- Email: [email protected]
Management of penetrating lung injuries in civilian practice
J
THoRAc CARDIOVASC SURG
1988;95: 184-90
Management of penetrating lung injuries in civilian practice Recent reports of military thoracic injuries have advocated early thoracotomy and aggressive management of pulmonary injuries with resection as opposed to the more conservative and traditional treatment with chest tube thoracostomy. A retrospective study was therefore performed to determine the incidence of thoracotomy and lung resection in civilian injuries and to evaluate the effectiveness of treatment of these injuries. Between 1973 and 1985, in a series of 1168 patients, there were 384 gunshot wounds and 784 stab wounds to the thorax. Two hundred eighty-three patients with a gunshot wound (74%) and 602 with a stab wound (77%) were treated with chest tubes alone. Sixty-eight patients (6% of the total) required operative repair of pulmonary hilar or parenchymal injury. Pulmonary resection was necessary in only 18 patients (nine with a gunshot wound and nine with a stab wound), and 10 patients had repair of hilar injuries (nine with a gunshot wound and one with a stab wound). Of patients requiring pulmonary resection, nine required wedge or segmental resection, six required lobectomy, and three patients required pneumonectomy. Mortality for aU thoracic injuries was 2.3 %: for those treated with chest tube alone, 0.7%; for pulmonary hilar injuries, 30 %; for pulmonary parenchymal injuries, 8.6 %; and for injuries necessitating lung resection, 28 %. Most civilian lung injuries can be treated by tube thoracostomy alone. Although relatively few patients with primary pulmoru.lry injury require thoracotomy, those that do are at significant risk and may require lung resection to control bleeding or hemoptysis or to remove destroyed or devitalized lung tissue.
Paul D. Robison, MD (by invitation), P. Kent Harman, MD" (by invitation),
J. Kent Trinkle, MD, and Frederick L. Grover, MD, San Antonio. Texas
Eetrating injuries to the chest in civilian practice result mainly from gunshot or stab wounds. Recent military reports have emphasized the need for early thoracotomy in patients having high-velocity missile and shrapnel injuries.!" In contradistinction, the experience of several large civilian centers continues to support the use of chest tube thoracostomy as the primary treatment of nonmediastinal chest injury with a low frequency of thoracotomy>" The following study was undertaken to analyze the frequency of thoracotomy and pulmonary
Table I. Hilar or lung repair/resection summary
From The University of Texas Health Science Center at San Antonio and the Audie Murphy Memorial Veterans Administration Hospital, San Antonio, Texas.
resection after penetrating nonmediastinal chest injury in our institution.
Read at the Sixty-seventh Annual Meeting of The American Association for Thoracic Surgery, Chicago, III., April 6-8, 1987.
Patients and methods
Address for reprints: Dr. Frederick L. Grover, Division of Cardiothoracic Surgery, The University of Texas Health Science Center, 7703 Floyd Curl Dr., San Antonio, TX 78284-7841. 'Present address: Northwest Surgical Associates, 1501 N.E. Medical Center Dr., Bend, OR 97701.
184
Gunshot wound In = 384)
Stab wound In = 784) No.
%
No.
%
Lung repair Lung resection Hilar repair
16 9 1
2 l.l 0.1
24 9 9
6.3 2.3 2.3
Total
26
3.3
42
10.7
From January 1973 through January 1986, 1168 patients were treated at Medical Center Hospital, San Antonio, Texas, for penetrating injuries to the thoracic cavity. Ninety-six percent of the patients were male, with an age range of 5 to 68 years and a mean age of 28.8 years. There were 784 stab wounds and 384 gunshot wounds in the series. A retrospective
Volume 95 Number 2
Penetrating lung injuries
February 1988
Pneumonectomy A-All 3 lobes PAl L-PA/PV (2) Total 18
Fig. I. Location of pulmonary injuries in patients requiring resection. PA. Pulmonary artery. PV. Pulmonary vein. review of the charts of these patients formed the basis of this report. Of those patients admitted, 37 had a stab wound and lO had a gunshot wound with minimal evidence of pneumothorax or hemothorax on chest roentgenogram; they were treated by observation alone. Additionally, II patients having a stab wound and two having a gunshot wound underwent creation of a subxiphoid pericardiaI window alone and had no chest tube placed. Eighty-three patients with a stab wound and 34 patients with a gunshot wound underwent exploration for mediastinal injury, and no evidence of significant lung injury was found. Twenty-five patients had injury to a chest wall artery or subclavian artery as a result of a stab wound and underwent thoracotomy for repair of those arterial injuries. Patients treated with observation alone, pericardial window, thoracotomy for subclavian artery injury, or median sternotomy for mediastinal injuries without lung injury are excluded from further consideration. Twenty-six of 784 patients with a stab wound (3.3%) and 42 of 384 patients with a gunshot wound (10.7%) had pulmonary injuries necessitating operation (Table I). Sixteen of the 26 stab wounds and 24 of 42 gunshot wounds necessitated only oversewingof the injury. Nine of the 26 (34.6%) patients with a stab wound who underwent operation required pulmonary resection; seven had wedge resection and two had lobectomy. Nine of 42 (21.4%) patients with a gunshot wound underwent lung resection: Two had wedge resection, four lobectomy, and three pneumonectomy (Fig. I). One patient with a stab wound and nine patients with a gunshot wound required repair of hilar vessels (Fig. 2). Additionally, three patients arrived at the emergency center in full cardiac arrest. These patients underwent thoracotomy in the emergency room for resuscitation, which was unsuccessful. At postmortem examination, two of these patients were found to have pulmonary hilar injuries with massive exsanguination, and the third patient was found to have a pulmonary parenchymal injury with massive exsanguination. Three of 28 patients who underwent pulmonary resection or hilar repair had associated cardiac injuries. Two patients had a subclavian artery injury, two patients were paraplegic postoperatively from bullet injuries to the spinal cord, and two patients had liver injuries. Additionally, one patient had injuries to the pancreas, colon, spleen, and stomach, and one patient had injury to the brachial plexus.
185
Hilar Repair-10 Pneumonectomy-3 1 PA/PV-Right 2 PA/PV-Le
Fig. 2. Location of hilar injuries in patients requiring repair or pneumonectomy. PA. Pulmonary artery. Pl/, Pulmonary vein. Clinical presentation. Patients requmng operation were frequently in shock and cardiac arrest because of hypovolemia. The hemodynamic status of patients requiring operation for primarily lung injury or chest wall arterial injury is shown in Table II. Additionally, three patients with a gunshot wound and one patient with a stab wound to the lung had evidence of cardiac tamponade and were found to have a cardiac injury during the operation. Significantly, three of 18 patients requiring lung resection had massive hemoptysis on admission. Of six patients requiring lobectomy, three had massive hemoptysis. Two of these six patients required a second operation after an inadequate first operation to control pulmonary hemorrhage. Diagnostic and therapeutic procedures. Twenty of 28 patients requiring pulmonary resection or hilar repair had a chest x-ray film taken preoperatively. Eight of 28 patients were in hemodynamically unstable condition and were rushed to the operating room or underwent thoracotomy in the emergency room before the chest x-ray film was made. The most common x-ray finding in the remaining 20 patients was hemothorax or hemopneumothorax, which was present in 12 of 20. Two patients had bilateral hemopneumothoraces and one patient a simple pneumothorax. Additionally, pulmonary contusions and hemothorax were diagnosed preoperatively in two patients (Fig. 3). Two patients had a missile in the mediastinum overlying the heart shadow, and one patient had an enlarged cardiac silhouette. Additional diagnostic and therapeutic procedures on the 28 patients undergoing pulmonary resection or hilar repair included central venous pressure lines in 10, diagnostic peritoneal lavage in three, intravenous pyelogram in three, exploratory laparotomy in two, pericardial window in two, and barium swallow and thoracic myelogram each in one patient. Blood loss after placement of the initial chest tube in patients undergoing operation for nonmediastinal chest injuries are shown in Table III. Significantly, chest tube drainage was not useful as a predictor of severity of injury in those patients requiring operation, nor was it an indicator of organ systems injured in this study, although patients with massive
The Journal of Thoracic and Cardiovascular Surgery
1 8 6 Robison et af.
Table II. Hemodynamic presentation of patients requiring operation for chest wall arterial, lung, or hilar injury Without lung resection
BP > 70 mm Hg systolic BP < 70 mm Hg systolic Arrest
With lung resection
Hilar repair
SW ICA/IMA
SW lung
GSW lung
SW
GSW
SW
GSW
14 8 I
9 6 I
12 8 4
5 2 2
4 4 I
0
3 2 4
I
0
Legend: SW, Stab wound. GSW, Gunshot wound. ICA, Intercostal artery. IMA, Internal mammary artery.
Table
m. Patients requiring operation for Repair
ICA or IMA alone Pneumonorrhaphy and ICA/IMA Pneumonorrhaphy Resection Wedge Lobectomy Pneumonectomy Hilar repair
noncardiac chest injuries: Initial chest tube-blood loss SWNo.
Blood loss (ml ± SEM)
GSW No.
Blood loss (ml ± SEM)
23 9 5
1995 ± 223 1633 ± 21 I 1650 ± 217
0 5 13
1500 ± 530 2122 ± 312
7 2 0 I
2257 ± 250 1175 ± 852
2 2 3 8
1400 Massive Massive 2688 ± 550
2000
Legend: SW, Stab wound. GSW, Gunshot wound. ICA, Intercostal artery. IMA, Internal mammary artery.
bleeding often had hilar injuries, Some chest wall injuries, however, also were associated with large blood loss. Indications for operation. The most common indication for operation in patients with pulmonary injury necessitating repair was continued hemorrhage or massive hemorrhageafter placement of the chest tube, which occurred in 43 of 68 patients, Blood lossin excess of 1500 ml after initial placement of the chest tube or continuing blood lossof more than 150 to 200 ml/hr after placement of the initial chest tube was an indication for operation, Twenty-three patients had shock that was not reversible by aggressive fluid replacement. Additionally, 12 patients in this category were in cardiac arrest on admission or shortly thereafter. Three patients had massive hemoptysis, four patients had evidence of cardiac tamponade, and four patients had missiles overlying the mediastinum suggestive of cardiac injury, The decision to proceed with pulmonary resection was made at the time of thoracotomy, In 14 of 18 patients continued hemorrhage was the primary indication for pulmonary resection. Four patients had severe parenchymal destruction, three had hemoptysis, and two patients had central pulmonary injuries, Several patients had more than one indication for lung resection.
Results Mortality from all causes of penetrating thoracic injury included 10 of 784 patients with stab wounds and 18 of 384 patients with gunshot wounds, or 2.4% of the total patient population. One patient of the 885 patients who underwent chest tube thoracostomy died of a pleural infection. Seven other patients who were treated primarily with chest tube thoracostomies died of associated injuries. Of the 40 patients who underwent over-
sewing of lung injuries during the operation, none died of a pulmonary cause. Three patients of the 40, however, died of associated injuries. Of 18 patients who underwent lung resection, five died, two of air embolus, two of exsanguination, and one of sepsis from pulmonary and intra-abdominal sources. Three of 10 patients undergoing hilar repair died, all of exsanguination. Six of seven patients who had cardiac arrest before the operation died. In addition, two of nine patients operated on in shock died; one had cardiac arrest before left thoracotomy while undergoing exploratory laparotomy. Three patients were admitted in full cardiac arrest and underwent thoracotomy for resuscitation but died. Three additional patients in the current series had cardiac arrest on the way from the emergency center to the operating room. These patients could not be resuscitated and subsequently died. It is difficult to ascertain from review of their records whether they might have been saved by an emergency center thoracotomy. In all three cases, the patients were in the operating room at the moment of arrest, and all three had massive exsanguinating injuries to the central lung or pulmonary hilum.
Complications Postoperative complications in the 23 patients who survived operation for pulmonary resection or hilar repair included hemoptysis in two patients undergoing
Volume 95 Number 2 February 1988
Penetrating lung injuries
I87
Fig. 3. A, Posteroanterior chest x-ray film taken on admission of a 21-year-old woman shot at close range with a 0.357 magnum handgun. Note the large pulmonary contusion and hemothorax. She underwent thoracotomy for massive hemorrhage and left pneumonectomy for central pulmonary destruction and pulmonary artery and pulmonary vein injury. B, Posteroanterior chest x-ray film taken immediately after chest tube insertion in a 24-year-old man who was shot with a police 0.357 magnum handgun. The chest was explored for excessive bleeding, which necessitated wedge resection for control.
wedge resection that necessitated subsequent lobectomy, bronchopleural fistula in one patient undergoing lobectomy and in one patient undergoing hilar repair, air embolism in one patient, sepsis in two patients, respiratoryinsufficiency necessitating postoperative tracheostomy in two patients, adult respiratory distress syndrome in one patient, and a significant gastrointestinal tract hemorrhage in one patient. Seven of 23 patients had no postoperative complications.
Discussion Recent reports of military experience in the management of penetrating thoracic injuries have emphasized early thoracotomy in the treatment of critical or decompensating casualties.!" In one series thoracotomy was performed in 71% of 1992 patients and definitive tube thoracostomy in only 29%.1 Of particular interest was thefact that 16% of the total group of patients and 22% ofthe operated group of patients underwent pulmonary resection for lung injury. The survival rate was 99.3% in the thoracic injuries treated by the thoracostomy alone and 98% in those managed by primary thoracotomy. The conclusion was drawn that early resuscitation combined with aggressive thoracic surgical procedures were life-saving in this group of patients. Injuries sustained in conventional warfare are more likely to be lethal than injuries sustained in civilian practice. The American experience in Vietnam indicates that about 90% of the total mortality occurs on the battlefield." The mechanism of injury among fatalities
from Vietnam showed 52% of deaths were caused by bullet wounds and 44% by fragmentary devices. The Israeli experience in Lebanon indicates that the probability of death as a result of gunshot wounds was 31%, and the probability of death from fragmentary wounds was 13%.14 In contrast, the reported probability of death from gunshot wounds in a large civilian practice is approximately 18%.13 Civilian gunshot wounds are usually due to low-velocity, low-kinetic energy missiles, versus the high-velocity, high-energy projectile missiles used in military situations. In our civilian practice, 885 of 1168 patients (76%) were able to be treated with chest tubes as the sole management of penetrating chest injuries. Nine of 602 (1.5%) patients with stab wounds and 16 of the 283 (5.7%) patients with gunshot wounds treated by chest tube thoracostomy required thoracotomy and decortication for empyema at a later date. This experience is similar to most civilian series, wherein tube thoracostomy and supportive measures have been adequate treatment for penetrating thoracic injuries in approximately 70% to 85% of patients with penetrating chest wounds.>" In our current series, 37 patients with a stab wound and 10 patients with a gunshot wound with penetration of the thoracic cavity and small or no pneumothorax were able to be treated by observation alone. These patients were admitted to the hospital overnight for repeat chest x-ray examination at approximately 12 hours. Other centers have also noted that a significant percentage of patients can be safely observed
The Journal of
1 8 8 Robison et al.
for evidence of progression of pneumohemothorax and that an observation period of 6 to 8 hours seems optimal. 15. 16 Only 5.8% of the total group in our series required operative repair of lung injury and only 1.5% of the total group required lung resection. This is in contrast to the Lebanon military experience of Zakharia,' in which 16% of the total group of patients required lung resection. Additionally, 0.9% of our total group of patients had pulmonary hilar vascular injuries that could be repaired primarily during the operation without pulmonary resection. Lung resections were required in only 2.3% of gunshot wounds to the thorax. Wedge resections were performed because of tissue destruction or hemorrhage in two patients. Massive tissue destruction, hemorrhage, or hemoptysis was noted in four patients requiring lobectomy. Three patients requiring pneumonectomy also had central vascular injuries that were judged irreparable at operation. Pulmonary injury after stab wounds tended to be isolated to the tract of the stab wound. Resections, if required, were limited to simple wedge resections in eight of nine patients, usually for continued hemorrhage after an attempt at oversewing the lung injury. In one of nine patients having stab wounds who required pulmonary resection, massive hemoptysis necessitated lobectomy to prevent further soilage of the tracheobronchial tree with blood. Other major civilian trauma centers have had experience similar to ours. Siemens and associates? in Louisville reported on 190 patients with penetrating thoracic injury, noting that only 28% required operation, 58% of those with cardiac injuries. No pulmonary resections were performed although several pulmonary injuries were oversewn. Oparah and Mandal" reported that 21% of patients with penetrating chest wounds required operation, with a mortality of 5.9% for stab wounds and 15.7% for gunshot wounds and an overall mortality of 10.4%.6 No lung resections were reported. Adkins, Whiteneck, and Weltering" reported on 270 patients with penetrating chest wall and thoracic injuries. Sixtynine percent of their patients required chest tube thoracostomy alone. Massive lung injury was the cause of death in seven patients, and pulmonary hilar injuries were noted in three. Six additional patients in their series died in the operating room from uncontrolled bleeding. Severe pulmonary injuries were present in four of these patients and pulmonary hilar vessel injury in two. Kish and colleagues'? reviewed the case histories of 180 patients with penetrating chest trauma and noted that only 12% required early operation and 4% required late thoracotomy. In their series, four of the patients bled significantly from gunshot wounds to the lung, requiring either oversewing or wedge resection.
Thoracic and Cardiovascular Surgery
There is, however, evidence supporting lung resection in selected patients with high-velocity missile injury as a means of decreasing postoperative morbidity and mortality. McNamara and associates' reviewed the case histories of 547 patients with thoracic injuries from the Vietnam conflict, of whom 14% required thoracotomy. They noted that two patients treated with chest tube alone after a single high-velocity missile injury to the lung died of pulmonary insufficiency. In two patients undergoing operation for high-velocity missile injury to the thorax, contused lung was identified, but no resection was performed. These two patients had aspiration of "copious bloody secretions" into the uninvolved lung and died of bilateral pneumonias. A more aggressive surgical approach with earlier thoracotomy and resection of contused lung was instituted. Five additional patients with similar injuries were then treated with pulmonary resection to relieve hemoptysis and remove the area of pulmonary contusion; they did not have aspiration or other major complications. Earlier thoracotomy was credited with improved survival, and resection of severely contused lung tissue was associated with much less morbidity and mortality than traditional conservative therapy. Fischer, Geiger, and Guernsey'? noted that 10 patients who had incurred pulmonary contusion from high-velocity missile injury in Vietnam were treated conservatively and died of respiratory failure or sepsis. Later, a group of nine patients similarly injured underwent early resection of damaged lung tissue with a subsequent mortality of II %. In six of these nine patients the only indication for thoracotomy was for resection of the severely contused lung. In another review of military injuries Wanebo and van Dyke' suggested pulmonary resection to decrease postoperative complications in lung contused by high-velocity missiles. Of 534 patients treated for penetrating thoracic injuries, 104 required open thoracotomy and 84 survived. The mechanism of injury was equally divided between high-velocity gunshot wounds and fragment wounds from high-explosive rockets. Sixty patients had severe pulmonary injuries necessitating resection, with only 35 patients surviving operation. Pulmonary complications occurred in 67 of 84 patients surviving operation, with only 39 patients surviving pulmonary complications. In conclusion, in civilian practice chest tube thoracostomy should remain by far the most common method of treating penetrating injury to the thorax, with only 15% to 30% of patients requiring thoracotomy. Among those patients requiring thoracotomy for pulmonary injuries, most frequently the injuries can be handled by simple oversewing of the lung. Some patients, however, may require pulmonary resection because of hemorrhage, tissue destruction, or air embolus. Lobectomy is per-
Volume 95 Number 2
Penetrating lung injuries
February 1988
formed for massive tissue destruction isolated to a lobe or a bronchovenous communication that is not controllable by local techniques or for pulmonary hemorrhage that can not be controlled by simple oversewing. Pneumonectomy is reserved for central irreparable pulmonary parenchymal, vascular, or bronchial injuries. The decision whether to perform pulmonary resection is made at operation and depends on the location and extent of the injury. REFERENCES 1. Zakharia AT. Thoracic battle injuries in the Lebanon. War: review of the early operative approach in 1,992 patients. Ann Thorac Surg 1985;40:209-13. 2. McNamara JJ, Messersmith JK, Dunn RA, et al. Thoracic injuries in combat casualties in Vietnam. Ann Thorac Surg 1970; 10:389-401. 3. Suleman NO, Rasoul HA. War injuries of the chest. Injury 1985; 16:382-4. 4. Wanebo H, van Dyke J. The high-velocity pulmonary injury: relation to traumatic wet lung syndrome. J THORAC CARDIOVASC SURG 1972;64:537-50. 5. Mattox KL, Allen MK. Penetrating wounds of the thorax. Injury 1986; 17:313-7. 6. Oparah SS, MandaI AK. Operative management of penetra ting wounds of the chest in civilian practice: review of indications of 125 consecutive patients. J THORAC CARDIOVASC SURG 1979;77: 162-8. 7. Ozgen G, Duygulu I, Solak H. Chest injuries in civilian life and their treatment. Chest 1984;85:89-92. 8. Adkins RB Jr, Whiteneck JM, Woltering EA. Penetrating chest wall and thoracic injuries. Am Surg 1985; 51:140-8. 9. Siemens R, Polk HC Jr, Gray LA, et al. Indications for thoracotomy following penetrating thoracic injury. J Trauma 1977; 17:493-500. 10. Kish G, Kozloff L, Joseph WL, et al. Indications for early thoracotomy in the management of chest trauma. Ann Thorac Surg 1976;22:23-28. 11. Demetriades 0, Rabinowitz B, Markides N. Indications for thoracotomy in stab injuries of the chest: a prospective study of 543 patients. Br J Surg 1986;73:888-90. 12. Borlase BC, Metcalf RK, Moore EE, et al. Penetrating wounds to the anterior chest: analysis of thoracotomy and laparotomy. Am J Surg 1986;152:649-53. 13. Bellamy RF, Meningas PA, Vayer JS, et al. Epidemiology of trauma: military experience. Ann Emerg Med 1986;15:1384-8. 14. Doleb E. Medical services in the Lebanon War, overview. Isr J Med Science 1984;20:300-70. IS. Ammons MA, Moore EE, Rosen P. Role of the observation unit in the management of thoracic trauma. J Emerg Med 1986;4:279-82. 16. Muckart OJ. Delayed pneumothorax and hemothorax following observation for stab wounds of the chest. Injury 1985;16:247-8. 17. Fischer RP, Geiger JP, Guernsey JM. Pulmonary resec-
I89
tions for severe pulmonary contusions secondary to highvelocity missile wounds. J Trauma 1974;14:293-301.
Discussion Dr. Watts, R. Webb (New Orleans, La.). Our experience at Charity Hospital in New Orleans is somewhat similar except that the percentage of gunshot wounds is reversed: We have had about twice as many gunshot wounds in recent years as stab wounds. I think our percentage of thoracotomies is perhaps slightly higher, although virtually always our thoracotomies are done for other injuries and not just for the lung injuries alone. We do a very rare thoracotomy just for lung injuries. In the civilian injuries usually the damage is not very great. By contrast, in military injuries there is a great deal of destruction. Most of these wounds, except hilar injuries, can easily be repaired, and even a lobe that has been sliced wide open can be repaired and not resected because the recuperative powers of the lung are tremendous. It is usual to have restoration of a nice functional unit. I would like to emphasize also our use of the thoracoscopy. The thoracoscope has been of tremendous value ot us in examining the lung and in seeing if other structures are damaged. We can see whether the bleeding is from the lung, the hilus, or from an intercostal vessel or an internal mammary artery. Perhaps more significantly, one can see if the diaphragm has been penetrated and thus know whether there is a thoracoabdominal injury. Thoracoscopy has influenced our therapy in at least a third of the cases and so has been of tremendous importance to us. I would just emphasize again one final point, that we resect only those injuries of the lung in which there has been complete maceration or destruction of the lung, not just penetration or laceration. Dr. William Heydorn (San Francisco, Calif). I noted that Dr. Robison alluded to recent reports of military thoracic injuries, including the paper by Dr. Zakharia in The Annals of Thoracic Surgery (1985;40:209-13), which related an experience in Lebanon. I am not aware that Dr. Robison related to a paper regarding the experience of our own army. At a military meeting recently where Dr. Zakharia's experience was discussed, some of my colleagues, who were in Vietnam, indicated that the majority of our own injuries were treated in much the same fashion as was just described by Dr. Robison. In Lebanon, there was a special situation in which injured persons were seen very rapidly in hospitals that were very close to the action. The result was a different type of patient and a different type of injury. In our military today, we teach treatment of the patient rather than the weapon. We would approve of the approach that was brought out by Dr. Robison. Dr. Juro J. Wada (Tokyo, Japan). The majority of thoracic injuries can be treated by an improved valved thoracostomy tube, which often helps in establishing an accurate diagnosis and indications for emergency thoracotomy. If surgical treatment is found to be not indicated, the patient can be placed on a program of early ambulation. Emergency equipment should include a disposable valved thoracostomy tube (which can be connected to three-bottle wall suction if so needed). Dr. Richard A. Rasmussen (Grand Rapids, Mich.). I would like to encourage the authors to continue their work.
19 0
Robison et al.
Such work should be presented to thoracic surgeons, who should treat the injured person rather than delegate this responsibility to emergency room technicians, physicians, or other less capable persons. This review is of penetrating wounds of the chest that are due primarily to the use of knives or pistols-the Saturday night specials-and that occur in the bush country or mainly in our larger cities. These wounds are usually clean and small but are potentially very dangerous. Our civilian experience in Grand Rapids over a 40-year period, following a military experience during World War II in our navy, is less in numbers and violence but is nevertheless similar to that presented. Some of our cases were incident to automobile collisions, glass injuries, and even skiing injuries. I believe that the treatment program should provide for rapid assessment in the emergency room and not allow the patient's condition to deteriorate there too long. The primary problems are bleeding and respiratory tract obstruction, both of which require prompt correction.
The Journal of Thoracic and Cardiovascular Surgery
One must prepare for the operating room early in any doubtful case. Thoracentesis and chest tube drainage, of course, are initiated first. Exploratory thoracotomy of an adequate nature should be immediately available. I would stress also that there is opportunity for blood salvage and retransfusion. That is, the blood can be evacuated into a large basin and maintained until the injury can be assessed. Then the bleeding can be controlled by suturing of the lung or otherwise as needed. Autotransfusion today is becoming more important. The surgeon general only recently mentioned this in one of his press conferences in connection with the acquired immunodeficiency syndrome program. One's own blood is far superior, in larger cities particularly, to any transfused blood, and this possibility of use in thoracic injuries should be emphasized. Otherwise, I think our treatment program coincides essentially with that presented.
THoRAc CARDIOVASC SURG
1988;95: 184-90
Management of penetrating lung injuries in civilian practice Recent reports of military thoracic injuries have advocated early thoracotomy and aggressive management of pulmonary injuries with resection as opposed to the more conservative and traditional treatment with chest tube thoracostomy. A retrospective study was therefore performed to determine the incidence of thoracotomy and lung resection in civilian injuries and to evaluate the effectiveness of treatment of these injuries. Between 1973 and 1985, in a series of 1168 patients, there were 384 gunshot wounds and 784 stab wounds to the thorax. Two hundred eighty-three patients with a gunshot wound (74%) and 602 with a stab wound (77%) were treated with chest tubes alone. Sixty-eight patients (6% of the total) required operative repair of pulmonary hilar or parenchymal injury. Pulmonary resection was necessary in only 18 patients (nine with a gunshot wound and nine with a stab wound), and 10 patients had repair of hilar injuries (nine with a gunshot wound and one with a stab wound). Of patients requiring pulmonary resection, nine required wedge or segmental resection, six required lobectomy, and three patients required pneumonectomy. Mortality for aU thoracic injuries was 2.3 %: for those treated with chest tube alone, 0.7%; for pulmonary hilar injuries, 30 %; for pulmonary parenchymal injuries, 8.6 %; and for injuries necessitating lung resection, 28 %. Most civilian lung injuries can be treated by tube thoracostomy alone. Although relatively few patients with primary pulmoru.lry injury require thoracotomy, those that do are at significant risk and may require lung resection to control bleeding or hemoptysis or to remove destroyed or devitalized lung tissue.
Paul D. Robison, MD (by invitation), P. Kent Harman, MD" (by invitation),
J. Kent Trinkle, MD, and Frederick L. Grover, MD, San Antonio. Texas
Eetrating injuries to the chest in civilian practice result mainly from gunshot or stab wounds. Recent military reports have emphasized the need for early thoracotomy in patients having high-velocity missile and shrapnel injuries.!" In contradistinction, the experience of several large civilian centers continues to support the use of chest tube thoracostomy as the primary treatment of nonmediastinal chest injury with a low frequency of thoracotomy>" The following study was undertaken to analyze the frequency of thoracotomy and pulmonary
Table I. Hilar or lung repair/resection summary
From The University of Texas Health Science Center at San Antonio and the Audie Murphy Memorial Veterans Administration Hospital, San Antonio, Texas.
resection after penetrating nonmediastinal chest injury in our institution.
Read at the Sixty-seventh Annual Meeting of The American Association for Thoracic Surgery, Chicago, III., April 6-8, 1987.
Patients and methods
Address for reprints: Dr. Frederick L. Grover, Division of Cardiothoracic Surgery, The University of Texas Health Science Center, 7703 Floyd Curl Dr., San Antonio, TX 78284-7841. 'Present address: Northwest Surgical Associates, 1501 N.E. Medical Center Dr., Bend, OR 97701.
184
Gunshot wound In = 384)
Stab wound In = 784) No.
%
No.
%
Lung repair Lung resection Hilar repair
16 9 1
2 l.l 0.1
24 9 9
6.3 2.3 2.3
Total
26
3.3
42
10.7
From January 1973 through January 1986, 1168 patients were treated at Medical Center Hospital, San Antonio, Texas, for penetrating injuries to the thoracic cavity. Ninety-six percent of the patients were male, with an age range of 5 to 68 years and a mean age of 28.8 years. There were 784 stab wounds and 384 gunshot wounds in the series. A retrospective
Volume 95 Number 2
Penetrating lung injuries
February 1988
Pneumonectomy A-All 3 lobes PAl L-PA/PV (2) Total 18
Fig. I. Location of pulmonary injuries in patients requiring resection. PA. Pulmonary artery. PV. Pulmonary vein. review of the charts of these patients formed the basis of this report. Of those patients admitted, 37 had a stab wound and lO had a gunshot wound with minimal evidence of pneumothorax or hemothorax on chest roentgenogram; they were treated by observation alone. Additionally, II patients having a stab wound and two having a gunshot wound underwent creation of a subxiphoid pericardiaI window alone and had no chest tube placed. Eighty-three patients with a stab wound and 34 patients with a gunshot wound underwent exploration for mediastinal injury, and no evidence of significant lung injury was found. Twenty-five patients had injury to a chest wall artery or subclavian artery as a result of a stab wound and underwent thoracotomy for repair of those arterial injuries. Patients treated with observation alone, pericardial window, thoracotomy for subclavian artery injury, or median sternotomy for mediastinal injuries without lung injury are excluded from further consideration. Twenty-six of 784 patients with a stab wound (3.3%) and 42 of 384 patients with a gunshot wound (10.7%) had pulmonary injuries necessitating operation (Table I). Sixteen of the 26 stab wounds and 24 of 42 gunshot wounds necessitated only oversewingof the injury. Nine of the 26 (34.6%) patients with a stab wound who underwent operation required pulmonary resection; seven had wedge resection and two had lobectomy. Nine of 42 (21.4%) patients with a gunshot wound underwent lung resection: Two had wedge resection, four lobectomy, and three pneumonectomy (Fig. I). One patient with a stab wound and nine patients with a gunshot wound required repair of hilar vessels (Fig. 2). Additionally, three patients arrived at the emergency center in full cardiac arrest. These patients underwent thoracotomy in the emergency room for resuscitation, which was unsuccessful. At postmortem examination, two of these patients were found to have pulmonary hilar injuries with massive exsanguination, and the third patient was found to have a pulmonary parenchymal injury with massive exsanguination. Three of 28 patients who underwent pulmonary resection or hilar repair had associated cardiac injuries. Two patients had a subclavian artery injury, two patients were paraplegic postoperatively from bullet injuries to the spinal cord, and two patients had liver injuries. Additionally, one patient had injuries to the pancreas, colon, spleen, and stomach, and one patient had injury to the brachial plexus.
185
Hilar Repair-10 Pneumonectomy-3 1 PA/PV-Right 2 PA/PV-Le
Fig. 2. Location of hilar injuries in patients requiring repair or pneumonectomy. PA. Pulmonary artery. Pl/, Pulmonary vein. Clinical presentation. Patients requmng operation were frequently in shock and cardiac arrest because of hypovolemia. The hemodynamic status of patients requiring operation for primarily lung injury or chest wall arterial injury is shown in Table II. Additionally, three patients with a gunshot wound and one patient with a stab wound to the lung had evidence of cardiac tamponade and were found to have a cardiac injury during the operation. Significantly, three of 18 patients requiring lung resection had massive hemoptysis on admission. Of six patients requiring lobectomy, three had massive hemoptysis. Two of these six patients required a second operation after an inadequate first operation to control pulmonary hemorrhage. Diagnostic and therapeutic procedures. Twenty of 28 patients requiring pulmonary resection or hilar repair had a chest x-ray film taken preoperatively. Eight of 28 patients were in hemodynamically unstable condition and were rushed to the operating room or underwent thoracotomy in the emergency room before the chest x-ray film was made. The most common x-ray finding in the remaining 20 patients was hemothorax or hemopneumothorax, which was present in 12 of 20. Two patients had bilateral hemopneumothoraces and one patient a simple pneumothorax. Additionally, pulmonary contusions and hemothorax were diagnosed preoperatively in two patients (Fig. 3). Two patients had a missile in the mediastinum overlying the heart shadow, and one patient had an enlarged cardiac silhouette. Additional diagnostic and therapeutic procedures on the 28 patients undergoing pulmonary resection or hilar repair included central venous pressure lines in 10, diagnostic peritoneal lavage in three, intravenous pyelogram in three, exploratory laparotomy in two, pericardial window in two, and barium swallow and thoracic myelogram each in one patient. Blood loss after placement of the initial chest tube in patients undergoing operation for nonmediastinal chest injuries are shown in Table III. Significantly, chest tube drainage was not useful as a predictor of severity of injury in those patients requiring operation, nor was it an indicator of organ systems injured in this study, although patients with massive
The Journal of Thoracic and Cardiovascular Surgery
1 8 6 Robison et af.
Table II. Hemodynamic presentation of patients requiring operation for chest wall arterial, lung, or hilar injury Without lung resection
BP > 70 mm Hg systolic BP < 70 mm Hg systolic Arrest
With lung resection
Hilar repair
SW ICA/IMA
SW lung
GSW lung
SW
GSW
SW
GSW
14 8 I
9 6 I
12 8 4
5 2 2
4 4 I
0
3 2 4
I
0
Legend: SW, Stab wound. GSW, Gunshot wound. ICA, Intercostal artery. IMA, Internal mammary artery.
Table
m. Patients requiring operation for Repair
ICA or IMA alone Pneumonorrhaphy and ICA/IMA Pneumonorrhaphy Resection Wedge Lobectomy Pneumonectomy Hilar repair
noncardiac chest injuries: Initial chest tube-blood loss SWNo.
Blood loss (ml ± SEM)
GSW No.
Blood loss (ml ± SEM)
23 9 5
1995 ± 223 1633 ± 21 I 1650 ± 217
0 5 13
1500 ± 530 2122 ± 312
7 2 0 I
2257 ± 250 1175 ± 852
2 2 3 8
1400 Massive Massive 2688 ± 550
2000
Legend: SW, Stab wound. GSW, Gunshot wound. ICA, Intercostal artery. IMA, Internal mammary artery.
bleeding often had hilar injuries, Some chest wall injuries, however, also were associated with large blood loss. Indications for operation. The most common indication for operation in patients with pulmonary injury necessitating repair was continued hemorrhage or massive hemorrhageafter placement of the chest tube, which occurred in 43 of 68 patients, Blood lossin excess of 1500 ml after initial placement of the chest tube or continuing blood lossof more than 150 to 200 ml/hr after placement of the initial chest tube was an indication for operation, Twenty-three patients had shock that was not reversible by aggressive fluid replacement. Additionally, 12 patients in this category were in cardiac arrest on admission or shortly thereafter. Three patients had massive hemoptysis, four patients had evidence of cardiac tamponade, and four patients had missiles overlying the mediastinum suggestive of cardiac injury, The decision to proceed with pulmonary resection was made at the time of thoracotomy, In 14 of 18 patients continued hemorrhage was the primary indication for pulmonary resection. Four patients had severe parenchymal destruction, three had hemoptysis, and two patients had central pulmonary injuries, Several patients had more than one indication for lung resection.
Results Mortality from all causes of penetrating thoracic injury included 10 of 784 patients with stab wounds and 18 of 384 patients with gunshot wounds, or 2.4% of the total patient population. One patient of the 885 patients who underwent chest tube thoracostomy died of a pleural infection. Seven other patients who were treated primarily with chest tube thoracostomies died of associated injuries. Of the 40 patients who underwent over-
sewing of lung injuries during the operation, none died of a pulmonary cause. Three patients of the 40, however, died of associated injuries. Of 18 patients who underwent lung resection, five died, two of air embolus, two of exsanguination, and one of sepsis from pulmonary and intra-abdominal sources. Three of 10 patients undergoing hilar repair died, all of exsanguination. Six of seven patients who had cardiac arrest before the operation died. In addition, two of nine patients operated on in shock died; one had cardiac arrest before left thoracotomy while undergoing exploratory laparotomy. Three patients were admitted in full cardiac arrest and underwent thoracotomy for resuscitation but died. Three additional patients in the current series had cardiac arrest on the way from the emergency center to the operating room. These patients could not be resuscitated and subsequently died. It is difficult to ascertain from review of their records whether they might have been saved by an emergency center thoracotomy. In all three cases, the patients were in the operating room at the moment of arrest, and all three had massive exsanguinating injuries to the central lung or pulmonary hilum.
Complications Postoperative complications in the 23 patients who survived operation for pulmonary resection or hilar repair included hemoptysis in two patients undergoing
Volume 95 Number 2 February 1988
Penetrating lung injuries
I87
Fig. 3. A, Posteroanterior chest x-ray film taken on admission of a 21-year-old woman shot at close range with a 0.357 magnum handgun. Note the large pulmonary contusion and hemothorax. She underwent thoracotomy for massive hemorrhage and left pneumonectomy for central pulmonary destruction and pulmonary artery and pulmonary vein injury. B, Posteroanterior chest x-ray film taken immediately after chest tube insertion in a 24-year-old man who was shot with a police 0.357 magnum handgun. The chest was explored for excessive bleeding, which necessitated wedge resection for control.
wedge resection that necessitated subsequent lobectomy, bronchopleural fistula in one patient undergoing lobectomy and in one patient undergoing hilar repair, air embolism in one patient, sepsis in two patients, respiratoryinsufficiency necessitating postoperative tracheostomy in two patients, adult respiratory distress syndrome in one patient, and a significant gastrointestinal tract hemorrhage in one patient. Seven of 23 patients had no postoperative complications.
Discussion Recent reports of military experience in the management of penetrating thoracic injuries have emphasized early thoracotomy in the treatment of critical or decompensating casualties.!" In one series thoracotomy was performed in 71% of 1992 patients and definitive tube thoracostomy in only 29%.1 Of particular interest was thefact that 16% of the total group of patients and 22% ofthe operated group of patients underwent pulmonary resection for lung injury. The survival rate was 99.3% in the thoracic injuries treated by the thoracostomy alone and 98% in those managed by primary thoracotomy. The conclusion was drawn that early resuscitation combined with aggressive thoracic surgical procedures were life-saving in this group of patients. Injuries sustained in conventional warfare are more likely to be lethal than injuries sustained in civilian practice. The American experience in Vietnam indicates that about 90% of the total mortality occurs on the battlefield." The mechanism of injury among fatalities
from Vietnam showed 52% of deaths were caused by bullet wounds and 44% by fragmentary devices. The Israeli experience in Lebanon indicates that the probability of death as a result of gunshot wounds was 31%, and the probability of death from fragmentary wounds was 13%.14 In contrast, the reported probability of death from gunshot wounds in a large civilian practice is approximately 18%.13 Civilian gunshot wounds are usually due to low-velocity, low-kinetic energy missiles, versus the high-velocity, high-energy projectile missiles used in military situations. In our civilian practice, 885 of 1168 patients (76%) were able to be treated with chest tubes as the sole management of penetrating chest injuries. Nine of 602 (1.5%) patients with stab wounds and 16 of the 283 (5.7%) patients with gunshot wounds treated by chest tube thoracostomy required thoracotomy and decortication for empyema at a later date. This experience is similar to most civilian series, wherein tube thoracostomy and supportive measures have been adequate treatment for penetrating thoracic injuries in approximately 70% to 85% of patients with penetrating chest wounds.>" In our current series, 37 patients with a stab wound and 10 patients with a gunshot wound with penetration of the thoracic cavity and small or no pneumothorax were able to be treated by observation alone. These patients were admitted to the hospital overnight for repeat chest x-ray examination at approximately 12 hours. Other centers have also noted that a significant percentage of patients can be safely observed
The Journal of
1 8 8 Robison et al.
for evidence of progression of pneumohemothorax and that an observation period of 6 to 8 hours seems optimal. 15. 16 Only 5.8% of the total group in our series required operative repair of lung injury and only 1.5% of the total group required lung resection. This is in contrast to the Lebanon military experience of Zakharia,' in which 16% of the total group of patients required lung resection. Additionally, 0.9% of our total group of patients had pulmonary hilar vascular injuries that could be repaired primarily during the operation without pulmonary resection. Lung resections were required in only 2.3% of gunshot wounds to the thorax. Wedge resections were performed because of tissue destruction or hemorrhage in two patients. Massive tissue destruction, hemorrhage, or hemoptysis was noted in four patients requiring lobectomy. Three patients requiring pneumonectomy also had central vascular injuries that were judged irreparable at operation. Pulmonary injury after stab wounds tended to be isolated to the tract of the stab wound. Resections, if required, were limited to simple wedge resections in eight of nine patients, usually for continued hemorrhage after an attempt at oversewing the lung injury. In one of nine patients having stab wounds who required pulmonary resection, massive hemoptysis necessitated lobectomy to prevent further soilage of the tracheobronchial tree with blood. Other major civilian trauma centers have had experience similar to ours. Siemens and associates? in Louisville reported on 190 patients with penetrating thoracic injury, noting that only 28% required operation, 58% of those with cardiac injuries. No pulmonary resections were performed although several pulmonary injuries were oversewn. Oparah and Mandal" reported that 21% of patients with penetrating chest wounds required operation, with a mortality of 5.9% for stab wounds and 15.7% for gunshot wounds and an overall mortality of 10.4%.6 No lung resections were reported. Adkins, Whiteneck, and Weltering" reported on 270 patients with penetrating chest wall and thoracic injuries. Sixtynine percent of their patients required chest tube thoracostomy alone. Massive lung injury was the cause of death in seven patients, and pulmonary hilar injuries were noted in three. Six additional patients in their series died in the operating room from uncontrolled bleeding. Severe pulmonary injuries were present in four of these patients and pulmonary hilar vessel injury in two. Kish and colleagues'? reviewed the case histories of 180 patients with penetrating chest trauma and noted that only 12% required early operation and 4% required late thoracotomy. In their series, four of the patients bled significantly from gunshot wounds to the lung, requiring either oversewing or wedge resection.
Thoracic and Cardiovascular Surgery
There is, however, evidence supporting lung resection in selected patients with high-velocity missile injury as a means of decreasing postoperative morbidity and mortality. McNamara and associates' reviewed the case histories of 547 patients with thoracic injuries from the Vietnam conflict, of whom 14% required thoracotomy. They noted that two patients treated with chest tube alone after a single high-velocity missile injury to the lung died of pulmonary insufficiency. In two patients undergoing operation for high-velocity missile injury to the thorax, contused lung was identified, but no resection was performed. These two patients had aspiration of "copious bloody secretions" into the uninvolved lung and died of bilateral pneumonias. A more aggressive surgical approach with earlier thoracotomy and resection of contused lung was instituted. Five additional patients with similar injuries were then treated with pulmonary resection to relieve hemoptysis and remove the area of pulmonary contusion; they did not have aspiration or other major complications. Earlier thoracotomy was credited with improved survival, and resection of severely contused lung tissue was associated with much less morbidity and mortality than traditional conservative therapy. Fischer, Geiger, and Guernsey'? noted that 10 patients who had incurred pulmonary contusion from high-velocity missile injury in Vietnam were treated conservatively and died of respiratory failure or sepsis. Later, a group of nine patients similarly injured underwent early resection of damaged lung tissue with a subsequent mortality of II %. In six of these nine patients the only indication for thoracotomy was for resection of the severely contused lung. In another review of military injuries Wanebo and van Dyke' suggested pulmonary resection to decrease postoperative complications in lung contused by high-velocity missiles. Of 534 patients treated for penetrating thoracic injuries, 104 required open thoracotomy and 84 survived. The mechanism of injury was equally divided between high-velocity gunshot wounds and fragment wounds from high-explosive rockets. Sixty patients had severe pulmonary injuries necessitating resection, with only 35 patients surviving operation. Pulmonary complications occurred in 67 of 84 patients surviving operation, with only 39 patients surviving pulmonary complications. In conclusion, in civilian practice chest tube thoracostomy should remain by far the most common method of treating penetrating injury to the thorax, with only 15% to 30% of patients requiring thoracotomy. Among those patients requiring thoracotomy for pulmonary injuries, most frequently the injuries can be handled by simple oversewing of the lung. Some patients, however, may require pulmonary resection because of hemorrhage, tissue destruction, or air embolus. Lobectomy is per-
Volume 95 Number 2
Penetrating lung injuries
February 1988
formed for massive tissue destruction isolated to a lobe or a bronchovenous communication that is not controllable by local techniques or for pulmonary hemorrhage that can not be controlled by simple oversewing. Pneumonectomy is reserved for central irreparable pulmonary parenchymal, vascular, or bronchial injuries. The decision whether to perform pulmonary resection is made at operation and depends on the location and extent of the injury. REFERENCES 1. Zakharia AT. Thoracic battle injuries in the Lebanon. War: review of the early operative approach in 1,992 patients. Ann Thorac Surg 1985;40:209-13. 2. McNamara JJ, Messersmith JK, Dunn RA, et al. Thoracic injuries in combat casualties in Vietnam. Ann Thorac Surg 1970; 10:389-401. 3. Suleman NO, Rasoul HA. War injuries of the chest. Injury 1985; 16:382-4. 4. Wanebo H, van Dyke J. The high-velocity pulmonary injury: relation to traumatic wet lung syndrome. J THORAC CARDIOVASC SURG 1972;64:537-50. 5. Mattox KL, Allen MK. Penetrating wounds of the thorax. Injury 1986; 17:313-7. 6. Oparah SS, MandaI AK. Operative management of penetra ting wounds of the chest in civilian practice: review of indications of 125 consecutive patients. J THORAC CARDIOVASC SURG 1979;77: 162-8. 7. Ozgen G, Duygulu I, Solak H. Chest injuries in civilian life and their treatment. Chest 1984;85:89-92. 8. Adkins RB Jr, Whiteneck JM, Woltering EA. Penetrating chest wall and thoracic injuries. Am Surg 1985; 51:140-8. 9. Siemens R, Polk HC Jr, Gray LA, et al. Indications for thoracotomy following penetrating thoracic injury. J Trauma 1977; 17:493-500. 10. Kish G, Kozloff L, Joseph WL, et al. Indications for early thoracotomy in the management of chest trauma. Ann Thorac Surg 1976;22:23-28. 11. Demetriades 0, Rabinowitz B, Markides N. Indications for thoracotomy in stab injuries of the chest: a prospective study of 543 patients. Br J Surg 1986;73:888-90. 12. Borlase BC, Metcalf RK, Moore EE, et al. Penetrating wounds to the anterior chest: analysis of thoracotomy and laparotomy. Am J Surg 1986;152:649-53. 13. Bellamy RF, Meningas PA, Vayer JS, et al. Epidemiology of trauma: military experience. Ann Emerg Med 1986;15:1384-8. 14. Doleb E. Medical services in the Lebanon War, overview. Isr J Med Science 1984;20:300-70. IS. Ammons MA, Moore EE, Rosen P. Role of the observation unit in the management of thoracic trauma. J Emerg Med 1986;4:279-82. 16. Muckart OJ. Delayed pneumothorax and hemothorax following observation for stab wounds of the chest. Injury 1985;16:247-8. 17. Fischer RP, Geiger JP, Guernsey JM. Pulmonary resec-
I89
tions for severe pulmonary contusions secondary to highvelocity missile wounds. J Trauma 1974;14:293-301.
Discussion Dr. Watts, R. Webb (New Orleans, La.). Our experience at Charity Hospital in New Orleans is somewhat similar except that the percentage of gunshot wounds is reversed: We have had about twice as many gunshot wounds in recent years as stab wounds. I think our percentage of thoracotomies is perhaps slightly higher, although virtually always our thoracotomies are done for other injuries and not just for the lung injuries alone. We do a very rare thoracotomy just for lung injuries. In the civilian injuries usually the damage is not very great. By contrast, in military injuries there is a great deal of destruction. Most of these wounds, except hilar injuries, can easily be repaired, and even a lobe that has been sliced wide open can be repaired and not resected because the recuperative powers of the lung are tremendous. It is usual to have restoration of a nice functional unit. I would like to emphasize also our use of the thoracoscopy. The thoracoscope has been of tremendous value ot us in examining the lung and in seeing if other structures are damaged. We can see whether the bleeding is from the lung, the hilus, or from an intercostal vessel or an internal mammary artery. Perhaps more significantly, one can see if the diaphragm has been penetrated and thus know whether there is a thoracoabdominal injury. Thoracoscopy has influenced our therapy in at least a third of the cases and so has been of tremendous importance to us. I would just emphasize again one final point, that we resect only those injuries of the lung in which there has been complete maceration or destruction of the lung, not just penetration or laceration. Dr. William Heydorn (San Francisco, Calif). I noted that Dr. Robison alluded to recent reports of military thoracic injuries, including the paper by Dr. Zakharia in The Annals of Thoracic Surgery (1985;40:209-13), which related an experience in Lebanon. I am not aware that Dr. Robison related to a paper regarding the experience of our own army. At a military meeting recently where Dr. Zakharia's experience was discussed, some of my colleagues, who were in Vietnam, indicated that the majority of our own injuries were treated in much the same fashion as was just described by Dr. Robison. In Lebanon, there was a special situation in which injured persons were seen very rapidly in hospitals that were very close to the action. The result was a different type of patient and a different type of injury. In our military today, we teach treatment of the patient rather than the weapon. We would approve of the approach that was brought out by Dr. Robison. Dr. Juro J. Wada (Tokyo, Japan). The majority of thoracic injuries can be treated by an improved valved thoracostomy tube, which often helps in establishing an accurate diagnosis and indications for emergency thoracotomy. If surgical treatment is found to be not indicated, the patient can be placed on a program of early ambulation. Emergency equipment should include a disposable valved thoracostomy tube (which can be connected to three-bottle wall suction if so needed). Dr. Richard A. Rasmussen (Grand Rapids, Mich.). I would like to encourage the authors to continue their work.
19 0
Robison et al.
Such work should be presented to thoracic surgeons, who should treat the injured person rather than delegate this responsibility to emergency room technicians, physicians, or other less capable persons. This review is of penetrating wounds of the chest that are due primarily to the use of knives or pistols-the Saturday night specials-and that occur in the bush country or mainly in our larger cities. These wounds are usually clean and small but are potentially very dangerous. Our civilian experience in Grand Rapids over a 40-year period, following a military experience during World War II in our navy, is less in numbers and violence but is nevertheless similar to that presented. Some of our cases were incident to automobile collisions, glass injuries, and even skiing injuries. I believe that the treatment program should provide for rapid assessment in the emergency room and not allow the patient's condition to deteriorate there too long. The primary problems are bleeding and respiratory tract obstruction, both of which require prompt correction.
The Journal of Thoracic and Cardiovascular Surgery
One must prepare for the operating room early in any doubtful case. Thoracentesis and chest tube drainage, of course, are initiated first. Exploratory thoracotomy of an adequate nature should be immediately available. I would stress also that there is opportunity for blood salvage and retransfusion. That is, the blood can be evacuated into a large basin and maintained until the injury can be assessed. Then the bleeding can be controlled by suturing of the lung or otherwise as needed. Autotransfusion today is becoming more important. The surgeon general only recently mentioned this in one of his press conferences in connection with the acquired immunodeficiency syndrome program. One's own blood is far superior, in larger cities particularly, to any transfused blood, and this possibility of use in thoracic injuries should be emphasized. Otherwise, I think our treatment program coincides essentially with that presented.