TBI (n = 2057) showed similar efficacy in seizure prevention (OR (95%CI) 1.02 (0.72–1.45); p = 0.89) with fewer adverse effects . However, the optimal dose of levetiracetam remains unclear and further prospective RCTs are required in this regard.
There is an ongoing debate within the neuroscience community about the role of osmotherapy in the management of TBI. There is a lack of clear outcome benefit for any individual agent studied (with hypertonic saline and mannitol both investigated extensively) but this remains an area of research activity, with seven meta-analyses published in the last 3 y. The most recent meta-analysis involved 12 RCTs containing 464 patients and showed no significant difference between hypertonic saline and mannitol in terms of mortality (RR 0.69 (95%CI 0.45–1.04); p = 0.08) or neurological outcome (RR 1.28 (95%CI 0.86–1.90); p = 0.23), despite both drugs reducing ICP and improving cerebral perfusion pressure . These findings were confirmed by a subsequent multicentre RCT of patients with moderate to severe TBI, in which continuous osmotherapy with a 20% hypertonic saline infusion did improve neurological outcome at 6 months . Considering these findings, osmotherapy should only be viewed as a temporising measure, not a definitive treatment, in patients with intracranial hypertension after TBI.
Novel therapeutic interventions
Several RCTs have investigated novel interventions in the management of TBI (Table 3). However, none have shown convincing evidence of clinical benefit and there are several methodological weaknesses in all the trials. It appears unlikely that any of the drugs tested will be utilised in clinical practice in the near future outside of research studies.
Table 3. Summary of recent randomised controlled trials investigating novel therapeutic interventions for the management of traumatic brain injury.
20 mg propranolol orally every 12 h up to 10 days or until discharge 
No significant difference in mortality (incident rate ratio (95%CI) 0.6 (0.3–1.4); p = 0.2) or long-term functional outcome
Reduces blood–brain barrier permeability; reduction of inflammation; anti oxidant
1 g metformin every 12 h for 5 consecutive days 
Serum concentration profile of S100b
Significantly lower S100b in patients allocated to metformin at 48 h, 72 h and 120 h post-treatment
No clinical outcomes recorded.
Decreases brain oedema; attenuates free radical damage; and reduces neuronal damage and inflammatory cytokine production.
Single dose of vitamin D 120,000 u vs. placebo 
No difference in mortality rates
Neurotrophic factor-like activity with pleiotropic neuroprotective activity
Cerebrolysin 50 ml daily for 10 days, followed by two additional treatment cycles with 10 ml daily for 10 days vs. placebo 
Complex multidimensional assessment of neurological outcome using 13 scales
Small to medium-sized effect size suggesting superiority of cerebrolysin
The mainstay of cerebral monitoring in patients with TBI remains ICP measurement, with the Brain Trauma Foundation recommending that this is maintained ≤22 mmHg . However, a study using data from the Collaborative European Neuro Trauma Effectiveness Research in TBI (CENTER-TBI) ICU cohort was able create patient-specific ICP thresholds by analysing the pulse reactivity index to determine when cerebral reactivity became impaired . Individual ICP thresholds could be determined in 128 of 196 patients, with the mean (SD) threshold 23.0 (11.8) mmHg. Compared with a threshold of 22 mmHg, time spent above the individualised ICP threshold was a better predictor of mortality (area under the curve 0.678 vs. 0.492, respectively; p = 0.035) and was similarly accurate for neurological outcome (area under the curve 0.610 vs. 0.456, respectively; p = 0.06). While these findings cannot be extrapolated to clinical care at present, the ability to individualise physiological targets for patients with TBI is a potentially exciting development.
In settings where ICP measurement is not available, Alali et al. proposed a clinical decision rule to predict the likelihood of intracranial hypertension based on a combination of a Delphi process and data from a Latin American trial . This suggested that one major criterion (compressed cisterns, midline shift > 5 mm or non-evacuated mass lesion > 25 cm3) or ≥2 minor criteria (GCS motor score ≤4; pupillary asymmetry; abnormal pupillary reactivity; and basal cisterns are present with midline shift ≤5 mm and/or high- or mixed-density lesion ≤25 cm3) predicted the likelihood of ICP > 22 mmHg with a sensitivity of 94% (95%CI 85–98%) and specificity of 42% (95%CI 32–54%). While primarily of interest in settings with limited healthcare resources, this tool may be of value in helping determine which patients need urgent ICP monitoring within high-income countries.
There has been interest in the use of optic nerve sheath measurement as a non-invasive assessment method for intracranial hypertension. This is discussed in detail in a review within this supplement  but in short, this technique shows promise with a recent systematic review suggesting 97% (95%CI 92–99%) sensitivity and 86% (95%CI 74–93%) specificity in the detection of raised ICP, which is similar to that seen with direct ICP-monitoring (sensitivity 93% (95%CI 88–96%) and specificity 85% (95%CI 72–93%)) .
Airway management and ventilation
Patients who have suffered acute brain injury often require ventilatory support on ICU for a prolonged period. The reasons for this are multifactorial and include: difficulty in determining prognosis for recovery in the early stages of illness; increased incidence of impairment of bulbar function; and a higher incidence of pulmonary infection due to poor cough and difficulty with physiotherapy engagement as a result of neurological impairment. A meta-analysis showed the pooled incidence of ventilator-acquired pneumonia in patients with TBI was 36% (95%CI 31–41%), and that this was associated with an increase in duration of mechanical ventilation and critical care stay . As a result, tracheostomy is undertaken routinely in patients with TBI, with a recent study showing around an insertion frequency of 32% . This study also showed a wide variation in timing of tracheostomy between different institutions. The optimal time to perform a tracheostomy has been the subject of much debate. A systematic review was unable to identify any validated model to predict which patients with traumatic injuries will benefit most from a tracheostomy . However, a further systematic review did identify seven studies that compared outcomes for patients with TBI who had early vs. late tracheostomy (mean (SD) 5.6 (0.34) days vs. 11.8 (0.81) days, respectively) . Both cohorts consisted of patients with severe TBI with a mean admission GCS < 5. Early tracheostomy was associated with a reduction in the incidence of ventilator-acquired pneumonia and duration of mechanical ventilation, critical care stay and hospital stay. Mortality was similar in both groups. These data suggest that early tracheostomy would appear to offer a meaningful clinical benefit. This is supported by a study by Wabl et al. who showed that among a small cohort of patients (n = 27) who underwent tracheostomy after severe brain injury, at 1-y follow-up only one patient had not been decannulated and over 50% of patients were able to walk and perform basic activities of living independently .
The potentially deleterious effects of both hypo- and hyperventilation in patients with TBI who have reduced intracranial compliance are well known. The importance of maintaining normocapnia is emphasised in TBI management protocols (Table 1), although these relate primarily to in-hospital care. A systematic review examined the association between ventilation strategies during the initial post-injury and resuscitation phase in patients with TBI . Normocapnia (the definition of which varied between the six studies) was associated with a lower mortality, although neurological outcomes were not assessed, and some studies measured end-tidal carbon dioxide which may not be reflective of PaCO2 . However, this suggests that pre-hospital emergency services that can undertake advanced airway interventions (such as tracheal intubation) may improve outcomes for patients with TBI, which is in line with the findings from retrospective analyses of UK trauma registries [56, 57].
While hypoxia has long been associated with worse outcomes in TBI, the influence of hyperoxia is uncertain. A systematic review of patients who were critically ill showed that hyperoxia (defined variably as PaO2 > 13.3 to > 64.9 kPa) was not associated with an increase in mortality (OR 1.23 (95%CI 0.91–1.67)) . In contrast with previous work, the findings were similar for patients with stroke and intracranial haemorrhage (OR 1.02 (95%CI 0.79–1.36)). However, in line with previous work, hyperoxia after cardiac arrest was associated with a higher mortality (OR 1.30 (95%CI 1.08–1.57)). Given the established dangers of hypoxia to the injured brain after TBI, it would appear to be reasonable to tolerate a degree of hyperoxia to allow adequate oxygen reserves for unanticipated events. However, this requires further investigation, especially with respect to the impact of hyperoxia on neurological outcome.
Fluid administration and blood pressure management
Episodes of hypotension after TBI may cause secondary brain injury and are associated with worse neurological outcomes. Intravenous fluids are commonly administered to maintain intravascular volume and cerebral perfusion. Traditionally, the fluid of choice in patients with TBI has been 0.9% saline, as this is relatively isotonic relative to normal plasma. However, there is increasing recognition of the deleterious effects of excessive administration of 0.9% saline, as this may result in hyperchloraemic acidaemia. As a result, there is research investigating the use of balanced intravenous solutions, such as Hartmann’s solution and Plasma-Lyte® (Baxter Healthcare Ltd., Norfolk, UK). A systematic review and network meta-analysis that investigated different types of fluid for resuscitation identified four studies that focused on patients with TBI . This suggested that 0.9% saline and low-molecular weight hydroxyethyl starch were associated with a lower mortality than balanced crystalloid solutions. However, it should be noted that neurological outcomes were not assessed. Similarly, a systematic review of pre-hospital fluid administration showed no improvement in mortality or neurological outcome with resuscitation using hypertonic saline compared with iso/hypotonic crystalloid solutions . Thus, at present it is not possible to recommend the use of one crystalloid solution. Further research is needed to determine not only the optimal intravenous fluid, but also how to assess precisely what volume to deliver and at what time-point.
Alongside intravenous fluid administration, vasopressor therapy is often initiated, with noradrenaline used most commonly in Europe and the USA. However, there is concern that, due to excessive vasoconstriction, noradrenaline may result in hypoperfusion and hypoxia in other organs. A systematic review investigated whether the administration of noradrenaline in patients with TBI resulted in improved neurological outcomes . The paucity of literature in this research area is illustrated by the fact that only two articles were identified, both of which were completed over 10 y ago. Neither study was able to show any outcome benefit associated with the use of noradrenaline but these data are unlikely to be generalisable to contemporary clinical practice. This illustrates one of the challenges facing future TBI-based research studies, with the need for long-term neurological status to be the primary outcome measure.
Other interventions for trauma
The benefits and efficacy of immobilisation of the cervical spine as a routine following trauma continues to be a matter of debate . A meta-analysis examined the effect of cervical collar application on ICP. Only five studies including 86 patients were analysed, but these showed that application of a cervical collar increased ICP by approximately 4.4 mmHg, despite the collars only being applied for a short time period (3–20 min). All included patients had suffered TBI, with a mean ICP before collar application of 13–21 mmHg. Given the lack of high-quality studies showing outcome benefits with cervical spine immobilisation , this work suggests that the routine use of cervical collars in patients with TBI should be reconsidered.
Over the past decade, there has been a move away from total definitive care for the patient with multiple injuries, with damage control surgery now favoured. Patients with TBI often have co-existing fractures and the optimal timing for surgical management of these is controversial; early surgery may offer a better chance of fracture union and improved function, but there is the potential risk of surgical complications (e.g. intra-operative bleeding and/or hypotension, systemic inflammatory response) either causing or exacerbating secondary brain injury). A review of 14 retrospective cohort studies involving patients with TBI (the majority of whom had moderate or severe brain injury) showed that extremity fracture fixation within 24 h of injury did not affect the risk of early mortality or neurological complications, but fixation beyond 14 days of injury was associated with increased rates of mal- and non-union . The optimal timing therefore remains unclear and should be determined on a case-by-case basis; however, fractures requiring surgical repair should ideally be addressed within 14 days of injury.
Future challenges and research
Many research studies have tried to identify a ‘magic bullet’ for the treatment of TBI, with little success. Indeed, there have been more negative than positive trials over the past 20 y, with the majority unable to demonstrate improvements in mortality or neurological outcome. This is unsurprising, given that TBI is not a single disease entity, but rather a collection of heterogeneous disease types, which will often have different causes and severity. For example, the term ‘TBI’ in an interventional study may include one or more of the following: subdural haematoma; extradural haematoma; cerebral contusional injury; traumatic subarachnoid haemorrhage; diffuse axonal injury; and intraparenchymal haemorrhage. Future research may need to focus on discrete types of cerebral injury in order to show any benefits. In addition, future studies should ensure that patients who are recruited reflect the baseline characteristics of the 21st century TBI population, with inclusion of older patients and/or patients with comorbid conditions. Outcome measures must be patient centred; for example, mortality is often of far less importance to patients and their families than long-term neurological outcome and functional ability.
Despite significant advances in the pre-hospital, peri-operative and critical care management of traumatic injuries, TBI continues to be associated with high mortality and significant long-term neurological impairment. At the present time, there are few clear therapeutic interventions that are associated with meaningful improvements in mortality, or more importantly, neurological outcome. The early administration of TXA appears to offer a clinically important early mortality benefit and early tracheostomy appears to be of value. In line with management strategies delivered on ICU for conditions such as acute respiratory distress syndrome, it is likely that the greatest benefits will be found from multifaceted interventions and consist of care bundles. Future research is needed and should focus on patient-centred outcome measures such as neurological recovery and quality of life for patients and their families.