counteropening was made in the chest-wall. Despite perfect drainage, secured by the daily passing of sterile gauze from one opening to the other, there ensued no resulting improvement. During the following three months there occurred morning remissions, with sharp evening exacerbations of temperature, with daily chills and sweats. There was marked emaciation, with weak and extremely rapid pulse. At this time I decided to substitute iodoform gauze for that previously used. This was drawn from one opening to the other and the pleural cavity freely packed. The temperature receded to normal upon the second day, and remained so for ten days. In order to avoid any misconception as to the effect of the iodoform, I reverted once more to the use of the sterile gauze. This was followed by an immediate elevation of temperature, which persisted several days and receded at once upon a return to the iodoform pack. It is, of course, recognized that a single case of this kind must not be accepted as establishing a direct relation of cause and effect, as regards the employment of the iodoform pack. Improvement was progressive and rapid.

Fig. 107.-Marked unilateral deformity following Schede operation for pneumopyothorax. Note lower edge of lung upon the right side. Compare with figures 75 and 76.

from this time. Nine months after the rib resection the patient, having gained thirty-five pounds in weight, a Schede operation was performed by Dr. Powers, which reduced the capacity of the pleural cavity to four ounces. The patient is now in excellent condition, the discharge being almost nil and the cavity holding but one and one-half ounces of salt solution. Figs. 107, 108, and 109 are of interest in showing the amount of deformity resulting from the multiple rib resection to effect an obliteration of the pus-secreting cavity. This would have been completely out of the question as an early operation.

The extent of the operation must depend largely upon the size of the cavity and the age of the patient. The operation is rarely demanded in children on account of the greater elasticity of the thorax. Among adults, however, the rigidity of the chest-wall presents an insuperable obstacle to the obliteration of the suppurating space, and in the absence of lung expansion the resection of ribs becomes the only rational procedure. The removal of a large portion of the bony thorax permits the

apposition and cicatrization of granulating surfaces. It is scarcely pertinent to the purpose of this book to enter into a discussion of the

Fig. 108. Same patient as Fig. 107, showing the lower edge of the lung in front and the swinging of the sternum toward the unaffected side.

comparative merits of the Estlander or the Schede operation, or to attempt a description of the technic.

Decortication of the lung, as introduced by Delorme, consists of the removal of the visceral pleura from the lower half of the collapsed and

Fig. 109.-Photograph of same patient as Figs. 107 and 108, showing extensive cicatrization, site of remaining sinus, and also boundary of lung in the lateral region.

atelectatic lung, which permits, to a very considerable extent, its subsequent expansion. This operation has been found to be much safer

and more easily performed than the attempted removal of adhesions between the costal and visceral pleura. The operation is sometimes practised in connection with Schede's multiple rib resection, in which event the skin-flap coalesces and cicatrizes with the denuded lung. In other cases a trap-door operation is performed, which includes the lifting of a portion of the chest-wall, permitting free access to the collapsed lung and the denudation of its pleura. Subsequently the trap-door is replaced and provision made for adequate drainage.

Discission of the pleura was devised and performed by Ransohoff, and has yielded fairly satisfactory results. This procedure consists of free incisions of the pleura carried downward to the lung proper.

SECTION III

TUBERCULOSIS OF THE PERICARDIUM AND PERITONEUM

CHAPTER LVII

TUBERCULOSIS OF THE PERICARDIUM

Etiologic and Pathologic Data. This condition, on account of its relative infrequency, is of much less importance than tuberculosis of the pleura or peritoneum. It is, however, considerably more common than generally supposed. The clinical evidences of tuberculous involvement of the pericardium are exceedingly obscure.

The position and size of the heart may be outlined with reasonable accuracy during life by a skilful examiner, and the results confirmed or modified by x-ray examination, but the only reliable data as to the frequency and nature of the pericardial complications in pulmonary tuberculosis are furnished by autopsy findings. To be of definite value it is obvious that the postmortem study should be conducted by a trained pathologist, and embrace a large number of autopsies upon tuberculous subjects. Only by means of such investigations are there afforded correct conceptions as to the prevalence of the condition.

During the first two years of the existence of the Phipps Institute for the Study of Tuberculosis there were conducted 143 autopsies, 88 of which were performed in the first year. Of the 88 autopsies, there was but 1 case of typical miliary tuberculosis with pericardial involvement. There were, however, 3 cases out of the 143 autopsies during the two years. Of the total number, there were reported 3 cases of local pericardial adhesions, 1 instance of general pericardial adhesion, 1 of thickened pericardium, and 8 of total obliteration of the pericardium. Only 1 case of acute serous pericarditis was found. Evidence of pericardial involvement of some kind was obtained in 18 cases out of 143, including both acute and chronic varieties. Actual tubercles were found in only 3 instances, and it is, therefore, more or less problematic whether or not the other cases were of actual tuberculous origin. It must remain a matter

of conjecture if the cases of chronic obliterative pericarditis and those with local adhesions were directly occasioned by tuberculous infection. It is reasonable to believe, however, that nearly all low-grade chronic inflammations of the pericardium among pulmonary invalids owe their existence to tubercle deposit. If the tuberculous character of nearly all idiopathic pleurisies is admitted, even among apparently non-tuberculous subjects, it is safe to assume that a similar involvement of the pericardium may occasionally take place among phthisical patients.

The autopsy report with study of heart lesions at the Phipps Institute was made by Drs. White and Norris.

In the third annual report, recently issued, there is contained a record of 57 autopsies conducted during the past year upon phthisical patients, with practically negative pericardial findings. Thus, out of 200 autopsies reported in three years, but 3 cases of miliary involvement of the pericardium were found, and 8, or 4 per cent., of obliterative pericarditis. During the first year White reports the latter condition in 3.4 per cent. of the cases recorded, and submits for comparison obliterative pleurisy of both sides in 4.5 per cent. of cases-the left side, 19.5 per cent., and the right in 17 per cent. Of the pericardial cases, one coëxisted with obliterative pleurisy upon each side, one with rightsided pleurisy, and one with general adhesions on both sides. No histologic investigation was made to determine precisely the tuberculous origin of the adhesions. Norris, in 1904, collected statistics concerning pericardial involvement in 1780 autopsies upon tuberculous subjects, and, exclusive of doubtful cases, found pericarditis, which was presumably tuberculous, to have occurred in 4.6 per cent.

Although the pericardium occupies a more or less protected position as far as opportunity for secondary infection is concerned, it is not surprising that tuberculous involvement should supervene on account of the existence of tuberculous processes in adjacent structures. The condition may originate from neighboring foci of infection in the pleura, lungs, mediastinal glands, and from caries of some part of the bony intrathoracic wall, notably the sternum, ribs, and dorsal vertebra.

In a large number of cases the symptoms of tuberculous pericarditis are entirely latent during life. At times there are present the usual manifestations of general miliary tuberculosis, without distinct evidence of pericardial involvement. It may exist clinically as the ordinary acute form of pericarditis, either of the dry variety or accompanied by effusion. Another group of cases may be expected to exhibit the symptoms of functional incapacity associated with dilatation and hypertrophy, which result in many instances from the existence of chronic pericardial adhesions. The two latter varieties are of especial interest, one pertaining solely to an acute pericardial condition, and the other relating to adherent pericardium, with possible changes in the size and position of the heart and accompanying circulatory disturbance.

Varieties. Acute tuberculous pericarditis which is likely to be overlooked clinically may exist in two forms-the dry plastic variety and the type characterized by effusion. The exudate may be serous, serofibrinous, purulent, or hemorrhagic. The more common form is the plastic pericarditis, which may be unattended either by subjective symptoms or physical signs. The serous membrane may present but a dull, slightly roughened appearance, or a shaggy, irregular fibrinous coating. The fibrinous exudation upon the internal pericardium varies

much in thickness, and successive layers of lymph sometimes completely cover the macroscopic evidences of tubercle deposit. The thickened membrane may be infiltrated with tiny yellowish-gray tubercles, or granulations may exist between the layers which later become confluent. Upon gross inspection there is often no evidence of tubercle deposit, but the tuberculous character of the process may be readily demonstrable, despite a normal macroscopic appearance.

Symptoms. The only subjective symptom of the acute plastic form of tuberculous pericarditis is pain, and this is not always present. It is rarely intense, although in exceptional instances quite distressing. The pain is usually referred either to the precordial region or to the tip of the ensiform appendix. It is sometimes sharp or stabbing, and occasionally synchronous with the cardiac pulsation. But little significance is to be attached to the presence of fever unless this develops in conjunction with pain and objective signs.

Upon palpation there is sometimes recognized, synchronous with the heart contractions, an appreciable fremitus to the left of the sternum between the third and fourth ribs. The important auscultatory sign is the to-and-fro friction-sound. This may be creaking, grazing, rubbing, or grating in character. It is more frequently a rub, and gives the impression of coming from directly under the stethoscope. These superficial sounds are usually intensified by increased pressure upon the skin with the bell of the instrument. They may be heard at the base, but the more frequent site is over the right ventricle. A peculiar feature of the friction-sounds is their inconstancy, as they are often recognized at one time and not at another. They are distinguished from the pleuropericardial friction-rub by the difference in rhythm and the influence of the respiratory movement. There is but little difficulty in differentiating the distinct rubbing character of the pericardial sounds from the soft blowing endocardial murmurs. This form of tuberculous pericarditis may terminate in effusion, or the condition may be changed into that of adherent pericardium, owing to the fusion of the serous surfaces through connective-tissue proliferation.

The symptoms of tuberculous pericarditis with effusion are extremely variable, and, as a rule, are unpossessed of special significance. In many cases they are entirely absent for prolonged periods, and the condition escapes recognition altogether. At other times the diagnosis is made purely through recourse to the objective signs. It is not unusual for symptoms previously latent, suddenly to assume an aggravated character. Marked dyspnea, pain, pallor, or cyanosis may quickly supervene upon a period of ill-defined malaise, slight shortness of breath, and tendency toward fatigue. Thus the development of the condition is occasionally found to be decidedly insidious, without exhibition of clinical manifestations, until the effusion has attained such size as to permit pronounced physical signs. On the other hand, the onset is not infrequently acute and attended by well-defined symptoms, even before the recognition of pericardial exudation. Often complaint is made of pain, which may range from a sharp, agonizing stab to a mere sense of oppression referred to the precordia. The pain may be increased upon pressure at the lower end of the sternum. A beginning shortness of breath in acute or subacute cases rapidly changes to a marked dyspnea, which necessitates a maintenance of the upright or semireclining posture. The patient is distinctly restless, and the expression anxious.