History

CHAPTER VI

Hemothorax and Hemothoracic Empyema

Thomas H. Burford, M.D.

GENERAL CONSIDERATIONS

At the beginning of World War II, there was little or norealization of the fact that hemothorax was the most frequent, and probably themost important, complication of wounds of the chest. A brief experience withthese wounds, however, compelled the recognition of both its frequency and itsimportance. The principles and techniques by which hemothorax was managedinfluenced the entire scheme of the treatment of chest wounds from forwardhospitals through chest centers.

In view of these facts, it is surprising to find that hemothorax hadattracted very little attention before World War II. Surgeons who dealt with itat all were divided into two schools. One group maintained that absorption ofthe sanguineous intrapleural material would invariably take place, and withreasonable promptness. The other group advocated the prompt evacuation of thepleural cavity, on the premise that blood, by its mere presence, was harmful.Even those who did not accept the theory that all hemothoraces were absorbed bynatural processes believed that this occurred in 90 to 95 percent of allthoracic wounds.

When the literature of the prewar period is examined, one cannot fail to beimpressed with the small number of observers who concerned themselves with thefate of the 5 or 10 percent of patients whose hemothoraces were not absorbed.This lack of concern was evident in the military manual on thoracic surgeryissued in 1943 under the auspices of the American Association for ThoracicSurgery and the National Research Council (1). Blood in the chest,according to this manual, tends to remain fluid, probably because it isdefibrinated by the motions of the lungs and the heart. A section was devoted tothe management of infected hemothorax, but no recommendations were made fordealing with the exceedingly important complication of clotted hemothorax. Theomission was typical. It was also unfortunate. The casualties whose hemothoracesclotted or did not absorb for other reasons constituted the group who developedempyemas that tended toward chronicity and who presented encased lungs orcontracted hemithoraces, with resulting impairment of respiratory function.

Until articles on unabsorbed hemothorax began to be published from theMediterranean Theater of Operations, U.S. Army, in World War II, not manyobservers even mentioned the possible clotting of blood in the pleural cavity.

238

It is also startling, in retrospect, to find in the prewarliterature almost no effort to distinguish postpneumonic and metapneumonicempyema from posttraumatic empyema, although the latter variety, if notcorrectly managed, gives rise to a far higher incidence of chronicity. Bothvarieties were discussed as if they were of the same origin and had the samecharacteristics. Studies on hemothorax as an antecedent to empyema wereremarkable for their paucity.

The recognition of hemothorax as perhaps the most important complication ofwounds of the chest was a development of World War II. Had the prewarlaissez-faire approach to it been carried over into military surgery, it wouldhave resulted in an intolerably high rate of pleural sepsis in World War II.

Nomenclature-Before proceeding with the discussion ofhemothorax and its sequelae, it is necessary to define the meaning andimplications of the terms used for this condition in World War II:

The term "simple hemothorax" implied the presenceof blood in the pleural cavity, without complicating features. This term wasused whether the blood was the result of injury to the chest wall alone or wasthe result of a penetrating or a perforating wound.

The term "infected hemothorax" implied the presence of bacteria inthe pleural fluid, plus the clinical manifestations of their presence.

The term "chronic hemothorax" or "chronic traumatic hemothorax"was used to indicate any collection of blood in the pleural cavity that couldnot be reduced by thoracentesis. The term implied that fresh blood had ceased toenter the pleural space, that hemostasis had been achieved, and that thecollection of blood was therefore residual. A chronic hemothorax was usually anorganizing hemothorax. It was the clotting of the blood plus its inevitablesequela, organization, that made aspiration no longer possible. A neglectedaccumulation of intrapleural blood that could be satisfactorily aspirated wasnot considered chronic because it could be evacuated at will by propertreatment.

There was general agreement that the term "chronichemothorax" was both inadequate and misleading. It did not indicate, oreven suggest, the really important element in the pathologic process. Thiselement was the presence of fibrin, whether it was derived from blood spilledinto the pleura, or from exudate formed by the pleura, or had been deposited insome other manner. Some such term as "fibrothorax" would have beenmore expressive and more correct, but the term "chronic traumatichemothorax" began to be used, and semantic changes were not practical inwartime.

INCIDENCE OF HEMOTHORAX AND ITS COMPLICATIONS

Hemothorax-Statements concerning the incidence ofhemothorax in combat-incurred wounds of the chest are necessarily imprecise.Probably every casualty with a perforating or a penetrating thoracic wound had ahemo-

239

thorax of some degree. In a certain proportion of injuries,perhaps 15 or 20 percent, the collection of blood was so small as to cause noclinical manifestations, and it was rapidly absorbed by natural processes.

Various studies from the hospitals in the Mediterraneantheater, including the chest centers, show the incidence of hemothorax to rangefrom 80 to 84 percent. Well over half of the casualties received in fieldhospitals after wounding already had enough blood in the pleural cavity to makeits removal an urgent part of the effort to restore the cardiorespiratoryequilibrium. More significant than the high initial incidence is the fact thatabout three-quarters of the casualties received in base hospitals arrived withsufficient intrapleural blood to require treatment.

In the great majority of hemothoraces, once the policy ofearly and complete thoracentesis had been established, recovery was prompt anduncomplicated in the absence of complications from other components of thethoracic injury. In 752 hemothoraces treated at the 300th General Hospital chestcenter in the Mediterranean theater, between 1 April and 20 December 1944, byMaj. Thomas H. Burford, MC, and Maj. (later Lt. Col.) Edward F. Parker, MC,there were no complications in 603 of 678 cases in which the blood could beremoved by aspiration. The average time required for complete clearing of thechest was from 10 to 14 days, and the average number of thoracenteses requiredwas from 3 to 4.

Clotting and organization-The hemothoraces that gave rise todifficulties were those in which clotting occurred. Exactly how often thishappened, it is not possible to say. During the Italian campaign, some forwardhospitals reported incidences of clotting ranging to 25 and 30 percent. Theselevels are exceptionally high. In one or two later studies, the incidence was 9or 10 percent, which is rather low. Clotting probably occurred in about 15percent of all hemothoraces, even when thoracentesis was undertaken early andwas repeated as necessary.

The incidence of clotting naturally varied with circumstances. It was morefrequent in severe wounds. It was also more frequent the longer thoracentesiswas delayed. A casualty who was brought promptly to a hospital near thefrontline and whose chest was evacuated as part of the regimen of resuscitationseldom developed an organizing hemothorax.

In nearly all instances of extensive clotting, some degree of organization ofthe clot followed. Infection, with its clinical manifestations, could and didoccur in many unclotted hemothoraces, but it was much more likely to occur inlarge clotted collections. Two other consequences of clotting were also serious:

1. If an organizing hemothorax was of any size, simpleremoval of the clots at thoracotomy or attempted evacuation of the pleuralcavity by water-seal intercostal drainage practically always proved inadequate,and the resulting pulmonary reexpansion was entirely unsatisfactory.

240

2. If expectant treatment was employed in organizing hemothorax, the periodrequired for resolution was always prolonged. In one group of 13 patientsobserved late in the war, the total time for clearing ranged from 41 to 125 daysand averaged 78 days. Clotting of a hemothorax was therefore an important causeof prolonged morbidity.

Infection and hemothoracic empyema-Most infected hemothoraces werepreceded by clotting, and most posttraumatic empyemas were preceded by infectedhemothoraces. There was seldom any doubt of the sequence of events. In a smallnumber of cases, even uninfected organizing hemothoraces produced extremedegrees of respiratory crippling, though the worst of these cases did notcompare in severity with the crippling observed in patients in whom infection ofa hemothorax had been permitted to become chronic. During World War I, chronicempyema was the most frequent complication of chest wounds (2). Almost aquarter of the thoracic casualties died, and from 25 to 30 percent developedempyema. If these same proportions had prevailed during World War II, theresults of chest wounds would have been disastrous.

In the early months of fighting in North Africa, before the management ofhemothorax had been standardized, the incidence of empyema was high, running to25 or 30 percent in some hospitals. Even in the chest center at the 53d StationHospital, Bizerte, which was equipped and staffed for the specialized care ofchest injuries, it was 22.6 percent. By the end of the war, the generalincidence in the theater did not exceed 10 or 11 percent, and more than onehospital had achieved an incidence as low as 7 percent. In the European Theaterof Operations, U.S. Army, which had the early experience of the Mediterraneantheater to guide it, chronic hemothoracic empyema was also infrequent.

An incidence of even 7 percent is much higher than theincidence of empyema after chest wounds in civilian practice. The explanation ofthe difference is simple: In civilian life, the agents of wounding are usuallyice picks and bullets, which have a limited and localized destructive capacity.In combat-incurred injuries, on the other hand, while bullets were frequently asinnocuous from this standpoint as they are in civilian life, most wounds werelarge wounds, and tissue destruction was extensive. Moreover, in combat-incurredwounds, the missiles often carried with them into the chest fragments offractured ribs, bits of clothing, and other debris. Retained metallic foreignbodies, unless they were very large, played no significant role in mostintrapleural infections (p. 327).

In World War II, the factors most frequently responsible forposttraumatic empyema were delayed and inadequate evacuation of the pleuralcavity, inadequate debridement, inadequate drainage of the pleural cavity, andexcessively prolonged treatment by aspiration and instillation of penicillinwhen it was obvious that radical therapy was required. The necessaryinterruption of professional supervision of casualties in the chain ofevacuation was probably responsible for a small number of cases early in thewar. How serious a part

241

inadequate early treatment plays in the development ofempyema is evident in figures reported by d'Abreu (3): When the Britishevacuated 88 soldiers of all nationalities who had been engaged in guerrillawarfare in Yugoslavia and who had had no early surgical treatment for theirchest wounds, they found that 48 of them had empyema.

The reduction in the incidence of posttraumatic empyema as the war progressedwas the result of several factors, the most important of which was theincreasing emphasis put upon early, complete evacuation of the pleural cavity,which was followed by prompt reexpansion of the lung. The advent of penicillin,with its beneficial effect on residual infection, also played a part. Penicillinrepresented an improvement over the sulfonamides, but it was not the majorfactor in the improvement. The most important single factor in the loweredincidence of posttraumatic empyema was the improvement in the surgery of chestwounds that came to pass with the increased experience of general and thoracicsurgeons and of anesthesiologists.

Figures for empyema supplied by the Medical StatisticsDivision, Office of The Surgeon General, appear in tables 13 and 14. Inconnection with table 13, attention is called to the fact that data availableonly for 1944 and 1945

TABLE 13. -Admissionsfor empyema of the chest in the U.S. Army, by area and year, 1942-45

[Preliminary data based on sample tabulationsof individual medical records]
[Rate expressed as number of admissions per annum per 1,000 average strength]

Area	1942-45		1942		1943		1944		1945
Area	Num- ber	Rate	Num- ber	Rate	Num- ber	Rate	Num- ber	Rate	Num- ber	Rate
Continental United States	406	0.03	121	0.05	175	0.03	75	0.02	35	0.01
Overseas:
Europe	100	0.02	2	0.02	13	0.05	20	0.01	65	0.03
Mediterranean¹	14	.01	---	---	5	.01	9	.01	---	---
Middle East	4	.03	---	---	4	.08	---	---	---	---
China-Burma-India	1	.00	---	---	---	---	1	.01	---	---
Southwest Pacific	15	.01	1	.01	2	.01	7	.01	5	.00
Central and South Pacific	25	.02	2	.01	11	.04	7	.02	5	.01
North America²	20	.04	5	.05	11	.06	4	.03	---	---
Latin America	4	.01	2	.02	2	.02	---	---	---	---
Total overseas³	201	0.02	12	0.02	48	0.03	51	0.01	90	0.02
Total Army	607	0.02	133	0.04	223	0.03	126	0.02	125	0.02

¹Includes NorthAfrica.²
Includes Alaska and Iceland.³
Includes admissions on transports.

242

show that the incidence of empyema during this period was nearly six timeshigher than the number of admissions due to the condition.¹

For 1944 (table 13), preliminary tabulations of individual medical recordsshowed 126 admissions and 805 secondary cases of empyema. The figure for thesecondary cases was derived from a 20-percent sample, consisting of 161punchcard records. One-third of the cases of empyema occurred among admissionsin the Zone of Interior and two-thirds among admissions overseas. In the lattergroup, more than 70 percent of the cases were of battle origin. Three-fourths ofthe remaining cases originated in disease and the other quarter in nonbattleinjuries. Nearly 90 percent of the empyemas secondary to battle in-

TABLE 14.-Distribution of a 20 percentsample¹ of secondary cases of empyema of the chest inthe U.S. Army, 1944,² by area of admission, type of admissionof primary condition, and evacuee status

[Preliminary data based on sample tabulationsof individual medical records]

Type of admission	Continental United States	Overseas
Type of admission	Continental United States	Europe	Mediterranean	Middle East	China-Burma- India	Southwest Pacific	Central and South Pacific	North America
	EVACUATED
Battle injury and wound:
Thoracic	---	36	8	---	1	4	1	---
Thoracoabdominal	---	5	8	---	---	---	---	---
Other regions	---	8	1	---	---	---	---	---
Total	---	49	17	---	1	4	1	---
Nonbattle injury:
Thoracic	---	1	---	---	---	---	---	---
Thoracoabdominal	---	1	---	---	---	---	---	---
Other regions	---	1	---	---	---	---	---	---
Total	---	3	---	---	---	---	---	---
Disease:
Pneumonia	---	1	---	---	---	---	---	---
Bronchiectasis	---	1	---	---	1	---	---	---
Other diseases	---	5	1	---	---	1	1	1
Total	---	7	1	---	1	1	1	1
Grand total	---	59	18	---	2	5	2	1

¹It should be pointed out again that the term "admissions" refers to patients admitted to medical treatment facilities for the specified condition as the primary cause of admission, while the term "incidence" includes not only these patients but also patients in whom the specified condition was reported as secondary to other admission diagnoses.

TABLE 14.-Distribution of a 20percent sample¹ of secondary cases of empyema of the chestin the U.S. Army, 1944,² by area of admission, type ofadmission of primary condition, and evacuee status- Continued

[Preliminary data based on sample tabulationsof individual medical records]

Type of admission	Continental United States	Overseas
Type of admission	Continental United States	Europe	Mediterranean	Middle East	China-Burma- India	Southwest Pacific	Central and South Pacific	North America
	NOT EVACUATED
Battle injury and wound:
Thoracic	---	1	---	---	---	1	---	---
Thoracoabdominal	---	---	1	---	---	---	---	---
Other regions	---	---	---	---	---	---	---	---
Total	---	1	1	---	---	1	---	---
Nonbattle injury:
Thoracic	4	---	---	---	---	1	---	---
Thoracoabdominal	---	---	---	---	---	---	---	---
Other regions	1	1	---	---	1	1	---	---
Total	5	1	---	---	1	2	---	---
Disease:
Pneumonia	26	3	4	2	---	---	---	---
Bronchiectasis	2	---	---	---	---	---	---	---
Other diseases	21	---	2	---	---	---	---	---
Total	49	3	6	2	---	---	---	---
Grand total	54	5	7	2	1	3	---	---

¹Two sample cases-oneadmission each in the European theater and in the Southwest Pacific Area-havebeen omitted from the table since they could not be located in the card file.²1944 refers to year in which primary cause of admission occurred. Theempyema may have occurred in that year or subsequent on 1944.

jury followed injuries of the chest or thoracoabdominal region. There were115 separations from service for empyema during the 1942-45 period.

PATHOGENESIS OF ORGANIZING HEMOTHORAX

Natural History of Simple Hemothorax

Origin of blood-Though it was not alwayspossible to determine the exact origin of the blood in the pleural cavity, byfar the most frequent source was wounds of the lung. Hemorrhage from the lung,however, tends to spontaneous arrest within a short time. Lacerations ofintercostal vessels were the

244

FIGURE 71.-Hemothorax with recurrentpulmonary hemorrhage. A. Posteroanterior roentgenogram showing extensive righthemothorax, with retained foreign body, 7 days after wounding. B. Same, 7 dayslater, after secondary hemorrhage from lung.

second most frequent source; in many instances, bleeding fromthese injuries was combined with bleeding from wounds of the lung. In a muchsmaller number of injuries, the source of the bleeding was the internal mammaryvessels; the vessels of the heart, the pericardium, or both; and the largervessels, such as the innominate and subclavian arteries and the azygos vein. Ashas already been pointed out, casualties with injuries of the great vessels ofthe chest seldom survived to reach a field hospital.

Secondary hemorrhage into the pleural cavity was extremely uncommon. It wasrecorded only 3 times in 752 cases of hemothorax studied by Major Burford andMajor Parker. In only one of the three cases did the hemorrhage recur from thelung (fig. 71), 2 weeks after wounding. In the second case, the recurrence wasfrom a laceration of the intercostal artery. In the third case, an unsuspectedtraumatic aneurysm of the right subclavian artery ruptured, and the patient diedalmost at once (p. 45). The roentgenologic examination (fig. 72) did not arousethe suspicion that a lesion of such potential gravity could coexist with thecharacteristic picture of clotted hemothorax.

Effects of blood on pleural structures-In spite of the problemhemothorax has presented, at least in combat wounds, there have been remarkablyfew studies of the intrapleural changes associated with it. The fluoroscopicstudies made by Edwards and Davies (4) in 1940 represent one of the firstserious efforts to ascertain the nature of the pleural response to blood in thepleural cavity. Before the war, opinions differed concerning its irritativeeffects. The World War II experience was not conclusive, and final judgment mustbe reserved.

245

FIGURE 71.-Continued. C. Same, 2 days later,showing residual clotted hemothorax after two thoracenteses. D. Same, afterthoracotomy for removal of foreign body and decortication of right lung. Atoperation, it was found that the secondary hemorrhage had occurred from alaceration of the lung at site of entry of missile.

Little evidence was produced in World War II to substantiatethe theory that unaltered liquid blood acted as an inflammatory agent. Thisgeneralization held, however, only while the blood remained unaltered. Thereseemed no doubt that it underwent rather profound physiochemical changes as soonas it left its normal endothelial-lined habitat and that these changes wereprogressive in both kind and degree. The first fluid aspirated after woundingwas pure blood or distinctly bloody, but its character changed as aspirationswere continued. It became serosanguineous, and then, before the chest wasentirely emptied, it became merely serous.

A fibrin layer was uniformly found on the pleural surface inall hemothoraces. It was always more pronounced in clotted than in unclottedcases. Since clotted blood not only binds fibrin but also represents a greaterphysiochemical departure from normal than a liquid hemothorax, in which fibrinis free, the inference was that the fibrin layer represented an inflammatoryresponse rather than a passive deposition of fibrin or a salting-out phenomenon.As would be expected, the response was proportional to the degree of alterationof the blood; that is, whether it was (1) clotted or (2) clotted andcontaminated.

In the hemothoraces controlled by thoracenteses, laboratorystudies showed that the pleural fluid was low in protein and relatively free ofred blood cells. Even in uncomplicated hemothoraces, as just pointed out, acertain amount of fibrin was deposited over the pleural surfaces, but the thin,shaggy coat was not demonstrable on roentgenologic examination, and the fibrincontent of the intrapleural fluid was so small that total aspiration wasentirely feasible (fig. 73).

246

FIGURE 72.-Posteroanterior roentgenogramshowing clotted right hemothorax. There is nothing in the film to suggest thetraumatic aneurysm of the first portion of right subclavian artery, rupture ofwhich was immediately fatal.

Theories of Origin of Organizing Hemothorax

The pathogenesis of organizing hemothorax was no clearer atthe end of the war than at the beginning. There were no facilities in a combatzone for experimental studies on why clotting occurred in some traumatichemothoraces and did not occur in others. Clinical and pathologic observationsin a large number of chest injuries do, however, permit a certain amount oflegitimate speculation.

Certain causes could be eliminated at once. An analysis ofthe missiles that caused the wounds associated with clotted hemothorax revealedno common denominator to explain what occurred. The degree of pneumothoraxpresent apparently played no part, as might have been expected; the presence orabsence of air does not influence the clotting of blood. The relative severityof the wounds of the lungs and soft tissues played no decisive role; hemothoraxwas more frequent in severe injuries because severe injuries were more frequent.Organizing hemothorax occurred in some cases in which the injury was limited tothe external chest wall and also occurred in some contusions and blast injuries.

With these obvious factors eliminated, it was necessary toturn to the mechanism of clotting for a possible explanation of clotted ororganizing hemothorax. There were three theories:

1. The first hypothesis was that bacteria might be present in the pleuralcavity and, in the presence of blood, might incite an exudative response, which

247

FIGURE 73.-Massive liquid hemothorax withoutinfection. A. Posteroanterior roentgenogram made shortly after wounding andshowing hemothrax. B. Same, 10 days later, after six thoracenteses, by which atotal of 3,150 cc. of blood and other fluid was removed.

would favor the formation of fibrin deposits, rather than atransexudative response. If bacteria were seeded generously over the pleuralsurfaces, a frankly suppurative response would be expected, and a truesuppurative pleuritis would develop. If, however, they were implanted in suchsmall numbers as not to incite a widespread pleural response, their presencemight still be sufficient to alter the response from transexudative to exudative,and the clinical picture would be dominated by the presence of fibrin depositsrather than by suppuration.

The sharp and often prolonged febrile responses seen inorganizing hemothorax furnished clinical evidence of the exudative character ofthe pleural response to blood. In uninfected hemothoraces, a moderate febrileresponse often occurred, but it was practically always of short duration,usually no longer than from 10 to 14 days. In infected cases, the clinicalmanifestations continued, and if suppuration were impending, they became morepronounced.

Bacterial seeding probably explained some instances oforganizing hemothorax, but it could not be accepted as a universal explanation.It was undoubtedly a factor in some hemothoraces which went on to empyema. Thefact that nonhemolytic Staphylococcus aureus was the most frequentcausative organism in empyema and that this organism does not produce fibrolysinlent some support to this theory. Clotted blood containing pathogenic bacteriaserved as an excellent culture medium, just as devitalized muscle served as aculture medium in all wounds. This was particularly true of anaerobes, whichdevelop rapidly under low oxygen tension. Clot formation about bacteria inhemothorax also had the effect of preventing penicillin or the sulfonamides fromreaching these entrenched organisms. They were not influenced by either local orsystemic therapy.

248

2. The second hypothesis was that clotting in hemothorax might be the resultof the simultaneous presence of blood and coagulative ferments in the pleuralcavity. This theory was not considered acceptable. For one thing, the actualamount of blood present seemed of no importance; the preponderance of white clotpresent in some cases pointed to an excess of plasma. There was also nouniformity in the time at which clotting occurred. Clotted blood was sometimesfound in the pleural cavity when thoracotomy was performed from 4 to 6 hoursafter wounding. Sometimes, the clot was so firm a few days after wounding thatit defied needle aspiration. In other instances, the blood might remain liquidfor many days, and all of it could be removed by thoracentesis.

There was no acceptable explanation for these variables. Itmight be that the churning movements of the diaphragm, heart, lungs, and chestwall had a defibrinating effect. Sellors (5) remarked that any surgeonwho opened a recent hemothorax must have been impressed "by the tumultuousway in which the blood is flung about the pleural cavity." Fibrin and bloodclots were often seen in the costophrenic sinus and the lower portion of thepleural cavity within a short time after wounding. Another consideration wasthat both the lungs and the muscles are ready sources of thromboplastin.

Even granting all of these facts, acceptance of the theorythat the presence of whole blood and coagulative ferments was responsible fororganizing hemothorax would leave many cases unexplained.

3. The most reasonable hypothesis was that, as the result of trauma, normalpleural fluid was so modified that it became exudative rather than serous. Ashas already been pointed out, a response by the pleura could be expectedwhenever blood was spilled into the pleural cavity as the result of wounding,even if the process went no further. Moreover, wounds of the lungs and pleuracannot be fundamentally different from wounds of other tissues. If thisreasoning is sound, the so-called edema of wounding might be expected to occurin these tissues just as in other damaged tissues. The presence of edema, whichreached its peak within 2 to 4 days after wounding, could reasonably be assumedto affect the character of the pleural effusion caused by trauma, in allprobability making it richer in fibrin. There would then be present in thepleural cavity an excess of coagulative ferments from the wounded chest wall orlung or both, combined with an excessive fibrin from the pleural effusion. Inother words, there would be added to the whole blood spilled into the pleuralcavity, which might itself clot by natural physiologic processes, incrementsfrom the pleural response to wounding that would enhance the natural process ofclotting.

It is easy to explain why early thoracotomy and evacuation of whateverintrapleural fluid was then present did not necessarily protect the patientagainst the development of a subsequent clotted intrapleural mass. Wounds of thethorax frequently resulted in the widespread changes known as wet

249

lung (p. 207). Wet lungs were heavy and soggy, and theyexpanded only sluggishly and never completely in the absence of vigorouscorrective measures. For a few days after wounding, therefore, dead space²in the pleural cavity permitted the accumulation of pleural exudate thateventually became a part of, or constituted, the clotted intrapleural mass. Ifall the blood in the pleural cavity had been evacuated at thoracotomy orotherwise, the pleural exudate that continued to accumulate under thesecircumstances was chiefly white. If all the blood had not been removed, the clotwould be layered or marbled.

The hypothesis seems reasonable that a change in thecharacter of the normal pleural fluid furnishes an explanation for the clinicalvagaries observed in clotted hemothorax. It also seems to explain the sequentialrelation of this process to the original trauma.

PATHOLOGIC CHANGES IN CHRONIC TRAUMATICHEMOTHORAX

The limited pathologic studies possible in a combat zoneseldom warrant generalizations on any subject. A large enough number ofhemothoraces were observed, however, to elucidate the pathologic changes inchronic traumatic hemothorax and to establish the rationale for its management.

Gross Pathology

All hemothoraces, as already pointed out, involved typicaland definite changes within the pleural cavity, slight at first but morepronounced as the process went on to clotting and organization.

The process of hemo-organization began with the laying downof a thin, loosely adherent film of fibrin and blood cells over both pleuralsurfaces. Within 2 or 3 days of its formation, angioblastic and fibroblasticproliferation began to occur.³ As early as from7 to 14 days after wounding (figs. 74 and 75), fibroblasts could be seen growingout from the older (pleural) side toward the younger portion. The transition toadult fibrous tissue was progressive and rapid (fig. 76). Capillaries grew infrom the pleura, and within 8 to 10 weeks, complete cellular symphysis hadoccurred in most cases. Decortication, which would have been simple earlier, hadnow often become impossible (figs. 77 and 78).

²Dead space may be defined as any air or fluid or matter without the confines of a normal anatomic habitat and without a blood supply.
³These observations were made on specimens secured at operation for bronchopleural fistulas, removal of foreign bodies, and similar conditions. Small clotted hemothoraces were often found at these procedures, and, when interest in this complication of thoracic wounds had developed, sections were regularly taken for study.

250

FIGURE 74.-Photomicrograph showing organizinghemothorax with replacement of acute fibrinopurulent deposit by young vascularconnective tissue at end of first week after wounding. The pleural aspect of thelayer of exudate removed from the visceral pleura is toward left and peripherytoward right.

The essential feature of chronic traumatic hemothorax was theformation of this fibroblastic and fibrinofibrous membrane or peel over thevisceral and parietal pleurae. Once it had formed, the hemothorax was convertedinto an enclosed hematoma of the pleural space (fig. 79).

It should be emphasized that, in the beginning, the pleura itself took nopart in the pathologic process. Until fusion between it and the rind occurred,it remained remarkably normal in appearance. The pleural thickening which wassometimes described on roentgenograms was never observed at operation. Thepractice of describing peripheral obscuration as thickened pleura was incorrectand misleading. In the early phases of an organizing hemothorax, examination ofsections of the encasing sac never revealed elastic tissue, although elasticfibers were always found in sections from the lung or the soft

251

FIGURE 75.-Photomicrograph showingfibroblastic membrane from clotted hemothorax 11 days after wounding. Note that,even at this early stage, the hemothorax has a definite capsulelike character.Capillaries have already begun to appear in older (pleural) side at bottom ofspecimen. Hematoxylin and eosin stain. (X 60)

tissue of the thoracic wall just beneath the pleuralmesothelium. If, however, the process went on without interruption, the membraneeventually underwent complete fibrous and vascular union with the pleura, whichthen lost its identity as a delimiting membrane.

The facts just stated were of great importance in theselection of the time for decortication (p. 286). The ease with which a cleavageplane could be established between the two layers (the rind and the pleura) atoperation was a direct function of the age of the limiting membrane (fig. 78).Experience showed that after 4 weeks from the date of injury the chances ofperforming satisfactory decortication dropped rapidly with each succeeding week.

The development of infection, even if it went on tosuppuration, did not accelerate the process of organization of the hemothorax toany appreciable degree. There was also no evidence that the union between thepeel and the underlying pleura was any firmer when infection was present.

The most prominent feature of the intrapleural mass was that it was alwaysrelatively or absolutely solid, because of the fibrin component. It was usuallythickest in the basal thoracic gutter posteriorly. The distribution of

252

FIGURE 76.-Photomicrograph showing section ofpeel of clottedhemothrax 4 weeks after wounding. Membrane has assumed adefinitely fibrous character. Capillaries are visible at their point of entrancefrom visceral pleura (lower left of section). Hematoxylin and eosin stain. (X100)

the fibrin thus followed the natural principles of gravity as they wouldoperate in any recumbent subject with an accumulation of fluid in the pleura.

In the typical organizing hemothorax, the mass was made up of three layers(fig. 79):

1. The outer layer was a tough organized membrane (rind, peel), the thicknessof which was related to the length of time that had elapsed since wounding. Atthe end of 4 or 5 weeks, the rind was usually from 2 to 4 mm. thick. Later, itmight be from 5 to 8 mm. thick, or even thicker. If the mass was composed ofsolid fibrin, the rind was always very thick.

This rind completely surrounded the original hemothorax and was intimatelyattached to the visceral and parietal surfaces of the pleura, though usually, inspite of the intimacy of the adhesions, it could be separated without difficultyfrom the visceral pleura, at least in early stages of the process. The

253

FIGURE 77.-Photomicrograph showing sectionfrom clotting hemothorax. Note fusion between dense layer of collagenousconnective tissue and subjacent pleura and lung (lung is to left).

rind was resistant, did not tear readily, and had the gross appearance offibrous tissue.

The rind usually passed directly from the visceral to the parietal pleura atall points at which these surfaces met, but attenuated projections frequentlyextended beyond these limits, along opposed pleural surfaces or into the interlobar fissures. The costophrenic gutter was almost consistently obliterated bythe elevated diaphragm and was sealed off by the reflection of the peel. Beyondthese points of symphysis, it was often surprising to find only a thin layer offilmy pleural adhesions. At a greater distance, entirely free pleural spacemight be encountered.

2. The middle layer of the chronic traumatic hemothorax was a soft,unorganized mass, with the consistency of fibrin. It differed from patient topatient in gross appearance, chiefly because of differences in color. Sometimes,it was almost white. Sometimes, it was the color of mahogany. Sometimes, it was

254

FIGURE 78.-Photomicrograph of section throughmembrane, pleura, and lung (lung is at lower margin) in posttraumatic empyema 12weeks after wounding. Note cellular intimacy. No line of demarcation can beseen, and surgically no line of cleavage could be established. This patient'schance for decortication was forfeited by delay in resorting to it. Van Gieson'sstain. (X 100)

definitely bloody. A certain amount of marbling was notuncommon, but it was more frequent to find the middle layer the same colorthroughout. Differences in color seemed to depend upon the relative proportionsof blood and plasma in the mass or upon the amount of pleural effusion orpleural exudate that had originally accumulated and that had been modified bywhatever degenerative changes might have occurred with the passage of time.

3. The third layer (center) of the intrapleural mass was typically lobulated(fig. 79). It might consist of a single large space or of a network of spaces.Sometimes, it was a mushy core from whose interstices fluid could be squeezed.The fluid looked like what it was, supernatant fluid from a mass of retractingfibrin. It varied in color from yellow to red, brown, or mahogany.

The formation of the central fluid-filled space or spaces was clearly theresult of retraction of the original clot. In fact, the whole pathologic processin organizing hemothorax did not seem to differ in any important way from thewell-known process of organization in intravascular thrombi.

The foregoing description bears out what has already been said (p. 238), thatthe term "organizing hemothorax" or "chronic traumatic hemothorax"

255

FIGURE 79.-Cross-sectional diagram showingpathologic process in clotting, organizing hemothorax. Membranous peel (rind),essentially a fibrinofibrous envelope, is shown investing visceral and perietalpleurae. In center is loculated purulent material. Loculation explains why inmany infected hemothoraces infection could not be diagnosed by thoracentesis,because aspirating needle did not reach center of process, which was recognizedonly at operation.

carried no implication of the origin or gross appearance of the intrapleuralmass. The nomenclature suggests that the mass might be the result of thepresence of clotted blood alone, as in some instances it was. The gamut was run,however, from such cases through (1) those in which a considerable portion ofblood was still present but in which there were also present increments of whiteclot (that is, plasma, fibrinous pleural effusion, or fibrinous exudate) to (2)those in which the entire mass seemed to consist of white clot. The variationsseemed due not so much to the amount of whole blood originally spilled into thepleura as to the pleural response to its presence.

Histopathology

Histologic examination of sections from an organizinghemothorax confirmed the gross observations (figs. 74-78 and 80-85). As alreadypointed out, as early as 7 to 10 days after wounding, there was microscopicevidence of fibroblastic and angioblastic proliferation from the pleuralsurfaces to the clot. The peel or rind thus formed increased progressively inthickness as the fibroblastic invasion continued.

256

FIGURE 80.-Photomicrograph showing later offibrin 3 mm. thick removed from visceral pleura in uninfected clotted hemothorax11 days after wounding. Note beginning organization in portion immediatelyadjacent to pleura (on left).

Within 3 or 4 weeks after wounding, adult fibrous tissue was demonstrable inthe outer layer of the organizing hemothorax. The fibers and nuclei werearranged roughly parallel to the surface. The long axes of the capillaries wereat right angles to the surface, and the capillaries were continuous through thepleura.

Within 4 to 7 weeks after wounding, small arterioles, with smooth musclefibers in their walls, could be demonstrated at or near the outer surface of thepeel. Active fibroplasia continued along the inner (younger) surface, withresulting increases in the thickness of the rind. In some instances, layers offat cells developed in the peel. along the outer surface. It was speculated, butnever proved, that this phenomenon might possibly represent a regressiveprocess, heralding eventual degeneration and absorption of the peel. For reasonsnot clear, the membrane was always thicker, more vascular, and more adherent onthe parietal than on the visceral pleura.

257

FIGURE 81.-Fibrotic process in organizinghemothorax. A. Photomicrograph showing section of fibrofibrinous layer removedfrom visceral pleura in uninfected clotted hemothorax. Note fibrosis in portionof rind adjacent to pleura at top. B. Photomicrograph showing more pronouncedfibrous changes at distal free margin of same section.

258

FIGURE 82.-Photomicrograph showing calciumdeposition in collagenous connective tissue layer of exudate on visceral pleurain organizing hemothorax.

Microscopically, the rind in hemothoracic empyema could notbe differentiated from the rind in organizing hemothorax without infection. Theonly evidence of inflammation seen under the microscope was observed on theinner surface of the rind, along which fibrin or blood clot was in process oforganization. It was limited to infiltration by varying numbers ofpolymorphonuclear leukocytes (fig. 85).

The second (middle) layer appeared on microscopic examinationto be merely a mass of fibrin with scattered, formed blood elements trapped inits meshes.

Pulmonary Effects of Organizing Hemothorax

When the process of organization was complete, a dense fibrousmembrane encased some portions of the lung, or the whole lung, thus limiting itsmobility and preventing its complete expansion. A contraction of the hemithoraxwas sometimes present also. Even in the early stages of an organizing hemothorax,it was easy to see that the tough, resistant, inelastic rind that invested thevisceral pleura effectively prevented pulmonary reexpansion. If the process werenot terminated by surgical measures, fibrosis increased, and the rind and thepleura became intimately adherent. As time passed, tough scar tissue extendedthrough the pleura into the interstitial pulmonary tissue. The degree of cel-

259

FIGURE 83.-Photomicrograph showing biopsiedsection of lung and pleura after removal of overlying exudate. Note that pleuraretains most of its normal characteristics.

lular intimacy made it difficult, and sometimes impossible,to develop a cleavage plane at operation, particularly if the pleura had beendenuded or the lung wounded. In such cases, the results of surgery were notlikely to be good: The lung, even when liberated by the procedure ofdecortication, could not expand because the fibrotic process had extended to it.When complete pulmonary reexpansion was impossible, the stage was set for thedevelopment of infection.

When the intrapleural mass was removed, it was frequentlyevident that the lung had not only been compressed by the intrapleuralaccumulation but had also been greatly displaced and had sometimes been infoldedalso (p. 183).

Resolution

Massive organizing hemothorax not treated by surgery couldterminate in two ways: (1) by the formation of a fibrothorax or pleuralcalcification (fig. 82) or (2) by the development of a suppurative process.Spontaneous resorption of small and even moderate amounts of accumulated clottedblood or small amounts of fibrin was conceivable. If resorption of largeramounts occurred, which was not usual, recognizable sequelae were almostinevitable.

The rapidity and extent of clearance in any given hemothorax were impossibleto predict. In some cases in which the initial observations gave promise ofcomplete and prompt resolution, the process might suddenly become static,usually about the fourth week. Perhaps the chief reason for the dis-

260

FIGURE 84.-Photomicrograph showingreplacement of fibrinous exudate, pleura, and adjacent lung (lung is at extremeleft) by collagenous connective tissue in organizing hemothorax.

cussion concerning the optimum timing of decortication (p. 286), whichcontinued throughout the war, was the lack of agreement concerning thepossibilities of resolution in an organizing hemothorax. The exact size and typeof an intrapleural mass in which spontaneous resolution might be expected werenever agreed upon, nor was the length of time in which expectant treatmentshould be practiced. Generally speaking, cavities involving less than 50 percentof the pleural space or less than 500 cc. in volume usually clearedspontaneously over a 6-week period, but this was a generalization to which therewere multiple exceptions. No doubt, the most important consideration was thethickness of the intrapleural mass. It was repeatedly observed thatroentgenologic clearing was confined to the inner portions of the clot. Thethickest areas, notably the portion in the posterior thoracic gutter, were stillclearly demonstrable on roentgenograms after almost complete clearing of otherareas had occurred.

261

FIGURE 85.-Photomicrograph showing infectedorganizing hemothorax with bacteria (Staphylococcus hemolyticus on culture) deepin unorganized portion of fibrin and growing in masses in its meshes. Noteconspicuous absence of leukocytes.

BACTERIOLOGY OF ORGANIZING HEMOTHORAX

Hemothoraces represent the largest hematomas with which the body must deal.They are usually far too large to vascularize rapidly, which probably accountsfor their tendency to infection, particularly when they are of the clotted type.This tendency was one of the many reasons why it was important thatthoracentesis be performed promptly and repeated as necessary.

All types of bacteria were isolated from hemothoraces,including a surprisingly large proportion of anaerobic streptococci andstaphylococci. Bacteriologic findings in 24 cases of hemothoracic empyema areshown in the following tabulation:

Bacteria	Number of cases
Proteolytic and saccharolytic clostridia	2
Micrococcus tetragenus	2
Hemolytic Staphylococcus aureus	1
Hemolytic Staphylococcus aureus, proteolytic Clostridium	1
Hemolytic Staphylococcus aureus, hemolytic Streptococcus, Pseudomonas aeruginosa	1
Hemolytic Staphylococcus aureus, Escherichia coli, Aerobacter aerogenes, nonhemolytic Streptococcus	1
Hemolytic Staphylococcus aureus, Escherichia coli, Aerobacter aerogenes, nonhemolytic Streptococcus	1
Streptococcus viridans	1
Hemolytic Streptococcus	1
Anaerobic Streptococcus	1
Nonhemolytic Streptococcus, Pseudomonas aeruginosa	1
Streptococcus viridans, proteolytic and saccharolytic clostridia, Bacillus proteus	1
Proteolytic Clostridium	1
Escherichia coli	1
Escherichia, type unidentified	1
Clostridium, unclassified	1
Cultures sterile	6
Pseudomonas aeruginosa	1

262

Clostridial infections, which were not numerous, were nearlyalways caused by non-gas-forming bacteria. Remarkable combinations of bacteriawere found in both the liquid and the unclotted hemothoraces. In the order offrequency, the offending organisms were streptococci, clostridia, staphylococci,Escherichia, and Pesudomonas. Regardless of the bacteriology, theclinical course was essentially the same, except that when Escherichia waspresent, the illness was usually more serious.

In some cases, bacteriologic confirmation was lacking in spite of clear-cutsigns that infection had occurred. There were two explanations:

1. Bacteria might not be present in demonstrable numbers in the fluidportions of the mass.

2. The needle might, by chance, have been introduced into loculations (fig.79) in which the fluid was still uninfected. If this was the explanation andinfection had actually occurred, repeated efforts to find organisms wouldeventually be successful as the infection spread to the entire mass or as morerewarding sites were selected for aspiration.

In some patients, for whom no explanation was forthcoming, signs of sepsiswere clear cut and organisms were found in fluid aspirated from the chest, butthere was virtually no leukocytic response (fig. 85). In this group of cases,the fluid was thin and practically without sediment, even when organisms werepresent in large numbers.

In some cases, suppuration, as evidenced by purulent materialon aspiration, did not manifest itself for 4 to 6 weeks. In a considerablenumber of these cases, it had been thought that resolution was occurring. Thefactor (or factors) which determined which cases would go on to promptsuppuration and in which suppuration would be delayed or would not be evident atall was never clearly demonstrated. Whatever the factor (factors) might be, theexplanation was probably the balance or imbalance between (1) the rapidity andextent of the clotting that trapped the infecting bacteria in the fibrous mass

263

and kept them from contact with the pleura and (2) theability of the pleura to control or eliminate the bacteria that were originallyin contact with it. If the organisms in an infected hemothorax were virulent andif seeding was heavy and early, a suppurative pleuritis was likely to occurpromptly. Both the liquid and the clotted contents of the intrapleural spacewould be involved in the pathologic process. If, however, for one reason oranother, the contaminating bacteria did not gain the ascendancy and insteadbecame entrapped within the clotted intrapleural mass, then the mechanism of adelayed or latent infection was set up.

PHYSIOPATHOLOGY OF HEMOTHORAX

The bony thorax imposes definite limitations on the size ofthe thoracic cavity. This means that any space-occupying medium can seriouslydisrupt the normal function of the heart and the lungs. From this point of view,pneumothorax and hemothorax produced almost identical physiopathologic results.By impeding cardiac return and pulmonary expansion, they tended not only toreduce the volume of circulating blood but also to decrease the degree of oxygensaturation. The response was an increased cardiac and respiratory rate, both ofwhich were almost invariably observed in any severe thoracic injury. The paincaused by damage to the thoracic cage and the voluntary restriction of motion tominimize the discomfort further added to the physiologic disturbances becausethe respiration was rapid and shallow and the pulse was fast. For these variousreasons, it was necessary to empty the chest of blood and air as rapidly aspossible.

CLINICAL PICTURE AND DIAGNOSIS

Simple hemothorax-The diagnosis of simple hemothorax almost neveroffered any difficulty. The history of wounding was clear cut. Hemothorax wasknown to be the most frequent complication of thoracic wounds. Finally, thepleural contents were readily aspirable. At first, the fluid was blood orbloody. Later, if resolution began to occur, the blood disappeared and the fluidbecame serous, as well as scantier, on successive aspirations.

Roentgenologic examination revealed intrapleural obscuration consistent withthe presence of fluid (fig. 86). The lung margin was characteristically distinctand scalloped. This finding was particularly suggestive if the lateral viewrevealed the process localized to the posterior half of the involved hemithorax.

Organizing hemothorax-The diagnosis of organizing hemothorax was notalways so simple. In the early stages, when clotting had occurred but fluidcould still be obtained, it was sometimes impossible to distinguish thecondition from a simple hemothorax. Later, it might be necessary to insert

264

FIGURE 86.-Spontaneous resolution of clottedhemothorax. A. Posteroanterior roentgenogram showing clotted hemothorax onright. B. Same, showing spontaneous resolution after 6 weeks.

FIGURE 87.-Posteroanterior roentgenogramshowing right clotted hemothorax. Note that border of partially collapsed lungis still fairly distinct. In liquid hemothoraces, borders of collection of bloodwere usually seen only as diffuse haze.

265

FIGURE 88.-Lateral roentgenogram showingclotted hemothorax over right lower and middle lobes. The sharp superior borderseen in this view and not seen in posteroanterior roentgenograms came to berecognized as typical of clotted hemothorax.

the needle into many different areas of the chest to obtainany fluid. The aspirated material consisted of only small amounts of serum,which contained fragments of clot and fibrin threads. Efforts at drainage byinsertion of an intercostal water-seal catheter were practically alwaysunsuccessful.

If organization had gone on to fibrosis, the clinical picturewas distinctive. A patient with fibrothorax was likely to show an overallcontracture of the affected hemithorax; narrowing of the intercostal spaces;elevation and more or less immobility of the leaf of the diaphragm on theaffected side; often considerable limitation of movement of the chest wall; and,not infrequently, an associated atrophy of the extracostal musculature, no doubtas the result of disuse.

Roentgenologic examination in organizing hemothorax (figs. 87and 88) revealed a dense opacity corresponding to the area of the originalhemothorax. Resorption and consequent improvement in the roentgenologicappearance (figs. 89 and 90) was observed in a few cases that had reached thisstage, but, as a rule, narrowing of the interspaces, as well as contracture andrelative immobility of the chest wall, was likely to remain unaltered.

Hemothoracic empyema-The clinical picture of hemothoracic empyema wasthat of typical empyema modified by the special considerations introduced

266

FIGURE 89.-Clotted hemothorax withspontaneous clearing. A. Posteroanterior roentgenogram showing clottedhemothorax 2 weeks after wounding. B. Same, 1 month later. Chest has almostcompletely cleared by natural processes. Neither thoracentesis nor surgery wasused.

by the chest wound. Fever was usually high, and the patientwas likely to become more toxic than he had been. There might be an associatedpneumonitis with productive cough.

The sudden spillage of purulent pleural fluids in large amounts into thebronchial tree made the diagnosis of empyema with bronchopleural fistula simple.If there was leakage of only a small amount, the diagnosis was more difficult.The finding of a fluid level, otherwise unexplained, on the roentgenogram wasalmost certain evidence of a fistula.

The diagnosis of empyema was confirmed by the aspiration offluid that was chocolate brown, foul smelling, or frankly purulent. In somecases, as just noted (p. 262), infection had been present for some time beforepurulent fluid appeared.

The degree of mediastinal immobility present in any given case depended uponthe degree of organization that had occurred in the preexisting hemothorax. Itwas therefore roughly proportional to the length of time that had elapsed sinceinjury rather than to the duration of the infection or the thickness of thepurulent exudate.

Differential diagnosis-Large extracostal or intrapulmonary hematomas, aswell as lobular or lobar atelectasis following bronchial obstruction, wereusually readily distinguished from chronic traumatic hemothorax. Pneumonia wasalso readily distinguished. Infolded lung sometimes offered special difficulties(p. 183).

267

FIGURE 90.-Clotted hemothorax withspontaneous clearing. A. Posteroanterior roentgenogram showing right-sidedmassive clotted hemothorax, without infection, 2 days after wounding. B. Same, 3weeks later. Complete spontaneous clearing occurred later in this case, in whichevidences of resorption of fluid were noted from beginning.

MANAGEMENT OF SIMPLE HEMOTHORAX

Rationale of Thoracentesis

Thoracentesis was part of the routine of resuscitation for four reasons, allconcerned with the restoration of normal cardiorespiratory physiology:

1. To relieve high intrapleural pressure.

2. To permit increased aeration of the pulmonary parenchyma and thusencourage prompt reexpansion of the lung.

3. To correct shock by relieving the hypoxia that was helping to cause it.

4. To relieve the shift of the heart and mediastinum and thus both improvethe venous return to the heart and increase the volume flow of blood (fig. 91).

There were three additional reasons for continuingthoracentesis after resuscitation as part of the management of simple hemothoraxin forward hospitals:

1. To decrease the possibility of later clotting.

2. To encourage apposition of the lung and the thoracic wall and thus limitthe extent of infection if hemothoracic empyema should develop.

3. To prevent late fixation and contraction of the chest.

Failure to carry out aspiration of a simple hemothorax in forward hospitalsoften presented surgeons working in rear areas with complicated problems oforganizing hemothorax and hemothoracic empyema, problems which could be solvedonly by radical surgery. The management of empyema was the responsi-

268

FIGURE 91.-Management of hemothorax bythoracentesis. A. Schematic showing of massive hemothorax with colapse ofinvolved lung and shirt of mediastinum with partial collapse of uninvolved lung.B. Expansion of involved lung after removal of blood by simple aspiration ofchest. This procedure, in addition to greatly increasing vital capacity, alsoindicates amount of intrapleural hemorrhage.

269

bility of surgeons in base hospitals and hospital centers, but its preventionwas usually the responsibility of surgeons in forward hospitals.

Very early in the war, the practice of air replacement afterthoracentesis was condemned, and all the attention was devoted to the policy ofearly, vigorous aspiration. Certain objections to thoracentesis that had beenadvanced before the war proved entirely theoretical:

1. The military experience showed that there was no risk ofcausing renewed bleeding by thoracentesis; no case is on record in which thishappened. This might have been expected. The two chief sources of hemorrhage inchest wounds were extrapulmonary (from the chest wall or the mediastinum) andpulmonary. The existence of a hemothorax, however, had no effect onextrapulmonary bleeding, while the reason that pulmonary bleeding usually ceasedspontaneously was not the effect of the hemothorax but the lowered pressure inthe pulmonary artery and its branches (p. 153). Many more errors were made inWorld War II by not aspirating the chest than by aspirating it and taking theremote and theoretical risk of recurrent hemorrhage.

Only one case is on record in the Mediterranean theater inwhich there seemed to be any connection at all between aspiration of ahemothorax and secondary hemorrhage, and, in this case, the relation wasentirely on a post-hoc-ergo-propter-hoc basis. Intrapleural bleeding occurred 10days after closure of a sucking wound of the chest, after 2,850 cc. of bloodyfluid had been aspirated over a 3-day period. Just as preparations were beingmade for exploratory thoracotomy to determine the source of the bleeding, thehemorrhage ceased as abruptly as it had begun, and recovery thereafter wasentirely uneventful.

2. The risk of infection after a properly performed thoracentesis provednegligible.

3. The concept that only small amounts of fluid could beremoved at one time proved completely false. Withdrawal of 1,000 to 1,500 cc. atone sitting usually caused no symptoms but, instead, produced relief ofsymptoms. Occasional patients complained of a tight feeling, pain in the chest,dyspnea, or dizziness after large amounts had been withdrawn, but these symptomsdisappeared as soon as aspiration was discontinued and seldom recurred when theprocedure was resumed 24 hours later.

If hemorrhage continued or recurred (fig. 92), thoracotomy was indicated.Reaccumulation of 1,000 to 1,500 cc. of blood within a short period after thefirst aspiration of the chest was likely to be associated with failure tomaintain a satisfactory blood pressure and with a falling hematocrit. In thesecircumstances, simple aspiration could not be expected to be effective.

Technique of Thoracentesis

Since the purpose of the first aspirations of the chest was to relievepressure and to aid in the restoration of normal cardiorespiratory physiology,it was not necessary to empty the chest completely. At the end of 48 or 72hours,

270

FIGURE 92.-Schematic showing of recurrenthemothorax after thoracentesis.

aspiration was carried out at least once daily, in order toempty the pleural cavity as rapidly and as completely as possible. Ifthoracentesis had been instituted as soon as the casualty was seen, not morethan three to five daily aspirations were usually necessary.

The patient was transferred to the base as soon as he was fever free and thechest had been completely or almost completely emptied. If he was received in afixed hospital with clinical or roentgenologic signs of fluid, aspiration wasresumed and was continued until the chest was empty or until it was clear thatother measures would be necessary.

A single negative aspiration was not accepted as conclusive.Often three or four punctures had to be made before a loculated pocket wasfound. If several punctures produced no results, another surgeon attemptedaspiration the next day. The effort was not abandoned in any case in which therewas roentgenologic evidence of fluid or a suggestive pleural shadow untilrepeated attempts to locate and remove the fluid had been made. The fact thatblood had clotted (fig. 93) did not necessarily interfere with its aspiration.

Results of Thoracentesis

The results of early, vigorous thoracentesis in hemothoraxwere excellent from the beginning. If this was the chief or the onlycomplication of a thoracic wound, the great majority of casualties who weretreated by this method could be returned to duty within 6 to 8 weeks afterwounding.

271

FIGURE 93.-Schematic showing of massiveclotted hemothorax. Aspiration of blood is still possible when process oforganization is still incomplete, but when physical signs and roentgenologicfindings point to a large accumulation of intrapleural fluid and only smallamounts of blood can be aspirated, further attempts at evacuating chest areunlikely to be successful. If patient is dyspneic or shows signs of earlyinfection, thoracotomy and evacuation of clots must be undertaken promptly.Usually, however, it is safe to wait for surgery (decortication) at the base.

Results improved as time passed. In August 1943, it was estimated that about6 percent of casualties with uninfected hemothorax observed at the chest centerat the 21st General Hospital, located near Naples, required decortication. Bythe end of the war, the only casualties in this group who required decorticationwere those in whom massive clotting had begun shortly after wounding. At thistime, the operation was chiefly used in complicated cases of hemothorax and inhemothoracic empyema. The most reasonable explanation of the improvement was theexcellent treatment of hemothorax in forward areas, including prompt, vigorousthoracentesis; measures to promote early pulmonary reexpansion; and a materialreduction in the number of thoracotomies performed at initial wound surgery forthe evacuation of hemothoraces.

While a hemothorax could become infected and empyema could develop earlyenough to require treatment in an evacuation hospital or even, occasionally, ina field hospital, these complications did not usually ensue until the patienthad reached the base section. Both complications are therefore discussedelsewhere (p. 275).

272

CLINICAL PICTURE AND DIAGNOSIS OF HEMOTHORACIC EMPYEMA

Hemothoraces, as already pointed out, represented the largesthematomas with which the body had to deal in combat-incurred wounds. The factthat they were far too large to vascularize rapidly might very well have playeda part in their pronounced tendency to infection.

When infection supervened in either the liquid or the clottedvariety of hemothorax, it was not always easy to make a distinction betweeninfected hemothorax and hemothoracic empyema,. The precise differentiationdepended on changes in the visible characteristics of the pleural fluid. Thesechanges did not always occur immediately (p. 266). The infection was classifiedas hemothoracic empyema as soon as gross pus or purulent exudate was observed inthe aspirated material.

There were distinct differences between the hemothoracicempyema of combat-incurred wounds and the postpneumonic or metapneumonic varietyof empyema. In the latter variety, two features were prominent:

1. Infection of the pleural space was initiated upon essentially normalpleural surfaces.

2. The preexistent pathologic changes were in the lung.

Hemothoracic empyema had the following characteristics:

1. The pleural space was seldom normal. It had been altered in varyingdegrees by the fibrinous deposit that resulted from the presence of blood in thepleural cavity. Posttraumatic empyema was therefore not primarily an infectionof the pleura per se but an infection originating in pleural dead space occupiedby a hematoma enclosed in a fibrinous capsule.

The antecedent status of the pleural space was a point ofgreat importance. Infection developing within a pleural cavity which has notbeen the site of a significant hemothorax is always a more favorable type thanhemothoracic empyema which develops in a preexisting pleural clot or a neglectedliquid hemothorax. In the former type, the restraining membrane or peel commonlyobserved in all pleural irritative phenomena dates from the onset of theinfection. In the hemothoracic type, the peel antedates the clinical evidence ofinfection and may be well on to maturity before treatment is started.

2. The lung was collapsed and compressed because of the presence ofintrapleural blood. There were no changes caused by disease in it, and by thetime empyema developed, healing had usually occurred in any pulmonary wound. Itwas repeatedly observed at decortication that the lung had returned to normal,as evidenced by its ability to expand, even when it had sustained extensivelacerations and contusions. There had been time for this recovery to ensue. Whenempyema developed in a liquid hemothorax, the average time between wounding anddemonstration of pus by thoracentesis or at operation was 15 days. In theclotted variety it was 18 days.

273

3. The constrictive pulmonary investment-that is, the rind which enclosed thehemothorax-was responsible for failure of the lung to reexpand.

4. The purulent exudate thickened more slowly in the hemothoracic than in thepostpneumonic or the metapneumonic variety of empyema.

5. Loculation was more frequent in the hemothoracic variety of empyema, andtotal empyema occurred more readily.

6. Even in patients with extensive hemothoracic empyema ofseveral weeks' duration, there was no reaction in the costal periosteum, theribs were never triangular, and the fingers were never clubbed as in the varietyof empyema in which there was no antecedent trauma.

All of the circumstances associated with a combat wound arefavorable for the development of empyema. They include the wound itself and thetissue destruction that accompanies it; the contamination introduced by themissile and the debris brought into the wounds with it; in many instances, thepresence of pathogenic bacteria; tissue necrosis; the presence of anintrathoracic foreign body; pneumothorax or hemothorax, with resulting deadspace in the pleural cavity; and the ideal culture medium furnished by a clottedor an organizing hemothorax. Neither chemotherapy nor antibiotic therapy couldin themselves overcome these circumstances. Only good surgery could meet thesituation.

EMPYEMA IN THE WORLD WARS

There was a vigorous discussion of empyema at the 1946meeting of the American Association for Thoracic Surgery (6), in thecourse of which much emphasis was put upon the differences in the infections ofthe pleural cavity encountered in the two World Wars. The presentations onchronic organizing hemothorax, hemothoracic empyema, and decortication wereprepared by the thoracic surgeons who had had the widest experience in thisfield during World War II. The discussion was participated in not only by thesesurgeons but also by a number of other surgeons, among them Dr. Evarts A.Graham, who had had the widest experience in chest trauma and postinfluenzalempyema in World War I.

Graham contended that no comparison was possible between theempyemas of World War I and World War II because there was no similarity betweenthem. The surgeons of World War I, he said, were called upon to treat empyemaswhich were, in the Hippocratic sense, true abscesses of the pleural cavity. Theempyemas encountered by the World War II surgeons were pleural infectionsattenuated by drugs. In other words, the World War II surgeons dealt withconditions never before seen. If Hippocrates could listen in on the discussion,Graham concluded, he would not know what the World War II surgeons were talkingabout.

The surgeons of World War II naturally took issue with this point of view.They contended that hemothoracic empyema was true empyema, and they

274

pointed out that their first patients, as well as some oftheir later patients, were treated without benefit of chemotherapeutic orantibiotic agents. Many of these casualties, they emphasized, were very ill andhighly toxic, and problems of management were quite as serious as thoseencountered in World War I.

The clinical and other distinctions between the empyemas encountered in thetwo World Wars were well stated by Dr. Carl Eggers:

1. Chronic empyema was the chief problem of thoracic surgeonsin World War I, while posttraumatic hemothoracic empyema was apparently thechief problem of these surgeons in World War II.

2. The patients with chronic empyema encountered in World War I were chronicinvalids. They had survived measles, influenza, lobar pneumonia, and the otherdiseases that were widespread in Army camps and that were frequently complicatedby bilateral bronchopneumonia. They had withstood prolonged sepsis. They hadusually undergone drainage operations for acute empyema. Structural organizingchanges had frequently taken place in the chest wall, the pleura, and even thelung. As a result, their resistance was materially lowered.

3. In World War II, patients who developed hemothoracic empyema after combatwounds of the chest were usually in good physical condition. They had normallyfunctioning organs. Their chest walls were usually intact or almost intact andwere therefore mobile. Their lungs were usually uninvolved; if there was anypulmonary involvement, it was usually traumatic and not infectious. Thesecasualties might have lost blood and been in shock, but their resuscitation hadbeen uniformly excellent, and their resistance was generally good. In short,there was a striking difference in the general condition and the surgical riskpresented by the casualties of the two World Wars.

In discussing the procedures possible in empyema in the two wars, Eggerspointed out that it would not have been possible, in World War I, to insert arib spreader after making the incision, which was part of the technique ofdecortication in World War II. It would have been impossible because the ribswere usually fused and were firmly embedded in the thickened parietal pleura.Resection of portions of all the ribs overlying the pleural cavity was thereforeprerequisite to mobilization of the chest wall. In World War II, the chest wallwas practically always mobile. In World War I, the parietal pleura was oftenhalf an inch or more in thickness. Mobilization of the compressed lung, the mostimportant surgical objective, had to be accomplished by reflection of thethickened tissue forming the angle of reflection from the parietal pleura ontothe lung. Then, if it were possible, the remaining layer of tissue compressingthe lung was peeled off. Often this was not possible. Then the solution was tomake crisscross incisions into it. Complete reexpansion of the lung occasionallyfollowed this procedure but, as a rule, success was only partial.

The degree of success achieved in individual patients in World War I, Eggersconcluded, depended not only upon the surgery performed but upon

275

the condition of the pulmonary parenchyma. If the lung wasfibrotic, as it frequently was, complete reexpansion could not possibly occur.From this standpoint, conditions in World War II were very much more favorable,a point of view which the World War II surgeons present did not accept.

MANAGEMENT OF INFECTED HEMOTHORAX ANDHEMOTHORACIC EMPYEMA

General Considerations of Treatment

Aspiration, with or without the instillation of penicillin, wasoccasionally used in small infected hemothoraces in the Mediterranean theater,but it was never a theater policy. This method was used only when cultures weresterile, compression of the lung did not exceed 25 percent, and the patient'scondition was good. If a purulent exudate appeared, drainage was instituted atonce. The unwise use of continued aspiration was an invitation to chronicempyema. In World War I, the patients who were thus treated endured months andyears of suppuration and, if they survived, eventually required multistageddeforming obliterative operations. Such patients were scarcely ever seen inWorld War II.⁴

In some European chest centers, a so-called irritating packwas occasionally used in chronic empyemas in which the cavities were small or ofmoderate size, in the belief that maintenance of an acute inflammatory statewould produce an obliterative pleuritis more rapidly. This method, which isusually employed only as a supplement to a small Schede thoracoplasty, was notused in the Mediterranean theater.

Intercostal catheter drainage as the sole method of treatmentof empyema was never considered adequate. The catheter was promptly blocked bylarge amounts of partially degenerated clot, fibrin, and exudate, and drainagesimply ceased. Early in the war, some surgeons employed catheter drainage as apreliminary to rib resection, but as experience increased, it came to beappreciated that if drainage was indicated, almost the only contraindication torib-resection drainage was that the patient was too ill to tolerate it as aprimary procedure.

Selection of Therapeutic Method

When treatment of infected hemothorax and hemothoracic empyema becamestandardized in the Mediterranean theater, three methods were employed:

1. Rib-resection drainage alone.

⁴The basic difference, of course, was that in World War I the empyemasencountered were of the synpneumonic or metapneumonic varieties. In World WarII, the empyemas were chiefly caused by infection of hemothoraces. They occurredin healthy pleural cavities and in association with healthy lungs.-F. B. B.

276

2. Primary decortication; that is, a procedure to eradicateinfection or empyema by securing immediate normal expansion of the lung, withoutpreliminary drainage of the pleural cavity.

3. Secondary decortication; that is, a procedure for the same purpose afteradequate rib-resection drainage for several weeks.

The choice of treatment depended upon several factors:

1. The size and location of the empyema cavity. If the cavity were small,unilocular, and basal, rib-resection drainage alone was usually adequate.

2. The percentage collapse of the lung and its distribution.To World War II thoracic surgeons, the status of the lung was of primaryimportance. Their position was based on the truism: "As goes the lung, sogoes the pleural space."

If collapse involved 50 percent or more of the lung, primarydecortication was the method of choice, particularly if the pulmonary apex wasinvolved in the collapse. Because of the increased technical difficulties ifapical collapse ensued after rib-resection drainage, collapse at the apex wasconsidered a more important indication for primary decortication than a similardegree of basal collapse.

The common method of estimating the percentage of pulmonary collapse wasbased upon the distance of the lung border from the lateral thoracic wall as itwas observed in posteroanterior roentgenograms. The measurement was frequentlyfound to be in error and was nearly always found to be a minimal estimate. Thelung was usually collapsed for an equal distance from the anterior and theposterior thoracic walls. In most instances, if the lung border did notapproximate the lateral thoracic wall by a third of the width of the hemithorax,the collapse was usually found to be 50 percent or more. A consistent lack ofcorrelation was found between the percentage collapse of the lung as determinedby roentgenograms and as observed at operation (fig. 94).

3. The status of the patient. A few patients were so seriously ill as theresult of massive pleural infection that a procedure of the magnitude ofdecortication could not be considered as a primary procedure. This operationsometimes lasted 3 hours or more, and there was a considerable loss of bloodduring decortication of the visceral pleural surfaces.

Most patients treated by primary decortication had had feverto 102? or 103? F. daily for several days or longer before operation. Whenother features of the illness were favorable, pyrexia was not regarded as acontraindication to its performance. Much more important was the estimation ofthe general status of the patient and his ability to withstand surgery as shownby simple observations of the color of his skin, his appetite and nutrition, hisgeneral strength and alertness, and his response to transfusion.

4. The results of rib-resection drainage when it had been used as a primaryprocedure. In these cases, it was observed that if satisfactory results were to

277

FIGURE 94.-Decortication in hemothoracicempyema. A. Posteroanterior roentgenogram showing original extent of empyema. B.Same, 2 weeks after thoracotomy and decortication, which resulted in completecure. At operation, degree of pulmonary collapse was found to be considerablygreater than indicated in figure 90A.

follow, they would be evident within the first 4 to 6 weeksafter operation (fig. 95). If the lung expanded rapidly within that period, sothat it seemed reasonable to anticipate continued complete obliteration of thecavity, no further surgery was considered necessary. If progress during thisperiod was not satisfactory (fig. 96) and the collapse of the lung remained at50 percent or more, secondary decortication was considered indicated, not onlyto prevent possible or probable chronic empyema but also to reduce the durationof pleural suppuration.

These considerations applied to all cases of empyema as wellas to infections of clotted hemothoraces. Because of the tendency to loculationin both lesions, there was little reason for assuming that all septa would bebroken and all loculi drained through the relatively small wound made at ribresection. This was true whether the purpose of the procedure was to provide anoutlet for frank pus or for drainage of a clotted hemothorax. Rib-resectiondrainage was a satisfactory procedure, in short, only when the cavity was sosmall that all parts of it were readily accessible.

Multiple empyemas associated with fistula formation in atraumatized lung that had fully reexpanded offered a real therapeutic challenge.The precise origin of the pathologic process in such cases was frequently notclear. The blowouts might have occurred primarily along the missile track.Another theory was that a primary vascular injury might have resulted in areasof pneumomalacia that later ruptured, thus giving rise to isolated fistulas withseparate encapsulated pockets of empyema.

278

FIGURE 95.-Management of hemothoracic empyemaby rib-resection drainage. A. Posteroanterior roentgenogram showing massive liquidhemothoracic empyema on left just before operation. B. Same, 4 weeks afterinstitution of drainage by rib resection. Lung is now almost completelyexpanded. This patient went on to recovery without further surgery.

Decortication was obviously not the solution in this group of cases, andgreat care was required in the institution of drainage. A misplaced incisioncould result in unnecessary mutilation or even introduce a real threat to life.Attempts to drain one cavity through another, however closely adjacent the twocavities might be, were usually ill-conceived and might well cause complicationsby tearing into a lung that was normal or only slightly involved. The besttechnique in this kind of case was drainage by multiple rib resections.

DECORTICATION OF THE LUNG

Development and Rationale

The first decortication recorded in World War II was performed in the NorthAfrican Theater of Operations, U.S. Army, in April 1943, by Major Burford, on apatient with an uninfected organizing hemothorax, 5 weeks after injury (vol. I).The smooth recovery and successful end result in this case elucidated manypreviously obscure problems of clotted hemothorax. Thoracotomy had sometimesbeen undertaken before this date for removal of the clot, but the operation wasseldom successful, as might have been expected, for it made no attack upon theessential pathologic feature of clotted hemothorax; that is, the fibroblasticmembrane or peel. It was the continued presence of this constrictive pulmonaryinvestment that prevented reexpansion of the lung compressed by it.

279

FIGURE 96.-Posteroanterior roentgenogram showing massiveliquid empyema on right side 4 weeks after institution of rib-resectiondrainage. Lung is still collapsed. Patient later required secondarydecortication.

The rationale of decortication was incontrovertible:

1. The continued presence of an organizing hemothorax prevents pulmonaryreexpansion.

2. The intrapleural mass must therefore be removed.

3. A line of cleavage exists between the pleural surfaces andthe membrane until such time as the process, if not interrupted, becomes chronicin the pathologic rather than in the chronologic sense.

4. Decortication, which removes the compressing membrane, if performed beforethe process becomes chronic, will result in immediate and total pulmonaryreexpansion and will also obliterate all pleural dead space.

In the beginning, decortication was undertaken only on the indication ofuninfected organizing hemothorax. At a number of operations, however, unexpectedloculations of pus were found, and inadvertent decortication for infectedhemothoraces and hematothoracic empyema proved the suitability of the operationfor both of these conditions. The extension of the operation to these conditionswas perfectly logical. The basic pathologic process was the same in both.Microscopic examination of the rind over the pleura in early empyema showed itto be identical, cytologically and chronologically, with the rind in noninfectedhemothoraces.

With the advent of penicillin and the demonstration of its usefulness by Maj.Champ Lyons, MC, the next step in the evolution of decortication was natural,the use of the operation without preliminary drainage in massive empyema. Thefirst deliberate primary decortication for hemothoracic empyema was performed byMaj. (later Lt. Col.) Paul C. Samson, MC, under penicillin protection, in March1944. It was a complete success.

280

Decortication was then extended to cases in which concomitant lesionsjustified operation even if the size of the clotted hemothorax or the empyemapresent would not, in themselves, have warranted the operation. In many residualempyemas, other persistent pulmonary processes required surgical correction,including obliteration of bronchial fistulas; wedge resections for multiplefistulas; or, occasionally, the same procedures for posttraumatic abscesses.

The rationale of early decortication in properly selected cases of sterileorganizing hemothorax, infected hemothorax, and posttraumatic or hemothoracicempyema rapidly led to the concept that it was not tenable to consider that apatient with massive hemothorax was convalescing satisfactorily merely becausehe showed no signs of infection. Such a point of view took no cognizance of theprimary concern of the surgeon; namely, the early restitution of pulmonaryfunction and the prevention of the crippling effects of chronic fibrothorax.

The most effective reply to the arguments of those who advocated a moreconservative approach to the therapy of posttraumatic empyema was the notinsignificant case fatality rate in adequately drained cases in which progresswas apparently satisfactory. Progress in even the most competently managed casewas unpredictable, and the road from drainage to complete cure was beset withmany dangers. Chronicity could develop at any stage, and at the best, theelapsed time between drainage and cure was a matter of many weeks, and often ofmany months.

The rediscovery and reapplication of total pulmonary decortication (vol. I)to organizing hemothorax and to hemothoracic empyema resulted in a promptdecrease in the duration of recovery after chest wounds and a reduction in thenumber of potential pulmonary cripples. In properly selected cases,decortication was almost invariably followed by prompt eradication ofintrapleural fibrosis and sepsis and prompt pulmonary reexpansion.

As Col. Edward D. Churchill, MC, pointed out in the discussion ofdecortication at the 1946 meeting of the American Association for ThoracicSurgery (p. 273), the principle for which the Empyema Commission fought in WorldWar I was to wait to establish drainage until a generalized empyema had becomelocalized. In World War II, the procedure was the exact reverse: It was to takea localizing empyema and convert it into a generalized pleural contamination inorder to achieve prompt reexpansion of the lung. When the principle was thusstated, Colonel Churchill concluded, it was as startling as it was important.

In the excitement of this discussion, three points of major importance in themanagement of the hemothoracic empyema encountered in World War II were notexplicitly stated:

1. There was no emphasis on the status of the mediastinum. As Graham hadpointed out in World War I, when the mediastinum had become fixed,

281

drainage was indicated and safe. When it was still labile, drainage was notindicated and could be lethal.

2. The status of the pleural space depended upon the status of the lung,according to the truism, already mentioned: "As goes the lung, so goes thepleural space."

3. The point emphasized in Colonel Churchill's discussion, that it was safeto take a localizing empyema and convert it into a generalized pleuralcontamination in order to achieve prompt reexpansion of the lung, was possibleonly because penicillin was available and was capable of protecting the pleuralcavity against bacterial invasion.

Indications

Before the concept of decortication for organizing hemothorax, infectedhemothorax, and empyema could be properly formulated, it was necessary to bearin mind certain fundamental facts:

1. The essential nature of the pathologic process. The demonstration of thefibroblastic membrane that invested the pleura explained the chronicity of manyhemothoraces with infection, made clear the object of treatment, and pointed tothe procedure that would achieve the desired results.

2. The efficacy of penicillin in protecting against and controlling invasiveinfection, so that surgery could be safely applied at the optimum time to securemaximum functional results.

3. The selection of cases on the basis of strict indications. This was not aprocedure for promiscuous use. After decortication was first introduced, theinitial wave of overenthusiasm, as often happens, led to its use in somepatients in whom it was not indicated. The original excess of enthusiasm wassoon tempered by experience, and the scope of the operation was properlydelimited and its indications defined.

There were four chief indications for decortication:

1. Uninfected organizing hemothorax.-Patients inthis group were treated conservatively for periods ranging from 3 to 6 weeks.During this time, repeated aspirations were carried out. Even if all the bloodin the pleural cavity could not be removed and some clotting had alreadyoccurred, the removal, from time to time, of as little as from 50 to 100 cc. ofcurrant-jelly-like material through a large caliber needle was found to beadvantageous. Breathing exercises (vol. I) were also instituted, in an effort toaid reexpansion of the lung.

Between the third and sixth weeks of conservative treatment,the patient's status was carefully appraised. If aspiration was yielding only afew cubic centimeters of blood and serum and if roentgenograms failed to showprogressive improvement, decortication was regarded as indicated if the patientpresented some or all of the following symptoms and signs: Diffuse pulmonarypain or discomfort; dyspnea on exertion; and poor thoracic expansion, with

282

FIGURE 97.-Management of clotted hemothoraxwith retained foreign body. A. Poisteroanterior roentgenogram showing extensiveleft hemothorax, with foreign body in left lung. B. Same, showing clottedhemothorax of moderate size after 2 weeks' trial of aspiration. A clottedhemothorax of this size would not, in itself, have constituted sufficientindication for decortication at end of 2 weeks, but surgery was consideredindicated because of missile in lung.

retraction and narrowing of the intercostal spaces. A generalized hazyappearance on the roentgenograms provided additional support for decortication.

Patients with organizing hemothorax associated with retained foreign bodies acentimeter or more in diameter often presented good expansion of the upper lobeby the third or the fourth week, although a lower lobe might still be partlycompressed by organizing hemothorax. In cases of this kind, the combination ofan organizing hemothorax, even of moderate degree, and of a retained foreignbody was regarded as sufficient indication for decortication and removal of themissile (figs. 97 and 98).

2. Infected organizing hemothorax.-Decortication was undertaken, assoon as the patient's condition permitted, in all cases in which the pulmonarycollapse was more than 25 percent, particularly if the collapse involved theapex. Although infection sometimes became evident by the 10th to the 12th dayafter wounding, it usually manifested itself at a somewhat later date. Positivesmears or cultures frequently could not be obtained, and presumptive evidence ofinfection had to be based on signs of increasing toxicity, mounting fever, orrapid enlargement of the hemothorax. Massive clotting was likely to be presentin these cases, and a policy of delay often led to the development of empyema.

Many surgeons took the position that bacteria were consistently present inhemothorax, even if cultures were negative and even if frank suppuration did notsupervene. If this concept were correct and the intrapleural mass

283

FIGURE 97.-Continued. C. Same, 2 weeks afterdecortication. Progress after operation was entirely satisfactory and patientwent on to complete cure.

were left in situ, it seemed logical to postulate that true latency mightresult from foci of infection left deep in the hemothorax.

If this latter hypothesis were correct, then it could befurther postulated that in the postwar period, patients with fibrothorax orneglected organizing hemothorax would present the persistent roentgenologicfinding that in the past has been erroneously termed "thickenedpleura." It is characteristically present in patients subject to recurrentfebrile episodes caused by recrudescence of activity in foci of infection in thechest, as well as in patients who have become chronic respiratory invalids.

3. Posttraumatic (hemothoracic) empyema.-As already pointed out, thedistinction between hemothorax with infection and frank hemothoracic empyema wasnot great in many cases. In frank empyema, the infection might have been presentfor a longer period of time or might be more virulent. In general, the patientswere more critically ill than those with infected hemothoraces, and theaspirated material was more grossly purulent. All the patients on whomdecortication was performed on the indication of empyema had clinical evidenceof infection; that is, fever, anorexia, malaise, and other signs of toxicity.The majority were acutely and seriously ill, and at operation all were found tohave purulent fluid in the pleural cavity.

The type of organism recovered from the pleural fluid was nota factor in the decision for or against decortication. Clostridial empyemas weretreated by the same routine as other empyemas. The percentage of negativecultures was probably no greater than would be encountered in any series ofempyemas. Negative cultures were to be expected after chemotherapy andantibiotic therapy had been employed.

Focal points of intense reaction were occasionally seen in hemothoracicempyema, and miliary abscesses were occasionally observed in the older portionof the organizing clot. On the whole, however, these layers (fig. 79) wereremarkably free from inflammatory reaction. For all practical purposes, thepleura and the layers of adjacent organized clot, with their blood supply, had

284

FIGURE 98.-Management of clotted hemothoraxwith retained foreign body. A. Posteroanterior roentgenogram showing clottedhemothorax on right, without infection, and foreign body in right lung. B. Same,2 weeks after decortication and removal of foreign body. There is no residualhemothorax, and lung is fully expanded.

cleared themselves of infection. This observation helps toexplain why such good results could be obtained by decortication even when therewas frank suppuration in the center of the organizing mass.

4. Other indications.-Persistent bronchopleural fistulas,esophagopleural fistulas, and transdiaphragmatic biliary fistulas were notfrequent, but all played their part in the genesis of chronic empyema andtherefore demanded energetic management. Early decortication, with excision andclosure of the fistula, gave excellent results. In a few cases, decorticationunder penicillin protection was combined with the excision and closure of smallintrapulmonary abscesses (up to 5 cm. in diameter) associated with empyema.

Penicillin, with its established bacteriostatic properties and its resultantability to diminish the dangers of invasive infection at operation in thepresence of acute intrapleural infection, proved to be the adjunct which hadlong been desired to render early surgery safe in such cases. Perhaps the mostvaluable role of penicillin was in the management of chronic empyema, in whichit afforded efficient protection against spreading infection when a radicaloperation such as decortication was undertaken for the obliteration of a chronicempyema cavity.

A precise estimate of its benefits was naturally notpossible. It seemed reasonable to assume, however, that, when the surgeon wasobliged to work in a potentially or actually infected field, penicillinprotection often prevented serious wound infections and other complications.When it became available, it was possible to pursue a rational plan of therapyin infected hemothorax and hemothoracic empyema, directed toward the achievementof (1) early

285

pleuropulmonary lysis, (2) total ablation of infected tissue,(3) complete pulmonary reexpansion, and (4) absolute obliteration of residualpleural dead space. Under penicillin protection, all of these objectives couldbe achieved by a procedure relatively free from the risk of septicdissemination.

When penicillin became available, in the spring of 1944, theindications for decortication in hemothoracic empyema became stabilized. Ribresection with drainage was considered wisest when the pleural infection waslimited to the basal portion of the thorax or when it was small and definitelyencapsuled. If these empyemas were drained adequately, persistent cavitationseldom developed. If complete obliteration of the cavity and complete pulmonaryreexpansion did not take place promptly, however, the patient was regarded as acandidate for decortication.

Preliminary rib-resection drainage was also used in patientswho were too ill and too toxic to be subjected initially to open thoracotomy anddecortication.

Decortication without preliminary rib-resection drainage had a number ofadvantages:

1. Complete pulmonary expansion occurred at once, with the immediateelimination of all intrapleural dead space.

2. Chronic empyema, which was a distinct possibility after rib-resectiondrainage, was prevented.

3. Weeks and months of disability were eliminated, as was the risk of seriousand even fatal complications, such as brain abscesses.

4. The patient was spared a great deal of discomfort andinconvenience, and the hospital staff was saved the time and effort required byrepeated dressings.

At least three factors were responsible for the poorer results in secondarydecortication:

1. The patients requiring preliminary drainage were, as agroup, more seriously ill than those who did not require it. In particular, theywere likely to be much more toxic.

2. The intrathoracic wounds were generally more severe.

3. The interval between wounding and decortication was longerin secondary decortications, because of the original poor condition of thepatients and the additional time necessary for the drainage operation andrecovery from it.

Two questions arose in connection with the selection of cases fordecortication:

The first question was whether it would be possible to develop a peel in theparticular case. This was not a question that could be answered positively.Timing of surgery played an important part in the possibility. Roentgenogramsdid not furnish a clue, nor did a study of the parietal pleura at a preliminarydrainage operation. As a matter of fact, the pleural membrane was

286

usually thicker, and the pulmonary collapse more pronounced, than the filmssuggested. Unsuspected pleural pockets were also often found.

The answer to the second question, whether or not the lungwould reexpand, depended upon the extent and type of pulmonary damage andwhether or not infection was present. The description written by the forwardsurgeon might be helpful in estimating the extent of the pulmonary injury. Ifthe injury was a contusion, a roentgenogram taken immediately afterward, beforethe development of hemothorax, was also helpful. Again, timing was a factor. Ifdecortication were performed too soon after wounding, the lung would not havehad time to heal after injury and regain sufficient elasticity to permitadequate reexpansion. If a homolateral pneumonitis was present, decorticationwas not performed until it had cleared up, not only to give the lung a chance torecover but also to avoid the risk of introducing organisms into the bloodstreamby opening the lung; septicemia could easily result. As these statementsindicate, the expansile ability of the lung could not be settled bygeneralizations but had to be estimated in the individual case.

Timing

As early as 1904, Violet (7) had stated that ifdecortication were to be employed, it should be employed promptly, as soon as itwas evident that an empyema would not undergo spontaneous cure and before thewalls of the cavity had become rigid and separated by too wide a space.Lilienthal (8), in 1915, had resorted to the operation in a matter ofweeks, but most American surgeons who used it believed that it should beemployed only after many months had elapsed and after other methods of treatmenthad failed. In 1928, Lockwood (9) reported that there were more failuresin cavities that had lasted from 6 to 18 months than in those of longerstanding. He had himself achieved partial success with decortication in anempyema of 21 years' standing, and he had had some good results in cases whichhad lasted for 16 years.⁵

The surgeons in the North African theater who first performeddecortication for organizing hemothorax originally thought that a period of atleast 8 weeks should elapse from wounding to operation. Their reasoning wasthat clearing of the pleural space would occur in a great many instances withinthis period. It proved fallacious as applied to all but small cavities.Eventually, by the trial-and-error method, it was found that the optimum periodfor operation in uninfected chronic traumatic hemothorax was between 3 and 5 to6 weeks after wounding. If infection occurred, especially if it were of therapidly developing type, operation was done earlier.

The timing was logical for several reasons:

1. If operation was deferred for 5 or 6 weeks, many patients who originallyhad seemed candidates for decortication showed satisfactory resolution. The

⁵These observers, it should be noted, were dealing with the type of empyema associated with the pneumonias, not with the hemothoracic variety of World War II.-F.B.B.

287

size of the clot played a part in this decision. In perhaps 90 percent of alldecortications, the intrapleural clot exceeded 1,000 cc. in volume. It was notconceivable that a mass of such size could be absorbed within 5 or 6 weeks, oreven within a much longer period.

2. In uninfected cases, a rind less than 3 weeks old wasusually so thin and so poorly defined that it had to be wiped off the pleuralsurface or stripped away piecemeal. It could seldom be removed in toto, as wasdesirable.

3. If operation was postponed for more than 5 or 6 weeksafter wounding, the increasing cellular intimacy between the rind and thevisceral pleura made the operation technically difficult. By the fifth or thesixth week, hemo-organization and fibroplasia had advanced to the point at whichthe membrane was substantial enough to be handled with ease. Extensions into thepleural and subpleural tissues, and sometimes complete synechiae between therind and the pleura, made clean separation difficult and even impossible, andtears of the lung were frequent.

The length of time necessary for complete definition of the rind varied frompatient to patient. In some instances, decortication was performed successfullywithin 10 to 14 days after wounding. At the other extreme, it was also performedsuccessfully from 12 to 14 weeks after wounding, and the time was extended toeven longer periods in some hospitals in the Zone of Interior (p. 405). On thewhole, the best results were obtained within 3 to 5 or 6 weeks after wounding.

An obvious advantage of performing decortication early in uninfectedorganizing hemothorax was that infection was prevented at the same time thatrecovery was expedited. The tendency to infection and prolonged disability inorganizing hemothorax weighed heavily in the decision to perform decortication.

Since the optimum time for decortication was within 3 to 6weeks after wounding, evacuation to the Zone of Interior was not usuallypossible, and decortication had to be performed on most patients in overseahospitals. It was preferably performed in a base hospital or chest center, butoccasionally, if fighting had slowed down and evacuation to the rear wasdelayed, it was performed in an evacuation hospital.

Preoperative Preparation

The excellent results secured by decortication in World War II might leadthose unfamiliar with the operation to conclude that it was not major surgery.The conclusion would be completely unwarranted. The operation was always one ofmagnitude and was never performed without careful preoperative preparation.

Preoperative measures included:

1. Identification of the infecting organism.

288

2. Complete roentgenologic investigation, to determine thepresence of retained foreign bodies, bronchopleural fistulas, lung abscesses,and other complicating lesions.

3. Replacement therapy. Daily transfusions were given, in amounts of 500 to1,000 cc., until the hematocrit level was 40 or higher.

4. Nutrition. The nutritional status was brought to as nearnormal as possible by a high-caloric, high-protein diet. The serum protein isalways low in any condition in which proteins are lost by exudation, and plasmainfusions, as well as a protein dietary supplement, were therefore employed asindicated.

5. Penicillin therapy. Penicillin was given intramuscularly in amounts of200,000 units every 24 hours (25,000 units every 3 hours), usually for 48 to 72hours before operation. Some patients were apparently safely prepared in 24hours, though this was probably not long enough in most cases, and otherpatients received penicillin for 14 days or more, which was usuallyunnecessarily long. It was useless to wait for defervescence. That could comeonly with the extirpation of the infected focus.

All of these measures were important. Compromise with any of them would haveput the morbidity and mortality of decortication on a prohibitively high basis.The unavailability of most of these supplemental measures, as pointed outelsewhere (vol. I), furnishes at least part of the explanation of the poorresults of decortication when it was first employed.

Principles

The principles upon which decortication was based were as follows:

1. The object of the operation was to peel off the membraneimmediately overlying the visceral pleura and to remove all blood clot andfibrin in the pleural cavity, thus permitting complete reexpansion of thecollapsed lung.

2. To accomplish this, adequate exposure was essential.

3. The first step of the procedure was to obtain a properplane of cleavage between the peel and the visceral pleura, beginning at thejunction of the collapsed lung with the chest wall or over the undamaged portionof the lung.

4. The peel was removed by blunt dissection, carried out digitally or withgauze.

5. The success of the operation depended upon complete freeing of the lungwherever it was directly adherent to the thoracic wall and the mediastinum, aswell as along the fissural margins, so that complete circumferential reexpansioncould be obtained. The apex of the lung had to be freed, so that it would fillthe cupola of the pleural cavity. The operation could not be regarded as afailure if the diaphragm was not decorticated and mobilized, but if it could bereleased, pulmonary efficiency was greatly enhanced. Particular attention waspaid to the redevelopment of the costophrenic sulcus.

6. Generally speaking, the greater the area of lung decorticated, the betterwere the results.

289

7. Bronchopleural fistulas had to be closed, especially fistulas in the upperpulmonary lobe. If this lobe remained collapsed, the operation was a failure.

8. If intrapleural bleeding occurred and complete hemostasis was notobtained, the status of the pleural cavity would be about as it was beforeoperation. It was best to leave the parietal pleura in situ if stripping itseemed likely to cause excessive bleeding. The membrane in this area was thickerand more vascular than elsewhere, and bleeding in it was not controlled bypulmonary reexpansion. It therefore became the practice to leave it in situ inmost cases, on the ground that removal increased the surgical risksunnecessarily and did not appreciably increase the benefits of operation.

Anesthesia

Decortication was performed under cyclopropane ornitrous-oxide-oxygen-ether anesthesia, always by the endotracheal technique. Theservices of a trained anesthesiologist, experienced in thoracic surgery, weremandatory.

It was part of his task to administer blood to the patientduring the operation, usually in the amount of 1,000 to 2,000 cc., according tothe indications. It was just as important for the patient to leave the operatingroom with the hematocrit approximately normal as it was for it to be raised tonormal level before operation. This aspect of blood transfusion was quite asessential in the end results of decortication as was its shock-preventingfunction.

Bronchoscopy was practically always performed at the end ofthe operation. It was never omitted if moisture was audible after catheteraspiration or if pulmonary reexpansion had been difficult.

TECHNIQUE OF DECORTICATION

Management of unhealed wounds.-Unhealed wounds on the hemolateralthorax were either closed before decortication was undertaken or were excisedand closed just before the incision for decortication was made. If the twooperations were done at the same sitting, fresh instruments were used fordecortication.

The detailed technique of decortication was as follows:

Incision and exposure.-The patient was placed in thelateral decubitus position, with the affected side uppermost. A posterolateralintercostal incision, just below the angle of the scapula, was most oftenemployed (fig. 99). It usually permitted excellent exposure, without section orresection of any ribs. There was no serious objection to removal of a ribposteriorly, but if an anterior incision was employed, costal resection was notnecessary. Most decortications were done without section or resection of a ribunless previous surgery or comminuted compound fractures made it advisable toremove all or part of one or more (fig. 100). Exposure was usually entirelyadequate without

290

FIGURE 99.-Technique ofdecortication.Position of patient on operating table and site of incision.

costal mutilation, and intercostal incision, with retraction of the serratusanterior, was the method of choice.

Entrance to the chest was usually gained through the sixth intercostalinterspace, but either the fifth or the seventh interspace could be used if amore generous intrathoracic exposure was thought desirable at either the apex orthe base. Selection of the incision depended somewhat upon the individualthoracic habitus but was chiefly dependent upon the extent and the location ofthe process to be dealt with.

The first view of the interior of the chest, after the retractors had beenslowly spread and full exposure had been secured, was of varying amounts ofliquid or clotted blood, fibrin, or purulent exudate.

Release of lung.-As soon as the pleural cavity was entered, the lungwas freed widely from the parietal pleural rind about the region of theincision. If this precaution was not taken, extensive damage to the lung couldbe caused when the ribs were separated. Great caution was necessary if therewere many adhesions.

Since the lung was often compressed into the posterior gutter, much of theseparation at this point was behind and mesial to the membranous envelope. Sincethe peel on the parietal pleura was usually left in situ (p. 289), it wasimportant during the process of paraincisional mobilization of the lung that theproper line of cleavage be established. This was between the layers of thefibroblastic membrane and not between the pleura and the peel (fig. 79).

As soon as the necessary dissection had been accomplished, rib spreaders wereintroduced and the blades were widened gradually. It was important that they bespread gradually, especially if a rib had not been sectioned or resected, toallow for muscular relaxation and prevent costal fracture, pleural tears, andmuscle avulsion.

Exploration.-All liquid exudate, fragments of blood clot, and massesof fibrin were evacuated, and all loculi were broken down. The cavity was thencarefully examined for retained foreign bodies, fragments of cloth and otherdebris, and bronchopleural fistulas. After it was certain that a fistula did notexist, the cavity was thoroughly irrigated with physiologic salt solution.

Further inspection was then undertaken. All anatomic landmarks wereobliterated, and the underlying lung was relatively immobile. Typically, the

291

FIGURE 100.-Techniques of decorticationshowing alternate methods of entering pleural cavity: Intercostal incision (a),and rib resection (b). The sixth intercostal space was the most common site ofentrance into the pleural cavity for decortication.

inner or younger surface of the peel had a brownish-red surface, with theappearance of shagreen leather. The contours were smooth.

Decortication.-The first step of decortication was a sharp incisionthrough the fibroblastic membrane down to the visceral pleura (fig. 101). Thisstep was facilitated if the anesthesiologist used moderate positive pressure, tobrace the lung against the incision. The incised edges of the membrane weregrasped by forceps and decortication was begun. As soon as the peel wascompletely disrupted in any segment, herniation of the underlying lung, underpressure, through the incision demonstrated the desired plane of cleavage (fig.102).

Once a cleavage plane had been established between the organizing peel andthe visceral pleura, dissection was continued digitally, or with small, firmlypacked gauze pushers, or with small bronchoscopic sponges attached to longhemostatic forceps (fig. 103). The main pressure was always directed toward therind. If the fingers were used, the palmar surfaces were turned toward the rind,and dissection was accomplished by sweeping motions. The pleura underlying thepeel was nearly always thin, transparent, and expansile, and beginningreexpansion of the lung could be observed almost as soon as the regional peelwas released.

Oozing was likely to be considerable from the visceral pleura, because of thenumerous capillaries torn during the dissection and the gaping ends left on

292

FIGURE 101.-Technique ofdecortication. Sharpincision through fibroblastic peel down to visceral pleura. Edges of incisionare now grasped with instruments, and dissection is begun.

FIGURE 102.-Technique ofdecortication.Herniation of lung through incision in peel as anesthesiologist uses moderatepositive pressure. Majority of lung is still compressed by intact peel.

293

FIGURE 103.-Technique ofdecortication.Removal of peel by blunt dissection. At this point in the operation, nearly allof the peel has been dissected free from visceral pleura and elevated. Exposedpleura is grossly normal, as is lung.

the pleural surface. It could usually be controlled by warm, moist packs,supplemented by gentle pulmonary expansion under positive pressure.

If the peel was several weeks old, it was usually laminated, and great carehad to be taken that the pleura was exposed in every area and that nolaminations were left in situ. Reexpansion was inhibited if this precaution wasomitted. These scattered islands, while thin, were surprisingly tough, and theirremoval was essential, for they caused crossings and infoldings of the lung thatmaterially hampered complete pulmonary reexpansion. Crosshatching of themembrane, with piecemeal removal, was not a satisfactory substitute. If islandsof thin, tough membrane remained in situ after removal of the primary peel, thecorrect procedure was to remove them separately.

Decortication was carried to the edge of the collapsed lung, where the peelwas reflected onto the parietal pleura. After the fissures and apex had beenfreed, the lung was separated from the diaphragm, mediastinum, and pericardium,down to the hilus and the pulmonary ligament. This was another extremelyimportant step of the procedure.

The diaphragm was almost always elevated and fixed. If it could not bedecorticated completely without difficulty, it was freed circumferentially andthus mobilized. Decortication was never omitted if diaphragmatic disruption orfistula formation was suspected.

294

FIGURE 104.-Technique ofdecortication. Lung completelyliberated. When inflated by positive pressure, it fills the hemithorax.

The costophrenic and cardiophrenic sulci were examined, to make sure thatthey were free of clots and membrane. Their reestablishment increased the motionof the diaphragm and aided in pulmonary reexpansion.

As already mentioned, no attempt was made to remove membrane adherent to theparietal pleura because of the risk of bleeding (p. 290). The costal segment ofpeel, however, was scraped and scrubbed with gauze, to remove fibrin and toprovide a smooth surface. Great care was taken in decorticating the fissuralmargins, about fistulas, and at the sites of healed pulmonary lacerations.Transdiaphragmatic fistulas were excised and closed.

Management of the lung.-After the initial release of the lung, itwas expanded at intervals throughout the operation under increased positivepressure. This gradual method was preferable to attempts at sudden, complete re-expansion at the end of the operation. Such attempts were always ill-advised andcould be harmful.

The visceral pleura was kept moist, and atelectatic areas of the lung weregently stroked as it was reexpanded. As a rule, the lung almost completelyfilled the pleural cavity before the chest was closed (fig. 104), though it wassometimes necessary to rearrange the lobes before they occupied their normalpositions.

After the lung had been completely freed, any necessarypulmonary surgery was carried out. Although every effort was made to protect thevisceral pleura, it was almost impossible to avoid small tears and areas oftraumatism.

295

FIGURE 105.-Management of clotted hemothoraxby decortication. A. Posteroanterior roentgenogram showing massive clottedhemothorax on right after 4 weeks of observation without improvement. B. Same, 4weeks after primary decortication. There is no residual hemothorax. Foreignbodies are in chest wall.

The leaks were plugged with fibrin foam or repaired withinterrupted sutures of fine silk on atraumatic needles. If tears wereoverlooked, or if plugs and repairs were not completely successful, the multipledrainage provided at the end of the operation could be depended upon to maintainexpansion until the openings were occluded.

Single superficial fistulas were closed after the edges had been freshened.Long sinus tracts were laid open as far as possible, and the lining membrane wascuretted before the lung was closed in layers. Superficial closure of sinustracts was not adequate.

Rib fragments in the lung, foreign bodies in the chest wall(figs. 105 and 106) and lung, and indriven rib fragments were also removed iftheir size warranted it or if they were readily accessible. It was seldompossible to identify intrapulmonary foreign bodies while the lung was stillcompressed by the rind of an organizing hemothorax, but they could usually belocated and removed without difficulty after decortication (fig. 107).

Infected hematomas and lung abscesses were usually opened, and the necroticlining membrane was dissected free and removed. The remaining dead space wasobliterated by closure in layers with fine silk, nylon, or cotton.

A routine search for areas of infolding (p. 183) was carried outat the conclusion of all decortications. Exploration was particularly necessarywhen, although the pleura appeared grossly normal, the lateral fringe wasabnormally blunt or the lingula was unusually short or a lower lobe wasrelatively small.

In all correctly performed decortications, the lung was freed from its pointsof attachment along the parietal or the diaphragmatic pleura; when this hadbeen accomplished, areas of infolding were promptly evident. The freeing

296

FIGURE 106.-Managment of massive liquidhemothoracic empyema. A. Posteroanterior roentgenogram showing hemothoracicprocess in left hemithorax. B. Same, 4 weeks after primary decortication.Foreign bodies are in chest wall. Complete cure resulted.

FIGURE 107.-Technique ofdecortication. A.Dissection of rind over upper lobe of lung. B. Lung incised over foreign bodywhich has been fixed by finger compression.

297

of points of attachment between the visceral and parietalsurfaces at the costophrenic fringe might not permit unfolding of the involvedfringe if basal distortion was extensive. In such a case, there might also be asymphysis between the opposed diaphragmatic surfaces of the visceral pleura. Theattachment of one visceral surface to another was usually much less firm thanthe attachment between visceral and parietal surfaces. The surgeon, therefore,could readily slip his finger into the groove. Here it would encounter anintrapleural tunnel, covered only by a veil of thin adhesions. Release of theseadhesions permitted unfolding of the involved portion of the lung andrestoration of its normal contour.

Complete reexpansion of the lung, with obliteration of intrapleural deadspace, was the best possible insurance against a complicating empyema. Anycompromise with this ideal was an invitation to failure. An operation that endedwithout bringing the lung into absolute contiguity with the chest wall andwithout making certain that it remained there had failed to achieve its purpose.It was a fallacy to assume that half measures in this direction would, by thegenerous intervention of Providence or something else equally unlikely, resultin cure. Unless a complete operation was performed, the patient might be left ina situation only slightly more favorable than his original state.

Final steps.-When all of these procedures had been accomplished, theinner thoracic walls were carefully inspected and palpated, to be certain thattheir surfaces were perfectly smooth. Tags and cuffs of tissue were removed asthey were encountered; they encouraged pocketing and the creation of dead space.

Two or three of the intercostal nerves above and below the incision wereinjected with procaine hydrochloride solution (1- or 2-percent). The pleuralcavity was thoroughly lavaged with at least 1,000 cc. of physiologic saltsolution and then was dried by aspiration. After it was certain that it wasempty and that there were no bleeding points, the lung was brought gradually tocomplete expansion and then was tested again for air leaks.

Modifications of technique.-The technique justdescribed was followed in most decortications. Some surgeons varied it in smalldetails. Thus, some thought it essential to free that portion of the lung which,while not encased, was slightly adherent to the chest wall beneath thereflection of the peel to the surface of the parietal pleura or was adherent tothe diaphragm and the mediastinum. An occasional surgeon removed all threeelements forming the capsule of the pleural pocket; that is, the visceral,diaphragmatic, and parietal elements. When this technique was followed, thevisceral element was completely removed before, by a return to the originalincision, a plane of cleavage was found between the parietal element and thechest wall. The separation, which was quickly accomplished with the finger or ablunt instrument, was continued until the previous field of dissection wasencountered along the line at which the visceral element was reflected onto theparietal pleura.

298

Drainage.-The sites for drainage were selected with the aim of obliteratingdead space and maintaining pulmonary expansion. Soft rubber tubing, with aninternal diameter of 10 mm., was used in all areas except the second interspace.Here, a de Pezzer catheter (No. 12 or No. 14) was used. The tip of the catheterwas cut off, so that just the flange remained, and the tube was brought outanteriorly in the midclavicular line. A second drainage tube was used in theeighth interspace in the posterior axillary line. It was introduced for 2?or 3 inches, and the proximal end was tacked to the parietal pleural wall, about6 cm. from the skin surface, by a single silk suture. In infected hemothorax orempyema, a third tube was used in the fifth or sixth interspace in themidclavicular line, and a fourth tube was sometimes used in the seventhinterspace in the midclavicular line.

If there had been a previous rib resection for drainage, thesite of election for the posterior drainage tube usually corresponded with thesite of previous drainage. In such cases, the margins of the wound werefreshened, the tube introduced through it, and airtight closure accomplishedabout it. The drainage tube introduced at the first operation was sometimesremoved 3 or 4 days before decortication, in anticipation of this procedure.

Closure.-Closure was accomplished in layers after the lung was completelywithin the chest. Sutures were placed in the inner layer of the incision butwere not tied until the ribs had been approximated with lion-jawed forceps orretractors. Pericostal sutures were not used.

Local penicillin instillation.-Penicillin solution, in roughly equal amounts,was introduced into each drainage tube in infected cases. Each cubic centimeterof the solution contained 1,000 units of penicillin. The total amount used wasfrom 25,000 to 50,000 units in 100 to 200 cc. of physiologic salt solution. Alltubes were clamped when the patient left the operating room. The posterior tubewas kept clamped for 4 to 6 hours, to permit contact between the penicillin andthe tissues for this period of time. The other tubes were connected withwater-seal bottles as soon as the patient was placed in bed.

POSTOPERATIVE MANAGEMENT AFTER DECORTICATION

General measures.-The postoperative management of a patientwho had undergone pulmonary decortication was much the same as the regimeninstituted after any other major thoracic operation. The reaction to operationwas usually minimal, and the temperature and the pulse remained generally levelin most cases (fig. 108). Even if empyema recurred, the temperature elevationwas usually minimal.

Replacement therapy, in the form of blood, plasma, ordextrose infusions, was employed as necessary in each case. Return to a fulldiet, high in calories and protein, was usually possible within 24 hours afteroperation.

299

FIGURE 108.-Typical temperature charts ofpatients with hemothoracic empyema treated by primary decortication withpenicillin coverage. Note absence of postoperative reaction and promptdefervescence in all cases.

300

As soon as he had fully reacted, the patient was elevated on a backrest, andthe importance of deep breathing, forced expiration, and voluntary coughing,which had been explained to him before operation, was reiterated. These were allessential measures in the maintenance of pulmonary reexpansion.

Many surgeons regarded early ambulation as an important phase of thepostoperative regimen. Some failures after decortication, in fact, wereattributed to the immobility required by management of fractures of theextremities in traction. Thereafter, orthopedic surgeons were requested tomanage these injuries in some way that would permit movement about the bed, atleast.

Postural consciousness was stressed during the period of bed rest, and motionof the shoulder and arm was encouraged in the early postoperative period, beforeformal exercises were begun. Return to full activity was as rapid as wasconsistent with progress.

Penicillin therapy.-Intramuscular injections of penicillin were given by thedosage schedule used before operation until the drainage tubes had been removedand the patient had been afebrile for at least 2 or 3 days. This usuallyrequired that it be given for 10 to 14 days after operation.

Drainage.-The drainage tubes required close supervision. Theywere useful only as long as they functioned to maintain pulmonary reexpansionand to obliterate intrapleural dead space. They were essential for thesepurposes. Once these purposes had been achieved, the tubes simply acted asunnecessary foreign bodies. If the tubes had been placed correctly andhemostasis had been satisfactory, so that there was no undue oozing, it was onlyoccasionally that any trouble was encountered in maintaining satisfactorydrainage.

If a fistula developed, the anterior tube was kept in place as long as theair leak persisted, which might sometimes be as long as from 10 to 14 days. Innone of these cases did intrapleural infection develop.

Complications.-Recovery was usually without incident, butfistulas and empyemas or recurrent empyemas were occasionally encountered. Thelow incidence of postoperative empyema was attributed both to good surgery andto penicillin protection.

If infection developed, as shown by a change from serous to purulentdrainage, the postoperative empyema was usually basal and small. Dependentdrainage was instituted at once. After a 2-inch resection of the eighth rib, alarge airtight tube was inserted in the posterior axillary line and was attachedto a water-seal bottle, to prevent secondary collapse of the lung.

DECORTICATION IN EMPYEMA OF NONTRAUMATIC ORIGIN

The experience with decortication in postpneumonic empyema in World War IIwas limited, one reason being (table 14) that the number of such cases waslimited.

The rationale of the use of this operation in this variety of empyema was,however, entirely clear. If one compares the description of the histologic

301

appearance of the fibrinofibrous membrane in such cases inWorld War I (2), it is evident that there is little difference between it andthe membrane observed in the posttraumatic empyemas of World War II. Thissimilarity, together with the fact that, at least in the early stages ofpostpneumonic empyema, the pleura is essentially normal, suggested the use inWorld War II of decortication in early, carefully selected cases of subacute orimpending chronic post-pneumonic empyema. When the indication for surgery wastotal or subtotal empyema with collapse of the apex, loculation of the exudatewas a further reason for considering decortication.

The thoracic surgeons who revived decortication in World WarII believed that it would be technically possible to employ it in early orsubacute tuberculous empyema, but their wartime experiences provided noopportunity to test the theory. It will be remembered that decortication hadbeen originally used by Delorme (10) in 1893 for a "tuberculous"abscess of the chest wall with intrapleural extension (vol. I).

Although postpneumonic empyema was only a minor problem inWorld War II, enough of it was observed to evaluate the role of penicillin inits management during the acute stage. Observations at the various thoraciccenters showed that this agent was a valuable adjunct, but that it did not takethe place of adequate surgical drainage or, in the selected cases justmentioned, of decortication.

SPECIAL STUDIES OF PULMONARY DECORTICATION

Approximately 1,500 pulmonary decortications were performed in theMediterranean and European theaters during World War II on the variousindications outlined earlier in this chapter (p. 281). It is known that thesurgical mortality rate was less than 2 percent. Lack of facilities and thegeneral circumstances of wartime prevented a followup of these patients, butanalyses of several series show that superior results were obtained. There wasno doubt that decortication gave better results than methods used earlier. Therewas also no doubt that the results in primary decortication without drainagewere much better than the results of secondary, postdrainage decortication (p.310).

As already pointed out (p. 240), the incidence of empyema in theMediterranean theater was far lower than even the most optimistic had dared tohope. The very favorable rate was the result of a totally integrated program ofmanagement of chest wounds that reflected credit on all echelons of medicalcare, particularly forward echelons.

The reduction in the incidence of posttraumatic empyema as the war progressedcould not be attributed to any single therapeutic advance. Instead, it was dueto a combination of improvements. The most important of these was the increasingemphasis put upon adequate debridement of chest wounds and

302

FIGURE 109.-Comparative results of managementof empyema in World Wars I and II. A. Typical patients with chronic empyema inZone of Interior hospitals in World War I (2).

upon early, rapid, vigorous evacuation of the pleural cavityand prompt reexpansion of the lung. Another extremely important considerationwas the decrease in the number of ill-advised thoracotomies in forward areas.The acceptance of the concept that a hemothorax is for all practical purposes ahematoma carried with it an implied warning as to how to avoid pleural sepsis, awarning that eventually was widely heeded. Fewer and fewer hemothoraces thatrequired multiple aspirations were encountered in base hospitals in the lastyear of the war. The use of pencillin in forward areas no doubt favorablyinfluenced the empyema statistics, but good surgery was the chief explanation ofthe improvement.

In base hospitals, the proper management of large clotted hemothoracesfurther tended to reduce significantly the incidence of empyema. Theprophylaxis of chronic empyema was, in fact, one of the most important functionsof a chest center. Early pulmonary decortication was of great value inaccomplishing this purpose as well as in curing empyema that had becomeestablished.

303

FIGURE 109.-Continued. B. Typical World WarII casualties after primary decortication for posttraumatic empyema. In each ofthese cases, recovery was without complications, and patients were left with noresidual disabilities.

The most striking and most convincing evidence of the value of decorticationwas the almost total elimination of the group of unfortunate victims of WorldWar I (fig. 109A) who spent the post bellum years doggedly undergoing one majorsurgical procedure after another or living through a shortened lifespan ofchronic invalidism due to recurrent episodes of suppuration. Many of thecasualties treated by decortication in World War II (fig. 109B) were sent backto full duty, and a considerable saving in human suffering as well as inmilitary manpower was thus effected.

Criteria of cure.-Several criteria of cure were set up in decortication fororganizing hemothorax, infected hemothorax, and hemothoracic empyema:

1. The lung must be fully expanded at operation or within 2 weeks thereafterand the pleural cavity must be completely obliterated within 2 weeks ofoperation, as demonstrated by roentgenologic evidence (figs. 110, 111, 112, and113) and by cessation of fluctuation in the water-seal drainage apparatus.

2. There must be no clinical or roentgenologic evidence ofoccurrence, persistence, or recurrence of empyema during the period ofpostoperative observation, which ranged from 4 to 12 weeks.

3. There must be no evidence of embolic spread of infection.

304

FIGURE 110.-Management of hemothorax bysecondary decortication. A. Posteroanterior roentgenogram showing massive liquidhemothorax on left before operation. Treatment by rib-resection drainage hadbeen ineffective. B. Same, 4 weeks after decortication. Primary cure occurred.

FIGURE 111.-Management of hemothoracicempyema by primary decortication. A. Posteroanterior roentgenogram showingmassive clotted right-sided empyema before operation. B. Same, showingpractically normal chest 3 weeks after primary decortication.

305

FIGURE 112.-Management of hemothoracicempyema by decortication. A. Posteroanterior roentgenogram showing early massivehemothoracic empyema just before operation. B. Same, 2 weeks after primarydecortication. Note that lung has completely reexpanded and there is no evidenceof residual empyema.

FIGURE 113.-Management of clotted hemothoraxby primary decortication. A. Posteroanterior roentgenogram showing large clottedhemothorax on right. B. Same, 3 weeks after decortication. Complete cureresulted.

306

Earliest Operations

An analysis of the first 25 decortications performed in NorthAfrica, all by surgeons of the 2d Auxiliary Surgical Group, showed that all thepatients were seriously ill and that the outlook in many of them was distinctlyunfavorable. The single death in the series occurred 8 weeks after operation andwas not related to it. In spite of the gloomy prognosis, there were 18 primarycures. In every instance, the temperature and the pulse returned to normallevels within 7 to 10 days after operation and there they remained. At the sametime, there was a striking improvement in every patient in general appearance,in appetite and nutrition, in a rapidly restored sense of well-being, and in thecapacity for a resumption of essentially normal activity.

One of the five recurrent empyemas was explained, at least inpart, by the fact that the patient was treated in traction for a fractured femurand was almost completely immobile throughout the postoperative course. In theother four patients with empyema, the original cavities were greatly reduced insize and became progressively smaller during the period of observation afteroperation.

Of the 24 survivors, 2 were returned to full duty and 7 to limited duty inthe theater, with the recommendation that their status be reviewed within 90days. The other 15 were evacuated to the Zone of Interior, 5 because of recurrent empyema and 2 because of concomitant wounds. All were in good conditionwhen they left the theater.

Decortication at the 21st and 300th General Hospitals

An analysis of 125 decortications performed at the 21st and 300th GeneralHospitals in the Mediterranean theater, between 1 April and 20 December 1944,showed no surgical deaths and only one complete failure in one of the earliestcases. There were no instances of invasive infection, and the only two woundinfections were superficial.

Primary healing, with complete pulmonary restoration,occurred in over 90 percent of the cases in which the lung was normal atoperation. The same results were accomplished in just over 75 percent of thegroup in which intrapleural infection had been present before operation. On theother hand, when the lung was involved in the original pathologic process,empyema was recurrent in 40 percent of the cases. Practically all of therecurrent empyemas, however, were basal and small, and they therefore respondedwell to rib resection and drainage. It seems fair to say that, with theexception of the single complete failure, all the patients in this series whodid not obtain a primary cure at least benefited materially from decortication.

Decortication at the 12th and 36th General Hospitals

In a followup study of 140 of the 160 decortications performed at the 12thand 36th General Hospitals in the Mediterranean theater, Maj. William M.

307

Tuttle, MC, Capt. (later Maj.) Hiram T. Langston, MC, and Maj. Robert T.Crowley, MC, published the following data:

There were no deaths in the 140 patients followed up, and, so far as could bedetermined, there were no deaths in the 20 not followed up.

In the 140 patients followed up, the hemothorax was organized but not grosslyinfected in 89. In all of the other 51 cases, infection either was known to bepresent before operation or was encountered unexpectedly at operation. In 90percent of the total (160) series, the intrapleural mass exceeded 1,000 cc. involume.

In 90 of the 140 patients followed up, pulmonary reexpansion wascomplete within the first 24 hours, and in 18 others, it was complete within thenext 24 hours. Twenty-six patients required 72 hours or more for completereexpansion, and in the 6 remaining patients, the results from this standpointwere not satisfactory.

Of the 89 patients with organizing uninfected hemothoraces, 8 developedempyema after decortication. Most of the infections were merely small pocketsthat responded readily to drainage. All were either completely healed when thepatients were evacuated or were progressing so favorably that it was thoughtthat no further surgery would be necessary.

Recovery was without incident in 25 of the 51 patients who had grosslyinfected hemothoraces before operation. In 23 cases, small residual empyemasrecurred, ranging in volume from 30 to 200 cubic centimeters. These pockets, asin the group just described, were either completely obliterated before thepatients were evacuated or were progressing so satisfactorily that no furthersurgery seemed likely to be required. In the three remaining cases, additionalsurgery seemed inevitable.

The only other complications in the series were two woundinfections. Although both patients had had preoperative empyema caused by Clostridium welchii, neither infection was severe. The absence of serious woundinfections in this series is rather remarkable when it is recalled that many ofthe operations lasted from 1 to 3 hours and that the infected material traversedthe wound during most of this time.

Penicillin became available while the operations in this series were inprogress and was used in 77 cases. They are not consecutive; even whenrestrictions on its use were removed, it was employed only on indications. Ofthe eight patients who developed empyema after decortication of a presumably uninfected organizing hemothorax, five received penicillin during the immediatepostoperative period and three did not. All 25 patients who recovereduneventfully after decortication performed in the presence of infection receivedit, which suggests the value of this form of therapy. Sulfonamides were not usedin any case in the series.

The duty status of 3 of the 140 patients was not known whenthe report was made. Six were returned to full duty and 64 to limited duty inthe theater, in a number of instances with the recommendation that their statusbe reviewed

308

in 90 days. Twenty-three other men could have been returned to duty in thetheater as far as their thoracic wounds were concerned, but associated woundsmade this impossible. The remaining 44 patients had to be evacuated to the Zoneof Interior because of the severity of their chest wounds.

Comparative Results of Rib-Resection Drainage andDecortication

An analysis of 94 cases of hemothoracic empyema treated by surgeons of the 2dAuxiliary Surgical Group after the indications for decortication had beenstabilized permits a comparison of the results obtained in hemothoracic empyemaby rib-resection drainage, primary decortication, and secondary decortication.

Rib resection was the method of treatment in 72 cases, in 6 of which secondarydecortication was necessary later. The other 22 patients were treated by primarydecortication.

Rib-resection drainage.-Excluding the 6 cases in which rib-resection drainagewas followed by secondary decortication and which will be discussed separately(p. 310), 66 patients, 5 of whom died, were treated by rib-resection drainage.

Of the 66 patients, 59 had small or moderately large empyemas, many of thembasal. The majority still had small residual cavities when they were evacuatedto the Zone of Interior, but progressive obliteration had occurred, and it wasnot thought that any of the patients would require further surgery.

The other seven patients had massive infections, and primary decorticationwould have been the treatment of choice had their condition permitted it. Insome cases, their serious status was due to associated wounds rather than toempyema. Satisfactory pulmonary reexpansion occurred in only four of the sevencases.

In one of the three failures, a large fistula in the right main bronchusopened into a persistent large empyema cavity. The fistula would ordinarily havebeen closed and secondary decortication performed, but an associated lacerationof the dorsal spine, with paraplegia, made the prognosis hopeless. In the tworemaining cases, exploratory thoracotomy was carried out, in the hope thatdecortication could be done, but in both, the infectious process was alreadychronic, and operation was not considered feasible. All three patients wereevacuated to the Zone of Interior, with recommendations for later surgery.

One of the 5 deaths in the 66 patients treated by rib-resection drainage wascaused by a metastatic brain abscess, which is one of the hazards of empyematreated by this method. Two deaths were due to overwhelming pleural infectionsecondary to esophagopleural fistulas, and two were due, in large part, tosevere associated wounds.

Primary decortication.-All 22 patients who underwent primarydecortication had massive involvement of the pleural cavity, 4 on the basis ofliquid hemothoraces and 18 on the basis of clotted hemothoraces.

309

In a few instances, decortication had to be deferred longerthan was considered desirable because of the gravity of associated extrathoracicwounds. In one case, the delay was 63 days. One or two patients did not comeunder observation until more than a month after wounding. Excluding thesepatients, the timelag between wounding and decortication ranged from 7 to 31days and averaged 20 days. This timelag included a purposeful delay of 2 to 4days after the diagnosis of empyema was made, to prepare the patients foroperation.

The results in these 22 decortications were as follows:

A prompt cure was obtained in 18 cases, in 16 of which the empyema hadoriginated in clotted hemothoraces. Included in these 18 cured patients were 2who had small lung abscesses, each of which was associated with a foreign bodyand a bronchopleural fistula at the site of the abscess. In each instance, theforeign body was removed at decortication, the walls of the abscess excised, andthe residual defect, including the fistula, closed in layers. There were nopostoperative complications in either case.

In the four remaining cases in which primary decorticationwas done, there was a postoperative recurrence of empyema. All four patientswere then treated by rib-resection drainage, with the following results:

In two cases, the recurrent cavity was small, and in both instances, it wasalmost completely obliterated when the patients were evacuated to the Zone ofInterior. One of these patients had a late bronchopleural fistula, which closedspontaneously. It was not thought that additional surgery would be necessary ineither instance. Why the empyema recurred in these two cases is not clear, forin both, complete expansion of the lung was obtained at operation.

In the third case, the cause of the recurrence was readily apparent. A largelung abscess was found at operation, surrounding a large foreign body and alsothe site of a bronchopleural fistula. To excise the abscess completely wouldhave required lobectomy, and the patient's condition was too critical for suchradical surgery. The fistula was closed, but the infection was not eliminated,and recurrence was regarded as inevitable. The recurrent cavity was small, butthe closure of the fistula broke down, and further surgery would be necessary inthe Zone of Interior.

In the fourth case, the cause of the recurrence was also apparent. Thispatient had a large lung abscess. It involved almost half of the lower andmiddle lobes on the right side and was also the site of a bronchopleuralfistula. Partial lobectomy was performed on each of the affected lobes, and thefistula was closed. The remaining lung reexpanded normally after decortication,but the recurrence was expected because of the large amount of dead space leftafter lobectomy. Both the original and the recurrent infections were caused by Escherichia coli. The patient was seriously ill, and it was debatable, beforethe first operation, whether primary decortication or rib-resection drainageshould be performed. He did not improve after decortication and lobectomy, anddeath occurred on the 30th postoperative day, from persistent and overwhelming

310

intrapleural infection. There was no response to either penicillin or thesulfonamides. This was the only death in the 28 patients treated bydecortication.

Secondary decortication.-In all six cases in which secondary decorticationwas employed after rib-resection drainage, the operation would have been doneprimarily had the patient's condition permitted it; in three cases, drainage wasdone for empyema necessitatis. In five of the six cases, the empyema haddeveloped in liquid hemothoraces.

In all six cases, although drainage was apparently adequate, there was eitherno reexpansion, or only unsatisfactory reexpansion, of the lung, anddecortication was resorted to because chronic empyema seemed to be impending. The duration of drainage before decortication ranged from 26to 53 days and averaged 37 days. The timelag from wounding to the secondaryoperation ranged from 44 to 64 days and averaged 52 days.

The outcome in thesesix cases shows again that primary decortication gives better results than thesecondary operation. Prompt recovery ensued in two cases, but empyema recurredin the other four. On the other hand, these four operations cannot be consideredtotal failures. They were performed because there was no significant pulmonaryreexpansion, in spite of adequate drainage, and because in each instance morethan 50 percent of the lung was collapsed. Even though the empyema recurred, itwas thought that the purpose of decortication had been accomplished: In twocases, chronic empyema was positively prevented, and the anticipated duration ofcontinued pleural suppuration was materially reduced. In the other two cases,while the situation was improved, it was thought that additional surgery wouldbe necessary.

Conclusions.--The results obtained with decortication in these various seriesmight be compared with the results reported by Nicholson and Scadding (11) in291 wounds of the chest treated in British hospitals in the Middle East betweenNovember 1941 and May 1943, before decortication came into use. Hemothoraxbecame evident in 187 cases and became infected in a third of these. The casefatality rate in 52 infected hemothoraces uncomplicated by pulmonary orsubphrenic suppuration and managed by standard conservative measures was almost20 percent, while at the time of the report, only 12 of the surviving patientswere known to be entirely well.

Two points concerning the advantages of decortication over drainageoperations for organizing hemothorax and hemothoracic empyema, although theyhave been made before, should be stressed again:

1. In favor of decortication was the short interval, usually from 6 to 8weeks, necessary to effect a cure, compared with the weeks of suppuration,changes of drainage tubes, and daily dressings necessary in other methods. Withtheir wounds cleanly and solidly healed, and their lungs fully expanded, thepatients were well on the way to recovery within 2 to 3 weeks after operation.

2. In pleural infections treated by drainage, the road from the institutionof drainage to a complete cure was not only long, it was also beset with dan-

311

FIGURE 114 (case 1).-Spontaneous clearing ofclotted hemothorax. A. Posteroanterior roentgenogram showing clotted hemothoraxof moderate size, without infection, 2 weeks after injury. B. Same, 1 monthlater. Chest was almost completely cleared without surgery or additionalthoracenteses.

gerous pitfalls. Chronicity might eventuate at any stage, anddeath might occur from a number of causes, including metastatic brain abscess.With this method, the outcome was uncertain as long as a vestige of the empyemacavity remained.

CASE HISTORIES

Case 1.-A soldier who sustained a perforating gunshot wound of the righthemithorax on 14 June 1944, with a laceration of the right lung, developed ahemothorax of considerable size. The wounds were debrided, and a sucking woundof exit was closed. A flapper valve intercostal tube was inserted after 800 cc.of blood had been removed from the chest by thoracentesis.

Drainage was unsatisfactory, and the tube was removed 48hours after wounding. Roentgenograms 14 days after injury showed a moderatelylarge, clotted right hemothorax (fig. 114A). There was no evidence ofintrapleural infection.

Repeated aspirations yielded only a few cubic centimeters of serum andfragments of clot, but, because of the moderate size of the hemothorax and theabsence of any evidence of intrapleural infection, spontaneous clearing wasallowed to progress. Recovery was smooth, and within 6 weeks after wounding thechest was almost clear (fig. 114B). The soldier was returned to duty on 17August 1944, 2 months after he had been injured.

Comment.-This case illustrates the successful outcome of a hemothorax ofmoderate size, without infection, treated conservatively. Clotting occurred inspite of the prompt removal of blood by aspiration.

Case 2.-When this soldier was wounded in action on 6 February 1944, hesustained a severe penetrating wound of the left thorax, with a laceration ofthe left lung. Initial

312

FIGURE 115 (case 2).-Management of clottedhemothorax by decortication. A. Posteroanterior roentgenogram showing massiveclotted hemothorax, without infection, but with no tendency toward clearingafter 6 weeks. Repeated thoracenteses were not effective. B. Same 10 days afterprimary decortication. This patient went on to complete cure and was returned toduty. Foreign body was in chest wall.

therapy consisted of debridement of the wound and left thoracenteses. Ametallic foreign body in the chest wall was left in situ.

Repeated thoracenteses were ineffective. Only small amounts of serum andfragments of clot could be obtained. Roentgenograms (fig. 115A) showedpersistence of pleural fluid, with no evidence of resorption after 6 weeks.Thoracotomy with pulmonary decortication was then performed. A complete cure wasobtained (fig. 115B), and the soldier returned to duty 3?months after wounding.

Comment.-On the surface, there seems little difference originally betweenthis case and the case just described (case 1). In both, clotting occurred, andadditional aspirations were unsuccessful. Conservative treatment was practicedin both, with practically complete resolution at the end of 6 weeks in the firstcase, while in the second, no tendency toward resolution was evident.Decortication produced highly satisfactory results, and might, perhaps, havebeen employed earlier. If no tendency toward clearing was observed at the end ofthe third week after wounding, it was the general experience that it wasunlikely to occur.

Case 3.-This soldier sustained a shell-fragment wound of the right chest on31 May 1944. He had a sucking wound, a laceration of the lung, and ahemopneumothorax, all on the right side. Thoracotomy was performed in a forwardhospital, with removal of the shell fragment in the lung, suture of the lung,and evacuation of the blood in the right pleural cavity.

When the patient was admitted to a chest center 11 days after wounding, hewas desperately ill from massive empyema. The thoracotomy incision was reopenedand the empyema drained through the bed of the rib that had been resected atinitial wound surgery.

Drainage was apparently adequate, but the lung did not reexpandsatisfactorily, and chronic empyema was evidently impending. Decortication wastherefore performed 6 weeks

313

FIGURE 116 (case 4).-Management ofhemothoracic empyema by primary decortication. A. Posteroanterior roentgenogramshowing right-sided hydropneumothorax, with 50-percent compression of lung.Fluid component secured by aspiration soon became thick and purulent. Ninth ribhad been partly resected when thoracotomy was done immediately after wounding inforward hospital. B. Same, 2 weeks after decortication. Regeneration is evidentin this roentgenogram. Complete cure was obtained, with total pulmonaryreexpansion and total obliteration of empyema cavity.

after the drainage operation. The lung was freed withoutdifficulty from the fibrinofibrous membrane which encased it. As soon as thiswas accomplished, reexpansion was entirely satisfactory. Convalescence waswithout incident, and the end result was excellent.

Comment.-This patient was far too ill for such a major procedure asdecortication when he was seen in the chest center 11 days after wounding.Rib-resection drainage was lifesaving, but it was not sufficient to accomplish acure. In World War I, this soldier would have become a chest cripple.Decortication saved him from this fate in World War II. Whether the thoracotomyperformed at initial wound surgery was necessary is at least debatable. He mighthave been better off without it.

Case 4.-This soldier, who was wounded in action on 4 June 1944, suffered asevere penetrating shell-fragment wound of the right thorax, with a lacerationof the right lung and right hemopneumothorax. A foreign body was seen in thelung. Thoracotomy was performed at a forward hospital on the day of injury. Theforeign body was removed and the pulmonary laceration sutured.

The immediate convalescence was uneventful, but a few days after operation, aclotted hemothorax was found on the right side (fig. 116A). The patient wastoxic, and his temperature rose daily to 101? or 102? F. Purulent fluid waseventually recovered from the chest, and Staphylococcus aureus was cultured fromit. Thoracotomy and decortication were done at once, without preliminarydrainage. Recovery was uneventful, and the end result was excellent (fig. 116B).

Comment.-This is another case in which prompt decorticationsaved the casualty from a long period of drainage and disability, and perhapsfrom chronic invalidism. It is also another case in which the patient might havedone better without thoracotomy as part of the initial wound surgery.

314

FIGURE 117 (case 5).-Management of clottedhemothorax by primary decortication. A. Posteroanterior roentgenogram 33 daysafter wounding, showing diffuse left-sided obscuration and narrowing ofintercostal spaces caused by organizing clotted hemothorax. Thoracenteses hadbeen ineffective. B. Same, 5? months after primary decortication, showingessentially normal left hemithorax.

Case 5.-A 27-year-old soldier sustained an accidental suckingwound of the left anterior thorax during bayonet practice on 11 June 1943. Atension hemopneumothorax was managed by water-seal drainage for 48 hours. Thesucking wound was trimmed and closed. Repeated thoracenteses produced no fluid.

When the patient was admitted to a thoracic surgery service on 10 July, theleft chest was flattened anteriorly, and there was a slight scoliosis concave tothe left. Expansion was appreciably limited. Horizontal expansion measurementsshowed the following differentials: Ensiform cartilage to spine on right, 1?,inches, on left ⁵/₈ inch; nipple to spine on right, 1inch, on left ? inch. Breath sounds were distant overthe entire left chest.

For the past 2 weeks, the patient had been complaining of increasing dyspneaon exertion. He was now incapacitated for any activity except the most minorexercise, and he was showing no appreciable improvement.

Analysis of serial roentgenograms showed a persistent diffuse obscuration onthe left, with the greatest density at the base. The roentgenogram taken on 14July (fig. 117A) showed diffuse clouding. The intercostal spaces were narrowed,and there was a slight scoliosis concave to the left. There had been noessential change in these findings for the preceding 2 weeks.

On 16 July 1943, 35 days after injury, thoracotomy anddecortication were performed. More than 500 cc. of clotted, degenerating bloodwas found occupying the space in the left hemithorax from the apex to the baseand forward to the midaxillary line. The peel adherent to the visceral pleurawas 3 mm. thick, and that adherent to the parietal pleura, which was left insitu, was 5 mm. thick. After decortication had been accomplished, there wasimmediate and almost complete reexpansion of the lung under slight positivepressure. A few bubbles of air escaped from a small tear in the visceral pleura,which was easily repaired.

Under the microscope, the visceral layer of rind showed adultfibrous tissue, with many penetrating capillaries, which were larger and morenumerous on the visceral pleural surface.

315

Frequent thoracenteses were carried out for the next 5 days.Roentgenograms 20 days after operation showed a small residual lateralpneumothorax, with complete pulmonary reexpansion at the base. Measurementstaken 30 days after operation showed the horizontal expansion from the ensiformcartilage to the spine to be 1? inches on the right and1? inches on the left. Expansion from the horizontalnipple to the spine was 1 inch on each side.

The patient was discharged to limited duty 60 days after operation. Afollowup roentgenogram on 6 January 1944 (fig. 117B) showed the left thorax tobe essentially normal.

Comment.-In this case, clotting apparently occurred almost immediately, forprompt and repeated thoracenteses produced no fluid. The patient was well on hisway to becoming a chest invalid when thoracotomy and decortication wereperformed. As this case shows, the risks of removing the thicker and morevascular parietal peel need not be taken. It was left in situ at operation, andthe patient made an excellent recovery. Repeated thoracenteses were employedafter operation because drainage had been omitted.

Case 6.-A 19-year-old German prisoner of war sustainedperforating wounds of the left chest from a high explosive shell fragment. Thesmall wound of entry was located in the second interspace in the midclavicularline and the slightly larger wound of exit, in the eighth interspace in themidscapular line. Neither was sucking.

The wounds were debrided 24 hours after injury. At this time,roentgenograms revealed a left-sided hemothorax that was causing approximately45-percent collapse of the lung. No fluid was obtained on thoracentesis.

For the next 8 days, the patient ran a febrile course, the temperature on oneoccasion reaching 103.2? F. Repeated aspirations during this period yieldedonly fragments of clot and shreds of fibrin. Smears and cultures were negative.

A roentgenogram 9 days after injury (fig. 118) showed theleft hemithorax obscured and the lung compressed. The following day, a catheter(No. 22) was inserted into the left pleural cavity, but neither fluid nor clotcould be evacuated.

Because of the continued fever and toxicity, it seemedprobable, in spite of consistently negative cultures, that infection wasdeveloping in an organizing hemothorax. Thoracotomy and decortication weretherefore performed 14 days after wounding. The pleural cavity was filled withclots and with loculated pools of thick, sirupy blood. The hemothorax was, asusual, larger than the roentgenograms had indicated. The lung was held immobileby a 2-mm. peel of fibrin, which was undergoing early organization.Decortication was technically tedious and difficult because of the thinness andfriability of the peel. It stripped easily, but piecemeal removal was necessary.The underlying visceral pleura was grossly normal and did not bleed. No attemptwas made to decorticate the parietal pleura. The wounds of entrance and exit inthe upper lobe were solidly healed, and the fibrin peel over both areas was leftundisturbed. When the diaphragm was freed, it immediately resumed its normalexcursion. Before decortication, positive pressure up to 15 cm. H₂Odid not result in any pulmonary expansion. After decortication, the lungexpanded promptly and filled the hemithorax.

Drainage was instituted through two intercostal tubes. The anterior tube wasremoved at the end of 24 hours and the posterior tube on the third day. Thepatient then became ambulatory. Roentgenograms on the seventh day afteroperation showed complete pulmonary reexpansion. When the patient was evacuatedto a prisoner-of-war camp on the 14th day after operation, he was in excellentcondition.

Cultures of the clot and peel removed at decortication were negative on bothaerobic and anaerobic media.

Comment.-This case is an illustration of the prophylactic useof decortication. Obviously, infection was impending in an organizing hemothorax,and while the operation is somewhat more difficult as early as 14 days afterwounding, because the rind is not yet well organized, it can be safelyperformed.

316

FIGURE 118 (case 6).-Posteroanteriorroentgenogram showing organizing clotted hemothorax on left side. Note apicalcompression and almost 50-percent pulmonary collapse.

Case 7.-A 34-year-old soldier sustained a sucking wound of the rightposterior thorax on 23 September 1943. The wound was closed immediately by anocclusive dressing, and 24 hours later, primary closure of the pleura wasperformed. Water-seal drainage was continued for the next 48 hours.

On 27 September, 1,800 cc. of air and fluid were aspirated from the righthemithorax. The following day, roentgenograms (fig. 119A) showed a large amountof blood in the right pleural cavity, with a 75-percent collapse of the lung. On30 September, 100 cc. of bloody fluid was aspirated, with considerabledifficulty. A second attempt to evacuate the chest with intercostal water-sealdrainage produced only 200 cc. of bloody fluid in 24 hours. On 5 October, allthat could be obtained on aspiration was 20 cc. of sterile fluid. At this time,a lateral roentgenogram (fig. 119B) showed extensive obscuration at the base andmultiple fluid levels, characteristic of intrapleural infection. The patientcontinued toxic and febrile.

On 7 October, right-sided posterior thoracotomy was performed after two smallgranulating wounds of the right posterior chest wall had been excised and closedwith a separate set of instruments. Short posterior segments of the seventh andeighth ribs were removed, and the pleural cavity was entered through the seventhintercostal space. The lung was found bound down with characteristic organizingexudate and approximately 90 percent collapsed. The pleural cavity was filledwith liquid blood, serum, masses of fibrin, and blood clot in various stages ofdegeneration.

When the peel was removed from the visceral pleura, there was immediate andalmost complete reexpansion of the lung under slight positive pressure. Thelacerations at the sites of the through-and-through pulmonary wounds wererepaired with fine silk. Drainage was instituted. Immediate bronchoscopy wasproductive of a moderate amount of bloody mucoid secretion.

317

FIGURE 119 (case 7).-Management of organizingclotted hemothorax by decortication. A. Posteroanterior roentgenogram 6 daysafter wounding showing extensive right-side hemothorax. Note compression of apexof lung. B. Lateral roentgenogram 3 days later showing extensive obscuration atleft base and characteristic multiple fluid levels. C. Posteroanteriorroentgenogram 30 days after decortication. Right lung has completely reexpanded, but there is still basal pleural reaction present.

Recovery was uneventful except for an unexplained, transient temperatureelevation to 103? F. on the third postoperative day. There was no woundinfection and no evidence at any time that an empyema was developing.Roentgenograms 30 days after operation (fig. 119C) showed complete pulmonaryreexpansion, with a moderate residual pleural reaction at the right base. Thepatient was fully ambulatory when he was evacuated on 11 November.

Pleural fluid cultured at operation showed a heavy growth of nonhemolyticstreptococci. Microscopic examination of the peel revealed dense fibrinopurulentexudate, beneath which was a characteristic thickened membrane of organizingfibrin, showing fibroblastic and angioblastic proliferation.

Comment.-In this case, thoracentesis was not begun until the fourth day afterwounding. In spite of roentgenologic evidence of hemothorax, aspiration wasincreasingly less productive, and there was clear clinical evidence ofinfection. This is another instance of the successful performance ofdecortication shortly after wounding as a means of avoiding the development ofempyema.

Case 8.-A 22-year-old staff sergeant sustained shell-fragment wounds of theanterior right thorax, right buttock, and right wrist on 10 November 1943. Hewas expectorating bright red blood when he was received in an evacuationhospital. Because of pronounced abdominal rigidity, a laparotomy was performed12 hours after wounding. No injuries

318

FIGURE 120 (case 8).-Management of clottedhemothorax with retained foreign body and bronchopleural fistula by primarydecortication. A. Posteroanterior roentgenogram 6 days after wounding showingextensive right-sided hemopneumothorax with compression of pulmonary apex. B.Lateral roentgenogram taken 6 days after wounding showing characteristicmultiple fluid levels and large intrathoracic shell fragment. C. Posteroanteriorroentgenogram 6 days after decortication, closure of bronchopleural fistula, andremoval of shell fragment. Lung has completely reexpanded.

were found. Thoracentesis immediately after operation yielded 500 cc. ofblood, which was used for autotransfusion.

For the next 10 days, bloody fluid and air were frequently aspirated from theright hemithorax. On 16 November, an abrupt attack of dyspnea was relieved byaspiration of 1,500 cc. of air. Clinically, there was no doubt of the existenceof a bronchopleural fistula. Roentgenograms (fig. 120A, B) showed a partiallycollapsed lung with numerous intrapleural adhesions and extensivehemopneumothorax with multiple fluid levels. A large intrapulmonary shellfragment was seen indistinctly at the inner end of the ninth rib posteriorly(fig. 120A).

When the patient was admitted to a thoracic surgery service on 30 November,he was comfortable except for moderate dyspnea. Shortly after admission, 1,250cc. of air was removed from the pleural cavity. Final manometric negativepressures of 12, 6 cm. H₂O were maintained in spite of episodes ofcoughing.

Daily thoracenteses for the next 3 days yielded a total of 1,200 cc. of airand of 500 cc. of dark brownish-red fluid. The temperature rose to 102? F.daily. Two smears and a culture revealed many gram-positive intracellular andextracellular nonhemolytic streptococci.

Because of increasing toxicity and other clinical evidences of infection,thoracotomy was performed on 4 December. The pleural cavity contained 600 cc. ofdegenerating

319

blood and exudate. The lung, which was partly collapsed, was immobilized by afirm organizing peel partly covered by yellowish-green exudate. The apex of thelung was encased in a tubular sheath that was attached to the cupola of thethorax. After complete visceral decortication, the shell fragment seen in theroentgenograms was easily found and removed from the right lower lobe; itmeasured 20 by 7 by 5 millimeters. A posterolateral, moderate-sizedbronchopleural fistula in the upper lobe was repaired with interrupted suturesof fine silk. Drainage was instituted before closure.

The lung was completely reexpanded within 48 hours. Roentgenograms on thesixth postoperative day (fig. 120C) showed no abnormality except a triangulararea of density in the right cardiophrenic angle. The collection of pleuralfluid which it represented was promptly absorbed.

The patient was ambulatory 7 days after operation and was asymptomatic whenhe was evacuated to the rear 6 weeks after wounding.

Comment.-In this case, because of increasing toxicity, decortication wasperformed 3? weeks after wounding. Thoracotomy wouldhave been indicated, in any event, for the repair of the persistentbronchopleural fistula; and the foreign body in the lung was of sufficientlylarge size to require removal also. The combined procedure was carried outexpeditiously and without difficulty, and convalescence was uneventful. This isanother instance of prompt recovery after decortication as compared with aprobable long period of invalidism and disability if conservative measures hadbeen used and if the infection of the organizing hemothorax had been allowed toprogress to empyema.

Case 9.-A 24-year-old Italian officer, a prisoner of war, on 10 July 1943sustained an extensive perforating sucking wound of the right posterior thorax,with compound comminuted fractures of the right scapula and the third, fifth,and sixth ribs. The wounds were debrided, but the pleura was not closed.

By 17 July, increasing pressure was evident in the right hemithorax, andlittle clinical relief was afforded by repeated aspirations. Roentgenograms(fig. 121A) showed complete collapse of the right lung, cardiac shift, andsubtotal compression of the left lung. Wide

FIGURE 121 (case 9).-Management of tensionhemopneumothorax and total hemothoracic empyema by secondary decortication. A.Posteroanterior roentgenogram 8 days after wounding showing enormous right-sidedtension hemopneumothorax, with mediastinal herniation; cardiac shift to left;complete collapse of right lung; subtotal compression of left lung. B. Same, 23days after institution of dependent rib-resection drainage. Right lung is stilltotally collapsed.

320

FIGURE 121.-Continued. C. Technique of rightpulmonary decortication: (a) Appearance of totally collapsed right lung, whichis completely immobilized by thick sheet of organizing fibrin and exudate, sothat individual structures are scarcely discernible. After initial sharpdissection, a small flap of thick, organizing peel has been elevated by bluntdissection. Positive pressure has caused a slight herniation of lung throughopening. (b) Mobilization of lung and diaphragm after complete decortication ofvisceral pleura. Visceral pleura is essentially normal and lung is expansile butnot yet completely reexpanded. The ridge represents line along which thickenedfibrous rind was reflected from visceral onto parietal pleurae. D.Posteroanterior roentgenogram 14 days after decortication. All drainage hasceased, and lung now completely fills pleural cavity.

mediastinal herniation and depression of the right diaphragm completed thepicture of massive tension hemothorax.

The patient's condition became increasingly serious. Theintrapleural tension was controlled between 20 July and 31 July by the use of anintercostal water-seal catheter, but the lung on the right continued collapsed,and a total empyema developed.

The patient was received in a thoracic surgery center on 2August. The following day, 6 cm. of the tenth rib was resected, and a largetube was placed in the pleural cavity and connected to a water-seal bottle. Forthe first 24 hours, there was evidence of a small bronchopleural fistula. Forthe next 14 days, negative intrapleural pressure of from 10 to 15 cm. H₂Owas constantly maintained. Total collapse of the right lung (fig. 121B)persisted for more than 4 weeks, in spite of the negative pressure, adequatedrainage, and frequent intrapleural irrigations with aqueous solution (1:3,000)of Azochloramid (chloroazodin). Bronchoscopic examination showed no bronchialobstruction.

321

Sequestrectomy on the scapula was performed on 24 August. On 2 September,right-sided posterior thoracotomy was performed, with complete visceraldecortication (fig. 121C). As roentgenograms had shown, the lung was foundtotally collapsed, and its margins could not be identified. It did not reexpandon positive pressure, and the diaphragm remained depressed and immobile. Bycombined blunt and sharp dissection, a proper cleavage plane was establishedbetween the visceral pleura and the adherent organized peel, which was 5 to 6mm. thick. The visceral pleura was normally thin, but slight loss oftranslucency, with subpleural scarring, was observed in occasional areas. Therewas considerable oozing from the pleural surface.

Removal of the investing peel from the diaphragm mobilized itcompletely. When the lung, including the entire mediastinal surface of the upperlobe, had been completely freed, slight positive pressure produced 80-percentreexpansion. Drainage was instituted.

By the fourth day after operation, both drainage tubes had been removed, andthe lung completely filled the pleural cavity. Convalescence was entirelyuneventful. Two weeks after operation, roentgenograms (fig. 121D) showed theright lung to be well aerated. There was considerable resorption of theextensive peripheral reaction during the next 6 weeks. The reaction was to beexpected, since no attempt had been made to remove the thick organizing peelfrom the parietal pleura.

When the patient was heard from by letter 3 months after operation, he hadno complaints referable to the chest and was in excellent general condition.

Comment.-This is another instance of a desperately ill patient who did notrespond to conservative treatment and who, as the result of decortication, wasspared weeks and months of invalidism. Had conservative treatment beencontinued, he might have lost his life. The conditions found at operation madeit clear that only decortication could have permitted the lung to expand.

Case 10.-A 26-year-old soldier, on 21 September 1943, suffered a perforatingsucking wound of the left posterior thorax, with compound comminuted fracturesof the seventh, eighth, and ninth ribs posteriorly. On the day of injury, thesucking wound was closed surgically by using a strip of latissimus dorsi toreinforce the intercostal muscles. The skin was left open. Five days later,although the patient seemed to be recovering satisfactorily, anteriorroentgenograms showed, in addition to the rib fractures, a probable encapsulatedarea of pleural fluid superimposed on the cardiac shadow. They also showed amoderate amount of pulmonary infiltration that was considered due to bleeding.

On the 10th day after wounding, the patient became suddenly and rapidlydyspneic and complained of a sensation of tightness in the thorax and abdomen.When he was received in a thoracic surgery center shortly afterward, histemperature was ranging from 100? to 102? F. The left hemothoraxhad increased greatly, and there was some cardiac shift to the right. Theimmediate aspiration of 1,500 cc. of dark blood produced great subjectiveimprovement. Smears and cultures were negative.

During the next 48 hours, although no additional blood couldbe aspirated from the chest, there was evidence of increasing intrapleuralpressure. Roentgenograms (fig. 122A) showed increased left-sided obscuration,with further cardiac and mediastinal shift.

Left posterior thoracotomy was performed on 5 October, 14 days after injury.Fragmented portions of the three fractured ribs were removed, and the pleuralcavity was entered through the eighth intercostal interspace. The lung was 75percent collapsed. The pleural space was filled with about 1,000 cc. of firmlyclotted blood; there was practically no liquid blood. A thin layer of fibrin,which showed early organization, was removed from the visceral pleural surface.

Posterolaterally, there was a laceration in the left upper lobe, 3 cm. deep,which contained four fragments of ribs. The lung bled briskly when these wereremoved. The laceration was repaired with fine silk. Anterior and posterolateralintercostal water-seal drainage was instituted. Closure of the chest wall wasdifficult posteriorly in the area in which the wound of entrance had beenexcised.

322

FIGURE 122 (case 10).-Management of infectedclotted hemothorax by early primary decortication. A. Posteroanteriorroentgenogram 17 days after wounding and 7 days after thoracentesis, showinglarge hemorrhage on left, with cardiac shift to right. B. Same, 6 weeks afterdecortication and 2 weeks after secondary rib-resection drainage of smallpostoperative empyema. Pulmonary field is now clear, and posterolateral empyemahas been well drained.

Both grossly and microscopically, the organizing exudate showed miliaryabscesses. Culture of the blood clot removed at operation was positive foraerobic hemolytic Staphylococcus aureus.

Within 48 hours, thin, purulent drainage was evident from the lower tube. Atthis time, roentgenograms showed the lung to be well expanded at the apex.Laterally, an oval area of obscuration was seen, which later proved to be asmall loculated empyema. By the seventh day after operation, the posteriorportion of the incision, in the area in which the entrance wound had beenexcised, had become grossly infected. By the 14th day, it was evident that thelocalized empyema just described was connected with a small bronchopleuralfistula and that intercostal drainage was not adequate. The following day, a15-cm. segment of the ninth rib was removed, and the empyema cavity was openedwidely and packed.

Convalescence thereafter was rapid. The small bronchopleural fistula closedspontaneously. When the patient was evacuated to the Zone of Interior 6 weeksafter the second operation, roentgenograms (fig. 122B) showed good pulmonaryexpansion and almost complete clearing of the lateral and basal obscuration.

Comment.-This case is an illustration of the difficulties andcomplications which could arise when thoracentesis was delayed. It was notresorted to in this case until the 10th day after wounding, when the patient wasseriously ill, and when, as later ineffectual aspirations showed, organizationof the clot had already begun. The seriousness of his condition, with evidenceof beginning infection, was considered a justification for early decortication,only 14 days after wounding. Miliary abscesses were already present atoperation, and empyema developed later, but the cavity was small and theresponse to standard measures was good. This is another patient who wouldundoubtedly have gone on to a long period of established infection ifdecortication had not been resorted to promptly.

323

FIGURE 123 (case 11).-Management of infectedclotted hemothorax, complicated by segmental pulmonary necrosis and retainedforeign body, by early primary decortication. A. Posteroanterior roentgenogram,10 days after wounding and after induction of diagnostic pneumoperitoneum forlocalization of foreign body. Note large left-sided hemothorax, which hasincreased in size in spite of repeated thoracenteses. B. Same, 8 weeks afterwounding and 6? weeks after primary decortication, segmental lobectomy, andremoval of shell fragment.

Case 11.- A 19-year-old soldier sustained a penetrating woundof the left thorax on the Salerno beachhead on 11 September 1943. There was nohemoptysis or dyspnea at any time. Roentgenologic examination on 16 Septembershowed a hemothorax of moderate size and a large shell fragment lying wellposteriorly at the level of the tenth interspace.

At this time the patient was somewhat toxic. His fever increasedprogressively, the highest elevation being 103.4? F. Repeated aspirations wereproductive of only small amounts of old, sterile blood. After diagnosticpneumoperitoneum on 20 September, roentgenograms (fig. 123A) showed that fluidwas increasing in the left thorax, in spite of repeated thoracenteses.

On 23 September, left posterior thoracotomy was performed through the sixthintercostal space, after removal of posterior segments of the sixth and seventhribs. The lung was collapsed, and there was extensive clotting of the hemothorax,with early organization. The shell fragment, which was intrapleural and had noclothing attached to it, measured 22 by 16 by 11 millimeters. It lay in aloculus of clotted blood mixed with greenish exudate, which had a foul odor.After the pleural cavity had been emptied, decortication showed the visceralpleura to be relatively normal. The lower half of the lower left pulmonary lobewas so badly lacerated and necrotic that segmental lobectomy was necessary.Drainage was instituted. Bronchoscopy at the end of the procedure producedbloodstained mucus.

Cultures of the organizing blood clot showed a mixture of staphylococci andanaerobic nonhemolytic streptococci.

After operation, the lung reexpanded well at the apex, but a loculated pocketof serosanguineous fluid developed in the axilla. It cleared with aspiration.

For the first 14 days after operation, the temperature ranged from 101? to103? F. By the end of this time, drainage from the posterior intercostal tubehad become purulent. When the tube was removed on 10 October, after drainage hadapparently ceased, there was an immediate febrile flareup, and roentgenogramsshowed increased basal obscuration.

324

On 16 October, open thoracotomy, with rib resection, was performed and tubedrainage was instituted. The small basal empyema was rapidly obliterated, androentgenograms taken on 13 November (fig. 123B) showed complete left-sidedpulmonary expansion, with a minimal pleural reaction at the base. The patientwas in excellent condition when he was evacuated to the Zone of Interior a fewdays later.

Comment.-In this case, decortication was performed in the face of existinginfection and pulmonary necrosis. Convalescence was stormy, but thepostoperative empyema was basal and readily cleared with the proper drainage.This case represents one of the few in the Mediterranean theater in whichsegmental lobectomy or any other excisional pulmonary operation was necessary.⁶

References

1. Neurosurgery and Thoracic Surgery. Prepared and edited bythe Subcommittees on Neurosurgery and Thoracic Surgery, Committee on Surgery,Division of Medical Sciences, National Research Council. Philadelphia andLondon: W. B. Saunders Co., 1943.

2. Dunham, E. K., Stevens, F. A., Graham, E. A., and Keller, W. L.: Empyema.In The Medical Department of the United States Army in the World War. Washington:Government Printing Office, 1927, vol. XI, pt. II, pp. 33-392.

3. D'Abreu, A. L.: War Wounds of the Chest. In War Surgery Supplement No. 3,War Injuries of the Chest and Abdomen, Brit. J. Surg., 1952, pp. 383-403.

4. Edwards, F. R., and Davies, H. M.: Traumatic Hemothorax. Lancet2: 673-675,30 Nov. 1940.

5. Sellors, T. H.: Late Results of Missile Injuries of theChest Treated in Britain. Penetrating Chest Wounds Treated at Leicester ThoracicUnit. In War Surgery Supplement No. 3, War Injuries of the Chest and Abdomen,Brit. J. Surg., 1952, pp. 403-408.

6. Samson, P. C., and Burford, T. H.: Total PulmonaryDecortication. Its Evolution and Present Concepts of Indications and OperativeTechnique. J. Thoracic Surg. 16: 127-153, April 1947.

7. Violet (de Lyon): De la d?cortication pulmonaire dans 1'empy?mechronique. Arch. g?n. de m?d. 193 (vol. 1, series 10) : 657-678, 1904.

8. Lilienthal, H.: Empyema: Exploration of the Thorax With PrimaryMobilization of the Lung. Ann. Surg. 62: 309-314, September 1915.

9. Lockwood, A. L.: The Empyema Problem. Arch. Surg. 16: 297-321, January1928.

10. Delorme, E.: Nouveau traitement des empy?mes chronique. Gaz. d' h?p.67:94-96, 25 Jan. 1894.

11. Nicholson, W. F., and Scadding, J. G.: Penetrating Wounds of the Chest.Review of 291 Cases in the Middle East. Lancet 1: 299-303, 4 Mar. 1944.

⁶The reader is referred to chapter XI (p. 441) for long-term followupstudies on casualties whose chest wounds were complicated by hemothorax andhemothoracic empyema.

RETURN TO TABLE OF CONTENTS