104

EMPYEMA )JTHORACISBy KENNETH S. MULLARD, F.R.C.S.

Thoracic Surgeon, Harefield Hospztal

Empyema has largely changed its character ofrecent years. While formerly the problem wasthat of the treatment of an acutely ill and toxicpatient, now, with effective chemotherapy, it hasbecome a technical matter of removing a collectionof fluid, with the least possible risk and with thespeediest possible return to full recovery, froma patient who is seldom in any immediate danger.Since the surgical treatment of empyema ceased tobe urgent and life-saving, more attention has beenpaid to the avoidance of chronicity, and to insis-tence on full functional recovery, matters whichreceived scant attention in the past. This articleis concerned, therefore, chiefly with the detailedmanagement of acute empyema and reviews brieflythe aetiology, pathology and diagnosis of empyema,together with chronic and tuberculous empyema.Aetiology

Most empyemata arise as a complication ofinfection in the underlying lung. This infectionis usually a primary pneumonia due to a specificorganism, the pneumococcus, streptococcus,staphylococcus, Friedlander's bacillus, etc. Thepurulent effusion in the pleural cavity may collectwhile the lung infection persists, or it may firstbecome apparent when the lung infection is sub-siding. This is as true of pneumonia treated bychemotherapy as it was of those not so treated, butthe division into syn-pneumonic and post-pneumonic empyema is of little practical valuetoday. Besides primary pneumonia, infection inthe lung may be due to inhaled material fromdental, sinus or pharyngeal infection; inhaledseptic material is far more likely to cause pneu-monia and empyema than the classical obstructivelung abscess. Lung infection may be secondaryto an inhaled foreign body or, most important,secondary to a carcinoma of brotchus. ,Empyemain association with carcinoma of bronchus isusually a late phenomenon, found at a stage whenthere is extensive infection ana destruction of lungdistal to a bronchus obstructed by a growth whichhas, by this time, spread widely in the mediastinumand elsewhere. This picture is so often seen thatempyema is sometimes regarded as contra-indicating pneumonectomy for carcinoma; butan empyema may occur at an early stage, when

resection is still practicable and worth while(Fig. i). Such an empyema is a token of theVirulence of the infection trapped beyond thebronchial obstruction, and is not, in itself, evidencethat the growth itself is far advanced. Empyemamay be provoked by an infected pulmonaryinfarct; it may sometimes occur in actinomycosis,and it may be secondary to a flare up of infectionin a bronchiectatic lung.So much for empyema secondary to lung

infection. The next main group consists ofempyema secondary to direct injury or infectionof the pleura. Penetrating or non-penetratingwounds and injuries may be responsible. Infectedhaemothorax is the usual precursor of frankempyema in this group. Empyema occurring asa complication of thoracic operations falls intothis group. After most intra-thoracic operationsthere is a pleural effusion, provoked partly bymechanical irritation of the pleura and partly bythe irritation of blood oozing from the operativesite. Rapid re-expansion of the lung to obliteratethe dead space, together with removal of theeffusion by aspiration or drainage, and adequatechemotherapeutic ' cover' are the best safeguardsagainst the development of empyema.

Traumatic rupture of a bronchus is a cause ofempyema which is often overlooked. A crushinjury of the chest may, in addition to causing ahaemothorax, shear across a bronchus. Theresulting empyema is, of course, quite intractableunless the causative lesion is recognized. Theempyema is due to the infection spreading fromthe lung, from which secretions cannot draineffectively. Direct contamination of the pleuralspace from the torn bronchus does not occur,since there is not complete disruption of, allcomponents of the bronchial wall ; the effect of theinjury to the bronchus is to cause a bronchialstenosis. Direct contamination of the pleuralspace from the bronchus does occur, however,as a complication of all forms of pulmonary re-section should the suture of the bronchial stumpbreak down.Empyema may be due to injury to the oeso-

phagus. When the mucosa alone is damaged,mediastinitis follows, but if gross rupture of alllayers, including mediastinal pleura, occurs, as in

copyright. on June 26, 2021 by guest. P

rotected byhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.27.305.104 on 1 March 1951. D

ownloaded from

http://pmj.bmj.com/

March 1951 MULLARD: Empyema Thoracis sOS

......

.s;......

FIG. I.-An empyema following an acute pulmonary infection in a man of 34, with no previous history ofchest disease. The empyema was due to a squamous-cell carcinoma of the right lower lobe bronchu3.

most cases of spontaneous rupture, and sometimesfollowing instrumentation, then empyema orpyo-pneumothorax results.A further group comprises empyema due to

causes other than pulmonary infection or directinfection of the pleura. A sub-phrenic abscesswill usually provoke a pleural effusion above it,which is at first sterile. If the sub-phrenic abscessremains untreated an empyema may develop,either from rupture of the abscess into the pleuralspace or by lymphatic spread. Liver abscessfollows a more chronic course and if it rupturesthrough the diaphragm it does so at a late stage,when the pleural space has been obliterated byadhesions, and the abscess discharges into the lung.Empyema thoracis has been defined as a collec-

tion of pus in the pleural space. It is well toremember that collections of pus may occur in theextra-pleural tissues. Breaking down mediastinalglands, or lower deep cervical glands, may causean extra pleural' empyema ' (Fig. 2). Osteomyelitisof a rib may do likewise (Fig. 3).The final group consists of tuberculous

empyema. Most of these cases occur as a com-plication of artificial pne.umothorax therapy. The

spontaneous rupture of adhesions, the rupture ofa cavity wall, and too extensive adhesion sectionin which tuberculous foci are opened are obviouscauses, but the great majority arise insidiouslyduring the course of treatment by artificialpneumothorax of pulmonary lesions in whichatelectasis of a segment, lobe or lung is a feature.The cause of the empyema in these cases is un-certain. Empyema is also likely to occur when apneumothorax is given up after some years oftreatment, an effusion developing which readilyacquires all the characteristics of a thick tuber-culous empyema. Contamination by faultyaspiration, or rupture of a pure tuberculousempyema into the lung with the production ofa broncho-pleural fistula, or ill-advised drainage,are the usual causes of a tuberculous empyemabecoming a mixed infection empyema.

PathologyThe striking feature of the microscopic patho-

logy of non-tuberculous empyema is that thepleura appears normal. To one side is lung tissue,which may show evidence of the infection whichgave rise to the empyema. To the other side is

copyright. on June 26, 2021 by guest. P

rotected byhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.27.305.104 on 1 March 1951. D

ownloaded from

http://pmj.bmj.com/

to6 POS'IGRADUATE MEDICAL JOURNAL March 1951

A

FIG. 2.-An extra-pleural ' empyema ' in a girl of25. Breaking down tuberculous glands in themediastinum and lower deep cervical groupcaused a cold abscess, which pointed in thesupra-clavicular triangle. Secondary infectiondeveloped, but the abscess never ruptured intothe pleural space or into the lung.

the layer of fibrin and pus cells. To the nakedeye, the fibrin layer may be thin and filmy, orhalf an inch in thickness and of the consistency ofcardboard or cartilage. The sharp angles whichare found in the normal pleural space, such asthe costo-phrenic angle and the angle between theaorta and paravertebral gutter, are filled withfibrin and obliterated, so that on opening anempyema widely it is difficult to make out land-marks. The fibrin layer passes smoothly fromthe lung to the diaphragm and chest wall. At themargins of the empyema there is an abrupt changefrom the fibrin-covered pleura to normal pleura.Unless an interlobar empyema is present, themargins of the fissures will be sealed by the fibrinlayer passing over them, but when this layer isdivided, the underlying fissure shows lightadhesions only. Similarly, the fringe of the lowerlobe is densely adherent to the diaphragm, butthe main diaphragmatic surface of the lobe is littleinvolved. The mediastinal surface of the lung isseldom involved, since an empyema usually arisesparietally, alnd the lung becomes lightly adherent

to the mediastinum before a dense fibrin layercan form on it.

This description almost holds true for tuber-culous empyema, except that a thick fibrin layeris commonly found. The important differencebetween the two is that in tuberculous empyemathe underlying lung frequently shows fibroussepta running at right angles to the pleura intothe parenchyma. Whether this feature is theresult of the empyema, or of the pulmonary tuber-culosis, is difficult to say. More probably it ispart of a fibrotic process following long-standingcollapse of the lung, as it is sometimes seen inchronic non-specific empyema. The visceralpleura is intact, as in non-specific empyema, butareas can be found where the tuberculous processin the lung has reached the surface and destroyedit. The contiguous parietal pleura may also beinvolved, so that there is no trace of the pleuralmembranes in these areas, all being replaced byfibrous tissue or tuberculous granulation tissue.These two points in the pathology are responsiblefor the difficulty in obtaining re-expansion of thelung following decortication, in those few cases inwhich re-expansion is desirable.

DiagnosisDiagnosis in the first place depends upon the

clinical finding of the signs of fluid in the pleuralspace. The next step is to obtain postero-anterior and lateral radiographs to confirm thesituation and extent of the effusion. The diag-nosis is made absolute by aspiration of some of thefluid, part of which is sent for cytological examina-tion and culture, part being retained for com-parison with later specimens. Having the know-ledge that an empyema is present, and with theradiographs at hand, the case should be reviewedand the exact cause of the empyema determined.An insidious onset will give rise to suspicion of anunderlying carcinoma, but a sudden and apparentlytypical pneumococcal pneumonia preceding theempyema by no means rules out carcinoma.Careful questioning as to any preceding change ofhabitual cough, chest pain or vague ill-health inthe months before the acute episode should neverbe omitted. Inspection of the radiographs maysuggest enlargement of the lymph glands at thehilum, or widening of the superior mediastinum.The mediastinum may not be shifted to the sideopposite the empyema, or may be drawn towardsthe side of the empyema, suggesting atelectasisof the lung beneath the effusion. If there is anysuggestion, from these points, that a carcinomamay be present, bronchoscopy should be carriedout. This will reveal a carcinoina, if present, inabout three-quarters of the cases. In the remain-der, the growth is situated peripherally in the

copyright. on June 26, 2021 by guest. P

rotected byhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.27.305.104 on 1 March 1951. D

ownloaded from

http://pmj.bmj.com/

March I951 MIULLARD: Empyema Thorac,'s I07

..:......

FIG. 3.-An extra-pleural 'empyema' in a girl of I3, secondary to staphylococcal osteomyelitis of the thirdrib. The sequestration of the rib, and the periosteal reaction is obscured by the density of the pus inthe normal film (a), but can be seen in the penetrating film (b).

lung, beyond the range of the bronchoscope. Inthese cases, a penetrating film may show a mass,or a film can be taken after as much fluid aspossible has been aspirated. Culture of empyemapus is not often helpful in these days of chemo-therapy, as it is usually sterile.A history of pain and a troublesome, querulous

patient complaining persistently of pain for whichthere appears insufficient reason, will, if ignored,cause difficulty in cases subsequently shown bypleural biopsy to be suffering from malignantendothelioma of the pleura (Fig. 4).

Radiographic evidence of collapse, with ahistory of chronic productive cough, suggestsbronchiectasis, and it can be anticipated thattreatment will be difficult.The diagnosis has to be made from lung abscess.

This is usually easily done on the history, signsand radiographic appearances. But a localizedempyema may simulate a peripheral lung abscess,and vice versa, in their radiographic appearances.This difficulty is most likely to arise in the dif-ferentiation between a localized empyema in theparavertebral gutter and a lung abscess in thedorsal segment of the lower lobe. In the last

resort, the diagnosis depends on the presence orabsence of lung sloughs in the cavity. Thepresence or absence of profuse purulent sputum isno certain guide, since a bronchopleural fistulamay develop from an empyema, and the sputumbe as profuse and foul as that from a lung abscess.

It may be necessary to distinguish betweena sub-phrenic abscess and a basal empyema.The two may co-exist. Drainage of the empyemaalone will not relieve the condition. A mostsatisfactory method of determining the presenceof a sub-phrenic abscess and its exact situationis to induce a pneumo-peritoneum, with 500 to1,000 cc. of air, and to take radiographs in thepostero-anterior and lateral planes, with thepatient in the upright position. Air will pass upunder the diaphragm and will separate the liverand diaphragm, anterior and posterior to theabscess, which will show as an area of adhesionbetween them. If no abscess is present, air willfill the entire sub-phrenic space, the liver willfall away, and the diaphragm, with the empyemaabove it, will be clearly outlined. There is nodanger that this procedure will rupture a sub-phrenic abscess into the general peritoneal cavity

copyright. on June 26, 2021 by guest. P

rotected byhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.27.305.104 on 1 March 1951. D

ownloaded from

http://pmj.bmj.com/

log POSTGRADUATE MEDICAL JOURNAL March 1951

..ni..

FIG. 4.-A small basal empyema from whichsterile pus was aspirated, in a man of 50. Per-sistent pain led to a diagnosis of pleural endo-thelioma, confirmed by pleural biopsy.

if it is carried out at the stage at which an empyemahas had time to develop, i.e. some weeks after theperitoneal infection first occurred.

It remains to consider the differential diagnosisfrom tuberculous pleural effusions and emn-pyemata. Nowadays most empyemata are foundto be sterile, thanks to chemotherapy, but pus cellswill be found, in varying quantity, depending onthe type and age of the empyema. . A cleareffusion, containing a few lymphocytes and a fewpolymorphs, is suspicious of tuberculosis. Theage of the patient, the Mantoux reaction, the re-sults of gastric lavage and laryngeal swab culturesfor tubercle bacilli, culture of the fluid itself fortubercle bacilli, the radiological appearances inthe lung after aspiration of the fluid, the tendencyor otherwise to reaccumulation of the fluid afteraspiration, and the general progress of the patientwill, between them, make the diagnosis, but thiswill only be done after some weeks in atypicalcases. In the diagnosis of the clear sterile effusion,it must be remembered that a sudden atelectasis ofa lobe, or even a segment, will usually provoke asmall sterile effusion no matter what the cause ofthe atelectasis.

TreatmentIn the introductory paragraph it was pointed out

that the danger to life in the acute stage of em-pyema has largely been removed by chemo-therapy. The good general condition of thepatient whose empyema has been sterilized leadsto a reluctance to impose upon him the discomfortof tube drainage, and there is a tendency to per-sist with aspiration for too long. Many empyematacan be treated by aspiration alone, but there aremany pitfalls. A normal temperature and pulsechart certainly shows that infection is controlled,but gives no guarantee that a small, untappedloculus will not flare up at a later date, some timeafter chemotherapy has ceased. These pockets arevery difficult to find with the exploring needle.They show their presence by rupturing into thelung, when their contents are coughed up causingsome pneumonitis in the process. All is thenquiet for a time after this spontaneous drainageuntil the pocket reforms and again ruptures intothe lung, until eventually the affected lobedevelops considerable bronchiectasis.

Radiological control of the progress of aspira-tion treatment is unsatisfactory, especially in thelater stages, when it is impossible to differentiatebetween fibrinous thickening of the pleura and asmall empyema too thick to be aspirated.By keeping specimens of the pus aspirated, as

advised by Barrett (1950), the process of thicken-ing can be watched, but it is always possible that athin loculus is being successfully aspirated and thata loculus of thick pus. which no longer com-municates with it, is being neglected. Strepto-kinase-streptodornase may be of value inpreventing the undue deposition of fibrin, but doesnot greatly affect that already laid down. Ingeneral, aspiration should not be continued oncethe pus has become thick, or does not flow readilythrough a needle. In practice this means thatempyemata containing thin pus which can beremoved entirely with a few aspirations spreadover two or three weeks at most are suitable foraspiration alone, but that for the remainder, someform of surgical treatment will be necessary.Aspiration is, further, of great value in infants andsmall children who do not tolerate tube drainagereadily.Some technical points in aspiration of the chest

are of importance. The patient must be com-fortably placed so that he will not tire and moveshould the procedure be prolonged. The positionmust also be comfortable for the operator so thathe can manipulate a syringe for a long time withoutmoving the needle through fatigue. All aspirationsshould be carried out with a syringe with a two-way tap. The joints between syringe, needle andtap must be of the locking type, not the ordinary

copyright. on June 26, 2021 by guest. P

rotected byhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.27.305.104 on 1 March 1951. D

ownloaded from

http://pmj.bmj.com/

March I95I MULLARD: Empyema Thoracis log

A (7I~UA~

Br

FIG. 5.-(a) A typical basal empyema whichshould heal within two or three weeks of opendrainage. (b) A total empyema, which wouldprobably take many weeks to close if treated bytube drainage; more rapid recovery can be7achieved by decortication.

push-on fitting. The needle should be of widebore. It is helpful to have an old artery forcepswhich can be clamped across the needle flush withthe skin as soon as the needle point reaches therequired depth and fluid appears in the syringe.The forceps prevents the needle being in-advertently introduced deeper, and should it bewithdrawn a little during the aspiration the causeof the sudden cessation of flow of fluid can beseen at a glance, the forceps now lying away fromthe chest wall. It is of the greatest importancethat no air should gain access to the pleural space,thus the insistence on the two-way tap with lock-ing connections. If air is allowed to enter it willrise to the apex of the empyema cavity and willreadily separate the pleural layers unless they arefortunately very firmly adherent. The pleurallayers above the empyema being separated, alocalized empyema is converted to a total pyo-pneumothorax. The simple act of disconnectingthe syringe from the needle during aspiration maybe the direct cause of prolonging the disabilityperiod from a matter of weeks to one of months oryears.

If surgical treatment is required it is necessaryto decide whether to employ open or closeddrainage or decortication. Drainage is simple andsafe; decortication (which is fundamentally anelaboration of closed drainage) involves a majorthoracotomy, and must always carry an added risk.

Drainage, however, requires skilled and perhapsprolonged after-care, while decortication, oncethe immediate post-operative period is over, isfree of the problems of dressings and tube manage-ment. Experience shows that the typical basalempyema treated by drainage is healed in fourweeks, and certainly within eight weeks. Thepatient, by the time the tube is removed, iscapable of full physical exertion and can returnimmediately to hard manual labour. Few patientswill resume full work within eight weeks of de-cortication, but should do so at about that time.A total empyema, or an apical empyema, treatedby drainage is likely to take longer than eight weeksbefore the tube can be removed, and is thereforesuitable for consideration for decortication in orderto reduce the disability period as much as possible(Fig. 5). Decortication is therefore to be con-sidered where a long disability period is anticipatedwith simple drainage. Decortication is useful invery fat patients in whom tube management isdifficult. It is useful also when patients have totravel long distances to obtain proper tube manage-ment. Rapid re-expansion of the lung followingeither drainage or decortication, depends moreupon the co-operation of the patient than on anyother single factor. It is a mistake to think that apatient who will not carrv out exercises, etc., willdo any better with decortication.

Exercises, best supervised by a trained physio-therapist, form an essential part of any treatmentof empyema. The exercises are designed to main-tain an erect posture and to prevent the appearanceof the earliest signs of the typical deformity of thechronic empyema. Breathing exercises are de-signed to obtain full recovery of respiratoryfunction. The exercises are localized expansionexercises against pressure, exerted by the physio-therapist's hand, or the patient's own hand, or bya belt. These localized exercises, in which atten-tion is focused on each part of the chest in turn,apex, axilla, base and diaphragm, are of thegreatest value in ensuring that the area of the chestaffected by the empyema is not neglected and thatproperly co-ordinated respiratory movements arerestored. In addition, patients should engage inthe most active pursuits possible, long walks,cycling, gardening, etc.

TechniqueAll empyemata should be drained in the sitting

position. It is only in this position that drowningin a sudden flood of pus into the bronchial treethrough a bronchopleural fistula can be avoided.Local or general anaesthesia can be used.

Closed intercostal drainage may rarely be neces-sary in an emergency and consists simply in intro-ducing through the appropriate intercostal space

copyright. on June 26, 2021 by guest. P

rotected byhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.27.305.104 on 1 March 1951. D

ownloaded from

http://pmj.bmj.com/

1o POSTGRADUATE MEDICAL JOURNAL March 1951

POST ANT. POSTAANL POST ANT.

POSAT, jj\A NT POST6B

POST ~ ANT. POST AT OTAT

FFIG. 6.-Progressive closing of empyema cavities, as seen on typical pleurogramns. (a), (b) and (c), lateral

views showing concentric shrinking of cavity towards the base; the tube is short and just enters the base ofthe cavity throughout. (d) and (e), paravertebral empyema. At the stage (e), it is necessary to lengthenthe tube so that it reaches within an inch of the apex of the cavity if the stage (f) is to be avoided.

the largest trocar and cannula possible. The trocaris replaced by a tube connected to a water seal andthe cannula withdrawn.

Almost always it is best to resect a rib to obtainadequate drainage. The situation of the mostdependent point of the empyema cavity can bedetermined by instilling some lipiodol into thecavity and taking radiographs in the uprightposition. But whether this is done or not, a testaspiration should be carried out, with the patientin position for drainage, immediately prior to theincision. If no pus can be aspirated, it is wise todefer operation. The lowest point of the cavitybeing found, a vertical incision, 2 in. in length, ismade over the rib selected. The muscle is divided,the periosteum incised and raised with an elevator.Two inches of rib are resected. The pleura is thenopened sufficiently to allow a sucker to be intro-duced and the cavity evacuated. But the cavitycannot be sucked out in this way unless the wallsof the empyema are firm and the mediastinumthus rendered stable. If these conditions do notobtain, opening of the pleura will cause a suckingpneumothorax and the patient's condition willdeteriorate rapidly. All that can or should be donein these circumstances is to introduce a tube intothe pleural cavity and to close the wound closelyaround the tube. This tube is connected to a waterseal. This procedure of closed drainage has its

prime indication in the conditions describedabove; that is, in the drainage of an empyema inwhich open drainage would result in an openpneumothorax.

Provided that the walls of the empyema are firm,no respiratory disturbance will follow opening thepleura. The cavity is sucked out and the pleurathen opened widely to allow fibrin clots to beremoved. The cavity is gently explored and, ifloculated, free drainage of the loculi ensured. Aroutine biopsy of the pleura should now be taken.A straight wide-bore tube (i in.) is then placed sothat its inner end projects 1 in. into the pleuralcavity. Penicillin-sulphathiazole powder is thendusted into the wound, which is closed looselyaround the tube. A safety pin is passed throughthe tube flush with the skin and secured with strap-ping. No dressings are placed between the skinand the pin. (If they are so placed, the tube isliable to be displaced when the dressings arechanged, or a greater thickness of dressings will bereplaced, so that the tube will be withdrawnsomewhat. It is simplest and best to have nodressings beneath the pin.) No side holes are cutin the tube, either at its first insertion, or subse-quently. Granulation tissue will grow into theside holes, so that pain and bleeding occur whenthe tube is removed for cleaning, and the button ofgranulation tissue will prevent the tube being

copyright. on June 26, 2021 by guest. P

rotected byhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.27.305.104 on 1 March 1951. D

ownloaded from

http://pmj.bmj.com/

MTarch 1951 MULLARD: Em??pyeena Tho; acis III

A

FIG. 7.-Decortication. (a) Beginning raising the fibrin layer from the visceral pleura. (b) The upper lobefreed. (c) Completed. The entire visceral pleura, including the fissures, has been freed, and the lungexpands readily. The diaphragm has also been freed.

replaced to its correct depth. At the time of opera-tion the tube can conveniently be left long, so thatit can be attached to a water seal for a few days.This will allow the heavy discharge of pus duringthis period to be conveyed out of the bed awayfrom the patient and will thus avoid frequentchanges of dressing. The gentle suction of thewater seal may help re-expansion of the lung, butin this case, where the empyema walls are firm,the water seal and closed drainage are a con-venience and a help, as distinct from the casedescribed above, of the prime indication forclosed drainage, where it is a vital necessity.

After a few days, when the heavy discharge hassettled, the tube is cut short, I in. from the pin,and the patient gets up and begins to acquire themobility and activity which is his best assurance ofrapid re-expansion and healing.The subsequent management of the case can be

divided into two parts. First is the encouragementof colrect breathing and posture, and secondly thedetailed management of the tube. If the em-pyema has been correctly drained, the dischargewill diminish to a very small amount of inoffensivepus. Provided this occurs the tube mav be left inposition for two to three weeks, when it is removedand the size and shape of the cavity determined byrunning lipiodol into the sinus. Films are taken intwo planes. The tube is replaced as soon as thefilms have been taken since the sinus contracts

very rapidly, and if the tube is left out for half anhour it may be difficult to replace. The site of theskin sinus should be marked with a radio-opaquemarker-a cuirtain ring serves-so that its positionis apparent on the films. Unsatisfactory pleuro-grams are obtained if the tube is left in while theoil is instilled. The pleurograms enable the sur-geon to keep the mouth of the tube just within themost dependent part of the cavity and also towatch the rate of re-expansion of the lung. Thepleurograms should be repeated every two or threeweeks. During the process of re-expansion thecavity may shrink concentrically towards its base,in which case little or no alteration in the lengthand direction of the tube will be necessary; butthe lung may move towards the chest wall in sucha wav that a loculus may become separated fromthe main cavity. In this case it will be necessaryto lengthen the tube so that the mouth enters thepocket so as to prevent it being nipped off from themain cavity (Fig. 6). Some complex empyemacavities may require two tubes and it may benecessary to use two or more separate entrysinuses, but the use of multiple rib resectionsdelays the return of full function. If a long intra-pleural tube has to be used the lung will eventuallyre-expand so that the tube is closely invested andthe cavity is converted to a tube track. When thisoccurs the tube is maintained 3 to i in. shorterthan the track. At the fortnightly pleurogram it

copyright. on June 26, 2021 by guest. P

rotected byhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.27.305.104 on 1 March 1951. D

ownloaded from

http://pmj.bmj.com/

112 POSTGRADUATE -MEDICAL JOURNAL March 1951

*.;.~ ~~ ..::......l;l..:.

..... .I....

FIG. 8.-(a) The deformity of chronic empyema.

will usually be found that the end of the track hasclosed by this amount and the tube and the trackare the same length. The tube is thereforeshortened by another similar amount untilfinally, when the tube passes no more than 2 in.into the pleural cavity, it can be removed. Thetotal length of the tube just before it is discardedis usually about 2 in. in the normal but more in theobese patient.The operation of decortication requires a full

thoracotomy, the usual practice being to resectthe greater part of the sixth rib from the tip of thetransverse process to the anterior axillary line.The pleural cavity is then entered and the chestwidely opened by a mechanical rib spreader. Pus,fibrin and debris are sucked out. The anaesthetistmaintains the lung as fully expanded as possibleand an incision is made, at a convenient point,through the fibrin layer, until the blue glisteningpleural surface is seen. The fibrin layer can thenbe peeled away from the pleural surface by blunt

and occasional sharp dissection. The entirevisceral pleura must be freed and as this is donethe lung progressively expands until it fills thechest (Fig. 7). It is difficult, if not impossible, todecorticate the parietal pleura, and it is notnecessary to do so. Difficulty in decorticating thevisceral pleura usually arises around the fringe ofthe lower lobe where the fibrin layer passessmoothly on to the diaphragm, and in clearing thediaphragmatic surface of the lower lobe. Themediastinal aspect of the lung is very seldomdensely adherent since it rarely forms part of thewall of an empyema cavity. Although the objectis to separate the fibrin layer from the visceralpleura this can seldom be achieved without tearingthe pleura. Multiple small alveolar fistulae are pro-duced, through which frothy blood is blown. Be-yond causing inconvenience to the operator theyare of no importance, and close sporntaneously. Thelung being fully expanded, the chest is closed withtwo or three closed intercostal drains connected to

copyright. on June 26, 2021 by guest. P

rotected byhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.27.305.104 on 1 March 1951. D

ownloaded from

http://pmj.bmj.com/

MULLARD: Empyema Thoracis

light suction pumps or water seals. It is essentialto maintain the lung fully expanded until the pleu-ral layers become adherent. There will be someoozing of blood from the operative area togetherwith a reactionary effusion during the first fewpost-operativre days, and if this and aol the residualair in the pleural cavity is not removed the lungwill fall away from the chest wall, which it reachedat the end of the operation, and the empyema willre-form. To avoid this it is customary to use atube at the apex of the chest, through the second orthird intercostal space, and a second at the base,posteriorly. A third may be placed in the axilla.These tubes are removed as and when they ceaseto drain fluid or air, which occurs when the lungin re-expanding seals them off. Drainage canusually be discontinued after four or five days.The post-operative period requires very closesupervision and frequent radiographic control.A development of the operation of decortication,

applicable to small localized empyemata, consistsof the excision of the empyema cavity intact withits contents. As before, a long piece of rib is re-sected, overlying the centre of the empyema. Theparietal pleura is not opened, however, but isstripped away from the chest wall by blunt dis-section until the edges of the hard empyema wallare felt to give way to the normal soft pleura. Theparietal pleura is incised around the edge of theempyema and the dissection continued along thedeep surface of the empyema, in the plane be-tween the visceral pleura and the fibrin layer.When this is completed the empyema can be re-moved like a cyst. The remainder of the pleuralspace is freed of light adhesions and the chestclosed with drainage as described above.

Chronic EmpyemaIt remains to consider the causes and treatment

of chronic empyema. By far the commonest causesare delayed or-inadequate drainage or a combina-tion of the two. Re-expansion of the lung is pre-vented by the thick fibrin layer deposited on thevisceral pleura; movement of the chest wall,essential if the lung is to expand, is prevented bythe similar deposit on the parietal pleura. The ribsfall together and eventually overlap, assuming atriangular shape on cross section. The inter-costal muscles atrophy. A scoliosis develops withthe concavity towards the side of the empyema.The shoulder on the affected side drops, the neckinclines to the empyema side and the head iscarried vertically but above the middle of theclavicle (Fig. 8). Prolonged physiotherapy can domuch to reverse these deformities provided that theempyema is properly drained. Other causes ofchronicity are the presence of a foreign body in theempyema cavity-a piece of drainage tube, etc.,

but this is not commonly found. The investiga-tion of the cause of chronicity in a well-drainedempyema is more fruitfully directed towards thestate of the bronchus and the lung. Bronchialobstruction, due to neoplasm, foreign body orstricture must be excluded by bronchoscopy, andif this shows no lesions then a bronchogram mayshow bronchiectasis.The treatment of chronic empyema-provided

the cause of the chronicity can be removed-consists firstly in a prolonged course of physio-therapy to obtain as much re-expansion and toreduce as much of the deformity as possible.Provided that the empyema is not so long-standing that fibrosis of the lung has occurred,decortication may be successful. All attempts tore-expand the lung having failed, it becomesnecessary to obliterate the space by some plasticoperation on the chest wall. A small empyemacavity may be obliterated by a muscle graftoperation, making use of the latissimus dorsi andthe intercostal muscles to fill the space. The bloodand nerve supply of the muscles used must bepreserved, and the muscle flaps must be carefullydesigned for this purpose. Large chronic cavitiescan often be closed by a postero-lateral thoraco-plasty, long segments of ribs being removed. Itis always necessary to remove the upper ribs,although they may lie above the cavity. If theyare left, the chest wall will not fall in adequately.The extent of the thoracoplasty towards the basedepends on the site of the lowest point reached bythe empyema cavity; one or two ribs below thislevel must be removed. The operation, consistingas it does of the total or subtotal resection of nineor ten ribs, is carried out in two or three stagesand is one of considerable difficulty, because theribs are triangular in shape, overlapping andbrittle. In addition, there is profuse haemorrhagefrom the chronically inflamed periosteum andchest wall tissues generally. Roberts' operationconsists essentially of raising the parietal wall ofthe empyema as a flap, which is mobilized byremoving from it the ribs and scar tissue. It isthen allowed to fall in and adhere to the visceralwall. A considerable raw area is left to heal bygranulation.

Tuberculous empyemaThe treatment of this condition is so closely

bound up with the treatment of the tuberculousdisease in the underlying lung, that it cannot beconsidered separately.. Since most tuberculousempyema arise in the course of artificial pneumo-thorax treatment, the empyema can usually becontrolled by repeated aspiration, and re-expan-sion of the lung. The immobilization of the lungprovided by the thickened pleura and somewhat

March I1951I II 3copyright.

on June 26, 2021 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.27.305.104 on 1 M

arch 1951. Dow

nloaded from

http://pmj.bmj.com/

114 POSTGRADUATE MEDICAL JOURNAL March 1951

sunken hemithorax may provide some insuranceagainst re-activation of pulmonary disease. Shouldthe state of the lung be such that re-expansion isundesirable or impossible, thoracoplasty willboth obliterate the empyema and allow the lung

disease to heal. If there is gross destruction ofthe lung, with an associated empyema, extra-pleural pneumonectomy, excising both the lungand the empyema, by dissection in the extra-pleural plane, may be indicated.

BIBLIOGRAPHYBARRETT, N. R. (ig5o), 'Techniques in British Surgery.' Phila-

delphia. W. B. Saunders.

ON THE TEACHING OF GERIATRICSBy T. N. RUDD, M.D.(London), M.R.C.P.Physician, the Belmont Hospital, Tiverton, Devon

Physicians who practise in hospitals devoted tothe care of the elderly will be in general agreementwith the view that many highly experienced andotherwise excellent doctors are nevertheless lackingin their approach to disease in old age. Theyadopt what will later be described as a ' negativeapproach,' characterized by an attitude of frustra-tion and laisser-faire. When the doctor concernedis a general practitioner his attitude will, inaddition,, be coloured by the great difficulties whichattend the care of the very old in their homes.There is, however, another and a positive approachto the subject, the adoption of which has resultedin many improvements, made during the lastdecade, in the care of the elderly in hospitals.More intensive study of old people's illnesses hasled to the development of a technique of geriatrics.Knowledge of this technique is at present in itsearly childhood and so far there has been, in thiscountry at least, little organized teaching of thesubject. Textbooks on the diseases of the agedhave appeared in America, where there is also a bi-monthly journal (Geriatrics) ' devoted to researchand clinical study of the aged and ageing.' Inthis country no textbook has been produced, whilesuch articles as have appeared under geriatrictitles have been usually of an administrative ratherthan a clinical nature. Clinical problems ofdisease in the elderly have received little attentionfrom British teaching centres, although at leastone teaching hospital in London has appointed alecturer in geriatrics. The urgency of the prob-lems of disease in the aged warrants a much moreintensive tea( hing of the subject.

VAhen tea hing of such a nature is considered,several great questions immediately present them-selves. 'I hese can be summarized under the

headings W"-hy? When? Whom? Where? What?and Vow? It is under these headings that theteaching of geriatrics will be considered.

Why Should Geriatrics be Taught?Geriatrics shzuld be taught for three main

reasons:(i) There is considerable ignorance among all

sections of the medical and nursing professions asto what can be achieved by a positive, activeapproa(h to therapy, even in the very old. Thevariations v h h diseases show, uhen they presentthemselves in aged subjects, are not, as yet, widelyunderstood and many disease conditions remainuntreated, either because cure at such an age isconsidered not worth while, or because the risk oftherapy has not been balanced against the dis-advantages of leaving the patient untreated. Muchmore treatment, both medical and surgical, couldadvantageously be given in the later years of life.The completion of second stages of prostatectomyoperations is a case in point.

(2) It is economically imperative to keep theelderly self-supporting physically, and indepen-dent of care by relatives and hospitals. Sheldon(1950) has pointed out that it is probable that thewage-earning population is already bearing itsmaximum load of c4ronic sickness and that itcannot carry any additional burden withoutreducing industrial output. Unless the elderly areless dependent in the future (when their numberswill be proportionately greater than today) thanthey are at present, the country will be faced witha grave problem.

(3) Preventive medicine in recent decades hasadded years to the life of the conimunity. Theseyears will be neither happy nor fruitful unless

copyright. on June 26, 2021 by guest. P

rotected byhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.27.305.104 on 1 March 1951. D

ownloaded from

http://pmj.bmj.com/