1. Standards
   Data are insufficient to support a treatment standard for fluid resuscitation in the patient with severe traumatic brain injury (TBI).
2. Guidelines
   It is customary to treat hypotension with fluids in patients with TBI. Inadequate data exist to support a specific target blood pressure. Inadequate clinical outcome data exist to prefer one resuscitation fluid choice over another; however, hypertonic saline and colloids offer clear logistical advantages over isotonic crystalloids in a combat environment. Hypertonic saline in the prehospital phase is safe in doses < 500 ml and can be used for hypovolemia.
3. Options
   Hypotension (systolic blood pressure < 90 mm Hg) in patients with TBI has an association with poor outcome. Fluid therapy can be used to maintain adequate cerebral perfusion pressure and limit secondary brain injury. Inadequate fluid resuscitation with aggressive diuresis can precipitate hypotension and should be avoided in the field setting. Hypertonic saline resuscitation, with or without dextran, has been used with some encouraging results compared to isotonic fluids. If a casualty requires additional fluids after the administration of 500 ml of hypertonic saline, isotonic fluids or colloids can be used.

The primary purpose of fluid resuscitation of TBI patients in the field is to treat shock and to prevent hypotension. Since there is an association of worse outcome in TBI patients with hypotension, it is thought that treatment with fluid resuscitation may potentially prevent secondary brain injury.

While the treatment of hypotension to prevent secondary brain injury may be intuitive, the quality of human data to demonstrate the cause and effect is lacking. However, it has been shown that even a single episode of hypotension can double mortality.1,2 Whether the worse outcome is due to the secondary TBI or is merely an association is unclear. Autoregulation sometimes fails following head injury, placing the brain at increased risk from hypovolemia as the compensatory mechanisms to maintain brain perfusion are disrupted. Although some animal data examine these phenomena, no research in humans exists.

In the civilian setting, a rapid infusion of 2 liters of crystalloid fluid (lactated Ringer's solution [LR] or normal saline) is customarily utilized to treat hypovolemia in adults.3 In casualties without head injury, there is concern that resuscitation without surgical control of the source of bleeding may increase blood loss by displacement of potentially hemostatic clots due to higher blood pressure. No clinical studies have demonstrated that prehospital fluid resuscitation is associated with improved outcome. Indeed, one randomized prospective study in hypotensive patients with penetrating torso trauma showed that patients treated with fluids had increased mortality.4 Studies have also demonstrated that the type of fluid used in the prehospital setting does not affect mortality. This is consistent in that if the use of fluids does not make a difference, then neither would the type of fluid.

It may not be legitimate to extend the findings on civilian prehospital fluid use to the combat setting. Transport times may be significantly longer on the battlefield compared to the urban settings of published civilian studies. The epidemiology and pathophysiology of combat-related trauma differs from civilian trauma with far less blunt trauma and more blast related and high velocity projectile penetrating trauma. In addition, the logistics of combat casualty care differ from civilian trauma care, with weight and volume of all emergency response equipment critically balanced with the weight and volume of gear required for mission accomplishment.

The primary goals of combat casualty care in the field are to control hemorrhage and to rapidly transport casualties to higher levels of care. Due to potential problems with aggressive fluid use, the current recommendation regarding fluid resuscitation for patients without head injury is to allow "permissive hypotension" by monitoring mental status and pulse character.5 In the noisy and potentially dangerous environment of first response and initial evacuation, sphyngomanometers are not likely to be effectively utilized. It is recommended that all casualties have intravenous access established when not under hostile fire. No fluids should be instituted in the presence of a strong radial pulse and normal mentation. If the mental status and radial pulse are not normal, fluids can be titrated to improve mental status and restore a weakly palpable radial pulse. This scheme of resuscitation is termed "permissive hypotension" and offers the potential benefit of minimizing uncontrolled blood loss and the logistical burden of fluid resuscitation on the battlefield.6 Since the vast majority of combat casualties do not require fluid administration, this has real logistical advantages.7 In addition, the use of oral hydration is recommended for the casualty without TBI, penetrating abdominal injury, or severe uncontrolled hemorrhage.

In the setting where the casualty has TBI, permissive hypotension and oral hydration are not yet recommended.8 Fluid resuscitation should be performed to establish normal pulse character or blood pressure in order to prevent possible secondary brain injury. On the battlefield, the optimal initial fluid of choice is (3-7.5%) hypertonic saline. There are ample human data to verify that hypertonic saline is safe as the initial fluid for resuscitation, although an outcome advantage over other types of fluid has not been proven definitively. Due to the logistical advantage of hypertonic saline being able to resuscitate equivalently to isotonic fluids with less weight and cube, it is the ideal fluid of choice. In addition to its ability to restore perfusion, it offers other proven and theoretical advantages such as its ability to reduce intracranial pressure and modulate the immune system to possibly reduce the inflammatory response which is often seen after severe injury.

In 1999, the Institute of Medicine recommended two 250 ml rapid infusions of 7.5% hypertonic saline as the initial fluid treatment of choice in both combat casualties and in civilian trauma.9 Since 7.5% hypertonic saline is not currently commercially available, alternative initial fluid choices are two 250 ml infusions of 5% hypertonic saline or two 500 ml infusions of 3% hypertonic saline. These latter solutions are commercially available, but they are not currently standard military field supplies.

Another option that has been recommended by the U.S. Special Operations Command and some military trauma experts is to use colloids as initial resuscitation fluid.10 Hextend is 6% hetastarch in LR and is currently available in the field. Two 500 ml boluses can be used in the forward area with an effect functionally equivalent to six bags of LR. If the casualty remains hypotensive after the infusion of hypertonic saline or colloids, the casualty should be assumed to have ongoing blood loss and continued efforts to maintain blood pressure should be weighed against logistical and tactical considerations. While the weight and volume of resuscitation fluids may not be as important in the setting of civilian trauma, it is of major concern on the battlefield. An option if hypertonic saline or colloids are not available is to use isotonic fluid, though this is less desirable for the reasons discussed above.

A MEDLINE search was conducted from 1978 to 2005 using the keywords "head injury," "field or prehospital," and "fluid resuscitation." The search turned up 150 references, 40 of which were relevant to fluid therapy for the patient with severe head injury. These were individually reviewed for content. The results were collated, and the analysis is presented here.

The traditional method of resuscitation of a hypotensive patient with TBI is with crystalloids. Although the scientific evidence still is not abundant, most textbooks and trauma courses such as the Advanced Trauma Life Support™ course recommend crystalloid use. While LR is the customary fluid in trauma, normal saline is preferred in the setting of TBI as the sodium content is higher, thus minimizing the potential for resuscitation with a hypotonic solution which could increase cerebral edema.

Because sodium is vital in casualties with TBI, the research regarding hypertonic saline is important. Hypertonic saline has multiple theoretical advantages. Resuscitation can be achieved safely with approximately one eighth the volume of normal saline and LR when using 7.5% hypertonic saline. While not very important in the civilian sector, it is critical in the combat scenario as the medics and corpsman have to carry the fluids to be used. Hypertonic saline may have immunomodulatory capabilities as the sodium affects neutrophils which have been implicated in the aberrant inflammatory response after trauma and massive resuscitation. Another potential benefit of hypertonic saline is the reduction of intracranial pressure as the high osmolarity produced from the sodium infusion reduces cerebral edema.

Clinical studies in patients with TBI have been performed examining the effect of hypertonic saline. In a multicenter trial, Mattox et al.11 demonstrated a higher systolic blood pressure in patients treated with hypertonic saline versus crystalloid resuscitation. Survival was significantly better in patients who required surgery, and the hypertonic saline group had fewer complications compared with the group receiving the standard isotonic crystalloid treatment. That trial did not mention head injuries specifically. Wade et al.12 performed a meta-analysis on published controlled studies of hypertonic saline/dextran, then abstracted the data on patients who had TBI (defined by an abbreviated injury score [AIS] for the head of 4 or greater). Survival to discharge was 37.9% for patients treated with hypertonic saline and 26.9% for standard therapy. These findings failed to reach statistical significance (p = 0.08). When logistic regression analysis was performed, the odds ratio was 1.92 for 24-hour survival and 2.12 for survival to discharge when hypertonic saline was compared with standard therapy. This was a statistically significant difference (p = 0.048). Wade concluded that patients who had TBI and received hypertonic saline/dextran were about twice as likely to survive as those who receive standard therapy.

Vassar and her colleagues13-16 published four prospective randomized double-blind trials between 1990 and 1993 concerning the use of hypertonic saline. In 1990, they compared two groups of head injury patients, one group receiving 7.5% hypertonic saline, the other receiving normal saline.13 Twenty-six percent of the head injury patients were found to have intracranial pathology with bleeding. No difference in outcome was found between the two groups. In addition, intracranial bleeding did not increase with either therapy. In 1991, Vassar et al.14 compared 7.5% hypertonic saline with LR in 166 patients, 32% of whom had severe TBI (defined as an AIS of 4 or higher). Crude mortality measurement was the same. When logistic regression analysis was used, hypertonic saline/dextran was associated with a statistically significant higher survival rate than isotonic crystalloid.

In 1993, Vassar et al.15 published a trial of 7.5% hypertonic saline versus 7.5% hypertonic saline dextran in 258 patients. Only 10% had severe TBI. However, in patients with a Glasgow Coma Scale (GCS) score < 8 and in patients with severe anatomic cerebral damage, survival with either agent was statistically significantly greater than what would be predicted with the Trauma Related Injury Severity Score (TRISS). The addition of dextran to the hypertonic saline did little to improve survival. In 1993, Vassar et al.16 also published a multicenter trial of 194 patients of whom 74% had severe TBI. There was no statistically significant increase in the survival in the overall patient population with the use of hypertonic saline. However, the survival rate in the hypertonic saline group was higher than in the LR group for patients with an initial GCS score of 8.

A recent study by Cooper et al.17 examined the use of 7.5% hypertonic saline in the prehospital setting in 229 hypotensive blunt trauma patients with TBI. This study was a well-performed prospective randomized trial to determine the affect of one 250 ml dose of 7.5% hypertonic saline or LR in the prehospital setting. The primary outcome variable was neurologic function as measured by the Extended Glasgow Outcome Score (GOSE, 1-8) at 6 months. Entry criteria in this study were hypotensive (systolic BP < 90 mm Hg) trauma patients with severe blunt traumatic head injury (GCS < 8). The control group (n = 115) received LR compared to 250 ml of 7.5% hypertonic saline (n = 115). The patients were otherwise treated identically in the field and in the hospital and were given as much LR or colloids as providers deemed necessary. This is an important fact as the effect of hypertonic saline may have been diluted by the liberal use of other fluids which were not controlled in this study. Cooper found that there was no difference in GOSE between those that received HTS compared to those that received LR at 6 months even though the patients were very similar in injury severity, pattern, and their demographics.

There is some criticism regarding this study that should be considered. The design of the study may have doomed it to fail. Hypotensive trauma patients with head injuries are one of the most serious of trauma patients as confirmed by this study. Approximately half the patients in this study died. In general, there are minimal proven treatment options to change the outcome after such a severe head injury other than providing an airway, controlling blood loss, decompression of the skull for mass effect, and aggressive supportive critical care to minimize secondary brain injury. Even in the optimal setting, these treatment options only affect a small minority of patients. Although intracranial hypertension is associated with poor outcome, the scientific clinical data to demonstrate that benefit of reducing intracranial hypertension or increasing cerebral perfusion pressure in humans are scarce.

Although the Cooper study can be interpreted as a failure of hypertonic saline to improve outcome, the converse interpretation is also valid. This study did demonstrate that the use of hypertonic saline is as safe as conventional fluid therapy in hypotensive trauma patients with severe TBI. Although the group that was treated with 7.5% hypertonic saline had a higher survival rate (55% vs. 50%, p = 0.23), it was not statistically significant. This study was not powered for survival as the primary outcome. This study also demonstrated that the mean ICP tended to be lower (10 vs. 15, p = 0.08) upon arrival to the intensive care unit and the duration of the cerebral perfusion pressure less than 70 mm Hg also tended to be shorter (9.5 hours vs. 17 hours, p = 0.06).

Pentastarch, another hyperosmolar solution, was tested in 1992 by Younes et al.18 in a Phase 2 clinical trial of 23 hemorrhage patients. Although that study did not state the number of patients with severe TBI, some of them almost certainly had head injuries because the average GCS score was 11 ± 5. Both Pentastarch and saline increased blood pressure equally, although the volume requirements with Pentastarch were less. No differences were found in complication rates in the two patient groups.

Mannitol is another therapy that has been proven to reduce ICP in hospital patients with intracranial hypertension. One concern is that mannitol may produce hypotension from volume deficits secondary to its osmotic diuresis. This could potentially produce secondary brain injury. One prospective randomized double-blind controlled trial investigated the prehospital administration of mannitol in head-injured patients, comparing mannitol with standard crystalloid resuscitation.19 The demographics in that study did not differ, nor did the overall head injury severity between the two groups. Mortality was the same in both groups. Importantly, systolic blood pressure did not change significantly in the mannitol group at the time of ED presentation. However, two hours after hospital arrival, systolic blood pressure was statistically significantly lower in the mannitol group when compared with the placebo group. Very few of these patients were hypotensive. In addition, since hypotension is treated with fluids, the avoidance of mannitol may potentially reduce the total volume of fluids required.

While mannitol can reduce ICP, it has not yet been conclusively been shown to effect outcome. The exception to this is three prospective randomized trials from one center in Brazil.20-22 These studies, however, used "high dose mannitol (~1.4 grams/kg)" compared to standard dose mannitol. The mannitol was given in the ED approximately 80-90 minutes after injury and the results are extremely impressive. They found improved survival and disability scores in the group that received the high dose mannitol. However, these studies should be examined with caution as the high dose mannitol required preemptive aggressive fluid resuscitation and aggressive invasive monitoring which included jugular bulb oxyhemoglobin saturation monitoring. Patients in these studies all had blunt TBI, not penetrating, and they had access to craniotomy within 4 hours of admission.

Hypertonic saline has also been shown to reliably reduce ICP. Human studies have also shown this effect as Hartl et al.23 demonstrated that hypertonic saline reliably reduces ICP in patients with TBI and intracranial hypertension. In a prospective randomized study in two centers, Shackford et al.24 demonstrated that the use of hypertonic saline in 34 patients lowered ICP. Two studies showed in children with severe head injury that hypertonic saline reduced ICP and increased cerebral perfusion pressure.25,26 Bentsen et al.27 also demonstrated that in seven critically ill patients with subarachnoid hemorrhage that needed urgent treatment for elevated ICP, the infusion of 7.2% hypertonic saline with 6% hydroxyethyl starch lowered ICP and elevated cerebral perfusion pressure.

There are also some studies comparing 7.5% hypertonic saline versus 20% mannitol. A randomized prospective crossover trial demonstrated that 100 ml of 7.5% HTS with 6% dextran (RescueFlow® - not approved yet by FDA in the U.S.) caused a significantly decreased ICP, and had a longer duration of effect than mannitol. However, this study was small and only had nine patients.28 A study by Vialet et al.29 on 20 consecutive patients demonstrated that 7.5% hypertonic saline was more effective than 20% mannitol in treating intracranial hypertension. Horn et al.30 showed that in patients with elevated ICP that was resistant to mannitol and barbiturates, 7.5% HTS was effective in decreasing ICP. De Vivo et al.31 evaluated the effectiveness of 3% hypertonic saline and mannitol in neurosurgery (supratentorial cerebral tumors) and found that hypertonic saline could be used safely to reduce ICP without reducing central venous pressures. Since hypertonic saline can reduce ICP while also increasing intravascular volume, it would make sense to avoid the use of mannitol in the field setting.

The deleterious association of hypotension in patients with TBI has been documented in the literature. While permissive hypotension is practiced in the field for penetrating torso trauma, it is not advisable to recommend this for patients with TBI at this point. Because the underlying cause of hypotension in TBI patients is almost always secondary to bleeding or other fluid losses, intravascular volume resuscitation seems to be the most efficacious way of restoring blood pressure. Isotonic crystalloid solution is the fluid most often used in the prehospital resuscitation of head injury patients.

There is Class I evidence that demonstrates that the use of hypertonic saline is a safe alternative method of treating hypotensive TBI without worsening outcome and there is lesser quality data to show it may have survival advantages in patients with TBI. Because hypertonic saline offers logistic advantage in terms of weight and cube in the field, it can be used in patients with TBI as it can reduce ICP while restoring intravascular volume. Two 250 ml bolus of 5% hypertonic saline or two 500 ml boluses of 3% hypertonic saline can be used as the initial resuscitation fluid. Colloids such as Hextend also offer weight and volume advantage compared to other fluids so it is also an alternative that can be used in the field setting. In patients with TBI that have no evidence of significant blood loss and have normal pulse character or blood pressure, there is no evidence to show that any fluid resuscitation is necessary. Mannitol in the prehospital/field setting has not yet been shown to improve outcome.

Research on fluid resuscitation in hypotensive patients with TBI has been very limited. There are little data to guide endpoints of therapy. One target blood pressure may be better than another, and MAP may be a better guide to therapy than systolic pressure, but these questions require investigation. In addition, the current concern that raising blood pressure may increase secondary blood loss, thus worsening cerebral hemodynamics, needs to be better validated in humans. Finally, more work must be done to elucidate the most effective fluid for resuscitation. The following specific questions should be studied in the future:

1. What is the optimal target blood pressure for resuscitation in both isolated TBI and the patient with multiple injuries?
2. Is mean arterial blood pressure a better endpoint than systolic blood pressure?
3. Is there a subgroup of patients in whom a lower volume of resuscitation fluid should be used?
4. What is the ideal resuscitation fluid for TBI patients in the prehospital setting?
5. Is there a role for large particle colloids in the prehospital setting?