of the great deal of gas and distress, crying most ANOMALIES OF TUBERCULOSIS IN THE and six stools a day, passing green, irritating day HIGHLANDS OF COLOMBIA which, under the microscope, were seen to contain con¬ siderable fat. The mother had considerable gas in the A NEW DIAGNOSTIC SIGN IN INCIPIENT CASES bowels. The analysis showed: fat, 1.4 per cent.; lactose, 8 per cent., and protein, 1.07 per cent. JORGE VARGAS S., M.D. Although no single component was excessively high, Professor of the General Pathologic Clinic, National University of the relative proportions of the fat, lactose and protein Colombia were abnormal. In this instance after weaning, the NEW YORK a of cow's baby straightened out on simple formula of the of observed milk. Physicians early part this century a curious evolution of tuberculosis in the Colom- In another instance a was being fed by a wet- great baby bian These have an elevation of nurse who also gave her own baby the breast after Highlands. highlands about 11,800 feet above sea level, and are inhabited the foster-baby received what it needed. It was by noticed that the wetnurse's an Indo-Spanish race which numbers very few Indians although baby gained and descendants of the it was uncomfortable and a good part of the many pure conquerors. rapidly, The an of time had stools. As time went on the highlands have average temperature undigested from 14 to 16 C. 57.2 to 60.8 with an inex- fosterbaby did not receive enough milk and was given (or F.), all of the wetnurse's milk. It then commenced to haustible fertility, and are not very thickly populated. Both men and animals in this have indigestion, and it was found on analysis that progress admirably the additional milk, namely, the end of the milk of region, notwithstanding deficient hygienic conditions the contained 8 per cent fat. After a wet- inherited from our Castilian fathers. nursing, I wish to describe the invasion of tuberculosis in nurse with poorer milk was obtained, there were no symptoms of indigestion, and the baby gained weight these places, which a few years ago were almost free in a normal manner. of the Koch bacillus. Today the region is infested SUMMARY with this terrible malady, but with pathologic peculiari- ties different from what in the zone. 1. Mixed of milk taken before and. after prevails temperate samples to tuberculosis is the con- or the entire amount of one breast, must be Immunity produced by nursing, stant of antibodies nodes or other a of milk for production by lymph taken to obtain characteristic sample tuberculous tissues. In the of It should be drawn at 9 or 10 a. m. highlands Colombia, analysis. certain forms of tuberculosis are most common because 2.. Accurate methods of chemical analysis must be the race does not possess to tuberculosis in used. Clinical laboratory tests cannot be depended immunity any The of these is to me of on because of their inaccuracies. degree. study problems the greatest importance. In the years from 1896 to 311 Beacon Street. 1899, I was shown in the clinic of my teacher, Dr. Josue Gomez, in the Hospital de Caridad of Bogota ABSTRACT OF DISCUSSION (Charity Hospital), as a pathologic rarity, one or two Dr. John Foote, Washington : I think all the great pedia- cases of pulmonary tuberculosis. On the other hand, trists have written about the necessity of milk examination. there were frequent cases diagnosed as tuberculous Soranus of Ephesus was the first He described the thumb¬ peritonitis. Numerous necropsies demonstrated the nail test in the second century, A. D. Breast milk varies truth of his assertions. Dr. Gomez attributed this on when the milk is taken, whether at the begin¬ depending to various causes. I believe that it was ning, the middle or the end of the nursing, and the quality phenomenon tuberculosis of the mesenteric in men who of the milk often depends on the time of day or whether it lymphatics inva¬ is taken from one breast or the other, so the matter of mak¬ did not have ancestral immunity. This made the ing a milk analysis involves many considerations. This was sion of the tuberculosis more rapid so that the action pointed out clearly by Dr. Abt in a splendid monograph a of the antibodies did not stop the rapid peritoneal inva¬ few years ago. The thumbnail test of Soranus was quite as sion. Probably, then, the lungs were kept almost accurate as some of our modern lactometer estimations of immune, the mesenteric ganglions being the place the quality of breast milk derived from a single sample. selected by the first tuberculous invasion. The inva¬ : am convinced that Dr. James D. Love, Jacksonville I sion in these organs not immune per continuitatem the routine examination of breast milk as commonly prac¬ occurred with a which did not the with a lactometer more-often leads us than not. rapidity permit tised astray of antibodies. A mother may state that her milk is thin and bluish in development I have served as appearance and she submits a specimen for exarrtination. We During the last four years pro¬ may find that this milk is deficient in fats and apparently is fessor in the same clinic and have observed hundreds not a milk on which the baby will thrive; and yet the baby of such cases. There were three things that called my is thriving and taking on weight. Too often we base our attention especially to the curious nature of the disease estimate on examination of a given specimen of milk. As in the.highlands of Bogota: the rapid cavity formation, Dr. Foote has said, the milk in the morning will differ very the forms, and the tuberculous meningitis. materially from an afternoon specimen. If the baby is nurs¬ miliary ing every two or three hours, the milk carries materially N. N., a man from the country, aged 24, had vomited blood more fat than if the baby is nursing every four hours. I two months before admission, while at his farm, without merely wish to make the point that unless an examination is feeling previously ill. He felt better a few days later, con¬ made properly and the milk is selected from at least two or tinued his work at the farm, and having vomited blood again three specimens in the course of the day, the routine exam¬ forty days afterward, went to Bogota and took a bed in the ination of milk as performed by the average physician will Hospital de Caridad. often lead to erroneous conclusions. While making the morning rounds, my intern and my chief Dr. Fritz B. Talbot, Boston : The appearance of the milk head of the clinic showed me that case as gastric hemorrhage really does not tell you anything. We have seen milks that with vomiting. The man had powerful muscles and a per¬ have been very yellow, looked very rich, and yet they have fect frame, with not a single degenerative stigma. He was had only 0.9 per cent, of fat in them. a beautiful example of that hardened and healthy type of our

Downloaded From: http://jama.jamanetwork.com/ by a Florida International University Medical Library User on 05/28/2015 country. On examining him, simple revealed to It is well known that to , inspiration is me the existence of a cavity larger than a dove's egg in the noisy and expiration silent. The murmur is produced, apex of the left lung. The auscultation confirmed my diag¬ not only by the expanding of the pulmonary vesicles, nosis. There was nothing else; not an enlarged lymph node; but because each column of air in the flute¬ not even a noise besides the breaking strange cavity phenomenon. like which forms each bronchial subdivision has Based on previous cases observed in my clinic I gave a fatal point to a musical prognosis. In fact, ten or twelve days later my patient died produce sound. This sound is helped also in a , and the necropsy revealed, as the only because the column of air subdivides itself in the thick tuberculous lesion, the cavity previously diagnosed. In this and narrow tubes. The normal exhaling, however, is case, as in eight or ten others, the man was free from silent. Now the of a person at rest, who of respiration hereditary tuberculosis, not having in any other part his breathes with the mouth open (as in the position of body a tuberculous lesion that might have produced anti- tuberculins. auscultation) and heard near the mouth, presents a phenomenon quite opposite : the exhaling is more noisy Tuberculous meningitis also belongs to the not than the inhaling. This exhaling murmur is produced uncommon form of acute tuberculosis generally in the larynx and pharynx, these organs being the first occurring in the sick—more often in women—while obstacles that the outcoming column of air finds in convalescing from great hyperpyrexia, principally from which it can produce a musical sound. We see, then, typhoid fever. It is not unusual, observing the cases that almost all exhaling reenforcement that is found in attentively, to find that the bed next to the meningitis auscultation is originated in the larynx, the trachea, and patient is occupied by a tuberculous patient, although the thick bronchial tubes. The ear that auscults hears on examining the meningitis patient no tuberculosis has them, thanks to the fact that they pass through pul¬ Here new been found. is ground, not previously monary tissues more dense which allow sounds to pass immune, in which, as in the case of a child, the malady better. takes its most virulent form. Laënnec showed that the first period of tuberculosis Miliary tuberculosis adopts forms similar to those is characterized by a peribronchitis ; that is to say, an already mentioned. I was able to observe its fre¬ infiltration of the interstitial peribronchial tissues. This quency after severe hemorrhages in patients not previ¬ sign is a physical phenomenon. The interstitial peri- ously subject to tuberculosis. bronchial inflammation increases the density of the There are, however, numerous cases of chronic pul¬ lung, and owing to the easy transmission of sound by monary tuberculosis. A very curious phenomenon is dense tissues, the and tracheobronchial noises the clear in tuberculosis contracted larynx diversity pulmonary are transmitted in a better way to the ear that auscults, in the and that contracted in sea level highlands through a tissue of greater density. This explains the regions. Even if there is no morphologic diversity respiratory inversion, that is, the reenforcement of the between the bacilli of one and the other, the clinical evolution of the disease is different. expiration. very When the is not Native and men afflicted peribrouchial injury sufficiently foreign with tuberculosis extensive; when the interstitial tissues of the have who come to the react to the lung highlands marvelously not increased to conduct with clearness to climate and well-directed treatment. been yet enough Having the ear that auscultates the tracheobronchial noise, it during several years physician for a corporation that is sufficient to make the repeat in a low voice imported into the country numerous tuberculous patient in each a word rich in consonants. In an patients, and which took great care in their treatment expiration instant the ear will the difference that of them, I had the advantage of seeing some remark¬ perceive great in the case of between the able cures. Only the tuberculous patients with cavities exists, incipient tuberculosis, and the affected side. With a who are so unfortunate (as many are) as to be subject healthy to arterial hypertension die rapidly with hemoptysis. carefully applied to the ear, isolating outside noises, On the other hand, persons acquiring tuberculosis in a very muffled murmur is perceived, at most, from the the highlands take a slow and sluggish reaction to all healthy side, while from the affected side each syllable treatment. Being acclimated, these patients do not is distinctly heard, with such force that it reaches some¬ benefit from the advantages of the climate. times the degree of whispered . I have followed the progress of 169 tuberculous THE A NEW SIGN FOR EARLY DIAGNOSIS OF patients from the beginning of the malady; in many INCIPIENT PULMONARY TUBERCULOSIS of them before the appearance of the slightest symp¬ The new sign which I describe has been found in tom. Always, sooner or later, this sign appeared 169 cases of incipient pulmonary tuberculosis. It con¬ before the respiratory reenforcement and other signs sists of the reenforcement of the whispered voice in the that, associated, did not leave doubt for the diagnosis. vertex of the affected lung. This sign appears before Naturally, in cases in which there already exists the disease reveals itself by any other symptom, respiratory inversion, the sign of the reenforcement of either auscultatory or percussional, before respiratory the whispered voice is much more accentuated, getting inversion and impairment in the supraspinal triangle of to be, sometimes, almost pectoriloquy. This sign the affected side, which the classics consider as first should not be confused with whispered pectoriloquy, signs of the malady, appear. A stethoscope and a which may accompany tuberculous pneumonia of the little practice are enough to find it. The mechanism apex of the lung, which has an acute progress, with : it be that of the sign is very simple might said it is the blowing and other signs. The cavities are accom¬ same as the respiratory inversion, and that, examining panied also by tympany to percussion. The Niess not the sign, respiratory noise, but the whispered voice, the consisting of whispered pectoriloquy in the space situ¬ physician resorts to a trick which allows him to confirm ated between the inside edge of the scapula and the the beginning of the tuberculous process, in a period columnar vertebrae, is found also in bronchial adenop- before the appearance of the principal respiratory athy. Its cause is the transmission of the tracheo¬ inversion. bronchial breath through the hypertrophie lymphatic

Downloaded From: http://jama.jamanetwork.com/ by a Florida International University Medical Library User on 05/28/2015 masses. It has nothing to do, therefore, with that greater precision in methods of examination. It has which I have described in this study. also been used by unscrupulous persons to discredit the As counter control of the sign of the reenforcement practice of medicine. Scientific medicine, however, is of the voice, I had 140 and too established to fear such attacks and can • whispered convalescing firmly any debilitated patients (with uncinariasis, anemia, chloro¬ only profit by the open mindednes's and critical self- sis, etc.) under observation, who did not present the examination of occasional reviews, such as that of sign. Cabot. The sign described appears at the beginning of the By fine examinations and strictest interpretations, the tuberculous lung process, when the roentgen rays do pathologic anatomist might demonstrate, perhaps, a not even give a distinct shadow and when there is not greater percentage of error on the part of clinical diag¬ yet any other sign either auscultatory or percussional. nosis than is indicated by Cabot, particularly errors of Moreover, one may easily become accustomed to find¬ omission, were he to include minor degrees of arterio¬ ing it. sclerosis, slight sclerosis of heart valves, cloudy swell¬ ing of parenchymatous viscera, agonal congestion and other similar lesions ; but fairness demands that he CLINICAL COMPARED DIAGNOSIS AS include only such apparent lesions as might reasonably WITH NECROPSY FINDINGS be expected to present definite in life. Such an attitude should be met in like spirit by IN SIX HUNDRED CASES the clinician, and studies such as the present one be accepted in the light of attempted friendly cooperation HOWARD T. KARSNER, M.D. rather than in that of the bitter and destructive criti¬ LEONARD ROTHSCHILD, M.D. cism all too frequently exhibited at the necropsy table. AND Our studies are of 600 cases equally divided between E. S. CRUMP, M.D. two large hospitals in Cleveland. Our series is too CLEVELAND small to permit of the same division as made by Cabot and still reach in some of the The in satisfactory percentages greatest strides the wonderful advances of less diseases. We have therefore decided to modern clinical medicine have been made virtue of frequent by classify by organs and systems rather than in and in in some by special investigations chemistry physics, lesions. There were, however, certain marked errors instances advances with which it has been impossible of diagnosis that any such classification. These for and to On defy pathologic anatomy histology keep pace. we have classified as gross errors, and we have not the other there are numerous well hand, established included any instances in which the clinical in from which the clin- diagnosis, conceptions pathologic anatomy even though in error, was of the same organ ical must the latter con- markedly diagnosis depart, naming the found diseased at necropsy. We have how¬ dition in a fashion of its included, descriptive pathologic ever, certain very obvious errors of omission. This Not methods of anatomy. infrequently, functional numbers of the 600 or 8 cent. have been in the clinics after group fifty cases, per diagnosis widely applied As for the hospitals, in one there was 6 per cent, of been verified at the table in a having necropsy only such errors, and in the other, 10 per cent. A few limited number of a fact too well known instances, only to illustrate the type of error included critical readers of recent medical literature. examples serving by in this group will be found in Table 1. It is not our intention to underestimate the value of on the other to clinical investigation, nor, hand, place TABLE 1.—ILLUSTRATIVE EXAMPLES OF THE TYPE OF too great a valuation on what may be learned from the "ERROR OF OMISSION" IN CLINICAL DIAGNOSIS postmortem study of patients, but to attempt to reem- the fact that in of the chemist's Clinical Diagnosis Pathologic Diagnosis phasize spite investiga- Lobar pneumonia.Cerebral tumor; no pneumonia tions in a case, the visualizing by roentgenograms of Acute endocarditis .Acute peritonitis from ruptured pus- tube; no endocarditis internal changes and the study of cardiac activity by Lobar pneumonia .Epithelioma of larynx; tuberculosis the and the as well as right upper lobe; no pneumonia polygraph electrocardiograph, Typhoid fever.Lobar pneumonia other methods, the vast of Carcinoma of stomach .Lobar pneumonia; no stomach lesion study by precise majority Cardiovascular disease .Carcinoma of stomach clinical diagnoses can be accepted or rejected only on Cerebrospinal syphilis .Acute pyelonephritis the basis of naked chemical and bac¬ Acute alcoholism.Fracture of skull eye, microscopic, Tuberculous meningitis .Bronchopneumonia examination of the after death. Such Aneurysm of descending aorta. .. . Atrophie cirrhosis; rupture esopbageal tériologie body vein control of diagnosis should serve as a stimulus rather Cirrhosis of liver .Lobar pneumonia; cloudy swelling of than as a deterrent to more exact the liver rendering study Carcinoma of stomach .Pulmonary tuberculosis; amyloid liver of the living patient, and it is with the hope of furnish¬ and spleen; no lesion of stomach ing such a stimulus that we present our case records. Richard Cabot's1 read article on widely "Diagnostic We have classified as "minor errors" those in which Pitfalls Identified a of Three Thousand During Study the affected organ was thus includ¬ out in manner the properly diagnosed, Autopsies" pointed very striking ing here the errors of commission. In addition, this factor of error in a list of lesions twenty-eight special group has been made to include the failure to observe and diseases. It occasioned resentment on the part of lesions of the organs. of these errors certain clinicians accustomed to the ex cathedra method special Many are not so serious as to inclusion in the of but in many others served to stimulate justify "gross diagnosis, errors ;" but in order to make the figures more accurate, From the Department of Pathology, Western Reserve University it is necessary to include with each organ cases which School of Medicine. also have been classified above as gross errors. We Read before the Section on Pathology and Physiology at the Seventieth Annual Session of the American Medical Association, have compared the number of correct diagnoses to the Atlantic City, N. J., June, 1919. total number of cases and have thus esti¬ 1. Cabot, R. C.: Diagnostic Pitfalls Identified During a Study of diagnosed, Three Thousand Autopsies, J. A. M. A. 59: 2295 (Dec. 28) 1912. mated the percentage of error of commission. It is to

Downloaded From: http://jama.jamanetwork.com/ by a Florida International University Medical Library User on 05/28/2015