EMERGENCY DEPARTMENT THORACOTOMY

Published on 10/03/2015 by admin

Last modified 22/04/2025

Print this page

This article have been viewed 1866 times

CHAPTER 19 EMERGENCY DEPARTMENT THORACOTOMY

Juan A. Asensio, Patrizio Petrone, Luis Manuel Garcí-a-Núñez, Dan L. Deckelbaum, Dominic Duran, Donald Robinson, Kenneth D. Stahl, Michael F. Ksycki, Francisco A. Ruiz Zelaya, Alexander D. Vara, John S. Weston, Allan Capin, Carl Schulman

Emergency department thoracotomy (EDT) remains a formidable tool within the trauma surgeon’s armamentarium. Since its introduction during the 1960s, the use of this procedure has ranged from sparing to liberal.¹ At many urban trauma centers, this procedure has found a niche as part of the resuscitative process.¹ Because of improvements in emergency medical services systems (EMS), many critically injured patients now arrive in extremis prompting trauma surgeons to perform this procedure to attempt saving their lives. This technically complex procedure should only be performed by surgeons familiar with the management of penetrating cardiothoracic injuries.

Indications for the use of EDT appearing in the literature range from vague to quite specific.¹ It has been used in a variety of settings including penetrating and blunt thoracic and/or thoracoabdominal injuries, cardiac and exsanguinating abdominal vascular injuries.¹^–⁴ It has also been used rarely in exsanguinating peripheral vascular injuries arriving in cardiopulmonary arrest and also in pediatric trauma. Many studies in the literature have also reported its use in patients presenting in cardiopulmonary arrest secondary to blunt trauma.

HISTORIC PERSPECTIVE

In 1874, Schift ⁵ was first to promote the concept of open cardiac massage. Rehn⁶ in 1896 reported the first successful repair of a cardiac injury, a stab wound of the right ventricle. In 1897, Duval⁷ described the median sternotomy incision widely used today. Igelsbrud⁵ in 1901 was the first to report successful resuscitation of a patient, sustaining a post-traumatic cardiac arrest patient with a thoracotomy and open cardiac massage. Spangaro ⁸ in 1906 described the left anterolateral thoracotomy widely used today for resuscitation as an intercostocondral thoracotomy.

Zolls ⁵ in 1956 was the first to introduce the concept of external defibrillation, and Kouwenhoven⁵ in 1960 described closed cardiopulmonary resuscitation. Beall⁹ et al. in 1961 was the first to propose that patients experiencing cessation of cardiac action should undergo immediate resuscitative thoracotomy and cardiac massage, whether in the ED, operating room (OR), or recovery ward, and was first to attempt this procedure. Similarly, in 1966, he advocated the use of immediate cardiorrhaphy in the emergency room and setting up an instrument tray; he was also the first to successfully perform this procedure.¹⁰

OBJECTIVES

Objectives of the EDT procedure include the following:

Resuscitation of agonal patients with penetrating cardiothoracic injuries (Figure 1A and B)

Evacuation of pericardial blood and clot to relieve cardiac tamponade

Control of thoracic hemorrhage

Perform open cardiac massage, which can produce up to 60% of the normal ejection fraction (EF)

Repair of cardiac injuries

Cross-clamp the pulmonary hilum to control pulmonary vessel hemorrhage

Cross-clamp the descending thoracic aorta (Figures 2 and 3)

Cross-clamp the pulmonary hilum and aspirate both the right and left ventricles to prevent and/or treat pulmonary embolism

Figure 1 (A) Suturing a stab wound to the heart, after emergency department thoracotomy. (B) Same patient with the suture completed (arrow).

Figure 2 The descendent aorta when entering the aortic hiatus. The aorta was dissected prior to being clamped. Note the esophagus superiorly.

Figure 3 Aortic cross-clamp in place. Note the size of the left hemithorax cavity, which can hold the entire patient’s volemia during exsanguinated injuries.

INDICATIONS

Indications for EDT can be subdivided into three categories: accepted, selective, and rare.¹^–⁴

Accepted Indications

Emergency department thoracotomy is best applied to patients sustaining penetrating cardiac injuries who arrive in trauma centers after a short scene and transport time with witnessed or objectively measured physiological parameters (signs of life), pupillary reactivity, spontaneous ventilation even if agonal, presence of a carotid pulse, measurable or palpable blood pressure, extremity movement, and cardiac electrical activity.

Selective Indications

Emergency department thoracotomy should be performed selectively in patients sustaining penetrating noncardiac thoracic injuries due to its very low survival rate. Because it is difficult to ascertain whether injuries are noncardiac thoracic versus cardiac this procedure may be employed to establish a diagnosis.

EDT should be performed selectively in patients sustaining exsanguinating abdominal vascular injuries due to its very low survival rate. Meticulous selection of patients should be exercised. This procedure should be used as an adjunct to definitive repair of abdominal vascular injuries.

Rare Indications

Emergency department thoracotomy should be performed rarely in patients sustaining cardiopulmonary arrest secondary to blunt trauma due to its very low survival rate and poor neurological outcomes. Extreme caution should be exercised in selecting patients for this procedure. It should be strictly limited to those who arrive with vital signs at the trauma center and experience a witnessed cardiopulmonary arrest. Most authors would caution against this indication.

TECHNIQUES FOR CARDIAC INJURY REPAIR

Incisions

There are two main incisions that are used in the management of penetrating cardiac injuries. Trauma surgeons should be aware that injuries caused by missiles can be unpredictable in their trajectory and that a missile injury that penetrates a hemithoracic cavity may not remain confined in the original area of entrance and may produce injury to the contralateral cavity. This will require the trauma surgeon to access the contralateral hemithoracic cavity.^67,^68,^70,^74,⁷⁷

Median sternotomy described by Duval ³⁷ is the incision of choice for patients admitted with penetrating precordial injuries that arrive with some degree of hemodynamic instability and may either undergo preoperative investigation with FAST and/or chest x-ray. It is also the incision of choice or those that are thought to harbor occult cardiac injuries. The left anterolateral thoracotomy is the incision of choice in the management of patients who arrive in extremis. This incision is used in the ED for resuscitative purposes.^67,^68,^70,^74,⁷⁷

The left anterolateral thoracotomy described by Spangaro ⁴¹ can also be extended across the sternum as bilateral anterolateral thoracotomies, if it is determined during the resuscitative period that the patient’s injury extends into the right hemithoracic cavity (Table 1). Extension into bilateral anterolateral thoracotomies is also is the incision of choice for patients that are hemodynamically unstable after incurring mediastinal traversing injuries. This incision allows full exposure of the anterior mediastinum and pericardium as well as both hemithoracic cavities. It is important to note that upon transection of the sternum both internal mammary arteries are also transected and must be ligated after restoration of perfusion pressure. Uncontrolled, they can serve as a significant source of blood loss. This is a frequent pitfall during the institution of damage control; as trauma surgeons may forget to ligate these vessels prompting return to the operating room for a patient that can ill afford it. For patients that sustain thoracoabdominal injuries, the left anterolateral thoracotomy is also the incision of choice if patients deteriorate in the OR while undergoing a laparotomy.^67,^68,^70,^74,⁷⁷

Table 1 Algorithm for Emergency Department Thoracotomy

Operator	Well-trained surgeon
Initial assessment and resuscitation	Endotracheal intubation
	Immediate venous access
	Rapid infusion
Position	Supine with left arm elevated
Incision	Left anterolateral incision
Incision	Fifth intercostal space from left sternocostal junction to latissimus dorsi m.
Procedure	Incision as above
	Sharp transection of intercostal m.
	Open pleura
	Place a Finnochietto retractor
	Open cardiac massage
	Elevate left lung medially
	Locate and dissect descending aorta
	Cross clamp aorta by Crafoord-Debakey clamp
If cardiac injury (bluish and tense pericardium)	Open pericardium longitudinally with preserving phrenic n.
	Evacuate blood clot
	Repair cardiac injury (mattress sutures of Halsted with Prolene 2/0)
If active bleeding at pulmonary hilum	Cross-clamp pulmonary hilum with Crafoord-Debakey clamp
If pulmonary parenchymal laceration	Clamp with Duval clamp
If associated injury in contralateral thoracic cavity	Extend incision to the contralateral side
	Transect sternum sharply
	Convert to bilateral anterolateral thoracotomy
If air embolism is suspected (air in coronary v)	Aspirate left ventricle
Miscellaneous	Ligate internal mammary a.
	Systemic or intraventricular epinephrine administration
	Internal defibrillation 10–50 joules
	Temporary pacemaker
Immediately transport to operating room after successful resuscitation

Adjunct Maneuvers

Trauma surgeons must possess several maneuvers in their armamentarium to deal with penetrating cardiothoracic injuries. The first adjunct maneuver dealing with these injuries was described by Sauerbuch ⁴² in 1907, as quoted by Brantigan. This maneuver entailed controlling blood flow to the heart by compression of the base. This maneuver is difficult to perform via a left anterolateral thoracotomy, has been abandoned, and is only mentioned because of historical interest only.

Total inflow occlusion to the heart is a complex maneuver that entails cross-clamping both the superior (SVC) and inferior vena cava (IVC) in their intrapericardial location to arrest total blood flow to the heart. Crafoord-DeBakey cross-clamps are employed, resulting in the immediate emptying of the heart. The trauma surgeon must recognize that cross-clamping the inferior vena cava intrapericardially at the space of Gibbons can be quite treacherous, as it is often fused with the posterior aspect of the pericardium. Inexperienced trauma surgeons will often force the cross-clamp in an attempt to rapidly achieve total occlusion leading to an iatrogenic injury of the intrapericardial IVC. Similarly, circumferentially dissecting this delicate vessel can also lead to iatrogenic injury. The clamp must be placed carefully and sometimes at an angle so as to totally occlude the intrapericardial IVC.^70,⁷⁷

Total inflow occlusion of the heart is indicated for the management of injuries in the lateral-most aspect of the right atrium and/or the superior or inferior atriocaval junction. Total inflow occlusion will lead to immediate emptying of the heart and allow the injury to be visualized and thus repaired. Frequently this procedure results in cardiopulmonary arrest, as tolerance by the injured, acidotic, hypothermic and ischemic heart is very limited. The safe period for this maneuver is unknown, although a 1–3 minute range is often quoted in the literature as the period of time after which clamps must be released. As the clamps are released, venous return fills the right-sided cardiac chambers and forward cardiac pumping motion will begin. More often than not, the heart will fibrillate requiring immediate direct defibrillation along with pharmacologic manipulation. This may be unsuccessful, particularly if a period of 3 minutes has been exceeded. Restoration of a normal sinus rhythm is often impossible.^70,⁷⁷

Cross-clamping of the pulmonary hilum is another valuable maneuver indicated for the management of associated pulmonary injuries, particularly those that have hilar central hematomas and/or active bleeding. This maneuver arrests bleeding from the lung and prevents air emboli from reaching the systemic circulation. However, one of its negative effects is responsible for significantly increasing the afterload of the right ventricle, as half of the pulmonary circulation is no longer available for perfusion. We recommend sequential de-clamping of the hilum to be carried out as expediently as possible along with a direct approach by stapled pulmonary tractotomy ^70,⁷⁴ for identification and control of hemorrhaging intraparenchymal pulmonary vessels. This will promptly unload the right ventricle. In the presence of acidosis, hypothermia, and ischemia, the right ventricle may not be able to tolerate this maneuver leading to fibrillation and arrest.^70,⁷⁴

Grabowski ¹ recently described a maneuver to facilitate exposure of posterior cardiac wounds by placing a Satinsky clamp at the right ventricular angle, which is formed at the acute anteroinferior margin of the right ventricle as it reflects on the right diaphragm. Grabowski¹ recommends that the clamp only grasp a small portion of the right ventricle. He recommends this maneuver for elevating the heart out of the pericardium to repair posterior injuries. We have no experience with this maneuver and cannot recommend it. We strongly feel that if used inappropriately, it will lead to the development of significant cardiac dysrhythmias.^70,⁷⁴

Maneuvers such as venting either the right or left ventricle postcardiorrhaphy are recommended to provide an avenue of egress for air emboli trapped in these chambers. This is usually accomplished by placing 16-gauge intravenous catheters. Theoretically, air should eject out of the repair chambers, thus preventing air emboli. Although the authors have used this maneuver successfully, little has been written in the literature describing its outcome.^70,⁷⁴

At times a trauma surgeon will need to elevate the heart out of the pericardium in order to repair certain injuries. Rapid and injudicious manipulation of the heart will often result in complex dysrhythmias that might include ventricular fibrillation and even cardiopulmonary arrest. Occasionally, given the degree of exsanguinating hemorrhage the heart must be extracted rapidly from pericardium in order to perform cardiorrhaphy. The trauma surgeon must communicate with the anesthesiologists whenever this maneuver is performed. If hemorrhage can be digitally controlled, gradual elevation of the heart by placing multiple laparotomy packs will allow better tolerance of this maneuver while decreasing the chances for the development of dysrhythmias.^70,⁷⁴

Recently described mechanical stabilizer systems to the heart have been utilized in the performance of conventional coronary artery bypass grafts (CABGs), which have traditionally used cardiopulmonary bypass to allow cardiac surgeons to operate on a motionless heart arrested by means of cardioplegic solutions. The deleterious systemic inflammatory effects of circulating blood through the extracorporeal circuit of the cardiopulmonary pump have prompted the development of mechanical stabilizer systems to allow off pump coronary artery bypass grafting (OPCABG) to be performed by cardiothoracic surgeons.

Waterworth ¹¹ recently reported the first case in which the Octopus IV Mechanical Cardiac Stabilizer was used on a 20-year-old patient who sustained a 2-cm stab wound in the right ventricular outflow tract approximately 1 cm below the pulmonary valve. According to the author, this area was difficult to suture without causing further tearing due to tachycardia sustained by the patient and the fragile nature of this area. After control of hemorrhage by direct pressure, the Octopus IV Mechanical Cardiac Stabilizer was placed, which provided for immobility to this area of the heart and thus facilitated repair. In this case report, the author describes the use of this device, suggesting that cardiac stabilization devices with adjustable suction foot blades may be used to control hemorrhage in addition to facilitating repair, particularly in areas difficult or dangerous to handle manually. The recommended positioning parallel to the direction of the wound and approximating the foot plates may result in closure of the wound, providing a clear field for repair. This case report by Waterworth¹¹ appears to be the first and only case reported in the literature utilizing a mechanical cardiac stabilizer in the management of a penetrating cardiac injury. Whether stabilizers will be routinely used in the management of penetrating cardiac injuries in the future remains to be seen.

Repair of Atrial Injuries

Atrial injuries can usually be controlled by placement of a Satinsky partial occlusion vascular clamp. Control of the wound will allow the trauma surgeon to perform cardiorrhaphy. We recommend utilizing 2-0 or 3-0 polypropylene monofilament sutures on an MH needle in a running or interrupted fashion. It is important to visualize both sides of the atrial injury, particularly those caused by missiles. Missile injuries can usually cause a significant amount of tissue destruction, which might require meticulous debridement prior to closure. Similarly, a portion of the atria may be resected and cardiorrhaphy performed utilizing a running suture of 2-0 or 3-0 polypropylene monofilament suture. The trauma surgeon must be aware that the atria have fairly thin walls and demand gentleness during cardiorrhaphy, as they can easily tear and enlarge the original injury. The use of bioprosthetic materials in the form of Teflon pledgets is not recommended for management of these injuries.^70,⁷⁴

Repair of Ventricular Injuries

Ventricular injuries usually cause significant hemorrhage. They should be occluded digitally while simultaneously repaired by either simple interrupted or horizontal mattress sutures of Halsted. Ventricular cardiorrhaphy can also be accomplished with a running monofilament suture of 2-0 polypropylene on an MH needle. Performing cardiorrhaphy for ventricular for stab wounds is usually less challenging than for gunshot wounds. Missile injuries often produce some degree of blast effect that causes myocardial fibers to retract. Frequently, missile injuries that have been successfully sutured and controlled enlarge, as the damaged myocardium retracts and becomes more friable. Frequently, these injuries require multiple sutures to control significant hemorrhage. In the presence of this scenario, bioprosthetic materials such as Teflon strips and/or pledgets are often needed to buttress the suture line. This is usually performed by fashioning a Teflon strip that may measure anywhere from 1 to 5 cm (Figure 4A and B). This strip is held by two straight Crile clamps held by an assistant. Simultaneously, the trauma surgeon may then place double-armed 2-0 polypropylene monofilament sutures on an MH needle, first through the strip, and then through both sides of the injury. A second strip is then held in a similar fashion so that the trauma surgeon then places both needles thru the second Teflon strip. The sutures are then gently tied against the Teflon strip and/or pledget, which will buttress and reinforce the suture line. This maneuver must be repeated until total control of ventricular hemorrhage is achieved. The authors have recently used commercially made fibrin sealants to seal complex ventricular injuries.^70,⁷⁴

Figure 4 (A) Patient sustained a cardiac apex penetrating injury. (B) Same patient with Teflon patches used to reinforce the sutures.

Coronary Artery Injuries

The repair of ventricular injuries adjacent to coronary arteries can be very challenging. Injudicious and/or inappropriate placement of sutures during cardiorrhaphy may narrow and/or occlude a coronary artery or one of its branches. Therefore, it is recommended that sutures be placed underneath the bed of the coronary artery. Coronary arteries are usually divided into three segments: proximal, middle, and distal. Injuries to the proximal segment of a coronary artery will usually require cardiopulmonary bypass for repair, although this is infrequently necessary. Injuries of the middle segment of the coronary artery may also require cardiopulmonary bypass, or if ligated in desperation, may result in immediate myocardial infarction at the operating table. These patients may benefit from the institution of intra-aortic balloon counterpulsation followed by aortocoronary bypass. Lacerations of the distal segment of the coronary artery particularly in the distal-most third of the vessel are managed by ligation.^70,⁷⁴

Use of Bioprosthetic and Autogenous Materials

Trauma surgeons are familiar with the use of Teflon pledgets and/or strips to buttress suture lines on friable myocardial tissue. Mattox ¹⁸ provided the first reference in the literature alluding to the use of this material. The authors strongly believe in the necessity to buttress complex suture lines and use Teflon when indicated. However, no studies have been performed to determine if the use of Teflon increases tensile strength of the repair. The use of autogenous materials such as the pericardium to bolster suture lines is also well known. A small flap is developed and excised from the pericardium to be used in a manner similar to use of Teflon pledgets. Inexperienced trauma surgeons will often suture the pericardium to a ventricular injury causing the chamber to be fixated, which leads to dysrhythmias. This is mentioned, as it is a pitfall that should be avoided at all costs.^70,⁷⁴

Complex and Combined Injuries

As trauma surgeons and trauma centers continue to develop greater expertise in the management of penetrating cardiac injuries, and patients are subjected to greater degrees of violence in urban arenas of warfare, a significant number of patients arrive harboring multiple associated injuries in addition to their penetrating cardiac injuries. Complex and combined cardiac injuries can be defined as a penetrating cardiac injury plus associated neck, thoracic, thoracic-vascular, abdominal, abdominal vascular, or peripheral vascular injuries. These injuries are quite challenging to manage. Priority should be given early to the injury causing the greatest blood loss or threatening the patient’s life.^70,⁷⁴

Wall and Mattox described 60 patients with complex cardiac injuries, which they defined as those beyond lacerations of the myocardium. These injuries were defined as those with concomitant coronary artery injuries, cardiac valvular injuries, intracardiac fistulas, and other unusual injuries. In this series, the authors described 39 coronary artery injuries; 2 valvular injuries; and 14 intracardiac fistulas including ventriculoseptal defects, atrioseptal defects (ASD), and another 10 injuries that they considered unusual ranging from ventricular false aneurysms to coronary sinus injuries and two patients who developed missile emboli to the heart. These types of injuries can also be considered complex and combined injuries.

RESULTS

The literature abounds with retrospective series describing the use of EDT. Great difficulties, however, exist in evaluating the results of these series. Close scrutiny reveals several flaws; most series have been retrospective reviews, many from institutions that employ this technique infrequently. Furthermore, many institutions report many overlapping studies that encompass the experience of many years. Whereas many series have selected physiologic parameters as predictors of outcome, none have statistically validated their predictive values. Invariably, these series omit data pertaining to the physiologic status of the patient upon initial presentation. To our knowledge, there is only one prospective study in the literature.⁶⁷ As a result, there are still many questions to be answered.²⁵

Which patients should be subjected to this procedure?

Are there any prospectively validated physiologic predictors of outcomes that can safely and accurately identify patients who will benefit from the procedure and also safely exclude those who will not?

What are the true survival rates of this procedure?

Of the surviving patients, how many survive with severe neurologic impairment or remain in a persistent vegetative state?

How can we ensure that individuals performing this procedure are qualified?

Well-known physiologic factors predictive of poor outcome include prehospital and ED absence of vital signs, fixed and dilated pupils, absence of cardiac rhythm and motion in the extremities, and agonal breathing. Similarly, the absence of a palpable pulse in the presence of cardiopulmonary arrest is also predictive of poor outcome. These factors have been validated by Asensio and colleagues,^12,¹³ who tracked cases in the field, during transport, and upon arrival in the ED in two prospective studies dealing with penetrating cardiac injuries. Interestingly enough, many of these physiologic predictors of outcome, as well as any data describing the physiologic condition of the patient prior to EDT, are often absent in many studies.^1,^2,²⁵

Previous work by Buckman et al.¹⁷ and Asensio and colleagues^12,¹³ applied and validated the cardiovascular respiratory score (CVRS) of the trauma score (TS). The cardiovascular respiratory component of the trauma score reflects individual elements of blood pressure, respiratory rate, respiratory effort and capillary refill. The highest possible CVRS is 11, which denotes a systolic blood pressure of greater than 90 mm Hg, respiratory rates that fluctuate between 10 and 24/min along with a normal respiratory effort and capillary refill. The lowest possible CVRS score is zero, and reflects an absence of blood pressure, no palpable carotid pulse along with an absence of breathing, respiratory effort, and capillary refill. This score has been statistically validated and applied to the only three perspective cardiac injury series reported in the literature.

Asensio and colleagues,² in the only prospective study on the use of EDT reported in the literature, analyzed parameters measuring the physiologic condition of patients incurring cardiopulmonary arrest in the field, during transport, and upon arrival at the trauma center. The CVRS, injury mechanism, and anatomic site of injury along with restoration of blood pressure were tracked prospectively with the objectives of identifying a set of parameters that would reliably predict mortality and exclude patients from EDT. This 2-year prospective study had a single inclusion criterion—cardiopulmonary arrest secondary to traumatic injury—as well as a single intervention, EDT for resuscitation. The main outcome of this study was survival at 1 hour and survival to discharge.

A total of 215 patients who sustained cardiopulmonary arrest were studied prospectively. Of this total, 167 (78%) sustained penetrating injuries, including 142 (66%) gunshot wounds, 21 (10%) stab wounds, and 4 (2%) shotgun wounds. In addition, there were 48 (22%) who sustained blunt injuries. The mean revised trauma score (RTS) was 0.6, the mean injury severity score (ISS) was 42, and the mean CVRS was 1, denoting a severely injured and physiologically compromised patient population. The mean duration of cardiopulmonary resuscitation (CPR) prior to arrival at the trauma center was 12 minutes.

A total of 162 patients (75%) succumbed in the ED. Fifty-three patients (25%) survived up to 1 hour, after successful ED resuscitation with some restoration of vital signs so that they could be transported immediately to the OR. Of the 215 patients, only 6 (3%) survived, all of whom sustained cardiac injuries. None of the 48 patients who were injured due to blunt trauma survived. Upon comparing patients succumbing in the ED with those who survived at least 1 hour, all physiologic parameters were predictive of outcome (p<0.001). When all patients who survived 1 hour were compared with overall survivors, none of the physiologic parameters predicted outcome. Duration of CPR (p=0.04), penetrating mechanism of injury (p<0.001), and exsanguination (p<0.006) were predictors of outcome. The CVRS showed a trend toward the prediction of survival (p=0.07). When all nonsurvivors were compared with survivors, restoration of blood pressure was a strong predictor of outcome (p<0.001).

The authors concluded from these data that physiologic parameters plus the CVRS score were predictive of survival for patients undergoing EDT. On the basis of these criteria, the authors estimated that 75% of these patients could be safely excluded from this procedure at a cost savings of over $500,000 at their institution, and recommended that EDT should be limited to patients sustaining penetrating cardiac injuries and should not be applied to patients sustaining cardiopulmonary arrest secondary to blunt trauma.

Precisely because of the lack of uniformity in the reporting process in many of the reports in the literature, Asensio and colleagues ²⁵ in the working group of the Ad Hoc Subcommittee on Outcomes of the American College of Surgeons Committee on Trauma closely scrutinized the literature to generate practice management guidelines for EDT.

In an extensive literature search, studies were classified into three classes. Class I comprises prospective randomized controlled trials and remains the gold standard of all clinical trials. In this category, the studies found were generally poorly designed, had inadequate numbers, or suffered from methodological inadequacies, rendering them clinically nonsignificant. In this group, no prospective randomized controlled trials were found. Studies in class II included clinical studies in which data were collected prospectively, as well as retrospective analyses based on clearly reliable data. Included here are observational, cohort, prevalence, and case-controlled studies. The authors found 29 studies that qualified for class II, three of which were prospective. Finally, for class III—defined as retrospectively collected data including clinical series, databases or registries, case reviews, case reports, and expert opinion—the authors located 63 studies.²⁵

Analysis was conducted by stratifying the series into series dealing with EDT, series reporting neurologic outcomes of patients subjected to EDT, series dealing exclusively with penetrating cardiac injuries, and series dealing with pediatric patients. In the 42 series ²⁵^–⁶⁷ dealing with EDT (Table 2), there were 7035 EDTs and 551 survivors, for a survival rate of 7.83%. When data were stratified according to mechanism of injury, there were 4482 thoracotomies for penetrating injuries; of these, 500 patients survived, yielding a survival rate of 11.16%. There were 2193 thoracotomies performed for blunt injuries; only 35 patients survived, for a survival rate of 1.60%.²⁵^–⁶⁷

Table 2 Emergency Department Thoracotomy

Of the 14 series reporting neurologic outcomes and their results,^27,^28,^30,^34–37,^41,^42,^46,^52,^56,^59,⁶⁵ a total of 4520 patients were subjected to EDT with 226 survivors, yielding a 5% survival rate. Of these 226 survivors, 34 (15%) survived with neurologic impairment. In the series dealing exclusively with EDTs performed to repair penetrating cardiac injuries (Table 3), in a total of 1165 EDTs, 363 patients survived, yielding a survival rate of 31.1%. Only four series were found that dealt exclusively with pediatric patients (Table 4). There were 142 EDTs performed. Of 57 thoracotomies performed for penetrating injuries, 7 patients survived, yielding a survival rate of 12.2%. Eighty-five thoracotomies were performed for blunt injuries; 2 patients survived, for a survival rate of 2.3%.