What Is the Best Approach to Fluid Management, Transfusion Therapy, and the Endpoints of Resuscitation in Trauma?

Exsanguinating hemorrhage is a major cause of death from trauma. Rapid fluid resuscitation accompanied by aggressive efforts at hemostasis is required to save lives. Many questions regarding fluid resuscitation remain. These include the choice of fluid, indications for blood products, and the goals for fluid resuscitation before and after hemostasis is achieved.

Modern fluid resuscitation in trauma began in the early 1960s with the work of Shires and colleagues. During hemorrhage, fluid shifts from the interstitial space to the intravascular space because of changes in compartment pressures. Likewise, fluid initially shifts from cells into the interstitial space. As cells become ischemic during severe hemorrhagic shock (HS), however, failure of membrane ion pumps leads to a shift of fluids back into cells with resultant cellular swelling. Consequently, the interstitial space further loses fluid. Shires postulated that resuscitation with crystalloids, which fill the vascular and interstitial spaces, would be beneficial. His animal studies demonstrated that survival improved with the addition of crystalloid (lactated Ringer’s [LR] solution) to re-infusion of shed blood. Crystalloid resuscitation was quickly adopted in the military for resuscitation of trauma victims during the Vietnam conflict. Although this approach decreased the renal failure that had been seen in previous conflicts, it may have contributed to a new finding, “Da Nang lung” or “shock lung,” which may have been acute respiratory distress syndrome (ARDS) or simply hydrostatic pulmonary edema from volume overload. Administration of LR solution quickly became a standard of the Advanced Trauma Life Support (ATLS) course and of care in prehospital and emergency department (ED) resuscitation of civilian trauma victims. Recent studies suggesting possible immunologic effects of LR solution have led to questions of this practice and a great interest in finding better alternatives as plasma substitutes.

Administration of blood to replace lost red blood cells has been another mainstay of resuscitation from HS. Although whole blood was initially used, blood banks have found that dividing the blood into packed red blood cells (PRBCs), fresh frozen plasma (FFP), and platelets is more efficient and economical. Recent studies have suggested that more rapid resuscitation with blood components that essentially reconstitute whole blood may be beneficial. Blood transfusions, however, have many potentially deleterious effects. Greater recognition of these effects has led to reconsideration of aggressive transfusion protocols once hemostasis has been achieved.

Endpoints for fluid resuscitation, similar to the fluids themselves, have undergone reconsideration in recent years. Although the trauma victim has ongoing hemorrhage (uncontrolled HS), normalization of blood pressure may increase bleeding and worsen outcome. Limited, or hypotensive, fluid resuscitation may be appropriate. Once hemostasis has been achieved, determining adequacy of resuscitation is critical. Standard parameters such as blood pressure, heart rate, and urine output are insufficient because many patients remain under-resuscitated, with “compensated hemorrhagic shock.” Adjuvant tests are necessary to recognize this condition and ensure restoration of homeostasis.

Pathophysiology of Hemorrhagic Shock

HS is characterized by acute blood loss leading to decreased oxygen delivery to tissues. Although blood pressure and pulse are typically the clinical parameters used to determine the severity of shock, they lack sensitivity. In general, patients need to lose at least 30% to 40% of their blood volume to be hypotensive. An individual’s response to hemorrhage may be affected by age, comorbid conditions, medications, and ingestion of drugs and alcohol. One of the most common mistakes of the novice clinician is to assume that the patient with a normal blood pressure is not in shock. Recognition and reversal of “compensated shock” is critical to achieve optimal outcomes.

Approximately 6% to 9% of trauma patients are in shock on admission. Of these, one third have exsanguinating hemorrhage, as evidenced by a lack of response to fluid resuscitation. These patients invariably require operative intervention and aggressive fluid resuscitation including blood products or they will die within minutes to hours. Another third of patients are classified as transient responders. They are initially hypotensive and improve with fluid resuscitation only to deteriorate again. They have less active bleeding than the first group, but their transient response can lull clinicians into a sense of complacency. Without ongoing resuscitation, operative intervention if necessary, and vigilance, they also have a high risk of dying or developing multiple organ dysfunction. The final third of these patients respond appropriately to fluid resuscitation and spontaneously achieve hemostasis. These patients are still at risk of hypoperfusion and organ dysfunction. In all trauma patients, early recognition of hypoperfusion, rapid restoration of homeostasis, and continued resuscitation to appropriate endpoints can reduce the risk of early cardiovascular collapse, development of organ system dysfunction, and death.

The inflammatory response to trauma may increase the risk of organ dysfunction and late death from trauma. Although laboratory studies have suggested therapies that could mitigate these deleterious cascades, none of these agents have made it to clinical use. Despite this, recent studies suggest that late deaths from multiple organ dysfunction and sepsis are actually rare.

Presentation of Available Data Based on Systematic Review

Choice of Fluid. Although the use of crystalloids for resuscitation from traumatic HS has become standard, it seems that these solutions are not as innocuous as originally believed. Laboratory studies have demonstrated that crystalloids may exacerbate cellular injury. LR solution can cause an increase in oxidative burst and expression of adhesion molecules on neutrophils in human blood and during HS in pigs. No clinical studies have yet compared different crystalloids.

Modifications of LR solution (e.g., substituting the l -isomer of lactate or substitution of pyruvate or ketone bodies [β-hydroxybutyrate] for racemic lactate) can decrease the neutrophil activation and apoptosis. In contrast, hypertonic saline (HTS) and fresh whole blood do not cause neutrophil activation. HTS can attenuate immune-mediated cellular injury after trauma.

Several small clinical trials have suggested a benefit of hypertonic solutions for resuscitation of trauma patients ( Table 77-1 ). These studies explored the use of HTS alone or with a colloid added (e.g., hypertonic saline-dextan [HSD]) to prolong the intravascular volume expansion. Multiple studies demonstrated that HTS or HSD increased blood pressure and volume expansion better than crystalloids but could not document improved survival. Mattox et al. and Wade et al. found that HSD improved survival in the subset of trauma patients who required operation, presumably more severely injured patients. Likewise, Bulger et al. found that HSD, compared with LR solution, improved ARDS-free survival only in patients who required more than 10 units of PRBCs.

Table 77-1

Summary of Clinical Trials

Study, Year	Number of Subjects (Intervention/No Intervention)	Study Design	Intervention	Control	Outcomes
Hypertonic Saline for Hemorrhagic Shock
2011	376 NS	DB	HTS/HSD	NS	No difference in 28-day survival.
	256 HTS
	220 HSD
2007	36/26	DB	HSD	LR	Inhibit CD11b.
2007	36/26	DB	HSD	LR	Trend increase IL-1β, IL-10.
2007	110/99	DB	HSD	LR	No difference ARDS-free survival. Improved ARDS-free survival if >10 U blood.
2006	13/14	DB	HSD	NS	Promotes a more balanced inflammatory response.
2003	120/110	DB	HSD	NS	Survival 83% vs. 76% overall (NS), 85% vs. 67% for patients requiring surgery ( P = .01).
1993	85 HTS	DB	HS or HSD	LR	HS improved survival compared with TRISS.
	89 HS
	84 NS
1992	35/35/35	DB	HS or HSD	NS	No difference in survival. Better BP and volume expansion. Less fluid needed.
1991	83/83	DB	HSD	LR	Improved BP. No change in survival.
1991	211/211	DB	HSD	Crystalloid	No difference in survival, except patient who required operation. Improved BP, fewer complications.
1990	32 HTS	DB	HSD	LR	No safety issues except mild hyperchloremic acidosis.
	23 HSD		HS
	51 LR		HS
1989	48		HSD	PlasmaLyte A	Feasibility study.
1989	32	DB	HSD	Crystalloid	No difference in survival.
Transfusion
2007	240/439	Retrospective	Leukocyte-depleted PRBCs	Standard PRBCs	No difference in LOS or mortality.
2006	286	Randomized	Leukocyte-depleted PRBCs	Standard PRBCs	No difference in infections, organ failures, mortality.
2006	93/117	Prospective Operation Iraqi Freedom	Transfused	Not transfused	Higher ISS, HR, lower Hct, increased infection rate, ICU, and LOS.
2005	102	Prospective, observational			The amount of transfused blood is independently associated with both the development of ARDS and hospital mortality.
2004	954/8585	Prospective	Transfused	Not transfused	Older, higher ISS, lower GCS, more SIRS, higher mortality.
2004	100/103 trauma patients	Prospective	Hb 7-9 g/dL	Hb 10-12 g/dL	No differences.
2003	15,534	Prospective	Transfusion	Not transfused	Increase mortality (OR, 2.8), ICU, and LOS.
2002	61	Prospective	Transfused		Older blood increased risk of infections.
Clotting Factor Replacement
2011	DCR	Prospective and retrospective	Permissive hypotension, less crystalloid	Standard care, historical controls	DCR resulted in less crystalloid, more FFP, improved survival (OR, 0.4; CI, 0.18-0.9).
2011	108/82	Prospective and retrospective	Permissive hypotension, less crystalloid	Standard care, historical controls	DCR was associated with increased survival (OR, 2.5; CI, 1.1-5.6).
2010	214 patients receiving massive transfusion	Prospective and retrospective	Massive transfusion protocol	Standard care, historical controls	Factors that influenced survival were FFP/PRBC, platelets/PRBC, ISS, age, and total PRBCs.
2009	442 patients receiving massive transfusion	Prospective and retrospective	Preemptive FFP and platelets	Standard care, historic controls	Mortality decreased from 31% to 20% long term. Thromboelastography was used to titrate.
2009	37/40	Prospective and retrospective	1:1.5 FFP/PRBC, more rapid product availability	Standard care, historical controls	No difference in ratios, but more rapid administration was associated with improved mortality.
Limited Fluid Resuscitation for Uncontrolled Hemorrhage
2011	44/46	Prospective, randomized, intraoperative	MAP >50 mm Hg	MAP >65 mm Hg	Lower MAP group required fewer blood products, had less coagulopathy, and decreased early death.
2002	55/55	Randomized	SBP >70 mm Hg	SBP >100 mm Hg	Survival 93% with no difference between groups.
1996	527	Retrospective	Rapid infusion system used	Historical controls	Increased risk of dying 4.8×.
1994	309/289	Randomized day of month	Delayed resuscitation	Immediate resuscitation	Improved survival: 70% vs. 62%. Decreased LOS.
1990	6855	Retrospective	Prehospital fluid	No prehospital fluid	No difference in mortality.
Endpoints of Resuscitation from Trauma
2002	18/18	Prospective, nonrandomized	DO ₂500	DO ₂600	Less fluid and blood needed; similar outcome.
2000	40/35	Prospective, randomized	Supranormal DO ₂	Normal DO ₂	Patients who achieve supranormal values increased survival, but no difference between groups in mortality, organ failure, or LOS.
1995	50/75	Randomized	Supranormal DO ₂	Normal DO ₂	Improved survival (18% vs. 37%) and organ system failures.
1992	33/34	Randomized	Supranormal DO ₂	Normal DO ₂	Decreased mortality, organ failure, LOS, ventilator days.
2006	5995	Retrospective			Lactate did not correlate with mortality.
2003	98	Prospective		Standard care	Admission lactate level correlates with ISS and 12-h lactate with survival.
1998	100	Prospective	High BD	Low BD	Increase MOF and mortality, low oxygen use.
1998	674	Observational			BD worse in nonsurvivors. No difference in pH.
1992	3791	Retrospective			BD, age, injury mechanism, and head injury were associated with mortality using logistic regression.
1988	209	Observational			Higher BD associated with lower BP and greater fluid resuscitation.