INCIDENCE

A recent analysis of the National Trauma Data Bank (NTDB) provides the most comprehensive perspective on the incidence of hypothermia among trauma patients.⁴ Of 1,126,238 injured patients, the admission body temperature was recorded in 701,491 (62.3%). A total of 11,026 patients (1.57% of all patients with a recorded temperature) were hypothermic, defined as a core temperature lower than 35° C.

MECHANISM OF INJURY

Trauma patients are disrobed in the emergency department, where most heat loss occurs, and are frequently administered cold intravenous (IV) fluids. Hypothermia is more common and more profound in the more seriously injured patients. Therefore, there is uncertainty over whether the increase in mortality is primarily attributable to the hypothermia itself, or to the underlying injuries. Some have proposed that hypothermia is actually protective in trauma patients, and that mortality rates would not be higher in cold patients if comorbid factors were equal.⁵ However, recent studies have documented an adverse effect of hypothermia on outcome, and a significantly improved likelihood of surviving initial resuscitation when hypothermia is aggressively treated.^1,3,4,6,7

EFFECTS ON COAGULATION

Perhaps the most serious effect of hypothermia in the trauma victim is its effect on coagulation. Uncontrollable hemorrhage, often compounded by coagulopathy, is the most frequent cause of early death in these patients. Dilutional thrombocytopenia is usually cited as the primary cause of coagulopathic bleeding when trauma victims undergo massive transfusion.⁸ However, a prospective, randomized, double-blind controlled clinical trial indicated that dilutional thrombocytopenia is relatively infrequent and that prophylactic administration of platelets was not beneficial.⁹ Consumptive coagulopathy appeared to be the more common problem associated with massive transfusion.

The extent to which hypothermia causes coagulation problems is often underestimated because of the multiplicity of potential etiologies for coagulation impairment that are usually present. These patients often have conditions such as acidosis, tissue trauma, shock, and dilution of the circulating blood volume with components containing reduced concentrations of clotting factors.

Systemic hypothermia affects coagulation in a variety of ways. Normal coagulation requires adequacy of vasoconstriction, platelet counts and function, and clotting factor levels and activity. Theoretically, hypothermia could affect blood coagulation in each of these domains.

EFFECTS ON VASCULAR PHASE OF COAGULATION

Systemic hypothermia has long been known to provoke cutaneous vasoconstriction. Because of this observation, the topical application of cold has been used as a method to stop external bleeding. The assumption was that because systemic hypothermia induced vasoconstriction, that local vasoconstriction might reduce bleeding. Yet, topical exposure to cold appears to be quite different from systemic hypothermia in that local cooling elicits skin and skeletal muscle vascular dilation at 33° C.

This phenomenon makes sense from a physiologic standpoint. With total body cold exposure, the risk of developing systemic hypothermia produces generalized vasoconstriction as a means of reducing heat and protecting core temperature. Topical administration of cold produces a regional hypothermia that provokes the flow of blood into the hypothermic region through vasodilation in order to reduce the risk of local tissue damage such as frostbite.

Thus, with systemic (but not local) hypothermia, peripheral blood vessels go into vasospasm, a process that limits external heat loss. While this could help coagulation seal bleeding vessels, it can only do so if the remaining components of the coagulation system are working effectively. However, this effect is often counteracted by the adverse impacts hypothermia has on the other components of the coagulation process.

EFFECTS ON PLATELET COUNT AND FUNCTION

During development of hypothermic cardioplegia for cardiac surgery, there was a surge of research interest in the effects of hypothermia on coagulation in the late 1950s. Experimental studies in dogs at that time demonstrated a reversible thrombocytopenia associated with systemic hypothermia.^10,11 However, the thrombocytopenia observed actually occurred at very deep levels of hypothermia, well below that typically seen in a trauma setting.^11–15 Yoshihara et al.¹⁶ reported that platelet counts dropped by only 20%-30% at an esophageal core temperature of 30° C.

In contrast, levels of hypothermia commonly encountered in clinical practice have been shown to have a significant effect on platelet function. Platelets experience a reversible inhibition of their function under conditions of local or systemic hypothermia, mediated at least in part through the temperature dependence of thromboxane B₂ by platelets.¹⁷ Thromboxane B₂ is a potent vasoconstrictor and platelet aggregating agent. Valeri et al.¹⁷ demonstrated this when they induced systemic hypothermia to 32° C in baboons, but kept one forearm warm using heating lamps and a warming blanket. Simultaneous bleeding time measurements in the warm and cold arm were 2.4 and 5.8 minutes, respectively. The authors concluded that warming to restore wound temperature to normal should be tried before resorting to transfusion therapy with platelets and clotting factors when treating hypothermic patients with nonsurgical bleeding.

EFFECT ON CLOTTING FACTOR LEVELS AND FUNCTION

Several studies performed on humans undergoing hypothermic open-heart surgery failed to demonstrate significant alterations in clotting test times except at extreme degrees of hypothermia (i.e., <20° C).^16,18–21 Yet, clinical experience suggests otherwise. Many patients with less severe degrees of hypothermia will have a serious coagulopathy that appears related to the presence of the hypothermia. This apparent inconsistency has been resolved by the realization that coagulation during mild hypothermia is disturbed more from enzymatic dysfunction than it is from altered clotting factor levels in blood. This explains the inability for the experimental data from the 1950s and 1960s to correlate with the clinical experience, as the clotting tests performed by the early experimenters were routinely performed at 37° C instead of at hypothermic temperatures.

In recent years, a number of studies have been performed wherein the clotting tests were performed at hypothermic temperatures. Bunker and Goldstein,¹⁸ in a study previously mentioned of controlled hypothermia in 10 patients, measured clotting tests at the hypothermic temperature of the patients as well as at 37° C. While they found no significant changes in clotting times when performed at 37° C, they state that “prolongation of the clotting times for all coagulation tests except whole blood clotting times was consistently observed when performed at the hypothermic temperatures.”

A detailed study of the kinetic effects of hypothermia on clotting factor function was undertaken by Reed et al.²² These studies were done by performing standardized clotting tests in a modified coagulation timer (fibrometer). Because the heat block of fibrometers used clinically are set by the manufacturer at 37° C, an external digital temperature controller was connected to the heat block power source to enable measurement of clotting times at the range of hypothermic temperatures typically encountered in trauma patients. Measurement of the prothrombin time, partial thromboplastin time, and thrombin time performed on assayed reference human plasma containing normal levels of all the clotting factors at temperatures ranging from 25° C to 37° C showed a significant slowing of clotting factor function that was proportional to the degree of hypothermia (Figure 1).

Figure 1 Comparison of effects of clotting test temperature with progressive degrees of hypothermia. Clotting times were performed on standard concentrations of assayed reference plasma using a fibrometer modified to enable control of the temperature at which the clotting test was conducted. PT, Prothrombin time; PTT, partial thromboplastin time; TT, thrombin time. *p, <0.001 vs. thrombin time prolongation. **p, <0.0001 vs. thrombin time prolongation.

(Adapted from Reed R, Bracey A, Hudson J, Miller T, Fischer R: Hypothermia and blood coagulation: dissociation between enzyme activity and clotting factor levels. Circ Shock 32:141–152, 1990.)

These results were later confirmed by Gubler et al.,²³ in a study using a similar modified fibrometer that demonstrated an additive effect of hypothermia on dilutional coagulopathy (Figure 2).

Figure 2 Prolongation of partial thromboplastin time (PTT) that results from cooling of the blood in samples with normal clotting factor levels, and in samples of blood with diluted clotting factor levels.

(Data from Gubler K, Gentilello L, Hassantash S, Maier R: The impact of hypothermia on dilutional coagulopathy. J Trauma 36:847–851, 1994.)

A subsequent study demonstrated that hypothermia could produce a coagulopathy functionally equivalent to a severe clotting factor deficiency, even at intermediate levels of hypothermia and even though there was no actual deficiency of clotting factors²⁴ (Figure 3).

Figure 3 Relative clotting factor activities at various temperatures expressed as percentage of normal clotting factor activity.

(Data from Johnston T, Chen Y, Reed R: Functional equivalence of hypothermia to specific clotting factor deficiencies. J Trauma 37:413–417, 1994.)

In summary, hypothermia does little to affect platelet and clotting factor levels, but it does a great deal to affect the function of these coagulation components. A recent analysis indicates that at mild temperature reductions between 33° C and 37° C, platelets are more profoundly affected than are clotting factors, although clotting factor dysfunction becomes increasingly severe as temperature cools further.²⁵ Because of the potent effect that severe hypothermia has on platelet and clotting factor function, it is essential that body temperature be normalized before exogenous platelets or clotting factors are administered. Even though clotting studies may demonstrate severe clotting factor deficiencies, there is no value in transfusing coagulation components to severely hypothermic patients. This is because normal levels of clotting factors fail to clot effectively in the setting of severe hypothermia. Thus, administration of platelets or clotting factors to moderately or severely hypothermic patients is essentially futile, as the coagulation components will not function in a hypothermic environment (i.e., below 34° C).

EFFECTS ON OTHER ORGANS

The organ systems that are most commonly affected by hypothermia include the circulatory, immunologic, neurologic, and coagulation systems. Cardiac function can be affected by hypothermia in the form of bradyarrhythmias and, at a core temperature below 28° C–30° C, ventricular fibrillation. The body’s attempt at restoring normothermia results in an elevation of oxygen that takes place primarily in muscles through shivering. Because of the excessive oxygen consumption required to maintain or restore normothermia in an environment with significant cold stress, organ dysfunction can occur due to a relative undersupply of oxygen, with a resultant increased risk of cardiac complications in elderly patients.²⁶

The potential immunologic consequences of hypothermia have been extensively studied. Because of the enzymatic nature of most immunologic functions, it makes sense that hypothermia would inhibit many of these processes. Moreover, our immunologic system is often pitted against bacteria that are not homeotherms as humans are, and may therefore not suffer as severe a functional deterioration in the presence of a hypothermic environment. Some relatively well-done clinical studies provide evidence that mild hypothermia is associated with increased risk for surgical site infection.^27,28 Laboratory studies of the neuroprotective effects of hypothermia appeared promising, but clinical trials have been disappointing, and the immunologic effects of hypothermia were associated with an increase in pneumonia and septic complications.^29–31

MANAGEMENT

The relatively high specific heat of the body makes hypothermia very difficult to treat. The rapidity and aggressiveness with which treatment is provided should be based on how severely the hypothermia is affecting the patient. There are a number of clinical studies that describe the efficacy of currently available rewarming techniques. However, many were conducted on healthy, nonvasoconstricted volunteers, and most did not take into account the patient’s initial body temperature and mass, the rate of endogenous heat production, and the presence or absence of anesthetics, vasodilating agents, shock, or shivering, all of which are important determinants of the rewarming rate.

Active Core Rewarming

Airway rewarming using humidified air at 41° C is one of the most frequently used core rewarming techniques. Fully saturated 41° C air can hold 0.05 ml H₂O per liter. At 30° C, air can only hold only 0.03 ml H₂O per liter. If a 30° C patient inspires a liter of saturated 41° C air, then 0.02 ml H₂O condenses within the airway when the air cools down to the patient’s temperature. With a ventilation of 10 l/min, 12 ml of H₂O will condense each hour. When water condenses heat is liberated at a rate of 0.58 kcal/ml H₂O (latent heat of vaporization). Thus, the amount of heat contributed by airway rewarming under these conditions will be only 7 kcal/hr (0.58 kcal/ml H₂O × 12 ml H₂O/hr). An additional 1–2 kcal will be transferred by the warming effect of the inspired air, independent of condensation. Since 58 kcal is required to increase core temperature by 1° C in a 70-kg patient, as with external techniques, airway rewarming has limited effectiveness.

Pleural or peritoneal lavage should be considered for use in unstable patients with a deleterious response to hypothermia. The amount of heat transferred depends on the difference between the inlet and outlet water temperature and the water flow rate. Since the specific heat of water is 1 kcal/kg/° C, if 1 liter of 42° C water that is infused into a body cavity exits at 35° C, 7 kcal of heat will have been left in the body. However, prolonging operative time in order to irrigate the open peritoneal cavity with warm fluids is counter-productive, as most of the heat that is lost from the water will be transferred to the 21° C operating room environment rather than to the patient.

The high specific heat of water makes it important to warm cold IV fluid prior to administration. A patient will have to generate 16 kcal to warm 1 liter of crystalloid infused into the body at room temperature (21° C). When patients are under anesthesia, their metabolic rate is relatively fixed. If they cannot increase their metabolic rate sufficiently to generate this additional heat, the loss of 16 kcal will decrease body temperature by 0.28° C, which is enough to cause vigorous shivering.

Warm IV fluids also provide a simple means of transferring significant amounts of heat to cold patients requiring massive fluid resuscitation. Warm IV fluids equilibrate with body temperature, liberating heat in the process. A 1-liter infusion of 40° C crystalloid infused into a 32° C patient is, in effect, equivalent to a transfusion of 8 kcal. Since hypothermic trauma patients frequently require massive fluid resuscitation, using warm IV fluids can provide a significant quantity of heat.

Rewarming with cardiopulmonary bypass is, in effect, a means of rewarming via the provision of a continuous infusion of warmed IV fluids. The limitations imposed by the patient’s fluid requirements are circumvented by recirculating the patient’s own blood. Continuous arteriovenous rewarming (CAVR) is a newly described means of performing extracorporeal circulatory rewarming that does not require a mechanical pump.^6,35,36 CAVR uses percutaneously placed 8.5-Fr femoral arterial and venous lines and the patient’s own blood pressure to create an extracorporeal AV fistula through the heating mechanism of a counter current fluid warmer. The tubing circuit is heparin bonded, and no additional heparinization is needed (Figure 4).

Figure 4 Diagram of continuous arteriovenous rewarming (CAVR) circuit.

Unlike cardiopulmonary bypass, this technique requires an intact circulation, and its effectiveness is limited when arterial pressure falls below 80 mm Hg. However, hypotensive patients generally require additional fluids, which can be “piggybacked” into the heat exchanger to supplement the fistula flow rate. The typical flow rate in normotensive individuals is between 250 and 350 ml/min. If the patient’s temperature is 32° C and blood is reinfused at a temperature of 39° C approximately 6 kcal of heat will be transferred every 3–4 minutes.

Rewarming efficacy can be analyzed using standard thermodynamic and heat transfer equations to provide a more accurate assessment various rewarming techniques. A mathematical model has been developed which takes into account body mass and surface area, the specific heat of tissues, the various conductivities of body tissues as a function of temperature, endogenous heat production, and the thermophysical properties of air, water, radiation, and other heat transfer media.³⁷ A computer simulation provides the expected rewarming rates based on the properties of the technique used (Figure 5).

Figure 5 Computerized simulation of rewarming rates using various clinical techniques. CAVR, Continuous arteriovenous rewarming; circ, circulating.

(Adapted from Gentilello LM, Moujaes S: Treatment of hypothermia in trauma victims: thermodynamic considerations. J Intensive Care Med 10(1):5–14, 1995.)

MORTALITY

Hypothermia has two well-known clinical effects: to preserve life and to kill. Which one of these properties is most active in the trauma patient has been debated for centuries. Hippocrates recommended packing injured soldiers in snow and ice. Baron de Larrey, a battlefield surgeon during Napoleon’s campaigns, noted that injured soldiers who sat closest to the fire were usually the first to die. Animal studies repeatedly demonstrate that hypothermic animals are better able to survive shock than normothermic counterparts.^38,39

Despite these observations, current recommendations for treatment of injured patients call for strict efforts to prevent hypothermia, and for aggressive treatment to reverse it once it has occurred.⁴⁰ These recommendations are based on findings of repeated clinical studies demonstrating that mortality is significantly higher in trauma patients who develop hypothermia.^1,2,4,41 One study controlled for magnitude of injury using the Injury Severity Score (ISS), the presence or absence of shock, and fluid and blood product requirements. Patients who became hypothermic had significantly higher mortality rates than similarly injured patients who remained warm. Mortality was 100% if core body temperature dropped to 32° C, even in mildly injured patients.¹ A large study analyzing the NTDB (National Trauma Data Bank) found that hypothermia was an independent predictor of mortality by using stepwise logistic regression (odds ratio 1.54, 95% CI 1.40-1.71) (Figure 6).⁴

Figure 6 Mortality at each admission body temperature determined from the National Trauma Data Bank (NTDB).

One study compared the mortality of hypothermic patients (<35° C) admitted over a 10-month period who were treated with a combination of airway rewarming, fluid circulating or convective heating blankets, an aluminized head covering, and warm IV fluids with a consecutive sample of patients who were rapidly rewarmed with CAVR.⁶ Time to rewarming (T > 35° C) was 3.23 hours with standard rewarming techniques and 39 minutes with CAVR. Rapid rewarming with CAVR resulted in a 57% decrease in blood product requirements, a 67% decrease in crystalloid requirements, and a reduction in mortality in trauma patients. In a more recent randomized, prospective clinical trial comparing slow versus rapid rewarming in critically injured patients, significantly more patients in the rapid rewarming (CAVR) group were able to be successfully resuscitated.⁷ Two additional prospective, but nonrandomized studies have demonstrated improvements in outcome in trauma patients when protocols designed to minimize heat loss were utilized.^42,43

CONCLUSIONS

The relatively high specific heat of the body makes hypothermia very difficult to treat. Early attention to the mechanisms of heat loss outlined previously remains the best form of therapy. Based on current data hypothermia has an adverse effect on outcome from trauma, and every attempt should be made to aggressively treat it once it has occurred.