In TBI, the increased intracranial volume caused by bleeding or tissue edema is initially compensated for by a decrease in cerebrospinal fluid (CSF) volume. However, there is a limit to this compensation, with further increases in space-occupying lesions increasing the ICP excessively, which restricts blood flow to the skull, resulting in tissue ischemia [7]. Elevated ICP is also transmitted to the brain parenchyma and can lead to uncal herniation [7].

Cerebral perfusion pressure (CPP) = Mean arterial pressure (MAP) – ICP

In order to maintain adequate CPP, monitoring and management of both systemic blood pressure and ICP is required. Both a decrease in MAP or an elevation in ICP will deleteriously alter the effective perfusion pressure.

Management of TBI includes maintaining adequate cerebral blood flow and oxygenation to avoid secondary brain injury. Previously, this therapy was directed at managing ICP, but there has recently been a shift towards strategies aimed at maintaining adequate CPP. Maintaining systolic blood pressure at ≥ 100 mm Hg for patients 50–69 years old and at ≥ 110 mm Hg or above for patients 15–49 or over 70 years old may be considered to decrease mortality and improve outcomes (level III) [6]. ICP monitoring remains a level IIB recommendation in the latest Brain Trauma Foundation guidelines [6]. Using information obtained by ICP monitoring for the management of patients with severe TBI is recommended to reduce in-hospital and 2-week post-injury mortality [8].

According to The Brain Trauma Foundation, ICP > 22 mmHg is associated with increased mortality (level IIB). There are no Class I study data that indicate an optimal CPP threshold or any specific optimal CPP in patients with TBI. It has been found that maintaining the CPP below 50 mmHg produces signs of ischemia and raising it above 60 mmHg avoids cerebral oxygen desaturation in the injured brain [9,10]; that is, it is suggested that the critical threshold for CPP lies between 50 and 60 mmHg and that CPP below 50 mmHg should be avoided. One study that proposed a critical CPP threshold of 60 mmHg suggested that the treatment goal should be to maintain the CPP at 70 mmHg to remain above the threshold [11]. Current guidelines recommend maintaining CPP at 60–70 mmHg and acknowledge that optimal CPP may vary depending on cerebral blood flow autoregulation [6].

Both invasive and non-invasive methods can be used to evaluate ICP [12,13]. Invasive methods generally involve the use of either external ventricular drains (EVDs) or intraparenchymal monitors. The EVD method serves the dual purposes of CSF diversion and continuous ICP measurement and is considered the gold standard [12]. Non-invasive methods for monitoring ICP include pupillometry, optic nerve sheath diameter measurement, and transcranial Doppler; however, these have various limitations with regard to intraoperative use. Although ICP monitoring in patients with TBI is a level IIB recommendation according to the Brain Trauma Foundation guidelines [6], ICP monitoring during extracranial surgery requires the patient to be equipped with a device, such as an EVD, that can be connected to pressure monitors used in the operating room. If there is no attached device, cooperation with a neurosurgeon is required, which will depend on the clinical environment at each institution.

The fastest way to decrease ICP is to allow CSF drainage, if possible. Another simple alternative to promote cerebral venous drainage is head-of-bed elevation to 30° (the reverse Trendelenburg position) while maintaining the patient’s neck in a neutral position. Hyperosmolar medications can lower ICP by allowing osmotic mobilization of water across the intact blood-brain barrier (BBB) into the intravascular space, thus reducing cerebral water content; commonly used agents include mannitol and hypertonic saline. If the patient is stable, 0.25–1 gm/kg of mannitol can be carefully administered slowly over 15 min to avoid intravascular volume depletion and hypotension due to urinary losses. Currently, there is not sufficient data to reach a definitive conclusion—although hypertonic saline showed a trend towards lower mortality and a more beneficial effect on CPP compared to mannitol, there were no statistically significant differences in mortality and neurological outcomes between hypertonic saline and mannitol administration [14].

Although the ideal serum sodium concentration is not well established, close monitoring of blood sodium levels is imperative to prevent hypernatremia, as rapid changes in serum sodium levels may be a causative factor for central pontine myelinolysis. A serum sodium concentration greater than 155 mmol/L has been suggested to be an independent predictor of acute kidney injury [15]. It should also be considered that the effects of administering a hyperosmolar agent will be altered in case of BBB rupture, which is common in patients with TBI [7].

Although decreasing ICP is the first-line therapy, increasing MAP to maintain CPP should also be considered. Patients with TBI who need to undergo non-neurosurgery may be in an acute state, in which systemic hemodynamics are unstable due to trauma, or a subacute state, in which patients are recovering from the trauma. Accordingly, blood pressure must be appropriately managed to maintain CPP even before the induction of anesthesia, and special attention is required to avoid exacerbating hypotension due to anesthetics or bleeding during surgery. If there is a modifiable cause of hypotension, its management should be prioritized, and vasopressor administration must always be considered. Selecting which vasopressor to use is less straightforward. A retrospective analysis showed that phenylephrine was the preferred vasopressor, with a greater increase in MAP and CPP after the start of infusion compared to norepinephrine or dopamine [16]. At equal MAP, using ephedrine resulted in better brain microcirculation and oxygen delivery than with the use of phenylephrine [17]. However, clear clinical evidence regarding the different effects of vasopressors in conditions with BBB disruption is still lacking and needs further investigation [18]. Given the ample evidence that autoregulation of cerebral blood flow in response to changes in CPP is impaired in both severe and mild TBI [19-21], careful attention to blood pressure management is necessary, because low blood pressure can be directly linked to reduced cerebral blood flow.

The incidence of cervical spine injury in trauma patients is reported to range from 3.5 to 6.2% [22,23]. Notably, cervical spine injury must always be suspected in patients with TBI, and caution should be exercised during endotracheal intubation and while repositioning patients. Video-guided laryngoscope with cervical immobilization was reported to reduce the upper cervical spine motion than Macintosh laryngoscope [24,25] and facilitated a more rapid tracheal intubation compared with the flexible bronchoscope [26]. Awake tracheal intubation using a flexible bronchoscope can minimize cervical spine movement [27], but requires sufficient procedural experience and may be difficult to perform depending on the level of consciousness of patients with brain trauma. In anesthetized patients without cervical immobilization, tracheal intubation using a video-guided laryngoscope resulted in greater cervical spine movement than a flexible bronchoscope, and jaw thrust during flexible bronchoscopy also causes movement of the cervical spine [28]. Therefore, an awake bronchoscopic approach is expected to be more effective for experienced clinicians when the situation, including the patient’s condition, allows, whereas using a flexible bronchoscope, lightwand, or video-guided laryngoscope with cervical immobilization may be a better option for inexperienced clinicians. Nasotracheal intubation is better to be avoided in case of patients with skull base fractures; in such cases, there is a risk that the tube may pass through the cribriform plate and enter the frontal brain region as it advances [29].

Tracheostomy is commonly performed on patients with TBI in the intensive care unit—in approximately 32% of cases—and is most frequently performed after the first week in the intensive care unit [30]. A recent systematic meta-analysis reported that early tracheostomy in patients with severe TBI contributes to reducing mechanical ventilation duration, intensive care unit and hospital stay duration, as well as the incidence of ventilator-associated pneumonia [31]. The Brain Trauma Foundation guidelines also recommend early tracheostomy when the overall benefits appear to outweigh the risk of complications [6].

Intravenous anesthetics, which are mainly used for the induction of general anesthesia, have various effects on cerebral blood flow and the cerebral metabolic rate of oxygen (CMRO₂). Propofol, thiopental, midazolam, and etomidate reduce CMRO₂, resulting in decreased cerebral blood flow and ICP. However, thiopental and propofol can cause hypotension and reduce CPP; therefore, caution must be exercised with regard to dose titration. Etomidate has a limited effect on MAP and is advantageous for managing CPP. Unfortunately, a single dose of etomidate is sufficient to cause adrenal insufficiency, and there are concerns that its use may be associated with mortality in patients with sepsis [32]. Even though a recent meta-analysis supports its use [33], the evidence is insufficient to guarantee the safety of etomidate use under septic conditions [34]. Ketamine has historically not been indicated for patients with brain trauma, based on the belief that it has detrimental effects on ICP. However, recent clinical data support the neuroprotective effects of ketamine via reducing glutamate levels and inhibiting cortical spreading depolarization [35,36]. A systematic review concluded that there was no evidence of harm due to ketamine use in patients with acute brain injury [37]. When inducing anesthesia in a patient with TBI, etomidate can be considered preferentially in patients with unstable hemodynamics, whereas ketamine can be an alternative option if septic shock is the cause of unstable hemodynamics.

There is no evidence to indicate that TBI outcomes can be improved based on the type of anesthetic agent used. Maintenance of anesthesia may be accomplished using inhaled or intravenous anesthetics, with careful consideration of the hemodynamic management goals. Total intravenous anesthesia using propofol and opioids is advantageous for cerebral hemodynamic management and neurophysiological monitoring and is frequently used in neurosurgery. As a highly selective alpha‑2‑adrenergic agonist, dexmedetomidine, which is expected to reduce ICP by reducing cerebral blood flow, has been reported to reduce anesthetic and opioid requirement and postoperative nausea and vomiting when used as an adjunct to general anesthetics [38], but evidence regarding its optimal dose as an adjuvant and its comparative effects versus other anesthetics is lacking [39]. Low concentrations of isoflurane and sevoflurane suppress brain metabolism and constrict cerebral blood vessels [40]. At higher concentrations, the direct vasodilatory effects of volatile anesthetics dominate, increasing both cerebral blood flow and ICP. Nitrous oxide is another cerebral vasodilator that increases cerebral blood flow and ICP; therefore, it should ideally be avoided [41]. When a volatile anesthetic is selected for general anesthesia, a minimum alveolar concentration of < 1 seems to be appropriate for patients with TBI [40].

Hypoxemia (generally defined as PaO₂ <  60 mmHg) is a major factor in the development of secondary brain injury and should be avoided [4,42]. Despite the evidence being insufficient to conclude whether it affects clinical outcomes, hyperoxia after TBI has been suggested to be associated with higher mortality [43]. Therefore, a balanced approach that avoids the higher and lower extremes is suggested for oxygenation management.

PaCO₂ is a potent mediator of cerebrovascular reactivity. Hypercapnia causes cerebral vasodilation via CSF acidosis; therefore, even a small increase in PaCO₂ may have deleterious effects on ICP in TBI patients with low intracranial compliance [44,45]. Hypocapnia causes cerebral vasoconstriction via an increase in CSF pH, and sustained reduction in PaCO₂ can lead to cerebral ischemia [44]. The Brain Trauma Foundation recommends avoiding hypocapnia [6] and hyperventilation in patients at risk of herniation, for short-term control of ICP, should be cautious. According to a recent meta-analysis, approximately 20% of TBI patients have acute respiratory distress syndrome (ARDS), which leads to worse neurological outcomes and higher mortality [46]; this indicates the need to use lung-protective ventilation during general anesthesia, especially in TBI patients with multiple-trauma, although concerns regarding resulting hypercapnia remain. Given the risks and benefits of lung-protective ventilation, the consensus when applying mechanical ventilation to patients with brain injury strongly recommends lung-protective ventilation in the presence of ARDS and no elevated ICP, and weakly recommends lung-protective ventilation in the absence of ARDS or elevated ICP [47]. Positive end-expiratory pressure (PEEP) can promote improved oxygenation and reduced intrapulmonary shunting; however, increased PEEP can lead to hypotension and alveolar overdistension with increased dead space, resulting in a higher PaCO₂ [48]. Furthermore, the increased intrathoracic pressure following elevated PEEP may reduce the pressure gradient of cerebral venous outflow and lead to an increase in ICP [49]. Therefore, as a practical strategy for safe PEEP titration in patients with ARDS and elevated ICP, it was suggested that the level of PEEP should be adjusted by monitoring the response of blood pressure and ICP such that increasing the PEEP does not increase ICP [50].

At the time of admission to the emergency department, coagulopathy is present in up to 25–35% of trauma patients [51]. During the initial hours of trauma-induced coagulopathy development, hypocoagulability is typically present, resulting in bleeding; later, coagulopathy is characterized by a hypercoagulable state associated with venous thromboembolism and multiple organ failure [52]. Coagulopathy is an integral component of a vicious cycle when combined with acidosis and hypothermia [53].

Ongoing intracranial bleeding can lead to increased ICP, brain herniation, and death. Tranexamic acid reduces bleeding by inhibiting the enzymatic breakdown of fibrin blood clots [54]. The role of tranexamic acid in the management of TBI remains unclear—a recent randomized controlled study showed that tranexamic acid administration within 3 h of injury reduced all-cause 24-h mortality in cases of mild to moderate injury, but had no effect on 28-day all-cause mortality [55], while a recent meta-analysis showing that tranexamic acid has no effect on mortality or neurological recovery, although its use probably does not increase the risk of adverse events [56].

Current approaches to trauma resuscitation focus on controlling bleeding and managing trauma-induced coagulopathy through the timely administration of hemostatic therapy [57]. Conventional coagulation tests include platelet counting and fibrinogen level and PT and aPTT measurement; however, they are relatively slow and do not provide a measure of platelet function or fibrinolysis [58,59]. Due to these shortcomings, anesthesiologists and trauma critical-care surgeons often prefer viscoelastic hemostatic assay-guided blood component therapy for evaluating bleeding patients in whom coagulopathy is common [58]. The two most commonly used viscoelastic hemostatic assays are thromboelastography and rotational thromboelastometry. Viscoelastic hemostatic assays provide clinicians with real-time data and a complete view of the coagulation process, from clot initiation and formation to clot stability and fibrinolysis measurements [57]. Although recent studies and reviews have described no differences in clinical outcomes between resuscitation guided by viscoelastic hemostatic assays and those guided by conventional coagulation tests [59,60], clinicians can use both methods to monitor hemostasis across heterogeneous groups of patients.

Trauma triggers an increase in stress hormone and cytokine levels, resulting in enhanced glucose production, reduced insulin production, and insulin resistance in peripheral tissues [61]. In the context of acute illness or injury, this has been termed stress-induced hyperglycemia and is defined as a transient plasma glucose level > 200 mg/dl in patients who are normally euglycemic [62]. There is no consensus on the best approach for glycemic control in patients with TBI. Studies on the association between stress-induced hyperglycemia and patient outcomes have consistently reported higher morbidity and mortality rates [63,64]. However, it remains unclear whether stress-induced hyperglycemia has a direct causative effect on worsening outcomes or is simply a marker of more severe disease [62]. The prospective, multicenter Normoglycemia in Intensive Care Evaluation-Survival Using Glucose Algorithm Regulation study reported a higher 90-day mortality with strict glucose control than with standard care, without any effect on organ failure or length of stay, suggesting that the target of glycemic control can be abandoned [65]. In a study that evaluated outcomes before and after strict glucose control therapy with a target glucose level of < 140 mg/dl in critically ill trauma patients, mortality, length of hospital stay and intensive care unit stay, and number of ventilator days did not differ between comparison periods [66]. These trial results do not allow for conclusive conclusions to be drawn regarding the exact glucose level that should be maintained. Tight glycemic control to maintain blood glucose levels below 110 mg/dl is likely not required and may even be detrimental to patient outcomes [62], including causing hypoglycemia [67]. A moderate level of glycemic control aimed at stabilizing glucose levels while reducing hyperglycemic and hypoglycemic events appears to be safe [68].

A systematic review published in 2003 showed that clinically induced hypothermia could reduce the risks of mortality and poor neurological outcomes in adults with TBI [69]. It has been hypothesized that hypothermia has a neuroprotective effect in TBI by disrupting post-injury biochemical and inflammatory cascades [70]. However, a multicenter randomized trial demonstrated that early prophylactic hypothermia did not improve neurological outcomes at 6 months compared to normothermia in patients with severe TBI [71]. Similarly, the results of another recent meta-analysis do not support the use of early prophylactic hypothermia (within 6 h after injury) as a neuroprotective strategy in adult patients with TBI, although they indicate that hypothermia could effectively reduce refractory high ICP [72]. Meanwhile, the incidence of central or neurogenic fever is considerably high in patients with severe TBI, especially in those with diffuse axonal or frontal lobe injury [73]; fever due to infection is also common in patients with severe trauma [74]. The detrimental effects of fever on the neurological recovery of patients with acute brain injury have been increasingly recognized [75-77]. Evidence supports the maintenance of normothermia in patients with TBI [78], and the Brain Trauma Foundation guidelines do not recommend prophylactic hypothermia [6]. Additionally, core body temperature is not a reliable indicator of brain temperature [79], and direct measurement remains the best way to monitor brain temperature in patients with brain injury, although brain temperature monitoring is currently not routinely applicable.

There is no evidence to specify the exact and safe time following TBI for elective surgery. Some studies recommend elective non-neurosurgery 6 months after recovery from TBI [80], based on the fact that patient symptoms generally do not improve or worsen 6 months after severe TBI [81]. However, in a meta-analysis comparing the effects of early and late fracture fixation on the prognosis of patients with limb fractures and concomitant TBI, late fixation (performed > 14 days after trauma) was associated with nonunion or malunion, and early fixation (within 24 h) did not affect the incidence of mortality, pneumonia, ARDS, or neurologic adverse events [82]. A retrospective cohort study reported that early orthopedic and facial fracture fixation (≤ 24 h after injury) under general anesthesia was not associated with worse neuropsychological or functional outcome than late surgery in multisystem trauma patients with TBI [83]. Currently, there are no contraindications to anesthesia in TBI, and once the decision to proceed with surgery has been made, steps should be taken to reduce the risks associated with surgery based on an understanding of the pathophysiology of TBI and the interactions between surgery and application of anesthesia [84].