HISTORY

Bailout/damage control surgery following trauma has developed as a major advance in surgical practice in the last 20 years. The principles of damage control surgery defied the traditional surgical teaching of definitive operative intervention and were slow to be adopted. Currently, techniques developed by trauma surgeons known as damage control surgery have been successfully used to manage traumatic thoracic, abdominal, extremity, and peripheral vascular injuries. In addition, damage control surgery has been extrapolated for use in general, vascular, cardiac, urologic, and orthopedic surgery.

In 1983, Stone was first to describe the “bailout” approach of staged surgical procedures for severely injured patients. This approach emerged after his observation that early death following trauma was associated with severe metabolic and physiologic derangements following severe exsanguinating injuries. Following massive transfusion exceeding two blood volumes in trauma and emergency surgery, severe physiologic derangement ensued and mortality was found to be greater than 60%. Profound shock along with major blood loss initiates the cycle of hypothermia, acidosis, and coagulopathy. It was at this time that hypothermia, acidosis, and coagulopathy were described as the “trauma triangle of death” or the “bloody vicious cycle.” A fourth component, dysrhythmia, which usually heralded the patient’s death, was later added by Asensio. Coagulopathy, acidosis, and hypothermia make the prolonged and definitive operative management of trauma patients dangerous. This approach, now called “damage control,” describes it as multiphasic, where reoperation occurs after correcting physiologic abnormalities.

METABOLIC FAILURE

Hypothermia is a consequence of severe exsanguinating injury and subsequent resuscitative efforts. Severe hemorrhage leads to tissue hypoperfusion and diminished oxygen delivery, which leads to reduced heat generation. Clinically significant hypothermia is important if the body temperature drops to less than 36° C for more than 4 hours. Hypothermia can lead to cardiac arrhythmias, decreased cardiac output, increased systemic vascular resistance, and left shift of the oxygen–hemoglobin dissociation curve. Hypothermia exerts a negative inotropic effect on the myocardium with depression of left ventricular contractility. The initial ECG change seen with hypothermia is sinus tachycardia, but as the core temperature decreases, progressive bradycardia ensues. The cardiac response to catecholamines may also be blunted in hypothermic hearts, and cold cardiac tissue poorly tolerates hypervolemia and hypovolemia. Hypothermia can also induce coagulopathy by inhibition of the coagulation cascade. Low temperature also impairs the host’s immunologic function. Hypothermia is aggravated by heat loss resulting from either environmental factors or surgical interventions. The multidisciplinary team caring for trauma patients must make every effort to prevent heat loss and help to correct hypothermia.

The causes of coagulopathy in patients with severe trauma are multifactorial, including consumption and dilution of platelets and coagulation factors, as well as dysfunction of platelets and the coagulation system. Clinical coagulopathy occurs because of hypothermia, platelet and coagulation factor dysfunction that occurs at low temperatures, activation of the fibrinolytic system, and hemodilution following massive resuscitation. Platelet dysfunction is secondary to the imbalance between thromboxane and prostacyclin that occurs in a hypothermic state. Hypothermia and hemodilution produce an additive effect on coagulopathy. After replacement of one blood volume (5000 ml or 15 units of packed red blood cells [pRBCs]), only 30%–40% of platelets remain in circulation. The prothrombin time (PT), partial prothrombin time (PTT), fibrinogen levels, and lactate levels are therefore not predictive of the severe coagulopathic state.

The predominant physiologic defect resulting from repetitive and persistent bouts of hypoperfusion is metabolic acidosis. Anaerobic metabolism starts when the shock stage of hypoperfusion is prolonged, leading to the production of lactate. Acidosis decreases myocardial contractility and cardiac output. Acidosis also worsens as a result of multiple transfusions, the use of vasopressors, aortic cross-clamping, and impaired myocardial performance. It is clear that a complex relationship exists among acidosis, hypothermia, and coagulopathy, and each factor compounds the other, leading to a high mortality rate once this cycle ensues and cannot be interrupted.

MODELS FOR DAMAGE CONTROL

Stone’s original work in 1983 only provided the intraoperative observation of coagulopathy as an indication for “bailout.” In this study, 17 patients underwent the “bailout” procedure, which included an initial laparotomy, followed by packing in patients with an observed clinical coagulopathy, and then completion of the surgical procedure once the coagulopathy was improved. This resulted in 11 survivors with a mortality rate of 35%. Subsequently, Rotondo et al. described the multiphase approach to the management of exsanguinating patients sustaining abdominal injury, but did not define any objective parameters during the intraoperative phase of damage control. The authors reported a survival rate of 77% in a very small subgroup of patients with major vascular injury and two or more physical injuries. Burch et al. proposed a model based on core temperature 32° C or less, pH 7.09 or less, and pRBC transfusion of more than 22 units that could predict 48-hour survival; the authors also described the “lethal triad.” In a study based on 39 patients, Sharp et al. defined a temperature 33° C or less, pH 7.18 or less, PT 16 seconds or higher, PTT 50 seconds or higher, and more than 10 units of pRBCs transfused as objective parameters to indicate the need for early packing.

Morris et al. described 107 patients who underwent staged laparotomy and abdominal packing. They proposed proceeding with damage control early in the course of operation based on patient’s temperature of less than 35° C, a base deficit greater than 14, and the presence of coagulopathic bleeding. Similarly, Moore described a progressive coagulopathy as the most compelling reason for staged laparotomy. A severe coagulopathic state was described as PT and PTT greater than two times normal, massive and rapid blood transfusion exceeding 10 units in 4 hours, persistent shock defined as a oxygen consumption less than 110 ml/min/m², lactic acid level greater than 5 mmol/l, pH under 7.2, base deficit higher than 14, and core hypothermia less than 34° C. It was postulated that the ability to predict the onset of coagulopathy would have significant implications for instituting damage control. Another predictive model for life-threatening coagulopathy included systolic blood pressure less than 70 mm Hg, temperature higher than 34° C, pH less than 7.10, and Injury Severity Score (ISS) of 25 or higher.

No single model has been able to accurately predict the timing for institution of damage control. A pH less than 7.1 or a core temperature of less than 33° C may indicate that the “bloody vicious cycle” is too far advanced and cannot be interrupted. Similarly, it is difficult to obtain intraoperative results for PT, PTT, fibrinogen, and lactate levels at all hospitals, or to place a Swan-Ganz catheter in the operating room (OR).

To define the patient at greatest risk for exsanguination and death, one must determine the threshold levels of pH, temperature, and highest estimated level of blood loss. Therefore, in an attempt to institute the development of intraoperative guidelines for “damage control/bailout,” Asensio et al. first retrospectively evaluated 548 patients over 6 years who were admitted to a large urban trauma center with the diagnosis of exsanguination. Inclusion criteria were intraoperative blood loss of 2000 ml or more, minimum transfusion requirement of 1500 ml pRBCs or greater during the initial resuscitation, and diagnosis of exsanguination. Data collected included demographics, prehospital and admission vital signs and physiologic predictors of outcome, Revised Trauma Score (RTS), Glasgow Coma Scale (GCS), Injury Severity Score (ISS), volume of resuscitative fluids, need for thoracotomy in the emergency department (EDT), volume of fluids in the operating room, need for thoracotomy in the operating room (ORT), and intraoperative complications. In this patient population, the Revised Trauma Score was 4.38 and the mean ISS was 32, denoting a physiologically compromised and severely injured patient population. There were 180 patients that underwent EDT with aortic cross-clamping, open CPR, 99 (55%) succumbed in the emergency department. In addition to the 81 patients that survived EDT, 117 required ORT for a total of 198 EDT and ORT, of which 56 (28%) survived to leave the operating room and the hospital. In this series, mean admission pH was 7.15, mean temperature was 34.3° C in the operating room, and these patients received an average of 14,165 ml of crystalloid, blood, and blood products. Overall, 449 patients survived to arrive in the OR with some signs of life and 281 patients died; 37% of these patients survived damage control. Table 1 shows the objective intraoperative parameters developed to predict outcome and provide guidelines on when to institute damage control based on these findings. This series also provided independent risk factors for survival, which included an injury severity score less than 20, spontaneous ventilation in the emergency room, no EDT or ORT, and the absence of abdominal vascular injury.

Table 1 Physiologic Guidelines That Predict Need for Damage Control

One of the natural sequelae in patients surviving damage control is an open abdomen. These guidelines were prospectively validated in a series of 139 patients who underwent damage control and had post-traumatic open abdomen. This study consisted of two groups of patients: 86 patients studied retrospectively prior to instituting the guidelines, and 53 patients studied prospectively after instituting the guidelines. The groups were comparable in all relevant parameters. Although there was no difference in the mortality rate between the two groups (24% for each), there were statistically significant differences in the number of intraoperative transfusions, less hypothermia and bowel edema, less postoperative infections and gastrointestinal complications, and shorter intensive care unit and hospital length of stay for the prospective group. Another significant finding in this study was that 93% of patients were able to undergo definitive abdominal closure in their hospital stay as compared with the historic 22%.

Awareness of potential triggers to initiate damage control is vital. A study of 68 patients who underwent damage control surgery found that the inability to correct pH greater than 7.21 and PTT greater than 78.7 seconds was predictive of 100% mortality. Delayed recognition of the need for damage control as well as poor communication with the anesthesia and nursing team are deleterious to the care of the multiply injured patient.

PATIENT SELECTION

Not all trauma patients require damage control measures. In addition to the physiologic guidelines for the institution of damage control (see Table 1), certain conditions and complexes of injuries assessed both preoperatively and intraoperatively require damage control (Table 2). Multiple mass casualties and the need for EDT predict the need for damage control. In the multiply injured trauma patient sustaining major abdominal injury, the need to evaluate for other extra-abdominal injuries in a timely fashion may also indicate damage control. The preoperative duration of hypotension (systolic blood pressure <90 mm Hg) was significantly different in those patients that exsanguinated as compared with survivors (45 minutes vs. 85 minutes). Therefore, in addition to other factors such as the preoperative assessment of hypothermia and coagulopathy, a period of sustained hypotension greater than 60 minutes would predict the need for damage control. Intraoperatively, certain complexes of injuries also predict the need for this technique. These injuries include major abdominal vascular, complex hepatic, major thoracic vascular injuries, and the need for intraoperative thoracotomy.

Table 2 Preoperative and Intraoperative States That Suggest Need for Damage Control

Patients with exsanguination are perhaps the best candidates to undergo damage control. Asensio has described an algorithm for the management of exsanguination that involves three phases (Figure 1). First, the patient is classified as exsanguinating; and second, resuscitation under the Advanced Trauma Life Support protocols is begun (see Figure 1). The third phase comprises rapid transport to the operating room (exsanguination from penetrating injuries is a dramatic ill-defined entity that requires leadership, fast thinking, aggressive surgical intervention, and a well-thought-out plan). Rapid damage control can lead to effective management of exsanguination and improve survival.

Figure 1 Algorithm for the management of exsanguinations. ATLS, Advance Trauma Life Support; ED, emergency department; OR, operating room.

TECHNIQUE OF DAMAGE CONTROL

The most important goal of early institution of damage control is patient survival. A four-stage damage control approach has been recently defined by Johnson in a study of 24 patients who underwent damage control and were retrospectively compared with patients who underwent damage control a decade earlier (Figure 2). The “ground-zero” stage includes the prehospital phase as well as early resuscitation in the emergency room. This ground-zero phase includes short paramedic scene times and identification of injury patterns in the ED that require damage control, as well as rewarming maneuvers that begin in the trauma bay.

Figure 2 Four stages of damage control. ICU, Intensive care unit; OR, operating room.

Damage control implies immediate control of life-threatening hemorrhage, control of gastrointestinal contamination with rapid resections or closures, the use of intraluminal shunts, and judicious abdominal packing with temporary abdominal wall closures. Specifically, for chest injuries, one should repair cardiovascular injuries, perform stapled pulmonary tractotomy, pack if needed, place chest tubes, and close the skin. For abdominal injuries, damage control can involve control of major hemorrhage, hepatic packing, pancreatic drainage, temporary hollow viscus closures, rapid stapled resections, splenectomy, nephrectomy, vascular pedicle clamping in situ, and the use of intra-abdominal vascular shunts. Frequently, these patients experience abdominal compartment syndrome. Therefore, the post-traumatic open abdomen with temporary abdominal wall closure is used as an extension of damage control.

The second stage begins in the ICU where the metabolic disorder is corrected. Rewarming the patient is a high priority, as coagulopathy and acidosis can only be corrected and maintained once the body temperature returns to normal. Further inspections are then made to identify injuries that may have not been detected in the initial survey. Twenty-four to 72 hours may be needed to correct metabolic derangements.

The last stage of damage control involves the timing of reoperation when definitive procedures are performed. Reoperation is considered early if major blood losses continue. Usually, there is a window of 36–48 hours after the initial injury between the correction of the metabolic disorder and the onset of the systemic inflammatory response syndrome and/or multiple organ failure. In this phase, definitive procedures are undertaken. Thorough re-exploration is made for any additional injuries and restoration of gastrointestinal continuity and vascular repair are done. Provisional feeding access may be placed, followed by washout of the abdominal cavity and attempts at definitive closure. The patient then returns to the intensive care unit for further care.

CONCLUSIONS

The management of exsanguination requires leadership, prompt thinking, aggressive surgical intervention, and a well-conceived plan. The exsanguinating trauma patient that requires massive transfusion incurs the greatest risk for the multifactorial interactions among acidosis, hypothermia, and coagulopathy. The ultimate goal is the prompt and effective control of the exsanguinating source; delays in the decision to perform damage control contribute to a higher morbidity and mortality. Therefore, damage control is a vital part of the management of the multiply injured patient and should be performed before metabolic exhaustion.