Vasoconstriction and Reduction of Capillary Hydrostatic Pressure

Peripheral and splanchnic vasoconstriction help to redirect blood flow away from muscle, skin, bowel, kidneys, and other nonvital organs. This is accomplished through the release of epinephrine, norepinephrine, and vasopressin. By increasing the peripheral vascular resistance, and effectively reducing the circulating volume through which the blood must pass, the mean arterial pressure is increased. Initially, during compensated shock, only the precapillary sphincter is constricted, reducing the capillary hydrostatic pressure and increasing the peripheral resistance. As the shock worsens, both the precapillary and postcapillary sphincters are constricted, leading to anaerobic metabolism, which results in excess lactate and acidosis. As the severity of shock increases, the postcapillary sphincter relaxes secondary to hydrogen ion buildup, releasing the metabolic byproducts into the circulation. The result is reperfusion injury. This is common when the system decompensates, and the patient shows signs of florid shock and global hypoperfusion.

The subsequent decrease in capillary hydrostatic pressure reduces the volume of plasma lost from the vasculature into the interstitial space, and increases return of interstitial fluid into the vascular space. This is an additional method of increasing venous return to the heart. Given the large capacitance of the venous system, larger volumes of circulating blood are stored within the veins compared to the arteries. According to the Frank-Starling laws of cardiac performance, the cardiac output is directly proportional to the venous return to the heart (preload) (Figure 1). As venous return decreases, so does cardiac output, and thereby arterial pressure to perfuse vital organs.

Figure 1 The effect of increased right atrial pressure on cardiac output. Increases in right atrial pressure are accomplished through either an increase in circulating blood volume, or a decrease in venous compliance. Cardiac output increases linearly in response to increased right atrial pressure.

(Adapted from Berne RM, Levy MN: Cardiovascular Physiology, 3rd ed. St. Louis, MO, Mosby, 1977.)

Hormonal Response

In response to hypotension, numerous hormonal pathways are initiated to restore circulating plasma volume. Vasopressin plays a major role in peripheral vasoconstriction. Additionally, it acts to increase free water absorption in the collecting ducts of the kidneys. In response to lower pressures at the afferent arteriole within the kidney, renin is released, triggering the renin/angiotensin/aldosterone cascade. The effect is to increase the absorption of sodium, which in turn increases water absorption. Additionally, angiotensin-II acts as a potent peripheral vasoconstrictor. Finally, cortisol acts to increase extracellular osmolarity and increased lymphatic flow.

The reduction of capillary hydrostatic pressure and the hormonal response increase venous return during the first 1–2 hours following injury, but this may be inconsequential in the setting of severe hemorrhage. Cessation or slowing of ongoing hemorrhage also contributes to a spontaneous rise in blood pressure by limiting ongoing fluid losses. The remainder of the response is carried out through vasoconstriction and the redistribution of blood flow, accounting for the majority of increased venous return. If prolonged, this can have a damaging, and sometimes irreversible effect upon the tissues supplied by the splanchnic circulation. This will result in ischemia and potential reperfusion injury.

Phases of Shock

Combined, the aforementioned physiologic effects produce the classic signs and symptoms exhibited by patients in hypovolemic shock. The progression from compensated to uncompensated shock gives rise to the physical findings of the systemic insult (Figure 2). Initially, compensated shock affects the periphery, manifesting the symptoms of cool extremities, cyanosis, mottling of the skin, and decreased capillary reperfusion (blanching). As the severity of shock progresses, vasoconstriction affects the torso, producing oliguria as the main clinical sign. Finally, in severe classes of shock, perfusion of the heart and brain are affected, producing cardiac dysrhythmias and a reduced level of consciousness. Over the hours following injury, these mechanisms are in constant flux, in efforts to restore homeostasis to the injured system.

Figure 2 Signs and symptoms of shock.

Lower cardiac output, variable tachycardia, oliguria, and decreased capillary pressure define the initial phase following hemorrhagic shock. The interstitial space is contracted in efforts to preserve circulating plasma volume. This occurs typically in the prehospital phase and extends into the early portions of the patient’s initial resuscitation. During this period active replacement of blood loss is achieved through either crystalloid infusion or blood transfusion to restore circulating blood volume. The second phase is marked by extravascular fluid sequestration. This often occurs following surgical intervention and massive resuscitation. The intracellular and interstitial spaces expand as resuscitation continues, and blood pressure typically stabilizes. Crystalloids are given to maintain plasma volume based upon the patient’s observed hemodynamic status. The third and final stage is marked by diuresis and mobilization of fluids with an increase in blood pressure. This response occurs only after appropriate volume resuscitation, surgical correction of the cause of the hemorrhage, and resolution of organ failure with return to a more homeostatic state.

Choice of Fluids

The use of crystalloids versus colloids has been an ongoing debate for decades. In the Vietnam War, isotonic crystalloids were used when laboratory work from the 1960s by Shires and others showed larger volume resuscitation with isotonic crystalloids resulted in the best survival. They also noted that extracellular fluid redistributed into both intravascular and intracellular spaces during shock, and rapid correction of this extracellular deficit required an infusion of a 3:1 ratio of crystalloid fluid to blood loss. Figure 3 shows the influence of fluids on the extracellular fluid compartments. Using this resuscitation strategy, the rate of mortality and acute renal failure decreased but a new entity of shock lung, now better known as acute respiratory distress syndrome (ARDS), was discovered. In reviewing studies comparing crystalloids and colloids with respect to pleural effusions and pulmonary dysfunction, two trials reported no differences. Two other series showed more pulmonary complications among patients resuscitated with colloid. When mortality was used as an endpoint, the use of crystalloids in trauma patients was associated with increased survival.

Figure 3 The influence of colloid and crystalloid fluids on the volume of the extracellular fluid compartments.

(Data from Imm A, Carlson RW: Fluid resuscitation in circulatory shock. Crit Care Clin 9:313, 1993.)

There are advantages and disadvantages to the use of both crystalloids and colloids. Crystalloids replace interstitial as well as intravascular fluid loss, do not cause allergic reactions, and are inexpensive. Possible disadvantages include limited intravascular expansion and tissue edema, which can impair gas exchange in the lungs and diminish bowel perfusion. Colloids have a longer intravascular half-life, which may improve organ perfusion and may cause less tissue edema. This may be a short-term effect as the interstitial oncotic pressure increases from diffusion of the colloid over time. Also, the choice of colloid may affect blood loss. Other possible disadvantages include increased incidence of allergic reactions, impaired blood cross matching due to dextrans, altered platelet function (dextrans, hetastarches), hyperchloremic acidosis due to high chloride content (hetastarch), and greater expense. A meta-analysis of clinical studies comparing crystalloids and colloids for fluid resuscitation failed to show any evidence of improved outcome with the use of colloids. In summary, there is no clear basis to give colloid products over crystalloid solutions for fluid resuscitation.

Crystalloid Solutions

The choice of lactated Ringer’s (LR) and normal saline (NS) appears to be less controversial. Both are hypo-oncotic, balanced salt solutions (Table 2). It is well known that large volumes of NS resuscitation can lead to hyperchloremic metabolic acidosis. While large volumes of LR infused rapidly can increase lactate levels, this is not associated with acidosis. Koustova et al. have shown that LR influences neutrophil function by increasing production of reactive oxygen species and affects leukocyte gene expression, but Watters et al. showed that NS and LR have equivalent effects on alveolar neutrophils in a swine model of hemorrhagic shock. Healy et al. showed, in a rat model of massive hemorrhage and resuscitation, animals were significantly more acidotic (pH 7.14 ± 0.06 vs. 7.39 ± 0.04) and had a significantly worse survival (50% vs. 100%) when resuscitated with NS compared to LR. NS has been theoretically preferred when giving blood because of the concern that the calcium in LR could exceed the chelating capabilities of the citrate, resulting in clots that may enter the microcirculation. However, Lorenzo et al. showed that when fluid is given rapidly, LR does not increase clots, and at present there is no evidence that this is a clinically significant issue. In theory, NS is favored over LR for the treatment of severe head injuries due to its higher sodium content, which would reduce intracerebral swelling. Despite this concern, there is no literature supporting the use of NS over LR in this setting.

Ringer’s ethyl pyruvate solution (REPS) is a calcium- and potassium-containing balanced salt solution. Ethyl pyruvate is derived from pyruvate, which is unstable in aqueous solutions, and acts as a potent reactive species of oxygen scavenger (e.g., H₂O₂, O₂, OH). It has been evaluated in several preclinical studies using animal models of mesenteric ischemia/reperfusion injury, hemorrhagic shock, and acute endotoxemia. In comparison to LR, REPS has been shown to improve survival and decrease expression of proinflammatory mediators. However, a recent study using a swine hemorrhagic shock model showed no short-term hemodynamic or tissue energetic advantage to using REPS as a resuscitation fluid when compared to LR. Further investigations are needed to evaluate the potential benefits of using REPS in hemorrhagic shock.

Hypertonic Saline

Hypertonic saline (HS) has osmotic properties that result in influx of fluid into the intravascular space. Because only small volumes are needed to achieve its desired effects, there has been significant interest in its use in the military and civilian settings. It is capable of rapidly expanding the intravascular volume and enhancing microcirculation by selective arteriolar vasodilatation. This occurs without causing swelling of red blood cells or the endothelium. Unfortunately, this improved microcirculation could also lead to increased bleeding. Adding dextran and hetastarch to HS can prolong its intravascular effects due to increased oncotic pressure. The volume of hypertonic saline solution that can be given is limited by the potential development of hypernatremia and intravascular volume overload. In a swine model of uncontrolled hemorrhagic shock, a single 250-ml bolus of 3% hypertonic saline plus 6% dextran produced an adequate and sustained rise in mean arterial pressure (MAP) and StO₂. More importantly, the increase in MAP was not associated with increased secondary bleeding. A meta-analysis by Wade et al. of both HS and HS with dextran (HSD) in the treatment of hypotension from traumatic injury suggests that HS is no different than isotonic crystalloids, but HSD did show a trend toward decreased mortality. In head trauma patients, its ability to draw fluid from the extravascular space can limit cerebral edema, lower intracranial pressure, and improve cerebral perfusion. Subgroup analysis in the previous study by Wade showed the greatest benefit with HSD to be in shock patients with concomitant severe closed head injuries. A recent prospective randomized study from Australia testing HS for field resuscitation in hypotensive and TBI patients failed to show any benefit in neurological outcome at 6 months when compared to LR. HS and HSD have also demonstrated a diminished inflammatory response, specifically neutrophil cytotoxicity, in animal models of hemorrhagic shock, ischemia/reperfusion, and sepsis.

Artificial Oxygen-Carrying Blood Substitutes

Although both crystalloids and colloids can replace intravascular volume, neither product restores the oxygen-carrying capacity of the lost red blood cells. Artificial hemoglobin products have the potential to improve oxygen-carrying capacity without the storage, availability, immune suppression, transfusion reaction, compatibility, or disease transmission problems associated with standard transfusions. Unfortunately, these products also fail to restore coagulation components, and hemostasis can be hindered with the loss of cellular elements that lower the viscosity of circulating blood. The use of a large volume of artificial oxygen-carrying solutions in severe hemorrhage has not been adequately studied.

Hemoglobin-based oxygen carriers have been derived from human blood, bovine blood, and recombinant DNA technology dating back to 1933. Carrier solutions need to be stroma-free polymerized or cross-linked hemoglobin tetramers with oxygen-carrying capabilities that remain within the intravascular space for a prolonged period of time. Early hemoglobin substitutes had a greater oxygen affinity because of the loss of 2,3 diphosphoglycerate (2,3 DPG). However, with pyridoxylation of the hemoglobin tetramer this problem was addressed. Hemoglobin-based oxygen carriers can also cause vasoconstriction. This occurs because tetrameric hemoglobin binds the nitric oxide in the vascular wall and results in unopposed vasoconstriction. Because of this phenomenon, blood pressures can be higher than expected for the level of intravascular volume replacement in hemorrhagic shock patients. Initial studies of these products in humans have been disappointing. A phase III trial of diasprin cross-linked hemoglobin was prematurely stopped due to an unexpectedly high mortality in the treatment group (46 vs. 17%). Polyheme®, glutaraldehyde-polymerized pyridoxylated human hemoglobin, has proven to be safe and effective in phase I and II trials and is currently being studied in a multicenter phase III prehospital trial.

Perfluorocarbons are chemically and biologically inert liquids that dissolve large amounts of gas. They require dispersion in plasma-like aqueous fluids such as albumin or in physiologic electrolyte solutions to be an adequate oxygen-carrying substitute. They have a lower oxygen-delivering capacity than normal blood and require an FiO₂ of greater than 70% to carry physiologically useful concentration of oxygen. The long-term biological effects of absorption, distribution, metabolism, and excretion, and the effects on the reticular endothelial system (RES) require further evaluation. Concern about toxicity to the RES and high FiO₂ requirements may limit their use in severe hemorrhage.

Blood Transfusions

The use of blood transfusions in resuscitation has been an ongoing debate since its initial use in World War I and in World War II, when it became the standard of care. There was an increase in early survival, but many casualties later died of acute renal failure. It is generally accepted that a patient in shock who fails to respond adequately to 2 liters of crystalloid is in need of blood transfusions. It is clear in unstable patients with active bleeding and hemorrhagic shock that blood should be given immediately. When larger volumes of crystalloid and red blood cells are given, fresh frozen plasma, platelets, or cryoprecipitate may also be needed to reverse the associated dilutional coagulopathy. One must also consider the inherent risks of blood transfusion, such as transfusion reactions, transfusion-related acute lung injury, infection, and immunosuppression. It has been observed that blood transfusions contain proinflammatory mediators that both prime and activate neutrophils. This has been proposed as a key mechanism in the development of multiple organ failure. Stored red blood cells also undergo substantial shape changes and impaired deformability by the second week of storage. This decreased deformity can lead to microvascular obstruction, and has been found to be associated with the development of splanchnic ischemia. Malone et al. reported that trauma patients who underwent blood transfusions within the first 24 hours of admission, independent of shock severity, were almost three times more likely to die than those who did not receive transfusions.

Early clinical studies concluded that hemoglobin (HgB) levels of 10 g/dl were optimal for shock resuscitation; however, recent consensus panels have suggested that a lower concentration is adequate. A prospective randomized trial showed that ICU patients who received blood transfusion for HgB less than 7.0 g/dl and were maintained at HgB 7.0–9.0 g/dl did as well and possibly better than patients who were liberally transfused for a HgB level less than 10 g/dl and maintained at 10–12 g/dl. All these patients were without ongoing bleeding, acute myocardial infarction, or unstable angina. Subgroup analysis looking at the safety of a restricted red blood cell transfusion strategy in trauma patients showed there was no statistically significant difference in mortality, multiple organ dysfunction, or length of stay, when compared to those managed with the liberal transfusion strategy. Randomized, controlled investigations need to be conducted to provide evidence to help dictate the use of blood in critically ill trauma patients.

Hypothermia

The definition of hypothermia is a core body temperature of less than 35° C. The internal core temperature is a net result of heat production and heat loss. Heat loss or production can be a result of evaporation, radiation, conduction, or convection. Heat production largely occurs as a result of cellular metabolism while heat loss occurs through the skin and respiratory system. Acutely injured trauma patients have numerous sources of potential heat loss. Exposure in the field can cause the patient to be hypothermic on admission. Further exposure per ATLS protocols causes further losses. Operations involving exposure of body cavities cause significant losses through evaporative and conductive means. Trauma procedures involving large laparotomies or thoracotomies are especially threatening. In terms of resuscitation, intravenous fluids present the highest potential for heat loss. This can be quantified by the following equation:

Given a specific heat of water of 4.19 kJ/kg/degree, 1 liter of 25° C crystalloid infused in a normothermic patient would result in a heat loss of 50.3 kJ. This heat loss exceeds the heat that can be returned to the patient by conventional methods in 1 hour. Given these findings, massive heat loss can occur in the setting of large-scale resuscitation.

The incidence of hypothermia in trauma patients during resuscitation has been described to be as high as 66%. Although Gregory et al. found that only 12% of patients arriving in the ED were hypothermic, 46% were hypothermic on arrival to the operating room (OR), and 57% were hypothermic when leaving the OR. This suggests that the majority of heat loss occurs in the resuscitation bay.

Hypothermia is commonly classified by severity. Mild hypothermia is a core temperature of 32°–35° C, and is characterized by tachypnea, tachycardia, hyperventilation, shivering, and impaired judgment. Increased urine output secondary to a “cold diuresis” has also been described. This is thought to be secondary to peripheral vasoconstriction, causing a temporary increase in the central intravascular volume. In moderate hypothermia, defined as 28°–32° C, heart rate and cardiac output are reduced. Decreased mental status and hyporeflexia are observed as well as decreased renal blood flow. Cardiac rhythm manifestations include atrial fibrillation and bradycardia. Severe hypothermia, that is, temperatures less than 28° C, can lead to pulmonary edema, oliguria, coma, hypotension, ventricular fibrillation, and asystole.

Hypothermia in the trauma patient is a poor prognostic indicator and it has been shown to be independently predictive of mortality. The mortality of victims of accidental exposure with moderate hypothermia has been documented to be 21%. However, trauma victims studied with similar core temperatures (<32° C) have 100% mortality. This difference has several explanations. Trauma patients do have similar losses secondary to exposure; however, they also have further losses secondary to hemorrhage and exposed body cavities. Beyond this, the decreased oxygen consumption secondary to blood loss and shock cause decreased heat production. Given these mechanisms, the mortality documented may be more a manifestation of the severity of injury rather than isolated hypothermia.

Because of the strong association of hypothermia to mortality and the other elements of the “lethal triad,” rewarming and prevention of ongoing heat loss should be a priority of resuscitation (Table 3). The first step is to obtain an accurate core temperature. Esophageal or bladder temperatures have been shown to be more reliable than rectal or axillary measurements. Given the high rate of heat loss with infusion of room temperature fluids, all resuscitation solutions should be warmed. Operating room temperatures should be elevated to minimize losses due to conduction and radiation. Once the secondary evaluation is complete, the patient should be covered. A Bair Hugger or similar device can be used to actively warm the patient. For profound hypothermia, active internal rewarming can be used. This is done either with lavage of a body cavity (peritoneal or thoracic) or by intravascular rewarming. Options for intravascular rewarming include veno-venous rewarming, arteriovenous rewarming, or cardiopulmonary bypass. Continuous arteriovenous rewarming (Figure 5) has been shown to decrease early mortality in critically injured hypothermic trauma patients.

Table 3 Approximate Rate of Heat Transfer with Available Rewarming Methods

Source: Adapted from Gentilello LM: Practical approaches to hypothermia. In Maull KI, Cleveland HC, Feliciano DV, Rice CL, Trunkey DD, Wolferth CC, editors: Advances in Trauma and Critical Care, vol. 9. St. Louis, MO, Mosby, 1994, pp. 39–79, with permission.

Figure 5 Continuous arteriovenous rewarming.

(From Gentilello LM, Jurkovich GJ, Stark MS, Hassantash SA, O’Keefe GE: Continuous arteriovenous rewarming: rapid reversal of hypothermia in critically ill patients. J Trauma 32(3):316–327, 1992, with permission.)

Coagulopathy

Hemorrhage is a major cause of early trauma deaths. Coagulopathies in trauma patients are common during major resuscitations. The mechanisms are thought to be related to hypothermia, metabolic disturbances, dilution, and disseminated intravascular coagulation. Most of these mechanisms can be traced in some way to resuscitation.

Dilution is a major cause of coagulopathy in resuscitated trauma patients. Intravascular fluid containing coagulation factors is lost and replaced with solutions lacking these products. The actual amount of coagulopathy is not easily predictable as the plasma shifts are quite complex. Coagulation factors are continually produced and sequestered in the post trauma setting. Dilutional coagulopathy is not thought to play a significant role until approximately one blood volume of replacement fluids is infused into a patient. Therefore, giving prophylactic products without data or evidence of bleeding is generally not recommended. However, there is increasing evidence that the early administration of coagulation factors is indicated.

Hypothermia has a profound effect on coagulation activity. Reed and Rohrer in different experiments showed that hypothermia resulted in prolonged partial thromboplastin time (PTT) and prothrombin time (PT) independent of the actual level of enzymes. Gubler et al. showed that the effect of a dilutional coagulopathy is additive to this hypothermic effect (Figure 6). In addition to these enzyme effects, fibrinolysis is thought to be enhanced by hypothermia. Animal studies have suggested that platelets are sequestered by the spleen in the setting of hypothermia. Platelet adhesion is also decreased in hypothermic patients. A recent in vitro study suggests that platelet effects cause the majority of hypothermic coagulopathy at temperatures above 33° C. In this trial, enzymatic dysfunction did not have a significant effect until the temperature was below 33° C. Studies have shown that acidotic and hypothermic patients with adequate blood, plasma, and platelet replacement still develop clinically significant bleeding.

Figure 6 The effect of hypothermia and dilution on prothrombin time. Filled squares, diluted specimens; open squares, nondiluted specimens.

(From Gubler KD, Gentilello LM, Hassantash SA, Maier RV: The impact of hypothermia on dilutional coagulopathy. J Trauma 36(6):847–851, 1994.)

Logically, one could conclude that this coagulopathy would result in more hemorrhage and poor outcomes. This has been substantiated in clinical reviews. A recent review of trauma patients revealed that 24.4% were coagulopathic on admission. Nonsurvivors had a coagulopathy rate of 46 versus 10.9% for survivors.

As with any disease, treatment of coagulopathy begins with recognizing the problem. Any trauma patient with evidence of significant tissue injury or ongoing bleeding should be screened with an (INR), PTT, platelet count, and fibrinogen level. In the setting of active traumatic bleeding, the platelet count should be kept above 50,000/mcl, the INR is less than 2, the fibrinogen greater than 100 mg/dl, and the PTT less than 1.5 times normal. Again, close monitoring of these parameters is recommended as empiric therapy often yields unpredictable results. In the absence of laboratory results, empiric therapy with platelets that contains plasma may be started.

Coagulopathy is also specifically linked to head trauma. The release of thromboplastin from injured brain tissue can cause severe consumptive coagulopathy leading to disseminated intravascular coagulopathy. More extensive coagulation monitoring is indicated in the setting of head trauma. There are reports and ongoing studies on the use of recombinant-activated factor VIIa in the setting of massive resuscitation and transfusion. This approach has not been widely accepted, but is likely effective in immediately reversing coagulopathy in the setting of continued bleeding.

Acidosis

Metabolic acidosis is commonly seen in trauma patients with hemorrhagic shock. This is postulated to occur secondary to tissue hypoperfusion in the setting of decreased cardiac output and oxygen-carrying capacity. This cascade of events eventually leads to anaerobic metabolism with the production of lactate. Massive resuscitation with crystalloid solutions has been associated with the development of a worsening metabolic acidosis. Following the Stewart model of acid base equilibrium, the administration of solutions with supraphysiologic levels of chloride relative to sodium results in a decreased strong ion difference (SID). (Na + K + Ca + Mg – Cl – lactate). This decreased SID causes further dissociation of H+ from H₂O to maintain charge neutrality, and therefore a decreased pH.

Although the presence of a hyperchloremic acidosis has not been associated with increased mortality, there are many potential hazards of profound acidosis. Acidosis has a depressant effect on the myocardium and increases ventricular dysrhythmias. The sympathetic-adrenal axis is stimulated in the setting of acidosis. However, the myocardium has decreased responsiveness to circulating catecholamines. The prolonged acidotic state also increases respiratory drive, increases intracranial pressures in head-injured patients, and worsens coagulopathy. A more common danger exists in misinterpreting a hyperchloremic acidosis for continued hypoperfusion and shock leading to unnecessary therapies.

In order to avoid hyperchloremic acidosis, fluids with supraphysiologic concentrations of chloride should be avoided. LR contains a more physiologic concentration of chloride (109 meq/l) compared to normal saline (154 meq/l). Several clinical series and animal studies document profound acidosis in the setting of large volume normal saline administration. Although less probable, this acidosis can be seen with LR, as its SID is less than physiologic. Careful monitoring of electrolytes with measurement of the anion gap can aid in guiding therapy. A normal or narrowed anion gap should be seen in the setting of an isolated hyperchloremic metabolic acidosis, as opposed to an elevated gap in lactic acidosis. Transitioning to fluids with less chloride (LR or Plasmalyte) or no chloride (Na acetate) can resolve the hyperchloremic acidosis.

The mechanism of coagulopathy in acidotic patients is a matter of active investigation. In vitro experiments have shown a decrease in the activity of factor VIIa-tissue factor and factor Xa-Va complexes. In vivo animal experiments have shown that acidosis independently decreases fibrinogen and platelet counts, increases PTT, and increases clinical bleeding time. In a study of patients undergoing massive transfusion, pH less than 7.10 independently predicted coagulopathy.