U.S. Army Medical Department, Office of Medical History
Skip Navigation, go to content







AMEDD MEDAL OF HONOR RECIPIENTS External Link, Opens in New Window






Chapter V



Wet Lung

Thomas H. Burford, M.D.


During the first year of fighting in the North African Theater of Operations, U.S. Army, it gradually became apparent that casualties with fluid in the lungs, in contrast to those whose lungs were dry, presented many difficulties of management. They required strenuous efforts at resuscitation if they were in shock, and they did not respond well to them. They presented poor surgical risks when emergency surgery was necessary. They were prone to develop postoperative pulmonary complications, especially if they had coughed up large amounts of blood before operation. Fatalities were proportionately more frequent among them than among casualties with similar wounds whose lungs did not contain fluid. When these patients came to autopsy in forward hospitals, pulmonary edema was the most prominent finding.

Although most surgeons in the theater were aware of the phenomena just described, the extreme seriousness of the problem was not immediately realized. There was nothing in the civilian experience to call attention to pulmonary edema as an immediate consequence of thoracic injuries.1 As a result, more than a year was to pass after the onset of fighting in North Africa before the concept of what came to be known as wet lung was developed and appropriate methods of treatment were devised.

The term "wet lung" and the concept which underlay it were first presented by Maj. Thomas H. Burford, MC, and Maj. Benjamin Burbank, MC, at a meeting of Fifth U.S. Army surgeons at the 38th Evacuation Hospital at Riardo in February 1944. At the same time, they discussed its etiology and outlined the principles of prevention and correction.

In their first presentation on this condition, Major Burford and Major Burbank described wet lung as a more or less specific response of the lung and chest wall to trauma. They postulated the operation of a somatovisceral reflex originating in the chest wall and mediating alveolar and bronchiolar

1In 1937, Betts and Overholt (1) described a syndrome which sometimes occurred from 24 to 48 hours after thoracic surgery, usually at night, and which consisted of slight cyanosis, noisy respirations, depression of the cough reflex, and a shocklike state. No matter how hard the patient tried, he could not raise the secretions whose retention was producing anoxia and anoxemia. The only way to break the cycle was to remove the retained secretions, for which bronchoscopic aspiration was necessary. In a number of respects, this syndrome seems to resemble the wet lung encountered in combat-incurred injuries.


responses. They particularly emphasized the part played by the painful chest wall in the pathogenesis of wet lung. It was the chief explanation of the patient's inability to raise the secretions formed in excess as the result of trauma. The immobilization of the chest wall made effectual coughing impossible, and the result was the saturation of the tracheobronchial tree with secretions which, under normal circumstances, would have been coughed up.

Major Burford and Major Burbank further suggested that a definite one-two relation probably existed between the wet lung of trauma and the massive pulmonary collapse which was so common in World War I. Their theory was that the wet phase, which was not recognized, precipitated the second phase, in which massive collapse occurred (p. 210).

The concept of wet lung, as presented by Major Burford and Major Burbank, was precise and specific. Later, Maj. Lyman A. Brewer III, MC, and other chest surgeons in the Mediterranean Theater of Operations, U.S. Army, expanded the original concept into a looser, broader concept of traumatic wet lung. In this later concept, the term "wet lung" was used for all forms of retained tracheobronchial fluid resulting directly or indirectly from trauma.

Recognition of the importance of wet lung in no wise detracted from the importance of the primary lesion or of associated injuries. On the contrary, it permitted a more rational approach to the entire problem of combat-incurred chest injuries. The important considerations stressed by Major Burford and Major Burbank were as follows:

l. The chest reacts to trauma in a way peculiar to itself, whether the injury is a slight contusion of the chest wall or a severe penetrating or perforating injury.

2. The reaction caused by trauma profoundly influences the whole pathologic process, the therapy applied, and the prognosis.


The pathogenesis of wet lung can be understood only if the normal physiology of the respiratory tract is borne in mind. In addition to its function as an airway, the tracheobronchial tree has a definite secretory function, combined with the capacity to rid itself of its own secretions and of aspirated material or other fluid substances which do not normally belong within its lumen. Evacuation of the tracheobronchial tree is accomplished (1) involuntarily, by action of the cilia and movement of the bronchial musculature, and (2) voluntarily, by coughing.

In spite of the extensive experience of World War II, little is yet understood concerning the effect of trauma to the chest on the function of the bronchial cilia and on other factors influencing the passage of secretions through the tracheobronchial tree. The presumption seemed warranted that after trauma, these functions became depressed or were perhaps entirely interrupted. Clinically, it was clear that the normal cough mechanism was


seriously impaired in the presence of moderate and severe trauma, and sometimes of slight trauma.

The lung, which is a specialized organ, reacts to trauma in just as specialized a manner as does the brain or any other specialized organ. In combat casualties with chest injuries, the pulmonary reaction took two forms:

1. The production of an excess of secretions or, in the broader concept of Major Brewer and his associates, of transudates and extravasated blood also.

2. The development of conditions which prevented the normal and adequate elimination of these fluids.

At present, there is insufficient evidence to permit formulation of a precise explanation for this dual process. Experimentally, de Takats and his associates (2) demonstrated that any appreciable trauma to the chest wall was followed almost immediately, in 60 percent of the animals, by widespread bronchial spasm and increased bronchial secretions. Bronchial spasm of the same degree could also be produced by stimuli applied within the abdomen, such as traction on the cystic duct or the mesentery.

It would be tempting to apply these experimental results to combat casualties and to postulate the existence of reflex bronchial and bronchiolar spasm originating after trauma as the result of a painful chest wall or pleura, but there are two reasons why the transfer would not be justified:

1. The whole case for bronchospasm rests on evidence that is still far from acceptable.

2. The hypothesis of bronchial spasm would explain only part of the phenomena observed in traumatic wet lung.

Even though it was not possible to explain the initial effect of thoracic trauma on the function of the tracheobronchial tree, the course of events afterward was quite clear:

1. The presence in the tracheobronchial tree of mucoid secretions in far more than normal amounts, as well as of transudates, blood, and aspirated material caused a mechanical obstruction and prevented adequate oxygenation of the alveoli.

2. With the suppression of the normal cough reflex, the tracheobronchial tree became increasingly less capable of eliminating secretions, and more fluid accumulated in it.

3. As these conditions continued, less oxygen became available to the pulmonary capillaries. It could be assumed from the work of Drinker and Warren (3) that anoxia increased capillary permeability and permitted more plasma to be lost into the tissues. Increased amounts of fluid extravasated from the pulmonary capillaries into the alveoli further decreased the oxygen available for the bloodstream.

4. The plugging of branch bronchi with secretions prevented the exchange of gases in the lobules supplied by these bronchial divisions, and lobular atelectasis resulted.


5. Plugging of a branch bronchus to a lobe or of a stem bronchus to a lung increased the anoxic anoxia that was the result of partial obstruction, for two reasons: (1) The alveolar bed available for the absorption of oxygen was far smaller than normal, and (2) sooner or later, a mediastinal shift resulted.

When blood was present in the tracheobronchial tree as the result of trauma, it could, in itself, cause bronchial obstruction. It also had another effect, that the irritation of the bronchial mucosa that it caused was responsible, in turn, for an increase of tracheobronchial secretions.

When wet lung appeared soon after wounding, its most potent effect was the promotion of anoxia. When it persisted or was allowed to progress without treatment to the point just described, it paved the way for subsequent pulmonary complications. A patient who was not too severely wounded and in whom other conditions were favorable might recover from his initial shock without complete removal of the fluid that was producing obstruction of the tracheobronchial tree. In such an instance, however, the causative factors were unchanged, and the stage was set for the development of tracheobronchitis, atelectasis, pneumonitis, and pneumonia. Wet lung sometimes seemed to be the chief factor upon which the outcome of an injury hinged. Failure to take active measures to combat it could lead to chronic pulmonary invalidism, and, before this complication was clearly understood, more than one patient with an uncontrolled wet lung died from this cause alone.

It was of the greatest importance that anoxia be relieved promptly. If it was not, serious changes occurred in other organs, particularly in the central nervous system, which is highly sensitive to an oxygen deficit. Severe cerebral anoxia, even if it persisted for only a short time, could result in serious cortical impairment. The psychotic manifestations and coma observed in some patients with chest injuries could reasonably be attributed to this cause.

There was clear clinical evidence of the relation of the pathophysiologic disturbances produced by wet lung to the patient's general status. In numerous instances, shock which had not responded to standard replacement therapy and other resuscitative measures, or which had become deeper as they were in progress, cleared up quickly, and even dramatically, when appropriate measures for the relief of wet lung were instituted.

As clinical experience accumulated, the theory seemed reasonable that a definite relation might exist between the wet lung of trauma as it was recognized in World War II and the massive pulmonary collapse described by Pasteur (4) in 1914 and described later, under the name of acute massive atelectasis, by Churchill (5). It was an attractive hypothesis to conceive of wet lung as the antecedent stage, and in a sense the precipitating cause, of silent massive pulmonary collapse.

Such an assumption would explain the extreme rarity of massive pulmonary collapse in the Mediterranean theater, where surgeons became increasingly alert to the recognition of wet lung and increasingly vigorous in its


treatment. Minor degrees of atelectasis were frequently observed in cases of wet lung which were inadequately treated.


The events just described were most often observed after chest injuries. Wet lung, however, was a possibility in other injuries, including thoracoabdominal injuries, abdominal injuries, and head injuries, particularly those that were associated with coma and unconsciousness.

Usually, though by no means always, the degree of wetness depended upon the type and severity of the primary lesion, the size and velocity of the missile, and whether or not the missile had struck a portion of the bony cage or only the soft tissues. The size of the intrapulmonary vessels that were damaged had a bearing on the amount of blood extravasated. Wet lung was present to some degree in practically all cases of thoracic trauma, though it might be transitory or so slight as to be almost unrecognizable clinically even when it was carefully searched for. Sometimes, too, the primary thoracic lesion was of such severity that it masked the underlying physiopathologic substratum. In many cases, on the other hand, wet lung was present to a far more severe degree than might have been expected from the relative mildness of the initial wounds.

A number of other factors played a part in the causation of wet lung:


Pain was an almost invariable component of chest trauma. Its exact influence in the production of wet lung was never fully clarified, but there was no doubt that it played an important part in both its inception and its progress. It had at least three harmful effects:

1. It induced a shallow type of respiration, for protective reasons.

2. The natural reaction to pain was voluntary splinting of the affected area.

3. The affected hemithorax therefore moved less than the unaffected side, with a resultant decrease in the movement of air back and forth in the bronchi. A decrease in the tidal respiration was probably of more importance than was at first recognized, as it lessened the amount of secretion that might be disposed of by evaporation.

Suppression of the Cough Reflex

The normal voluntary cough mechanism consists of four steps:

1. The air is drawn voluntarily into the lungs by raising the ribs and lowering the diaphragm.

2. The breath is held momentarily by closing the glottis.


3. The abdominal muscles are tensed, the intercostal spaces are narrowed, and the diaphragm is relaxed to provide explosive force.

4. The glottis is then suddenly opened, and the bechic blast required to raise the sputum is thus released.

A number of conditions present in chest trauma prevented the succession of steps just listed, all of which are necessary for effective coughing. Pain arising in either the chest wall or the pleura prevented both expansion and forceful contraction of the thorax. Rib fractures were frequently associated with severe pain when either coughing or deep breathing was attempted. If there were multiple fractures of the ribs or of the sternum, with a resulting flail chest, the paradoxical movement of the chest wall made effective coughing impossible. On inspiration, it moved outward. Attempts at coughing simply aggravated these paradoxical movements.

Wounds that caused abdominal pain, whether they were located in the abdomen or in the chest, prevented the tensing of the abdominal muscles necessary for an effective cough. The pistonlike action of the diaphragm, which is a part of the act of coughing, was impossible in a diaphragmatic injury. Ileus, acute dilatation of the stomach, and the presence of free fluid in the peritoneal cavity also had a more or less inhibitory action on the cough mechanism.

Sedation and Anesthesia

Oversedation from the too-liberal use of morphine was a fairly frequent contributing cause of wet lung. As their experience increased, shock officers in field and evacuation hospitals learned to be constantly on their guard against giving morphine to freshly wounded casualties. Often the patients did not need it at all. They were simply apprehensive as the result of anoxia and disturbed and uncomfortable from the ambulance ride. When morphine was really needed, it was given intravenously, not subcutaneously, in -gr. or at most -gr. doses.

Early in the war, as pointed out elsewhere (vol. I), and much less often later in the war, casualties with fractured ribs, flail chests, and other painful thoracic and abdominal wounds often arrived in the shock wards of field and evacuation hospitals with overdosages of morphine compounding their original difficulties. For a variety of reasons they had received too much medication. Respiratory distress was sometimes as inherently frightening to the inexperienced medical officer as to the casualty, and efforts made by these officers to relieve the patient were sometimes as frantic and ineffectual as the patient's own efforts to get his breath.

Whatever the background, oversedation invariably led to an increase in the accumulation of secretions in the tracheobronchial tree and a decrease in the cough reflex. As a result, shock was increased, and preparation of the casualty for emergency surgery or for further evacuation became more difficult.


Prolonged administration of an inhalation anesthetic to a patient in poor condition, whose reaction was delayed after operation, often led to the retention of bronchial secretions and to later pulmonary complications. This was particularly true if the airway had not been kept clear by catheter suction during the operation and if bronchoscopy had not been performed at the conclusion of the procedure when there was any doubt at all about the efficacy of suction.

Preexistent Respiratory Infection

It was not unusual for secretions to be present in the bronchial tree before injury. Soldiers who had undergone exposure during combat, particularly during the cold, wet months of winter, frequently had upper respiratory infections or frank bronchitis, associated with mucopurulent secretions. In the Mediterranean theater, preexistent respiratory infections were often aggravated by the circumstances of evacuation. During the mountain fighting in the winter of 1943-44, casualties had to be brought down the mountain by litter carry. Often as many as 8 to 12 relays were required between the battalion aid station and the collecting station. As a result, exposures of 12 hours or more after wounding were not unusual.


No matter how urgently replacement therapy might be needed, it had to be instituted judiciously in patients whose cardiorespiratory physiology was disturbed after wounding (vol. I). The administration of too much plasma in battalion aid and clearing stations, or even of too much blood in field and evacuation hospitals, might place such an extra load on the circulatory system that pulmonary edema might develop and contribute to the etiology of wet lung arising directly from the wound.


Wet lung usually appeared shortly after wounding, sometimes within a few hours. Less often, it did not develop for as long as 5 or 6 days after injury.

The presence of abnormal amounts of fluid in the tracheobronchial tree in thoracic injuries was manifested by certain cardinal symptoms and signs, the early recognition of which prevented the development of graver complications. Cough and dyspnea were the chief symptoms, and rales were the predominant physical finding.

Symptoms-The cough associated with wet lung was usually described as a wet cough because of the rattling or gurgling element always present in it. It might be hacking, harassing, continuous, or paroxysmal. Small amounts of sputum were often raised by constant coughing, but the cough remained moist; because there was little expulsive force behind it, it was never fully productive.


Inexperienced medical officers might receive the impression that, because some sputum was raised, tracheobronchial drainage was good, but the persistence of wheezes and rattles in the chest proved that it was still poor. As a matter of fact, only the most superficial secretions were raised, just as small amounts of fluid are splashed out from the top of a full cup that is continuously replenished.

The dyspnea characteristic of wet lung could be explained in a number of ways. It might be the result of painful, jerky, shallow respirations, each inspiratory effort being limited by the pain that the preceding effort had caused. When trauma had been considerable, the explanation might be hemothorax or mediastinal shift. When damage had been minimal, the most reasonable explanation was partial bronchial obstruction with resulting anoxia. Some patients seemed to be having typical asthma. Apprehension was an occasional cause of dyspnea.

When dyspnea was extreme, orthopnea, restlessness, excitement, and disorientation were present as the result of cerebral anoxia.

Signs-The patient's appearance varied according to the severity of his injury and whether or not he was in shock. Fever and evidences of toxicity were constantly present in late cases.

Respirations were frequently grunting because they were painful. Motion was usually restricted over the involved area of the chest. Breath sounds were reduced in intensity, but if hemothorax, pulmonary hematoma, or some similar complication were not associated with the injury, the percussion note and the tactile fremitus were unlikely to be materially altered.

The most characteristic physical finding in wet lung was the presence of rales, bronchial in character. They were sometimes heard on both sides of the chest; their bilateral presence was a valuable diagnostic sign. More often, they were more intense on the injured side. They ranged from high-pitched wheezes, which some surgeons considered indicative of associated bronchospasm, to medium or coarse bubbling rales or rhonchi. Rales that were wheezing and high-pitched sometimes had a dry quality. Sometimes fine rales that were predominantly moist or bubbling were so numerous as to suggest the type heard in simple, nontraumatic pulmonary edema or in bronchial asthma. These rales were frequently bilateral. Bubbling, sonorous rales were sometimes heard in combination with a classical tracheal rattle; in such cases, the sputum was predominantly liquid.

Rales were often so loud that the stethoscope was not needed to identify them. They could be heard at the bedside, or even at some distance from the patient, just as in typical bronchial asthma. While all varieties and gradations of rales might be present in any single chest, diffuse moist rales were especially characteristic of the primary phase of traumatic wet lung.

Rales were sometimes not demonstrable unless the examination was made immediately after the patient had coughed. They might not be heard at all if sticky mucus was attached to the tracheal or bronchial wall or if a branch


bronchus was completely blocked. In such cases, fine crackling sounds could usually be heard if the stethoscope was placed over the patient's mouth.

As already indicated, a coexistent hemothorax, pneumothorax, or some other finding might so alter or mask the typical clinical symptoms and signs of wet lung that its existence would not be suspected unless the medical officer bore the possibility in mind.


Clinical observations-In early stages of traumatic wet lung, the evidence of moisture obtained by auscultation was the most important diagnostic sign. The presence of rales anywhere in the chest, even over portions not involved in the trauma, was a valuable clue, indicating as it did that wet lung was either present or in process of development.

In advanced cases, the diagnosis, as already stated, could sometimes be made without the use of the stethoscope because dyspnea was so obvious and rales were audible even at a distance from the bed. The physical findings in late cases were those ordinarily observed in atelectasis and pneumonia. Distant bronchial breathing was often heard over small areas of the chest and might represent either patchy lobular atelectasis or a pneumonic process. The diagnosis of wet lung was established if the intensity or location of bronchial breathing changed with forced coughing.

The character of the secretions raised was also of diagnostic significance. Hemoptysis usually denoted pulmonary injury; the blood was fresh-looking when hemorrhage was recent and was clotted and mixed with mucus when it was not. If the secretions were the result of reflex stimuli from trauma to the chest or irritation from aspirated material, or were part of a very early infection, they were likely to be mucoid. As infection developed in the bronchial tree, they became increasingly purulent. The presence of purulent sputum could sometimes be explained on the basis of a preexisting respiratory infection and sometimes by the development of a new infection in the stagnated bronchial secretions. Thin, yellow, pink or colorless fluid, often frothy because of the admixture of air, indicated the pulmonary transudation and exudation (edema) typical of more advanced wet lung. Then, serosanguineous fluid suggested hemothorax with bronchopleural fistula. Similarly, thin, seropurulent fluid suggested early empyema with bronchopleural fistula. Most often, the sputum represented a combination of the types described rather than a single, clear-cut type.

Roentgenologic examination-While roentgenologic examination was highly desirable in all instances of chest trauma, it was always deferred until the patient was brought out of shock and steps had been taken to dry out the tracheobronchial tree.

The roentgenograms, which were taken, whenever possible, in at least two planes, were frequently difficult to interpret and were not always useful. In the initial stages of wet lung, roentgenologic changes were minimal, even when


there was clinical and auscultatory evidence of a considerable degree of obstruction and moisture. Patchy lobular atelectasis in the early stages might be indistinguishable from the shadows cast by a pulmonary hematoma (p. 165) or the pathologic process present in pulmonary contusion (p. 4). When intrapulmonary bleeding had occurred, the shadows were likely to be round or oval. When a hematoma was present, there might be evidence of small, loculated air pockets. Later, when lobular or total pulmonary atelectasis had occurred, the collapse of the affected parts of the lung produced the classical signs of mediastinal shift and narrowing of the intercostal spaces.

On the whole, while roentgenologic examination was sometimes useful and was never omitted because of the chest injury, the diagnosis of wet lung in the early stages was a clinical matter.

Differential diagnosis-As a practical consideration, the question of differential diagnosis seldom arose in connection with traumatic wet lung. Theoretically, the condition had to be distinguished from bronchial asthma, pulmonary edema of cardiac origin, and pulmonary edema associated with peripheral vascular failure.

The differentiation was not difficult. The history of trauma and the absence of any history of previous asthmatic attacks promptly ruled out bronchial asthma. Since the young soldier was almost never a cardiac subject, the necessity of differentiating traumatic wet lung from pulmonary edema of cardiac origin almost never arose. Occasionally, however, patients with thoracic injuries had had such vigorous intravenous fluid therapy that the blood volume had become too much for an already embarrassed respiratory mechanism and right heart failure had supervened. In such cases, the manifestations of wet lung were interwoven with, and intensified by, those of right heart failure.

Pulmonary edema associated with peripheral vascular failure either from shock due to severe trauma or from an overwhelming toxemia was sometimes difficult to differentiate from wet lung, particularly when shock was associated with the latter. The nature of the exciting lesion, the pallor, low blood pressure, tachycardia, and weak, thready pulse had to be evaluated, as well as the response to attempts to clear the tracheobronchial tree. In wet lung, the response was usually rapid. If the pulmonary edema was associated with peripheral vascular failure, it would not respond to the therapy indicated for wet lung because its etiology was different.


As their experience accumulated, medical officers came to realize that it was the part of wisdom to regard as a potential candidate for wet lung every casualty with a chest wound, a thoracoabdominal wound, a severe abdominal injury, or a head injury. On their admission to forward hospitals, all patients in these categories, particularly those who complained of thoracic pain or who presented dyspnea, wheezing, hemoptysis, and a cough, were carefully examined


for evidences of fluid in the pulmonary tree. The following routine was carried out:

1. Unless the patient was in shock or unconscious, he was placed in Fowler's position, so that the abdominal contents would gravitate into the pelvis. The diaphragm could thus act more effectively, with corresponding improvement in respiration and in the efficacy of the cough. The position was changed frequently if the patient could not turn himself. This measure was not always stressed as it should have been.

2. If the patient was dyspneic or cyanotic, oxygen was administered by nasal catheter or by the Boothby-Lovelace-Bulbulian mask.

3. If his breathing was at all wet, the patient was encouraged to cough (fig. 67A), the chest being supported manually whenever personnel could be spared for this purpose. The physical support of a painful chest was helpful, and the psychic effect was excellent; the patient was impressed with the importance of coughing because someone took the time and trouble to help him to do so. If he was sufficiently alert to comprehend, as already mentioned, the importance and rationale of coughing were explained to him, which made him much more willing thereafter to cough, in spite of any associated discomfort.

4. If the patient was conscious, complained of pain, and had evidently had no previous sedation, a small dose of morphine sulfate, preferably not more than gr. 1/8, was given by vein, so that it would act promptly and its effect on the painful chest could be more accurately evaluated.

5. The usual measures employed in all chest wounds were carried out. If the dressing covering a sucking wound was not already airtight, it was rearranged. The indicated resuscitative measures were instituted, with great care not to overload the circulation with fluids. The gastric tube was inserted if there was evidence of ileus or gastric distention (fig. 67B). If a hemothorax or tension pneumothorax was present, aspiration was employed to increase the vital capacity.

If wet lung had not yet become established but was merely impending or incipient, these measures were usually effective and the patient rapidly became a fit candidate for surgery in a forward hospital or could be safely evacuated to a fixed hospital.



As already pointed out, the various factors responsible for wet lung constituted a vicious circle. They included accumulation of secretions in the smaller bronchi; limitation of normal respiratory motion because of pain; the suppression of the cough reflex, as well as of the desire to cough, because of pain; and the resultant effect of the wet lung on the oxygenation of the blood.


FIGURE 67.-Management of patient with wet lung. A. Clearance of chest by coughing, with patient supporting his own chest. B. Insertion of Levin tube for control of gastric dilatation. C. Application of elastic binder for stabilization of flail chest or external fixation of fractures. D. Intercostal nerve block (for details of technique, see fig. 68). E. Aspiration of bronchi by catheter passed transnasally. F. Bronchoscopy, which is resorted to if catheter aspiration is not successful. G. Administration of intermittent positive pressure oxygen therapy by manual compression of ventilatory bag of field hospital portable anesthetic machine.


Breaking this vicious circle by attacking its various components was the basis of the active therapy of established wet lung. In its active management, as well as in its prophylaxis, the principal aims of therapy were:

1. The maintenance of an open airway, so that inspired oxygen could reach the alveolar capillaries.

2. The prompt correction of conditions that were contributing to dyspnea, anoxia, and shock.

To achieve these aims, it was necessary to control the production of fluid in the lungs and to promote adequate bronchial drainage, as well as to institute such other measures as were necessary to correct the disturbed cardiorespiratory physiology.

Relief of Pain

Relief of pain and physiologic remobilization of the chest wall after thoracic injury were the sine qua non of a successful therapeutic effort in wet lung. In prewar civilian practice, the desired results had been achieved by adhesive strapping and the administration of morphine. These measures had never been physiologically sound. In combat-incurred injuries, both were ineffective, and even in stove-in chest, they were not necessary.

In severe flail chest with extreme paradoxical respiration, strapping of the chest or the use of a firm binder was occasionally necessary to prevent ballooning out of the chest wall (fig. 67C). The method was never officially forbidden, but with these very occasional exceptions, its use was avoided.

The administration of morphine was equally undesirable. It made the patient less aware of his discomfort, it is true, but it did not make the cough mechanism painless, and it diminished the cough reflex. Furthermore, by dulling the sensorium, it made the patient less aware of his own responsibility for overcoming his respiratory difficulties.

Intercostal Nerve Block

The most effective method of relieving chest pain and making it possible for a casualty with a chest wound to evacuate the tracheobronchial tree by coughing was intercostal nerve block at the site of injury (fig. 67D), or, alternatively, paravertebral sympathetic nerve block. One reason that it was necessary to employ both the intercostal and the paravertebral techniques of nerve block was that certain wounds were so situated that simple intercostal block was not always feasible.

Whether intercostal nerve block did more than abolish pain and effect an interruption of the reflex are was a matter of speculation. It was, however, a logical procedure. According to Lewis (6), afferent pain impulses from the thoracic abdominal wall are transmitted by way of the sympathetic nerves, whose fibers accompany the intercostal nerves as far as the intervertebral foramina, whence they depart as rami communicantes to form part of the paravertebral sympathetic chain. There seemed little difference, therefore, between


interrupting the pain pathway by blocking the intercostal nerves or blocking the paravertebral chain through which pain impulses pass.

The nerves could be anesthetized at any point posterior to the lesion. At least three nerves were always injected, to provide a generous margin of anesthesia around the traumatized area. As many as 10 intercostal nerves or paravertebral ganglia could be blocked with safety, but when more than 4 or 5 nerves were to be injected, it was the policy to give a small preliminary dose of some barbiturate, to lessen the risk of a procaine hydrochloride reaction.

Local infiltration, which produced its effects by blocking the sensory end organs, was effective in simple, uncomplicated rib fractures with contusion of the chest wall, but it was contraindicated in casualties with open wounds, because all such wounds were potentially if not actually contaminated. Furthermore, the involved area in casualties with open wounds was frequently too extensive for local infiltration to be practical. This was sometimes, however, a valuable supplement to an extensive blocking procedure after which a small localized area of pain occasionally persisted, though an otherwise effective block had been obtained.

If the site of injury were so located posteriorly that even paravertebral block would require injection through contaminated tissue, paravertebral block incorporating at least two nerve segments above the superior margin of the injury was useful, for it blocked an appreciable number of the pain impulses which travel cephalad in the paravertebral sympathetic chain.

Technique of nerve block-The needles used for intercostal nerve block were ordinary intravenous needles, 1 inches long. A 1-percent solution of procaine hydrochloride was usually used, though an occasional surgeon preferred a 2-percent solution. An intradermal wheal was raised over the chosen rib at its angle over the inferior border of the rib between the costal angle and the wound or fracture site (fig. 68). Then the needle, without the syringe attached, was introduced through the wheal until the point had impinged lightly upon the inferior margin of the rib. From this point, with the bevel directed inferiorly, the needle was moved until the tip just cleared the inferior edge of the same rib. After the needle had been advanced about 0.5 cm., aspiration was carried out in two planes, to exclude the possibility that it might have entered a blood vessel. When this point had been settled, from 4 to 6 cc. of 1-percent procaine hydrochloride solution was introduced into the tissues. The total number of thoracic segments injected was determined by the number of ribs fractured or, if the lesion was a simple contusion, by the area involved.

Needles 3 or 3 inches long were preferred for paravertebral nerve block. A small piece of loose rubber on the shaft indicated the depth to which the needle was to be inserted.

Paravertebral injection of the thoracic sympathetic ganglia was best performed with the patient in the lateral recumbent or the prone position. The sites for injection were marked on the skin, opposite the spinous processes and about 4 cm. lateral to the midline on the affected side. These points lie directly


FIGURE 68.-Intercostal nerve block in management of painful wounds of chest wall. A. Schematic showing of effects of painful chest wall on retention of secretions and on effectiveness of cough. B. Infiltration of skin and muscles with procaine hydrochloride. Long needle is gently manipulated until the tip comes into contact with the rib. C. Injection of procaine hydrochloride (3-5 cc. of 1-percent solution) beneath rib four fingers from spinous process. Care is taken not to inject solution into intercostal vessels.


over the transverse processes. The best results were achieved when one or two ganglia cephalad to the area of involvement were also incorporated in the block.

The needle, without the syringe, was introduced perpendicularly through the skin until contact was made with the dorsal surface of the transverse process; it was seldom necessary to introduce it more than 4 centimeters. The rubber marking was readjusted on the shaft from 3 to 4 cm. from the skin surface before the needle was withdrawn slightly. Then, with its bevel directed medially, the needle was redirected anteromedially and passed just inferior or just superior to the transverse process. It was slid along the body of the vertebra to the depth indicated by the marker. At this location, the tip of the needle lay in the immediate vicinity of the sympathetic chain.

If, during its insertion, blood or spinal fluid appeared, the needle was withdrawn and was reinserted in a slightly different direction. If blood or spinal fluid did not appear, the syringe was attached to the needle, and aspiration was carried out in two planes, to be sure that the needle did not lie in a blood vessel, in a prolongation of the subarachnoid space, or in the pleural cavity.

If aspiration was negative, 6 cc. of 1-percent procaine hydrochloride solution was injected into the tissues. If the needle had been introduced correctly, relief from pain was almost immediate, only just enough time being required for infiltration of the solution through the tissues surrounding the sympathetic chain.

As soon as either intercostal or paravertebral nerve block had been accomplished, an attendant forced the patient to cough.

If open thoracotomy was necessary in a patient who had not previously been submitted to nerve block, advantage was taken of the opportunity to block the nerves under direct vision by injecting procaine hydrochloride solution. Crushing the nerves with a fine hemostat was a much less desirable procedure.

Results of nerve block-Blocking out of the pain stimuli by injection of the intercostal nerves or by paravertebral block produced prompt and often dramatic results. Pain and discomfort almost invariably disappeared. The patient was willing to cough because it was no longer painful to do so, and evacuation of the fluids in the tracheobronchial tree was effected.

It was impossible to overlook the possible influence of nerve block on constriction of the bronchial tree and secretion of mucus. Clinical evidence supported the hypothesis of its role. Frequently, casualties were seen with respiratory distress that could not be accounted for by the severity of the thoracic injury, pain in the chest wall, shock, the presence of blood or other fluid in the bronchial tree, aspiration of gastric contents, infection, or exposure to noxious gases. Coughing was apparently free and forceful, but wheezing persisted, and sputum was not raised. Once the intercostal nerves were blocked, however, the picture changed. Wheezing disappeared promptly, sand sputum was raised freely. The assumption in such cases was that reflex bronchial spasm probably played some role in the previous symptoms and signs.


Relief of pain and discomfort after intercostal nerve block lasted for varying periods of time but almost never for less than 24 hours. The block could be repeated as necessary, but frequently a single injection gave permanent relief. Why relief of pain extended beyond the period of the pharmacologic action of the drug is difficult to explain. Possibly relaxation of muscle spasm and improvement of the blood supply to the traumatized area played some part in the results.

The beneficial effect of paravertebral block usually lasted from 2 to 4 hours. One injection frequently gave permanent relief, and more than three were almost never necessary.

An effective nerve block, in addition to rendering anesthetic the area innervated by the somatic nerves blocked, sometimes also provided immediate relief of pain in a referred area, occasionally on the side opposite the trauma.

When the nerves were blocked at thoracotomy, the stimuli arising from the chest wall and parietal pleura were diminished, and the patient was usually free from pain during the most important part of the postoperative period. Nerve block at operation also eliminated the necessity for later intercostal nerve block.

Catheter Aspiration

If, after a reasonable period of time, a patient with traumatic wet lung did not improve under the measures just outlined, combined with the usual regimen of resuscitation, mechanical measures had to be employed. The circumstances of the individual case determined whether catheter aspiration (fig. 67E) should be tried or bronchoscopy (fig. 67F) resorted to at once. The length of the trial period, which might be as short as 5 or 6 hours or as long as 20 to 24 hours, depended upon the patient's original injury, his clinical status, the roentgenologic findings, and his progress under treatment. Some patients were too exhausted to cough effectively without aid, and some were uncooperative for other reasons.

Whether catheterization or bronchoscopy was used, the aims were the same:

1. To remove excess fluid in the lower airway.

2. To loosen deeper secretions by local manipulation.

3. To promote more efficient cough.

A long delay was not justified if the situation was complicated by a communication between the tracheobronchial tree and a pleural cavity containing fluid.

Catheter aspiration of the tracheobronchial tree had so many advantages that it was employed in all cases in which bronchoscopy was not clearly indicated (p. 227):

1. It was preferable to bronchoscopy in patients who were desperately ill but conscious. Experience showed that in such cases the manipulations required in bronchoscopy might be extremely dangerous.


2. Catheter aspiration was particularly valuable in acute emergencies which permitted no delay.

3. The equipment was almost always readily available from standard hospital supplies.

4. The technique was so simple that it could be readily mastered, in contrast to the elaborate training necessary in bronchoscopy.

5. Catheter suction could be repeated as often as necessary.

Technique-The technique generally employed for catheterization (fig. 69) was a modification of the method described by Haight (7) in 1938. Suction was usually provided by ordinary ear, nose, and throat suction machines, which delivered from 15 to 20 pounds of negative pressure. Connecting tubing and a Robinson ureteral catheter or a similar catheter, size No. 16 or No. 18, with one or two openings, constituted all the equipment necessary. If electricity was not available, as it sometimes was not in combat areas, an attachment to the windshield wiper or gas intake manifold of a motor vehicle served as an emergency source of power. The portable hand suction machine devised by Major Brewer was often used in forward installations in which electrical suction machines were not always at hand.

Anesthesia was not required unless the gag reflex was hyperirritable. Then 2 percent Pontocaine hydrochloride (tetracaine hydrochloride) or 5 percent cocaine was sprayed onto the pharynx and painted onto the hypopharynx with a curved swab or applicator. When local analgesia was necessary, a small dose of some barbiturate was given before it was applied, and fluids were withheld until the gag reflex had returned.

The patient was placed in the semi-Fowler position, with the neck flexed. If he was unconscious, the neck was fixed by supporting the raised head on a pillow or a folded blanket. If he was conscious, the tongue was pulled rather sharply forward, to anchor the larynx. Occasionally an epiglottis that was unusually flaccid so covered the larynx that catheterization was impossible without a laryngoscope to provide direct vision.

The catheter was introduced through the nostril and advanced until the larynx was reached. If the patient was unconscious, the index finger of the

FIGURE 69.-Technique of tracheobronchial catheter aspiration. A. Robinson type of rubber catheter after sterilization by heat on bent catheter guide, to provide for slight curve of tip. The catheter is sterilized in 70 percent alcohol or some similar sterilizing solution when it is used, to assure maintenance of curve.


FIGURE 69.-Continued. B. Forward fixation of larynx with left hand, by holding tongue forward with gauze-covered fingers. With the right hand, the catheter is rapidly advanced through the nares and past the glottis  as the patient inspires deeply. If he is comatose, a mouth gag is used, and the epiglottis is picked up with the left forefinger. The catheter is then guided through the larynx. C. Advancement of catheter into trachea by intermittent suction over Y-tube. Once it is in trachea, it is moved back and forth to stimulate coughing.


FIGURE 69.-Continued. D. Introduction of catheter into right main stem bronchus. When the patient's head is turned sharply to the left, the right stem bronchus is more directly in line with the trachea. The catheter is then advanced and intermittent suction employed. E. Introduction of catheter into left stem bronchus, by reversing the procedure described in D. The head is turned markedly to the right in this stage of the procedure.


left hand engaged the tip of the epiglottis, so that the catheter could be passed through the larynx without difficulty.

When the catheter had entered the larynx, it was withdrawn for about 1 centimeter. If the patient was conscious, he was asked to take a rapid, deep breath. During the inspiration, the catheter was quickly advanced into the trachea. In an occasional case, it was easier to introduce it during the expiratory phase of a cough. The successful entrance of the catheter into the trachea was signified by involuntary coughing, passage of air through the trachea, or sudden hoarseness. The irritation caused by the catheter was likely to stimulate all but moribund patients to cough, but hoarseness was the most certain indication that the catheter was in the trachea. Occasionally, when the tube had been accidentally introduced into the esophagus, a free flow of air about it simulated respiration. If the accident was not recognized and oxygen therapy was used, gastric dilatation of serious proportions might occur.

Suction was not applied until the catheter had been advanced into the trachea for several centimeters. When the manipulations were discontinued, the patient was urged to cough. This he usually did forcibly-in fact, he found it impossible not to-because of the stimulation of the tracheal mucosa by the catheter. Often more sputum was coughed up around the tube than was aspirated through it. Suction, coughing, and manipulation of the catheter were alternated until all portions of the tracheobronchial tree were apparently free of fluid. Continuous periods of suction never exceeded 5 seconds, and they were briefer if the patient seemed exhausted by them or if he became cyanotic.

It was simple to introduce the catheter into the right stem bronchus; the patient's head was merely turned well to the left. To enter the left stem bronchus was somewhat more difficult (vol. I). The head had to be turned far to the right and the chin elevated. The tube was advanced as far as possible into each stem bronchus. After aspiration, the patient was rolled on his side, with the more involved lung uppermost, to permit drainage of the small bronchi by gravity.

After catheter aspiration, most patients coughed with greater efficiency because of the improved ventilation of the pulmonary tree. Because of the absence of appreciable trauma to the larynx or bronchi, the procedure could be repeated as often as every 2 to 3 hours if this was necessary. In the occasional case, bronchial secretions and pulmonary transudates formed with such rapidity that aspiration was necessary at 30-minute intervals. In such cases, the catheter was left in the trachea, sometimes from 8 to 12 hours. When this was necessary, 100 percent oxygen was administered continuously through the catheter between aspirations, at the rate of 1 to 2 liters per minute.


Bronchoscopy played a valuable role in the management of war wounds of the chest. At no time was its use routine. It was properly considered a highly technical procedure, requiring skill and experience for its safe use and


to be used only on definite indications. It was made more widely available as the war progressed, however, because thoracic surgeons already trained in its use taught the technique to selected general surgeons and qualified anesthesiologists.

The decision to employ bronchoscopy rather than catheter aspiration was made on the indications of the individual case. It was sometimes preceded by a trial of catheter aspiration but in other instances was resorted to immediately. Its chief advantage over catheter suction was that it was more thorough. Because both stem bronchi were visualized, aspiration could be carried out with more assurance that all secretions and obstructing mucous plugs had been removed. When it was indicated, bronchoscopy also permitted the direct application of procaine hydrochloride, cocaine, or epinephrine solutions to the swollen bronchial mucosa. The resulting shrinkage of the mucosa further increased the lumen of the airway.

Indications-No hard-and-fast rules were formulated, but in general, bronchoscopy was regarded as the procedure of choice in wet lung under the following circumstances:

1. When tracheobronchial aspiration by catheter had been ineffective because of failure of the lung or segments of the lung to reexpand with positive pressure, and moisture persisted in the chest.

2. When mucous plugs or blood clots were thought to be present in the branch bronchi.

3. When obstruction of the trachea or bronchi had been present for a considerable time. Under these circumstances, sudden loosening of large amounts of secretion from the obstructed bronchus might flood a normal bronchus, with grave consequences.

4. When the cough was weak and ineffectual and there was danger of flooding the contralateral bronchus by sudden release of large amounts of secretion.

5. When a patient was admitted to the hospital completely exhausted, with the tracheobronchial lumen brimming with secretions that he did not have the strength to cough up. In this type of patient, bronchoscopy was mandatory. In fact, the more critical his condition, the more urgent was the indication.

6. When lobar or total pulmonary atelectasis had occurred, because of the dangerous reduction in the vital capacity.

7. When bronchial obstruction recurred in spite of repeated tracheobronchial catheter aspirations. The cause of the obstruction was usually incomplete removal of obstructing material by catheter suction.

8. When vomitus had been aspirated, or was thought to have been aspirated, into the lungs. This was an imperative indication, because of the highly deleterious effect of gastric secretions on bronchial and pulmonary tissue.

9. When emergency thoracotomy was necessary. Bronchoscopy immediately before operation, in addition to emptying the tracheobronchial tree and


improving the cardiorespiratory status, made the administration of the anesthetic simpler and safer.

Techniques-The standard technique of bronchoscopy was employed. The procedure was usually carried out in the operating room but could be performed by a skilled bronchoscopist without moving the patient from the stretcher or bed.

General anesthesia was contraindicated if there was an unusual amount of fluid in the tracheobronchial tree. Topical analgesia was employed in most cases, but in numerous semicomatose or apparently moribund patients, no anesthesia was used. Oxygen (100 percent) was administered continuously to all patients who showed any evidence of anoxia.

The operation was performed as rapidly as possible, so that a severely wounded patient, in shock, would not be exposed to the fatiguing effects of prolonged coughing.

Some surgeons, after the trachea and main bronchi were cleared, palpated the chest carefully, in an endeavor to find secretions still in situ in spite of apparently successful aspiration. When such areas were found, the tip of the aspirator was applied directly to the orifice of the affected lobe, to excite coughing and raise the remaining secretions.

Oxygen Administration

Oxygen inhalations by means of the nasal catheter or the Boothby-Lovelace-Bulbulian mask were used promptly whenever cyanosis accompanied wet lung or dyspnea was severe. Some patients were disturbed at first by the face mask, but with sufficient encouragement, they all became reconciled to its use.

If moisture persisted in the lungs after intercostal nerve block and catheter aspiration or bronchoscopy, the oxygen was delivered under positive pressure (figs. 67G and 70). This was a technique described by Barach and his associates (8) shortly before the war, to relieve the pulmonary edema associated with pneumonia, gas poisoning, and cardiac disease. When the technique was first applied to combat-incurred wounds in the winter of 1943-44, considerable difficulty was encountered. A simple and effective solution of the problem was the use of a to-and-fro anesthetic system, with a soda-lime canister and rebreathing bag. Positive pressure was maintained manually on the bag, care being taken to avoid pressures higher than from 2 to 6 cm. H2O. Pressures as high as 10 cm. H2O or more were dangerous; they could cause a considerable rise in venous pressure; a fall in systolic, diastolic, and mean blood pressures; and a decreased blood flow.

Fluid that was constantly forming in the alveoli and bronchioles could not be completely removed by suction, but suction, with nerve block and the other measures described, removed enough fluid for oxygen under positive pressure to keep the bronchioles patent. The use of oxygen under pressure also opposed the hydrostatic pressure of the blood in the capillaries and increased the vital capacity. If the lungs were not too severely damaged and if shock, tension


FIGURE 70.-Administration of oxygen under intermittent positive pressure in management of wet lung, Cassino, Italy, 1943. A. Administration with portable circle filter anesthetic machine: Face mask; tubes in circle filter; respiration bag, which was manually compressed to supply positive pressure oxygen to face mask; portable anesthetic machine with circle filter and soda-lime canister; water manometer to determine amount of positive pressure; and large oxygen tank.

pneumothorax, and other adverse factors could be controlled, this method was practically always successful in drying up wet lung caused by tracheobronchial transudation and exudation. It was of particular value when an excessive and injudicious use of blood and plasma was a contributory factor in the edema and when ventricular failure was impending.

Oxygen was sometimes administered under positive pressure in the early stages of traumatic wet lung. Its prophylactic value was thought to be consid-


FIGURE 70.-Continued. B. Administration with portable to-and-fro soda-lime canister anesthetic machine: Face mask; soda-lime canister; respiration bag, which was manually compressed to supply positive pressure oxygen to face mask; portable anesthetic machine with valves and bubbler for ether (only water was used in oxygen therapy); and oxygen tank.

erable, but the indications for its employment on this indication did not become fully established during the war.

Other Measures

Atropine administration-If moisture continued to be present in the lungs and vigorous measures to empty the tracheobronchial tree proved ineffective, atropine was sometimes given intravenously, in 1/100-gr. doses. If reflex- spasm of the bronchus were initiated by thoracic injury, as the experimental work of de Takats and his associates (2) had suggested, one might expect atropine to abolish vagal reflexes and reduce tracheobronchial secretions. De Takats himself, however, found that this method prevented bronchospasm in less than half of his experimental animals, even when amounts up to gr. 1/8 were employed. Atropine was not widely used during the war, and the clinical experience with it was too conflicting to permit any definitive statements concerning its value except that when wet lung was fully developed, this drug was not effective.


Carbon dioxide inhalations-Carbon dioxide inhalation was another controversial measure occasionally employed in wet lung. This gas is a true expectorant, and hyperventilation with it had the mechanical effect of loosening viscid secretions and propelling them upward along the tracheobronchial lumen. If hypoxia was present, most observers regarded the method as equivalent to whipping a tired horse.

The preferred method was to use 100 percent carbon dioxide and to administer it by means of a catheter and small funnel as the patient took two or three deep breaths. Forced coughing during and following the brief period of hyperpnea thus produced was of the greatest importance and was not always sufficiently stressed. Unless the patient coughed, the method lost much of its effectiveness. If carbon dioxide was not available, rebreathing into a paper bag was often employed as a substitute, usually very effectively.


Once proper measures had been instituted to relieve pain in the chest wall, clear the airway, and permit effective coughing, the entire picture of a patient with wet lung usually changed promptly. Results were often dramatic. A cyanotic, comatose patient, with wet, rattling respirations, often became alert and oriented, with good color and almost normal respirations, within a matter of minutes.

Once the wet lung was controlled, it was possible to evaluate the primary injury. Hemothorax, shattered ribs, and the shell fragment in the lung looked less ominous and often seemed less urgent. The lung had been restored to normal, or almost normal, capacity to cope with the injury or with the surgery required to correct the injury. Evaluation of the total patient was possible, in short, under circumstances which permitted a deliberate rather than a frantic approach to the problem.

An undoubted result of the recognition and vigorous management of wet lung was apparent in the Mediterranean theater in a number of respects, including (1) the low case fatality rate and low morbidity of chest injuries, (2) the infrequency of lung abscess after injury, and (3) the almost complete absence of massive atelectasis.


The following case histories illustrate a number of the points in the foregoing discussion of traumatic wet lung:

Case 1-A 23-year-old sergeant was received in an evacuation hospital 3 hours after he had sustained penetrating wounds of the chest and buttocks from a high explosive shell fragment. The pulse was 110, the respiration 30 and noisy, and the blood pressure 110/80 mm. Hg. The patient was extremely dyspneic and complained of severe pain in the chest and abdomen.

Evidence of a large amount of fluid in the left chest almost obliterated all other physical findings. Loud musical wheezes were heard over the right chest anteriorly and posteriorly. Roentgenologic examination confirmed the presence of fluid in the left chest. The


left eighth rib was fractured, and a small foreign body was seen above the left diaphragm behind the heart.

Morphine (gr. ) and atropine (gr. 1/100) were given by vein. The pain was somewhat relieved, but not enough to permit effective coughing. Complete relief followed blocking of the fifth through the tenth left intercostal nerves. The patient immediately coughed up a large amount of mucus mixed with fresh and old blood, after which he was able to breathe deeply and without discomfort. The lungs were dry on auscultation. Debridement of the chest wall was then carried out without difficulty under Pentothal sodium (thiopental sodium) anesthesia.

The patient continued to raise bloody sputum for a week, but there was no reaccumulation of fluid in the tracheobronchial tree and no recurrence of chest pain. After a total of 2,730 cc. of bloody fluid had been removed from the left chest by repeated thoracenteses, the left lung was almost completely reexpanded, and the remainder of the patient's convalescence was uneventful.

Comment.-In this case, retention of blood in the bronchial tree as the result of intrapulmonary hemorrhage was chiefly responsible for the wet lung, though an excess of mucus was also a factor. Severe chest pain, which prevented effective coughing, was not relieved by a large dose of morphine but was promptly relieved by blocking the appropriate intercostal nerves. Immediately thereafter, the patient, by his own efforts, raised the blood and mucus that had accumulated in the tracheobronchial tree, and recovery was without further complications.

Case 2-A 21-year-old sergeant was received in a field hospital in severe shock 3 hours after he had sustained a sucking wound of the left chest and penetrating wounds of the right chest, thigh, forearm, hand, and cheek. The pulse was 140, and the blood pressure could not be obtained. Breath sounds were diminished on the right side of the chest, and signs of fluid were elicited on the left side.

The patient had had no medication since he was injured and was in severe pain. He was therefore given morphine gr. 1/8 by vein after an occlusive dressing had been placed over the sucking wound. Pain was completely relieved after a second injection of morphine, in the same amount and also by vein, but only small quantities of bloody sputum were raised. After the administration of two units of plasma and 500 cc. of blood, the pulse fell to 132 and the blood pressure was 104/80.

Two hours later, the patient's condition suddenly became critical. He was comatose and cyanotic. The pulse was weak and thready. Rales were audible in the trachea. Catheter suction was instituted at once, and large amounts of frothy yellow sputum were aspirated. The response to this measure was dramatic. The patient immediately regained consciousness, and his color and pulse improved. The catheter was kept in place for 8 hours, and suction was repeated every 20 to 30 minutes. Between aspirations, 100 percent oxygen was administered through the catheter.

At the end of 8 hours, during which time the patient's condition steadily improved, roentgenograms of the chest showed fluid in the left chest and foreign bodies in both lung fields. Since the tracheobronchial tree was now apparently dry, the catheter, which meantime had slipped into the right main bronchus, was removed. After a transfusion of 500 cc. of blood and a unit of plasma, 100 cc. of air and 200 cc. of bloody fluid were aspirated from the left chest.

Seventeen hours after admission, when the blood pressure had reached 108/74 mm. Hg, bronchoscopy produced a considerable amount of bloody fluid from both main bronchi. Careful inspection showed that the prolonged intubation had apparently caused no trauma to the larynx or the trachea.

Immediately after bronchoscopy, operation was done under intratracheal ether-oxygen-anesthesia. It included debridement of the sucking wound of the chest wall; ligation of the internal mammary vessels, which were lacerated though not actively bleeding; suture of the laceration of the upper lobe of the left lung; removal of several small metallic foreign


bodies from the pericardium and another from the right chest wall; and intercostal drainage. A foreign body in the lower lobe of the left lung and another in the right lung were left in situ. While the chest operation was in progress, another surgical team debrided the other wounds. At the conclusion of the operation, bronchoscopy was repeated, and a moderate amount of bloody mucus was aspirated from both stem bronchi. The patient received a unit of plasma and 500 cc. of blood during the operation.

The immediate postoperative status was satisfactory. The following day, the patient became disoriented, presumably as the result of cerebral anoxia, though he had been receiving oxygen continuously, by nasal catheter, since operation. The temperature was 99.4 F., the pulse 150, and the respiration 48. Coughing produced thin yellow fluid, but the pulmonary tree appeared to be filling up with moisture.

Tracheobronchial aspiration by catheter somewhat improved the patient's critical condition. Tracheal rhonchi and coarse bronchial rales disappeared, but medium and fine rales could still be heard diffusely over all lobes of both lungs. Fluid was apparently being formed more rapidly in the peripheral portions of the pulmonary tree than it could be removed by natural processes. Since the clinical picture resembled that seen in pulmonary edema accompanying heart disease and gas poisoning, positive pressure oxygen was begun. Within 15 minutes, a notable improvement occurred. The pulse fell to 120. The respirations fell to 32 and became considerably less labored. Only a few fine rales could be heard at the bases of both lungs. At this point, positive pressure oxygen was discontinued, and administration by the Boothby-Lovelace-Bulbulian mask was resumed.

During the next 24 hours, the patient had two additional attacks of pulmonary edema, each one resembling the acute attack just described. In each one, the effect of oxygen administration under mild positive pressure was as striking as in the first attack. Recovery thereafter was uneventful.

Comment-This case is a striking example of the therapeutic problems presented by traumatic wet lung in a patient with multiple chest and other wounds. Immediate tracheobronchial catheter aspiration, intratracheal oxygen administration, and cautious fluid replacement over an 8-hour period made this soldier a suitable candidate for surgical closure of a sucking chest wound and the intrathoracic surgery required. Bronchoscopy just before operation aided in the maintenance of a patent airway during operation, and its repetition at the conclusion of the procedure simplified the first 24 hours of the postoperative course. Three attacks of apparently true pulmonary edema over the next 48 hours were controlled by the administration of oxygen under mild positive pressure. Why edema developed in this case after operation is not clear, but anoxia caused by severe trauma to both lungs was probably an important factor, as was tracheal obstruction from mucus and blood.

There seems little doubt that this patient survived because of the therapy instituted for the control of the wet lung that followed his combat-incurred injuries, one of which was a sucking wound of the chest.

Case 3-A 27-year-old lieutenant sustained a sucking wound of the chest that was treated immediately by an occlusive dressing and 24 hours later by suture; the skin was left open. For the next 3 days, small amounts of pure blood were expectorated.

When the patient was admitted to a thoracic surgery center 5 days after wounding, he was comfortable while at rest but moderately dyspneic on exertion. The cough was wet but was productive of only small amounts of blood-tinged mucoid sputum. A few coarse rales were heard parasternally on the right. Roentgenograms taken the following day showed a right-sided hydropneumothorax without cardiac shift. Aspiration of the chest on the second and third days after admission produced 360 cc. and 390 cc., respectively, of old blood and air from the right pleural cavity. On each occasion, highly negative intrapleural pressure prevented further aspiration.

The patient's wet cough continued in spite of these thoracenteses, frequent changes of position, and the administration of carbon dioxide. On the third day after admission,


roentgenograms of the chest showed a minimal residual hemopneumothorax, an expanded upper lobe, and atelectasis of the middle and lower lobes on the right side. Bronchoscopic aspiration, instituted immediately, produced large amounts of blood clots and mucus, amid which partly formed bronchial and bronchiolar casts were discernible. Roentgenologic examination 18 hours later showed almost complete reaeration of the atelectatic lobes. The cough became loose and remained productive for 4 days after bronchoscopy, then ceased altogether. On the 12th day after wounding, roentgenograms showed the right lung to be completely expanded and almost normal in appearance.

Comment-In this case, although roentgenologic evidence of atelectasis was obscured in the first roentgenograms by the overlying hemothorax, certain points in the history and physical findings pointed to the diagnosis. They included continued wet cough with inadequate expectoration, extensive pneumothorax without cardiac shift, and the development of pronounced evidence of negative pressure after aspiration of relatively small amounts of fluid from the pleural cavity. The aspiration of old blood clots and mucus by bronchoscopy resulted in the prompt reestablishment of a patent airway and prompt reexpansion of the lung.

Earlier attention to the wet lung in this case would have permitted the use of simpler methods of management and might have prevented the development of the complicating atelectasis.

Case 4.-When this patient was seen 2 hours after he had fallen off a motorcycle, he was complaining of severe pain in the left chest and dyspnea. He had a paroxysmal, ineffectual wet cough. There was extreme tenderness over the left fifth and sixth ribs, and a flail segment in this area was undergoing paradoxical oscillations. Numerous wheezes and rhonchi were present bilaterally. Roentgenologic examination revealed anterior and posterior fractures of the left fifth and sixth ribs.

Intercostal block of the third through the ninth ribs, which resulted in only partial relief of pain, was supplemented by paravertebral block of the sympathetic ganglia of the same segment. Pain was promptly relieved, and paradoxical oscillations of the flail segment ceased. There was still some moisture in the lung, but the remaining wheezes and rhonchi disappeared, and respirations became entirely normal, within 5 minutes after the intravenous administration of 1/150 gr. of atropine sulfate.

On the third day after wounding, a moderate recurrence of pain and wetness was immediately controlled by repetition of the paravertebral block.

Comment-This case well illustrates the effectiveness of nerve block in the management of flail chest of limited extent, as well as the (occasional) effectiveness of atropine as an adjuvant measure in the control of wet lung.

Case 5-This patient was seen in a shock tent on 4 February 1944, 43 hours after wounding. He had sustained a penetrating wound of the left anterior chest and was complaining of hemoptysis; dyspnea; a painful wet cough; and pain in the left shoulder, left chest, and hepatic area. He was in critical condition, extremely dyspneic and cyanotic, and showers of moist rales were heard throughout both lungs. Examination also revealed tenderness and rigidity in the left upper quadrant of the abdomen. Roentgenologic examination revealed a retained missile in the region of the left diaphragm and an extensive pathologic process in the left lung and left pleural cavity.

An hour after the patient had been received in the shock tent, intercostal block of the left fourth and fifth intercostal nerves was carried out, and 200 cc. of air and blood were aspirated from the left pleural cavity. There was immediate improvement in the pain, dyspnea, and cyanosis originally present. The left upper quadrant of the abdomen was no longer rigid. There was also an improvement in the wet lung. Two hours later, the patient again complained of pain, and there was a recurrence of the pulmonary difficulties. A paravertebral sympathetic block (D4-9) was immediately performed, and a large amount of blood and mucopurulent sputum was aspirated through a tracheal


catheter. Pain was promptly relieved, and pulmonary moisture was greatly reduced. Shortly afterward, the wound was debrided.

Recovery was smooth except for slight pain and moderate dyspnea the day after operation. Both symptoms were promptly relieved by intercostal block of the fifth through the ninth intercostal nerves. A week after his admission, the patient was so much improved that he could be evacuated.

Comment-When this patient was first seen, his condition was so critical that it did not seem that he could possibly survive. Nerve block effected considerable improvement, but when his pulmonary difficulties recurred, he seemed so exhausted that mechanical aspiration of the tracheobronchial tree was resorted to at once. In this case, the correction of wet lung was clearly lifesaving.

Case 6.-A 21-year-old soldier, injured in an automobile accident, complained of hemoptysis, persistent pain on coughing, and pain on deep breathing. When he was admitted to a thoracic surgery service 48 hours after the accident, he was dyspneic and complained of discomfort and pain in the left lower chest. There was no external evidence of injury. The percussion note was resonant bilaterally, but there were diffuse wheezes and rhonchi throughout both lungs. Pain was so severe that coughing was impossible, in spite of urging. Pressure over the seventh, eighth, and ninth ribs in the left anterior axillary line elicited exquisite tenderness, but rontgenograms of the chest revealed no abnormalities.

In spite of failure to demonstrate any rib fractures, an intercostal block of the area of tenderness completely relieved the pain. Shortly afterward, the patient coughed up several tablespoonsful of thick, white, tenacious mucus. Two hours later, physical examination revealed entirely normal findings in both lungs. Recovery was uneventful thereafter.

Comment-In this case, nerve block permitted effectual coughing by relieving the pain that had prevented it previously. There was only mucus in the tracheobronchial tree, and it may be that this is an instance of an abnormally large secretion of mucus because of reflex stimulation via the intercostal nerves. If this reasoning is correct, one effect of the nerve block may have been to reduce the secretion of mucus.2


1. Betts, R. H., and Overholt, R. H.: The Prevention and Treatment of Postoperative Pulmonary Complications by Bronchoscopic Aspiration. S. Clin. North America 17: 885-893, June 1937.

2. De Takats, G., Beck, W. C., and Fenn, G. K.: Pulmonary Embolism. An Experimental and Clinical Study. Surgery 6: 339-367, September 1939.

3. Drinker, C. K., and Warren, M. F.: The Genesis and Resolution of Pulmonary Transudates and Exudates. J.A.M.A. 122: 269-273, 29 May 1943.

4. Pasteur, W.: Massive Collapse of the Lung (Syn. Active Lobar Collapse). Brit. J. Surg. 1: 587-601, April 1914.

5. Churchill, E. D.: Pulmonary Atelectasis, With Especial Reference to Massive Collapse of the Lung. Arch. Surg. 11: 489-518, October 1925.

6. Lewis, Thomas: Pain. New York: The Macmillan Co., 1942.

7. Haight, C.: Intratracheal Suction in the Management of Postoperative Pulmonary Complications. Ann. Surg. 107: 218-228, February 1938.

8. Barach, A. L., Martin, J., and Eckman, M.: Positive Pressure Respiration and Its Application to the Treatment of Acute Pulmonary Edema. Ann. Int. Med. 12: 754-795, December 1938.

2The reader is referred to chapter XI (p. 441) for long-term followup studies on casualties whose chest wounds were complicated by the wet lung syndrome.