THE TREATMENT OF DIPHTHERIA BY ANTITOXIN. 
By William H. Welch, M. D., 
Pathologist to the Johns Hopkins Hospital and Professor of Pathology, 
Johns Hopkins University. 
[From The Johns Hopkins Hospital Bulletin, Nos. 52-58, July-August, 1895.] [From The John* Hopkins Hospital Bulletin, Noa. 52-53, July-August, 1895.] 
THE TREATMENT OF DIPHTHERIA BY ANTITOXIN.* 
By William H. Welch, M. D., 
Pathologist to the Johns Hopkins Hospital and Professor of Pathology, 
Johns Hopkins University. 
I shall endeavor in this paper, after a brief historical intro- 
duction, to present some of the more important general con- 
siderations hearing upon the treatment of diphtheria by anti- 
toxic serum, together with statistics of results already reported, 
with the expectation that those who are to follow in this dis- 
cussion before the Association will be able to offer the results 
of personal experience in the application of the new remedy. 
In July, 1889, Babes and Lepp, in an article entitled 
“Becherches sur la Vaccination Antirabiques,”f published 
results of experiments undertaken to solve the question 
“ whether the fluids and cells of animals which have been 
rendered by vaccination immune have not become vaccines 
and capable of protecting also other organisms.” The results 
of these experiments showed that the blood of dogs thoroughly 
vaccinated against rabies, when injected into susceptible 
animals, conferred a certain amount of protection against the 
effects of subsequent inoculation with the rabid virus, and 
appeared capable of preventing the development of rabies 
| * This paper is based upon the address at the opening of the dis- 
cussion on this subject before the Association of American Physi- 
cians at the meeting held in Washington, D. C., May 31, 1895. I 
have endeavored to bring the paper up to the date of sending it to 
the printer (July, 1895). 
\ Babes and Lepp, Annales de l’Institut Pasteur, July, 1889. 
Richet and Hericourt are sometimes quoted as the first experi- 
menters to show that the blood of animals is capable of conferring 
protection upon susceptible animals, but their work has no refer- 
ence to modern serum therapy, as their experiments were made 
with the blood of dogs which had not previously been vaccinated 
or treated in any way. 2 
even when the injection of the immune blood was made imme- 
diately after the reception of the virus. The authors con- 
cluded that “one must admit the possibility of vaccinating 
with the fluids and cells of animals which have been rendered 
refractory to the disease.” 
The first publication clearly demonstrating the principles 
of serum therapy was made by Behring and Kitasato on 
December 4, 1890, in an article in the Deutsche medicinische 
Wochenschrift entitled “Ueber das Zustandekommen der 
Diphtherie - Immunitat und der Tetanus - Immunitat bei 
Thieren.” Although in this article the immunizing and 
curative property of the blood and blood serum of artificially 
immunized animals was demonstrated only for tetanus, the 
application of the same principle to diphtheria was indicated 
in the same article and in a second paper by Behring in the 
following number of the same journal. 
The first public announcement of the demonstration of the 
power of the blood serum of animals artificially immunized 
against diphtheria to protect and cure susceptible animals 
inoculated with the diphtheria bacillus or its poison, was made 
by Behring in the report of experiments made by himself and 
Wernicke, and communicated to the Seventh International 
Congress of Hygiene and Demography held in London in 
August, 1891. There followed in 1892 the article by Behring 
and Wernicke,* in which these experiments were fully 
described and which sets forth the fundamental principles 
underlying serum therapy of diphtheria. 
The first trial of immune serum in the treatment of human 
diphtheria was made in von Bergmann’s clinic in Berlin in 
the autumn of 1891. This trial, together with those made in 
1892 by Henoch in Berlin, by Heubner in Leipzig, and in the 
Institute for Infectious Diseases in Berlin, were of a tentative 
nature and made with weak serum and insufficient doses. 
It was not until early in 1893 that Behring succeeded in 
obtaining anti-diphtheric serum equaling the strength of 
even his so-called normal serum, of which sixty times the 
strength is that of the weakest Behring’s serum at present in 
* Behring und Wernicke : Ueber Immunisirung und Heilung von 
Versuchsthieren bei der Diphtherie. Zeitschrift fur Hygiene, 
Bd. XII. 3 
use. In April, 1893, Behring referred to 30 cases treated with 
this normal serum. Of these cases, 11 treated in the Institute 
for Infectious Diseases were reported in detail by Kossel. 
From this period on Behring and Ehrlich succeeded in 
obtaining healing serum of greater and greater strength, until 
in August, 1893, Ehrlich and Wassermann obtained from 
goats healing serum twenty to sixty times the strength of 
Behring’s normal serum. At the Eleventh International 
Medical Congress held in Rome (March 29 to April 5, 1894) 
Heubner reported the results of his experience with the serum 
treatment of human diphtheria. His observations, however, 
were made on cases treated with much weaker antitoxin than 
is now recognized as suitable. 
In April, 1894, Ehrlich, Kossel and Wassermann reported 
briefly the results of serum treatment of 220 cases in six Berlin 
hospitals, the inception of the treatment in these cases dating 
from June, 1893, but the great majority of the cases occurring 
after December, 1893. These cases, with additional ones mak- 
ing a total of 233 cases with a mortality of 23 per cent., were 
reported more fully in an article by Kossel in the Zeitschrift 
fur Hygiene in July, 1894. The era of serum treatment of 
human diphtheria by approximately sufficient doses of anti- 
toxin really begins with this publication of Ehrlich, Kossel 
and Wassermann in April, 1894, although even in this series 
of cases, according to later statements of Ehrlich and Kossel, 
a large number of the cases were treated with quantities of 
antitoxin which we now consider to be insufficient. 
In an address before the Berlin Medical Society on June 27, 
1894, Katz reported the results of antitoxin treatment dating 
from March 14, 1894, with Aronson’s serum from horses on 
128 cases of diphtheria in Baginsky’s service. In the dis- 
cussion on this address four weeks later Baginsky completed 
the series of cases up to 163 with a mortality of 12.9 per cent., 
and Aronson stated that similarly favorable results had been 
obtained by Ganghofner in Prague and Escherich in Graz. 
In 1893 and the first half of 1894 various articles appeared 
concerning the preparation of antitoxin, the best methods of 
estimating its strength, the proper immunizing and thera- 
peutic doses, and similar questions. Since August 1, 1894, 
Behring’s serum prepared at Hochst has been for sale. 4 
It is evident from this brief historical summary that the 
general principles of serum therapy of diphtheria were fully 
established and its application to human beings in active 
operation before Roux delivered his memorable address on the 
subject at the Eighth International Congress of Hygiene and 
Demography held in Budapest, September 1-9, 1894, three 
years after Behring’s original communication to the preceding 
Congress in London. Roux, however, presented the subject 
with such clearness and force, and with such an array of con- 
vincing and carefully analyzed statistical evidence, that the 
attention of the great body of physicians throughout the 
world, who had paid little heed to the previous work, was 
arrested, and the question of the healing power of diphtheria 
antitoxin became and has continued to be the foremost medical 
question of the day. From September, 1894, onward the supply 
of antitoxin from various sources (not all of equally trust- 
worthy character) has become more and more accessible to 
physicians, and each succeeding month has given birth to a 
large number of articles on the serum therapy of diphtheria 
from various parts of the world. 
Unless one denies absolutely the causal relation of the 
Loffler bacillus to diphtheria, it must be admitted that the 
treatment of this disease by antitoxin rests upon a sound 
experimental basis. The only notable opponent of the view 
that the Ldffler bacillus is the cause of diphtheria is Hanse- 
man. His arguments, which have been well answered by C. 
Fraenkel, are equally applicable to the acceptance of the 
etiological relations of the cholera bacillus, the tubercle 
bacillus and many other specific bacteria of infectious dis- 
eases. It is not probable that any one here sides with Hanse- 
man in this matter, so that it is unnecessary to rehearse the 
arguments, which in my judgment are conclusive, that the 
Loffler bacillus is the cause of diphtheria. 
The laboratory does not furnish any more impressive ex- 
periments than those which demonstrate the power of anti- 
toxic serum to prevent and to cure the disease caused in 
animals by inoculation with the diphtheria bacillus or its 
poison. The serum arrests the spread of the local process 
and abates the symptoms of general toxaemia. These experi- 
ments prove beyond question that this healing serum pos- 5 
sesses properties which are directly and powerfully antag- 
onistic to the toxic action of the diphtheria bacillus, and 
there is no good reason to doubt that under similar circum- 
stances this antagonistic power, so readily and surely and 
uniformly demonstrable in the case of lower animals, will 
manifest itself also in human beings. The only question, and 
that of course an important one, in this connection is: To 
what extent the conditions in the treatment of experimental 
diphtheria by antitoxin are or can be made similar to those 
in the therapeutic application of the same agent to human 
diphtheria ? 
Although it is true that the lower animals are not sus- 
ceptible, or only very exceptionally susceptible, to natural 
infection with the Loffler bacillus, still there is in my opinion 
identity in essential points, anatomical, clinical and etiological, 
between experimental diphtheria and uncomplicated human 
diphtheria. The assertion sometimes made that spreading 
pseudo-membranous inflammations resembling those of diph- 
theria cannot be produced experimentally in animals by inocu- 
lation with the Loffler bacillus is an error, as I have repeatedly 
had opportunity to demonstrate by intratracheal inoculations 
of kittens and rabbits. It is rarely in our power to reproduce 
experimentally in one species of animal the exact counterpart 
of a disease caused in another by natural modes of infection, 
but in the case of diphtheria the resemblance is closer than 
in most of our attempts to reproduce such diseases by inoc- 
ulation of their specific germs. 
But even if the anatomical and clinical characters of ex- 
perimental diphtheria are believed by some to differ more 
widely than I think they do from those of human diphtheria, 
there remains as the most important point, regarding the 
matter here under discussion, the conclusive demonstration 
that in uncomplicated human diphtheria no less than in 
experimental diphtheria the local inflammation at the site of 
infection is caused by the growth of the Loffler bacillus, and 
the lesions of internal parts and the systemic symptoms are 
due to the absorption of a toxic substance or of substances 
formed by this bacillus. It would be difficult to understand 
why an agent with the specific property of neutralizing in the 
bodies of animals the effects of these toxic substances should 6 
be unable to neutralize in human beings similar effects of the 
same toxic substances, provided this agent can be administered 
in the proper dose and at the right time. 
Dosage and timely administration are factors of prime im- 
portance in determining the efficacy of antitoxic treatment. 
It is our inability to conform to the demands of these factors 
which has rendered thus far the treatment of tetanus in 
human beings by antitoxin disappointing. The tetanus anti- 
toxin can be produced by methods similar to those employed 
in making the diphtheria antitoxin, and of a power expressed 
in immunizing units greater than that of the diphtheria 
antitoxin. No less striking than in diphtheria are the 
laboratory experiments in the prevention and cure of arti- 
ficial tetanus by administration of the tetanus antitoxin, but 
in this case the dose of antitoxin required to check the dis- 
ease increases so enormously with increase in the size of the 
animal, on the one hand, and with the lapse of time after 
reception of the virus on the other hand, that we meet herein 
most serious obstacles to the successful application of this 
agent in the treatment of human tetanus. 
It has been shown experimentally by Behring, Boer, Roux 
and others that as regards both of these points the conditions 
are far more favorable for the treatment of diphtheria by its 
antitoxin than in the case of tetanus. In an animal, at a 
certain time after reception of the tetanus poison, the theor- 
etically efficacious dose of the tetanus antitoxin may be a 
million-fold greater than that required for simple immuniza- 
tion, a quantity too large to administer; whereas the effective 
dose of the diphtheria antitoxin at relatively the same period 
may be increased only eight or ten-fold. Doubtless the great 
advantage which we have in the treatment of human diph- 
theria by antitoxin as contrasted with tetanus is that we are 
able to recognize and treat the former disease before the pro- 
duction and absorption of a serious quantity of poison. 
Only clinical experience can determine what practical 
difficulties there may be in the way of the successful em- 
ployment of antitoxic serum in the treatment of human 
diphtheria, but there is no doubt in my mind that the results 
derived from experiments on animals justify, nay, demand, 
the most careful and thorough trial of the new method of 
treatment upon human beings. We have no certain knowledge as to the nature of the sub- 
stances called antitoxins nor as to their mode of action. This 
is not, however, an argument against their therapeutic employ- 
ment, for we have no positive knowledge as to the mode of 
action of many of our therapeutic agents. There are two 
prominent theories as to the mode of action of the diphtheria 
antitoxin. The one may be called the chemical and the other 
the vital theory. The chemical theory is that the antitoxin 
directly neutralizes in a chemical sense the toxins. This 
seemed to be the natural interpretation of the fact that the 
injection into susceptible animals of a mixture in suitable 
proportion of the antitoxin and the toxin is harmless, but 
Buchner and Roux have shown that this earlier view is incor- 
rect, and that by selecting animals of greater susceptibility or 
by increasing the natural susceptibility of an animal, the 
presence of active toxin in the mixture can still be demons- 
trated. The experimental evidence, therefore, is in favor of 
the other theory, viz. that the antitoxin acts through the 
agency of the living body, and probably in the sense that it 
renders the cells tolerant of the toxin. 
The results of the treatment of human diphtheria with 
antitoxin speak also in favor of this vital theory. 
If, as seems probable, the curative effects of the healing 
serum are brought about through the agency of the living 
cells of the body, we can understand why these effects will not 
follow the introduction of the serum with the certainty and 
precision of a chemical reaction. The cells must be in a con- 
dition to respond in the proper way to the introduction of the 
antitoxic serum. For one reason or another “this responsive 
power may be in abeyance. It may be weakened by intense 
or prolonged action of the diphtheria poisons, or by other pre- 
vious or coexistent disease, or by inherent weakness, or there 
may even be some individual idiosyncrasy which hinders the 
customary response of the cells to the antitoxin. Clinical 
experience shows that cases of diphtheria inherently refrac- 
tory to timely treatment with antitoxic serum are most excep- 
tional, if indeed they occur at all. 
There is some evidence in favor of the view that while anti- 
toxin may exert its protective action upon certain groups of 
cells, other cells, as for example the nerve cells, may, either 8 
by their nature or on account of such influences as I have 
mentioned, not he equally protected against the toxin. There 
is also the possibility that antitoxin may neutralize the effects 
of certain toxins and not of others present in diphtheria. 
Antitoxic serum exerts no bactericidal effect upon the diph- 
theria bacillus, although, when administered in proper quan- 
tity, sufficiently early in the disease, it arrests the spread of 
the local inflammation which is caused by the bacillus. Viru- 
lent bacilli may persist in the throat days and even weeks 
after recovery following injection of antitoxin. 
One of the most important characters of antitoxin is that it 
requires a definite quantity of this substance to neutralize the 
effects of a definite quantity of toxin. In animals the cura- 
tive dose of antitoxin stands in a definite quantitative relation 
to the size and susceptibility of the individual and to the 
amount and intensity of the poison in the system. We have 
no precise method of determining how much and how virulent 
the poison may be in a given case of human diphtheria nor 
how susceptible to the toxin the patient may be. The dosage 
of antitoxin, therefore, in human diphtheria is empirical, the 
main factors determining it being the age of the patient, the 
assumed duration of the disease up to the time of administer- 
ing the remedy, and the apparent severity of the disease. As 
the healing serum is expensive and is capable of inducing 
unpleasant symptoms, it is desirable not to give an excessive 
quantity. Under these circumstances it may readily happen 
that an insufficient dose is given and that the administration 
must be repeated. The general rules regarding the dosage of 
antitoxin are sufficiently well known not to require mention 
here, and I speak of this matter only to indicate that because 
a patient may have received a dose or even two or more doses 
of antitoxin, this furnishes no absolute guarantee that a quan- 
tity of antitoxin adequate to neutralize the effects of the toxin 
has been given. We now know that in the early period 
following introduction of the treatment entirely insufficient 
doses were given. 
Both experiments on animals and clinical experience de- 
monstrate that the earlier antitoxic serum is administered 
after the inception of the disease, the better are the chances 
of recovery. It is usually impossible to rescue the lives of 9 
guinea-pigs by means of antitoxin if the treatment is de- 
layed longer than forty-eight hours after inoculation with an 
amount of diphtheria poison fatal to these animals in four or 
five days, although the duration of life may he considerably 
prolonged. In human beings the conditions are different, 
but, as will appear from the statistics to be presented, the 
evidence is conclusive that the superiority of serum treatment 
over all other methods is most strikingly manifested in the 
results of the cases in which the antitoxin is given not later 
than the third day of the disease. Although in many cases 
the treatment is beneficial when the antitoxin is administered 
in larger doses at a later period of the disease, the importance 
of beginning the treatment at the earliest possible date, with- 
out waiting to determine by cultures whether or not the 
Loffler bacillus is present, cannot be too strongly emphasized. 
It is of course often impossible to meet this demand for 
early treatment, as cases of diphtheria are frequently not seen 
or recognized by the physician, particularly in hospital prac- 
tice, until after several days’ duration of the disease and when 
grave symptoms have already developed. It is, moreover, in 
many cases difficult or impossible to determine how long the 
disease has existed when it is first seen by the physician. 
The fact that the benefits of antitoxin treatment become 
more and more doubtful the further the disease has pro- 
gressed and the graver the lesions and symptoms, renders 
more difficult the collection and analysis of absolutely con- 
vincing statistics in favor of the treatment. The accusation 
is sure to be brought that many of the cases which have 
responded promptly to early treatment, and these for reasons 
which have been stated will form a large contingent of the 
successful cases, were mild cases which would have recovered 
equally by other methods of treatment. This objection can 
be fully met only by large series of statistics collected from 
many epidemics, at different times and in various localities. 
The bacteriological study of human diphtheria has dis- 
closed several points important to bear in mind in determining 
the value of antitoxic treatment. The Loffler bacillus has been 
found in healthy throats, although only very exceptionally 
unless the person has been exposed to diphtheria. This same 
bacillus may cause all grades of inflammation of the throat, 10 
from a mild erythematous angina to the gravest pseudo-mem- 
branous inflammations. There has resulted a conflict, not yet 
settled, between the clinical and the bacteriological diagnosis 
of diphtheria. As regards these diversities of effect, however, 
the conditions pertaining to the diphtheria bacillus are in no 
way different from those relating to many other pathogenic 
bacteria, as for example the pneumococcus, the streptococcus, 
the cholera bacillus, and even the tubercle bacillus, all of 
which may be found on healthy mucous membranes and may 
exert their pathogenic activity with all degrees of intensity. 
Inconvenient as these facts may be, they must be recognized, 
and they require a readjustment of previously adopted boun- 
dary lines of diagnosis. It would, of course, be absurd to 
say that a person who harbors in his healthy throat Loffler 
bacilli has diphtheria, just as it would be equally ridiculous 
to consider a person infected with the pneumococcus or the 
streptococcus when these latter bacteria are present under 
similar conditions, but it is no less absurd to limit the appli- 
cation of the term “ diphtheria ” only to those higher degrees 
of pathogenic action of the Loffler bacillus characterized by 
spreading pseudo-membranous inflammations and general 
toxsemia. 
But while the boundaries of the domain of diphtheria have 
thus been widened by the inclusion of cases not presenting 
the ordinary clinical characteristics of diphtheria, in another 
direction they have been restricted by the exclusion of some 
cases which on clinical grounds would be diagnosed as diph- 
theria but which by bacteriological examination are proven 
to be caused by other bacteria than the Loffler bacillus. 
The statement is sometimes made that twenty-five to 
thirty per cent., or even a larger percentage of the clinical 
diphtherias are not genuine diphtheria in the bacteriological 
sense, but this statement is quite misleading. These figures 
are based upon the bacteriological examination of large num- 
bers of cases in which there was simply more or less suspicion 
of diphtheria. They do not relate generally to a large number 
of cases presenting unmistakable anatomical and clinical 
characteristics of diphtheria. They are derived from the 
routine examinations for Boards of Health and children’s 
hospitals of suspected cases of diphtheria. When one con- 11 
siders that in some cases of diphtheria repeated, painstaking 
examination, microscopical and cultural, by a skilled bacteri- 
ologist, is required for the detection of the diphtheria bacillus, 
it is evident that less reliance is to be placed upon these sta- 
tistics en gros than upon many smaller series reported by 
bacteriological experts. Of the statistics of the latter charac- 
ter there are many which show that in the series of cases 
examined (including in each series from a dozen to over three 
hundred cases) from ninety to one hundred per cent, of the 
clinical diphtherias are due to the Loffier bacillus. Our expe- 
rience in Baltimore has been that not over five per cent, of 
the cases which the clinician would confidently diagnose as 
diphtheria are false diphtheria or diphtheroid. These latter 
figures relate, of course, to primary diphtheria and not to the 
pseudo-membranous anginas complicating scarlet fever and 
other infectious diseases, a large proportion of which are not 
referable to the Loffier bacillus. 
I shall consider subsequently in this article the influence 
which the control of the clinical diagnosis of diphtheria by 
bacteriological examination is likely to have upon fatality 
statistics of this disease. 
There is an important difference between experimental 
diphtheria and many cases of human diphtheria, a difference 
of great significance in determining the scope of efficiency of 
treatment by antitoxic serum. Our experimental diphtheria 
is a pure, uncomplicated infection in which only the diph- 
theria bacillus and its toxins are concerned. On the other 
hand, in many cases of human diphtheria there are complica- 
tions and mixed infections due to other micro-organisms 
against which, when fully developed, the diphtheria antitoxin 
is powerless. The most common and dangerous complicating 
micro-organism is the streptococcus pyogenes. Bacteriologi- 
cal examinations of fatal cases of diphtheria demonstrate in a 
large proportion of cases the invasion and pathogenic effects 
of this most common of all secondary invaders. The confi- 
dence with which some observers, particularly of the French 
school, classify their cases of diphtheria into pure and mixed 
infections, on the sole basis of the bacteriological examination 
of the exudate in the throat, does not seem to me justifiable. 
The complete microscopical and cultural examination of this 12 
exudate will in practically all cases reveal the presence of other 
bacteria, and usually of streptococci, besides the Loffler ba- 
cillus. But as these other bacteria are common or regular 
inhabitants of the healthy throat, their mere presence in this 
situation is not conclusive evidence that they are engaged in 
pathogenic action. The abundance of these other bacteria 
may afford some indication as to their role, but of greater 
importance is their demonstration in situations where they 
are not normally present. 
Reiche,* in 42 autopsies on cases of diphtheria in which 
the Loffler bacillus had been demonstrated during life, made 
cultures from the kidney and spleen. In 64.3 per cent, of 
these cases streptococci and staphylococci were found in the 
kidney or spleen, and in 45.2 per cent, streptococci were found 
alone. These cocci must have reached these organs through 
the circulating blood. He found streptococci in the kidney 
in one case which died on the second day of the disease, and 
positive results were obtained also on the third and fourth 
day. These results are evidently of much significance in 
indicating the frequency and the earliness of invasion of 
complicating micro-organisms in diphtheria and the resulting 
obstacles to uniformly favorable results of antitoxin treat- 
ment. 
But the chief evidence in favor of mixed infection must be 
sought during life, in the character of the lesions and symp- 
toms, although these may be misleading. There is also 
evidence that the failure of a case of diphtheria to respond in 
the usual way to the timely injection of a sufficient dose of 
antitoxic serum is an indication of complications and mixed 
infection. 
The opinion is entertained by Roux, Martin and other 
French writers that broncho-pneumonia, one of the most 
common and serious complications of diphtheria, is due to a 
large extent to local unhygienic conditions which can be 
guarded against. Thus they attribute the frequent occur- 
rence of broncho-pneumonia in some groups of their cases to 
the infection of the hospital wards with the bacteria causing 
pneumonia, and claim that by improved sanitary conditions 
*Reiche : Centralblatt f. innere Medicin, 1895, No. 3. 13 
this complication may be to a large extent eliminated. Fur- 
ther investigations are needed to determine to what extent 
this view as to the causation of broncho-pneumonia is justi- 
fied, but it can scarcely be doubted that this complication is 
often the result of invasion of the lower air passages and the 
lungs by bacteria which are regularly present in the throat, 
and whose activity is likely to be manifested in this way in 
many cases of diphtheria, independently of the local sanitary 
conditions. 
Without doubt the remedial role of diphtheria antitoxin is 
materially restricted by its inability to combat developed 
streptococcus sepsis, broncho-pneumonia and other complica- 
tions referable to secondary infection, or to stop impending 
suffocation by immediate removal of mechanical obstacles in 
the form of false membranes in the air passages, but the anti- 
toxic serum is the most powerful agent which we possess to 
prevent the development of these complications and secondary 
infections. The timely administration of the healing serum, 
by antagonizing the effects of the Loffler bacillus, antagonizes 
in large part the causes of the increased susceptibility to secon- 
dary infections and thus greatly lessens the frequency of their 
occurrence. 
In considering the obstacles in the way of cure of diph- 
theria by antitoxin, the self-evident fact should not be for- 
gotten that this remedy cannot restore cell life which has 
already been seriously damaged by the action of the diphtheria 
bacillus or its poison. The researches of Oertel upon human 
diphtheria, and those of Flexner and myself upon experi- 
mental diphtheria, demonstrate that the toxins of the diph- 
theria bacillus are most powerful poisoners of cells, the 
internal lesions of pure diphtheria being especially character- 
ized by widely distributed areas of cell death. We have no 
way of gauging accurately at any given period of the disease 
the extent of the damage already inflicted upon the cells of 
the body. If the nerve cells or their axis cylinders have 
already been so damaged that paralysis must follow, or the 
cardiac nerve cells or muscular fibres have been similarly 
injured, or the renal epithelium so affected that degeneration 
and nephritis ensue, the administration of antitoxin cannot 
restore these cells which are already on the way to degenera- 
tion and death. 14 
This irretrievable damage to cell life may be present for a 
considerable time before we are able to recognize its effects. 
P. Meyer detected pathological changes in the peripheral 
nerves as early as the third day after the onset of diphtheria 
and before paralysis was manifest. The occurrence of pa- 
ralyses, including cardiac paralysis, after antitoxin has been 
administered even thus early in the disease, cannot therefore 
necessarily be attributed to failure of this agent to neutralize 
toxin developed after its injection. 
Having now considered the experimental basis and the 
theories of action of antitoxic treatment, the importance of 
early administration and sufficient dosage, and certain etio- 
logical and pathological characters of human diphtheria to 
be borne in mind in estimating the scope of the treatment, 
let us turn to the examination of the evidence which has 
hitherto been published concerning the efficacy of the anti- 
toxic treatment of human diphtheria. This evidence is of 
two kinds, first, the general impressions of clinicians who 
have had opportunity to observe the effects of antitoxin ad- 
ministered in a number of cases of diphtheria, and second, 
the fatality statistics of cases treated with antitoxin. 
Unquestionably great value attaches to the impressions and 
conclusions of careful clinical observers as to the merits of 
therapeutic agents. Baginsky has said that naked figures are 
so little the expression of the endless variations of clinical 
observation, of all those fortunate and unfortunate accidental 
circumstances which pertain to the constitution and nutrition 
of the patient, and of the complications and difficulties which 
may bring danger in a mild attack, or lead to a successful 
issue an apparently severe attack, that to the clinical observer 
such figures appear of little value in comparison with the 
treasure-house of his accumulated experience. And it is to his 
experience of many years in the same hospital and on similar 
clinical material that Baginsky repeatedly recurs in his mono- 
graph “ Die Serum-therapie der Diphtheric,” in support of 
his favorable conclusions as to the healing power of antitoxin, 
and this in spite of the fact that his statistical results, lead- 
ing to the same conclusions, are based upon a larger number 
of cases than those of any other single observer yet published, 
and are among the most convincing of the statistical reports. 15 
In explaining why at the end of ten months’ trial of anti- 
toxin he has determined to commit himself to a definite judg- 
ment in its favor, he says: “ The reasons for this are to he 
found in the continual repetition of improvement and recovery 
of severe cases which previous experience indicates would 
have terminated fatally; and furthermore, in the outcome of 
an involuntary experiment with interruption of the use of the 
serum for a period on account of failure in its supply. During 
this period the mortality of our patients immediately rose 
again to its former height.” “ The improvement in the gen- 
eral condition of the patients imparts to our diphtheria wards 
an entirely different character from the former one. That 
this is not due to any change in the character of the clinical 
material, to milder forms of the disease, was unfortunately 
demonstrated by the observations in the months of August 
and September, when, as by a single blow, we were trans- 
ported back to the old times, to the same melancholy picture 
of children deeply prostrated and often in vain struggle with 
death.” In August and September the supply of antitoxin 
failed. 
The published testimony of those who have had the largest 
opportunity to study the therapeutic effects of antitoxin is 
overwhelmingly in its favor. In no less favorable terms than 
those of Baginsky are expressed the opinions of such observers 
of high reputation and extended experience as Heubner, von 
Widerhofer, von Banke, Ganghofner, Escherich, Bokai and 
the physicians of the Hopital des Enfants Malades and 
Hopital Trousseau in Paris. These observers have reported 
already in detail over 2300 cases of diphtheria treated with 
antitoxin. 
Many of those who have reported smaller series of cases, and 
a few who have reported as many as a hundred cases, have 
expressed themselves with much caution or have not ventured 
any final judgment, although in most of these reports the 
results appeared to be favorable to the new treatment. An 
example of this conservative position is that of Yierordt, who 
says that a final decision as to the value of antitoxic serum is 
not to be expected in the immediate future, as such decision 
requires a long series of observations in different epidemics 
and on varied clinical material. 16 
Antitoxic serum is a new and strange remedy, but the 
effects which follow its injection in individual cases are not 
new and strange. Nothing happens which the physician 
may not have occasionally seen to happen in cases treated in 
the ordinary way. In severe as well as in mild cases of 
diphtheria he may have seen an apparently progressive local 
process quickly arrested and the general symptoms promptly 
abated. But why should anything new and strange happen 
after the administration of antitoxin ? Cure by antitoxin is 
cure by nature’s own remedial agent. That which is new and 
strange is the frequency with which in case after case the 
timely injection of antitoxin promptly arrests the local in- 
flammation and checks the constitutional disturbance. 
Recovery following treatment by antitoxin is such a natural 
kind of recovery that in any given case the physician may 
readily have the feeling that the same thing might have hap- 
pened without the use of the remedy. We can, therefore, 
understand why it should be those with the largest experience 
in the treatment of diphtheria by antitoxin who are most 
decided in expressing their opinion as to its beneficial effects. 
The very fact that the mode of cure is such a natural one 
and unattended by peculiar phenomena is an obstacle to draw- 
ing positive conclusions from a small number of observations, 
even if these appear to be most favorable. 
That there should be wide diversity in the percentage of 
cures in reports of different observers is of course to be 
expected when we consider the varied character of the cases 
treated and the importance of early administration of anti- 
toxin. It may happen that a series of cases is made up so 
largely of advanced and complicated diphtherias at the time 
when the antitoxic treatment is begun, that the beneficial 
effects of the treatment are not apparent. It is on the whole 
remarkable that there should have been so few reports in 
which the fatality has not been materially diminished during 
the period of administration of antitoxin. 
There are only very few writers who on the basis of per- 
sonal experience (and this in no instance a large one) have 
expressed an opinion unfavorable to antitoxin. Kohts may 
be mentioned as one who on the basis of 47 cases treated with 
serum, with 29.1 per cent, deaths among the tracheotomized 17 
and 7.6 per cent, among the non-tracheotomized, finds such 
apparently favorable results no better than by other methods 
of treatment. 
So far then as the testimony of physicians based upon their 
clinical experience is concerned, this, as I have already said, is 
overwhelmingly in favor of the antitoxic treatment, wherever 
their experience in its employment has been a large one. 
Those with less experience are often even more enthusiastic, 
hut many of these, in view of their limited experience, are 
wisely conservative and a few are hostile to the new treatment. 
But general clinical impressions, convincing as they may he 
to the individual receiving them, may not be equally con- 
vincing to others. They do not furnish any strict scientific 
proof of the value of a therapeutic agent. If antitoxin really 
exerts any specific curative action in diphtheria this must be 
apparent in the figures of fatality statistics of this disease, 
and it is only by such statistics, much as they may he decried 
by some, and difficult as it may be to guard them from errors 
of interpretation, that a strictly scientific demonstration of 
the efficacy of antitoxin in the treatment of diphtheria can be 
brought. 
The possible fallacies of interpretation belonging to fatality 
statistics in general apply in no small measure to those of 
diphtheria. The case mortality from diphtheria varies within 
wide limits according to the more or less severe character of 
the prevailing epidemic, according to the season of the year, 
according to the age, according to the method of treatment, in 
cities and in country districts, etc. Statistics of case mor- 
tality from hospital practice will differ widely from those 
from private practice, and each of these will differ from the 
general case mortality returns from cities. Nor does each of 
these three classes of statistics represent a uniform material. 
The material of one hospital may consist very largely of cases 
of diphtheria admitted in an advanced stage of the disease, or 
of laryngeal cases sent for operation, while that of another 
hospital may contain a much larger proportion of cases 
admitted in early stages of the disease. In general the fatality 
of diphtheria in hospital practice is higher than that of private 
practice, as would be expected from the later stage of the dis- 
ease in which the patients generally enter the hospital; hut to 18 
this rule there are many exceptions. In some hospitals the 
patients are all children, in others there may be a considerable 
proportion of adults with diphtheria. In private practice 
among the poor, patients may be first seen by the physician 
frequently in as advanced stages of the disease as in hospitals, 
and the conditions for successful treatment, and particularly 
for intubation or tracheotomy, are less favorable for this class 
of private patients than for hospital patients. 
Still other reasons might be given for the lack of uniformity 
of diphtheria statistics from different sources, but enough has 
been said to show that as regards the question which interests 
us here, each report of a series of cases treated with antitoxin 
requires its own special consideration and analysis and is not 
comparable with reports from other sources relating to a 
different class of cases. 
The larger the number of cases embraced in the statistical 
tables, the greater becomes the mutual compensation of such 
differences as those mentioned, and therefore the more trust- 
worthy are the conclusions derived from the statistics; but in 
collecting the statistics of the general fatality of diphtheria 
treated with antitoxin it has seemed to me important, for the 
reasons which have been mentioned, that the tables should 
contain for each report, as far as possible, statements of the 
total number of cases treated with antitoxin, of the num- 
ber and percentage of deaths, of the previous or simulta- 
neous fatality, and of the class of cases, whether in hos- 
pital or in private practice. I have also analyzed the cases 
so far as practicable according to the ages of the patients and 
according to the day of the disease on which antitoxin treat- 
ment was begun. It has seemed to me of especial interest to 
consider the fatality in operated and not operated cases. 
There are of course many other points of view which it would 
have been interesting to consider in the statistical study of 
the cases reported, but it has seemed to me that the analysis 
already indicated should suffice to determine the main ques- 
tion at issue, namely, the specific curative power of antitoxin, 
as well as certain other questions. 
It is scarcely ten months since antitoxin has been used by 
more than a very few favored physicians, and it is a much 
shorter time since its use has become at all general. In this 19 
comparatively short time there have, however, been published 
more or less definite reports of the results of the treatment in 
at least 15,000 cases. I have collected 82 reports from 80 dif- 
ferent sources containing 7166 cases. These are presented in 
Table I. This collection of cases is by no means complete, as I 
have consulted only the more readily accessible journals, but 
it is believed to include all of the more important reports. I 
have not included any reports of single cases, as these are 
often to illustrate some special point, nor any reports of series 
of cases less than ten. Indeed only four reports with less 
than 12 cases in the group have been included in the tables. 
Nor have I made use of such merely general published state- 
ments without detail as that there have been treated in France 
up to the end of December 2700 cases with a mortality of 16 
per cent., in Austria outside of Vienna 950 cases with a mor- 
tality of 15.7 per cent., in Croatia and Slavonia 428 cases with 
a mortality of 10.8 per cent., in Berlin hospitals 1500 cases 
with a reduction in fatality of one-half, etc. 
Eulenberg has recently (July 15, 1895) made a provisional 
report concerning the collective investigation inaugurated by 
the Deutsche medicinische Wochenschrift by sending out 
cards to be filled out by physicians regarding their results in 
the treatment of diphtheria with and without serum. Up to 
the date of the report the cards returned embraced 10,240 
cases of diphtheria, of which 5790 were treated with serum 
and 4450 without serum. The total fatality of the former 
group was 9.5 per cent., that of the latter group 14.7 percent. 
No further details of this investigation have as yet been pub- 
lished, and these cases are not included in my tables. 
I have entirely avoided the duplication of cases, so far as 
I can determine.* There has been no selection whatever of 
* This duplication of cases appears in several of the published 
statistics, especially of the Berlin statistics. Thus the cases reported 
by Schubert, Yoswinckel, Canon and Weibgen appear in the report 
of Ehrlich, Kossel and Wassermann, and partly in the reports of 
Korte, Sonnenburg and Hahn, sometimes twice repeated. Most of 
these cases are included without duplication, in my table, although 
they were treated with insufficient doses to a large extent. The 
cases attributed to Virchow, Aronson and Katz in some statistics 
are included in those of Baginsky. 20 
cases. All of the reports of the characters described which 
came to my notice are included, although many of the early 
Berlin cases (contained in statistics of Korte, Sonnenburg and 
Hahn in my tables) and some of the others were treated with 
entirely insufficient doses of antitoxin, and some observers 
have purposely selected, especially at the time when little 
serum was to be had, only severe cases for treatment. Kohts’ 
47 cases could not be inserted, as in the report which I have 
seen he does not give the number of deaths. 
The reports are of unequal value. Some present full and 
precise details of each case or of the group of cases, with state- 
ments as to previous or simultaneous fatality in the same class 
of cases in the same locality, whereas others are meagre and 
unsatisfactory. 
Some reports are based upon the bacteriological control of 
the clinical diagnosis, others upon the clinical diagnosis 
uncontrolled by bacteriological examination. In general the 
statistics from the larger hospitals relate to cases in which 
the Loftier bacillus was demonstrated, whereas many of the 
reports from private practice are without bacteriological exam- 
ination. A few, notably Leichtenstern and Wendelstadt, 
purposely base their observations upon cases in which the 
diagnosis is purely clinical without bacteriological control. 
I shall take this opportunity to consider the influence which 
the requirement of the bacteriologist, that the clinical diag- 
nosis of diphtheria should be controlled by a bacteriological 
examination in testing the efficacy of antitoxin treatment, 
is calculated to have upon the characters of statistics intended 
to show the value of the treatment. It is a favorite criticism 
of these statistics that the bacteriological, as distinguished 
from the purely clinical, diagnosis of diphtheria will operate 
in favor of a low fatality in antitoxin statistics, and that there- 
fore it is unfair to compare these statistics, with those which 
are based upon the uncontrolled clinical diagnosis of diph- 
theria. Some of the critics would have us believe that the 
antitoxin statistics on the one hand contain a large proportion 
of cases of mild inflammation of the throat with Loffler 
bacilli, but which no clinician would recognize as diphtheria, 
and on the other hand exclude a large proportion of fatal 
pseudo-membranous inflammations of the throat and air pas- 
sages which clinically would be regarded as diphtheria. 21 
In most of the statistical reports from hospitals on anti- 
toxin treatment the statement is expressly made, and it is 
apparent from the description of the cases, that they do not 
represent anything else than the usual run of cases of diph- 
theria as they have regularly for years past presented them- 
selves at the same hospitals. The mild diphtheric sore 
throats without clinical evidences of ordinary diphtheria are 
not likely in any large number to he recognized at all as diph- 
theria, and still less likely to find their way into general 
hospitals, from which most of the statistics are derived. 
Where, as in the statistics of Baginsky and others, sufficient 
detail concerning each case is reported to enable the reader to 
form an intelligent estimate of the character and severity of 
the case, it is evident that affections without the customary 
anatomical and clinical characters of diphtheria do not enter 
into the statistics. 
It is erroneous to say that the antitoxin statistics are not 
based upon the clinical diagnosis of diphtheria. The diag- 
nosis is clinical, but with subsequent bacteriological control. 
The cases are admitted to the hospital with the clinical diag- 
nosis of diphtheria, and the healing serum is or should be at 
once administered without waiting for the result of the cul- 
tures from the throat. Soltmann has been justly criticised 
for delaying the injection of antitoxin until after the bacterio- 
logical examination was completed. 
The assumption that non-membranous anginas and tonsillitis 
containing Loffler bacilli figure to any appreciable extent in 
these statistics is without warrant of facts. 
There are treated of course together with severe cases many 
mild cases with small patches of membrane on the tonsils or 
in the throat, but such cases are clinically diphtheria, or cer- 
tainly ought to be suspected of diphtheria by the clinician. 
It is important that such cases, when caused by the Loffler 
bacillus, especially in young children, should be treated by 
antitoxin, for not a few such cases when untreated develop 
into severe cases, sometimes suddenly into laryngeal diph- 
theria. Kurth, for example, relates a case in which a twin 
brother of a child ill with diphtheria was found to present 
small membranous patches on the tonsils which during two 
weeks of observation would at times disappear and which did 22 
not apparently make the child ill. Loffler bacilli were demon- 
strated, but the parents would not consent to the injection of 
serum. At the end of fourteen days, laryngeal diphtheria 
suddenly developed. The injection of antitoxin was followed 
by recovery in four days. This is simply a type of not a few 
cases which are regarded as suddenly developed laryngeal diph- 
theria. 
If, as is doubtless true, in some hospitals a larger number 
of cases are now received for serum treatment in earlier stages 
of diphtheria than formerly, this is not because the bacterio- 
logical diagnosis has supplanted the clinical, but because the 
importance of early inception of serum treatment has been 
justly emphasized. The recognition of mild and very mild 
cases of diphtheria is not a discovery of the bacteriologist, but 
has long been known to physicians, nor is it a peculiarity of 
the fatality statistics of cases treated by antitoxin that such 
mild cases are included in the statistics. They appear equally 
in previous fatality statistics of diphtheria. Mosler, for 
example, reports 313 cases of diphtheria with a fatality of 
14.5 per cent, treated during a year in the Grreifswald clinic 
before the introduction of serum, and there are numerous other 
statistics showing that mild cases often preponderate in previ- 
ous fatality statistics of diphtheria. Nor is the comparison 
in all of the reports with statistics in which the diagnosis is 
without bacteriological control. In several reports the com- 
parison is with the results of cases in which the Loffler bacillus 
was demonstrated, but which were not treated with serum. 
As regards the exclusion from antitoxin statistics of cases 
presenting the clinical characters of diphtheria without the 
Ldffler bacillus, it is evident, from what has previously been 
said, that, with thorough bacteriological tests, this can affect 
only a very small number of cases of unmistakable primary 
clinical diphtheria. Of the cases concerning which the 
clinician is in doubt, a considerable proportion are not diph- 
theria by bacteriological examination, which alone can decide 
the question. Although some of the non-diphtheric, pseudo- 
membranous cases are very grave affections, their general 
fatality is much lower than that of genuine diphtheria. 
The exclusion from the fatality statistics of diphtheria of the 
pseudo-membranous cases without Loffler bacilli is, therefore, 
the exclusion of a generally milder class of cases, as has been 23 
repeatedly demonstrated, and the result is to assign a higher 
and not a lower fatality to the remaining cases. A few 
examples taken from reports in Table I will suffice to demon- 
strate this. From the statistics of Eoux, Martin and Cbaillou 
128 pseudo-membranous cases treated with serum were thrown 
out because they were proven subsequently to be devoid of 
Loffler bacilli. The fatality of these cases was only 8.5 per 
cent., whereas in the 300 remaining cases which contained 
Loffler bacilli and were injected with serum the fatality was 
26 per cent. In Sevestre and Meslay’s statistics 29 cases with- 
out Loffler bacilli, but treated with serum, gave a fatality of 
3.4 per cent, as opposed to a fatality of 10 per cent, in the 
treated cases containing Loffler bacilli. A similar difference 
appears in other reports. The serum has no curative influence 
on pseudo-membranous inflammations not caused by the 
Loffler bacillus. 
In the best reported statistics information is afforded as to 
these various points, and the reader can learn the ratio of 
apparently mild cases and the number and results of the diph- 
theroid cases. 
Although only those statistics which are based upon the 
thorough bacteriological examination of the cases treated can 
lay claim to entire accuracy, the benefits of antitoxic treat- 
ment are clearly apparent in reports based upon the uncon- 
trolled clinical diagnosis of diphtheria. Of course in ordinary 
general practice it is not to be expected that the diagnosis will 
rest upon a bacteriological examination, but it should be 
understood that in the absence of such examination there 
must be occasional instances of apparent failure of antitoxin 
which woiUd be found explicable had a bacteriological exami- 
nation been made. 
In many reports the percentage of deaths in the cases 
treated with antitoxin is corrected by excluding cases evidently 
hopeless on admission or dying within twenty-four hours after 
commencement of the treatment. These corrected percentages 
are usually very materially lower than the rates based on all 
of the deaths. For example, if the cases dying within twenty- 
four hours after injection of antitoxin be excluded, the per- 
centage of deaths in Koux, Martin and Chaillou’s cases 
becomes 21.5 instead of 26; in Baginsky’s 13.5 instead of 15.6; 
in von Widerhofer’s 14.3 instead of 23.7; in Vierordt’s 14.6 24 
instead of 25; in Lebreton and Magdelaine’s 10.8 instead of 
12; in Moizard and Perregaux’s 11.3 instead of 14.7; in 
Sevestre and Meslay’s 6.6 instead of 10; in Bokai’s 18.3 
instead of 25.5, etc. I have, however, not used these reduced 
percentages, although in many instances it might with pro- 
priety have been done. The statistics in my tables, therefore, 
do not give in many instances as favorable percentages for 
antitoxin as may justly be claimed, but on the other hand 
they are more properly comparable with the previous or simul- 
taneous fatality rates from diphtheria, these being based upon 
the total number of deaths in all of the cases treated. I have 
aimed to avoid the accusation of selection of cases or of unfair 
manipulation of the figures. 
In the majority of the reports the cases treated are all or 
nearly all of the cases of diphtheria which were admitted to the 
hospital or came under observation during the period of treat- 
ment, but in some it is expressly stated that in consequence of 
the cost and the scarcity of the healing serum mild cases and 
evidently hopeless cases did not receive the serum. There is 
no evidence of selection of mild cases in order to obtain 
results favorable to antitoxin. 
The percentages in the column headed “ Previous Fatality ” 
are those given by the writers for diphtheria not treated with 
antitoxin. In some instances they relate to the average 
fatality for a series of preceding years, in some to the minimum 
and the maximum fatality for several years, in some to the 
simultaneous fatality or the fatality during a period of inter- 
ruption in the supply of serum. In all instances they are the 
case mortality rates of diphtheria in the hospital or locality 
from which the cases treated with antitoxin are derived. The 
arrangement of the reports is only in part chronological. 
Following the references to the articles, it is stated, so far as 
could be ascertained, whether the cases were in hospital or in 
private practice. 
It appears from Table I* that of 7166 patients with diph- 
* The report of Fiirth concerning the results of serum treatment 
in the medical and surgical clinics at Freiburg was published too 
late to be included in my tables. 100 cases were treated with a 
fatality of 12 per cent. During the five preceding years the fatality 
from ordinary treatment fluctuated between 31 and 49 per cent., 
averaging 39 per cent. The same average existed during the seven 25 
theria treated with antitoxin 1239 or 17.3 per cent. died. 
Among these cases are included many treated during the 
early period after the first introduction of the treatment, with 
entirely insufficient doses. There are also included a large 
number of cases dying from complicating diseases not refer- 
able to diphtheria, or dying within twenty-four hours after 
beginning the treatment, cases which cannot properly be 
regarded as indicating failure of the serum treatment. The 
great bulk of the statistics come from children’s hospitals. 
Under these circumstances, indeed from any point of view, 
the fatality derived from Table I of cases treated with anti- 
toxin is very low. 
There is, however, no standard of comparison for the 
fatality in this entire group of cases. It cannot be compared 
with fatality statistics from hospitals nor with those from 
private practice. The table contains at least five or six times 
as many cases from hospital practice as from private practice. 
The ratio of deaths to all cases, therefore, is greater than 
could be expected from the returns of all cases similarly 
treated in cities, but even in comparison with municipal 
fatality statistics of diphtheria during the prevalence of mild 
types of the disease, the percentage of deaths is very low. 
This strikingly low fatality in itself speaks strongly in favor 
of the efficacy of the serum treatment. 
In 46 reports contained in the table the previous or simul- 
taneous percentage of deaths from diphtheria not treated with 
antitoxin is given for the same hospital or locality in which 
were the cases treated with antitoxin. These reports contain 
5406 cases of diphtheria treated withantitoxin, with 1008 deaths 
or 18.6 per cent. If we calculate the number of deaths in 
each series of these cases upon the basis of the previous 
fatality, selecting the lowest percentages given, we have 2279 
deaths or 42.1 per cent. There is, therefore, on this estimate 
by the use of antitoxin an apparent reduction in the number 
of deaths of 55.8 per cent. 
months of the year corresponding to the period during which anti- 
toxin was used. There was laryngeal involvement in 43 cases and 
tracheotomy was performed in 31 with 11 deaths (35.4 per cent ). 
In previous years tracheotomy was required in 46.2 per cent, of the 
cases, with a mortality of 70.4 per cent. (Abstract in the Medical 
News, August 17, 1895.) TABLE I.—FATALITY OF CASES OF DIPHTHERIA TREATED WITH ANTITOXIN. 
Number of Cases Treated with Antitoxin, the Number and Percentage of Deaths, and the Previous Fatality, in 82 Reports. 
Reporter. 
Cases. 
Deaths. 
Previous Fatality. 
References. 
1. Roux, Martin and Chaillou. 
300 
78 (26 per cent.) 
50 per cent. 
Annales de l’lnstitut Pasteur, Sept., 1894. (Cases in HOpital des Enfants- 
Malades.) 
2. Kossel. 
117 
13 (11.1 per cent.) 
52-61 per cent. 
Deutsche Med. Wochenschrift, 1894, p. 946. (Hospital.) 
3. KOrte. 
121 
40 (33.1 per cent.) 
45.1 per cent. 
Berliner Klin. Wochenschrift, 1894, p. 1039. (Hospital.) 
4. Sonnenburg. 
95 
16 (16.8 per cent.) 
27.6 per cent. 
During period of in- 
terruption of serum 
treatment. 
Deutsche Med. Wochenschrift, 1894, p. 930. (Hospital.) 
5. Hahn. 
205 
49 (24 per cent.) 
41 per cent. 
Ibid. 1895, Vereins-Beilage, p. 2. (Hospital ) 
6. Baginsky. 
525 
83 (15.6 per cent.) 
41 per cent. 
Die Serumtherapie der Diphtherie. Von Dr. Adolf Baginsky, Berlin, 1895. 
(Hospital.) 
7. Heubner. 
207 
27 (13 per cent.) 
48-52 per cent. 
Reported at 13ter Congress fiir Innere Medicin, Miinchen, April 2, 1895. 
Miinchener Med. Wochenschrift, April 9, 1895. (Hospital.) 
8. v. Widerhofer. 
300 
71 (23.7 per cent.) 
50 per cent. 
Ibid. (Hospital.) 
9. v- Ranke. 
96 
19(19.7 per cent.) 
42.2-56 per cent. 
Ibid. (Hospital.) 
10. Stintzing. 
59 
12 (20.1 per cent.) 
25 per cent. 
Ibid. (Hospital.) 
11. Rauchfuss. 
100 
34 (34 per cent.) 
55 per cent. 
Ibid. (Hospital.) 
12. v. Mering. 
74 
4 (5 per cent.) 
30 per cent. 
Ibid. 
13. v. Noorden. 
81 
19 (23 per cent.) 
45 per cent. 
30-37 per cent. 
Ibid. (Hospital.) 
14. SchrOder. 
63 
8 (12.7 per cent.) 
Ibid. (Hospital.) 
Deutsche Med. Wochenschrift, 1895, p. 169. (Hospital.) 
15. Vierordt. 
63 
16 (25 per cent.) 
41-67 per cent. 
16. Rumpf. 
22 
2 (8 per cent.) 
13-28 per cent. 
Miinchener Med. Wochenschrift, Nov. 20, 1894. (Hospital.) 
17. Lebreton and Magdelaine. 
258 
31 (12 per cent.) 
50 per cent. 
Le Bulletin Medical, 1895, No. 10. (HOpital des Enfants-Malades.) 
18. Le Gendre. 
17 
3 (17.6 per cent.) 
50-60 per cent. 
Bull, et Memoires de la Soc. Med. des HOpitaux de Paris, Dec. 20, 1894, 
(HOpital Trousseau.) 
19. Moizard and Perregaux. 
231 
34 (14.7 per cent.) 
50-60 per cent. 
Jour, de Medecine et de Chirurgie, Dec. 15,1894. (HOpital Trousseau.) 
20. Sevestre and Meslay. 
150 
15 (10 per cent.) 
50-60 per cent. 
Le Bulletin Medical, 1895, No. 18. (HOpital Trousseau.) 
21. Ganghofner. 
110 
14 (12.7 per cent.) 
43.6-78.2 per cent. 
Prager Med. Wochenschrift, 1895, Nos. 1, 2 and 3. (Hospital.) 
-'2. Soltman. 
89 
13 (14.6 per cent.) 
27.2 per cent. 
Deutsche Med. Wochenschrift, 1895, No. 4. (Hospital.) 
23. Bokai. 
176 
42 (24 per cent.) 
53.5-67.5 per cent. 
Ibid. 1895, No. 15, and Wiener Med. Presse. 1895, No. 12. (Hospital.) 
24. Escherich. 
51 
5 (9.5 per cent.) 
Abstract in Miinchener Med. Wochenschrift, 1895, No. 7. (Hospital.) 
25- Gnandinger. 
27 
11 (40.7 per cent.) 
36-45.5 per cent. 
Wiener Klin. Wochenschrift, 1895, No. 1. 
26. Monti. 
25 
1 (4 per cent.) 
Wiener Med. Wochenschrift, 1895, Nos. 4 and 5. (Hospital.) 
27. Heim. 
27 
6 (22 per cent.) 
52.5 per cent. 
Ibid. 1895, No. 4. (Hospital.) 
28. Unterholzner. 
31 
8 (25.8 per cent.) 
66.7 per cent. 
Ibid. (Hospital.) 
29. Baumler. 
26 
2 (7.7 per cent.) 
Miinchener Med. Wochenschrift, 1894, p. 1062. (Hospital.) 
30. Borger. 
30 
2 (6.66 per cent.) 
20 per cent. 
Deutsche Med. Wochenschrift, 1894, No. 48. (Hospital.) 
31. Hilbert. 
11 
0 
Ibid. 1894. Vereins-Beilage. p. 142. (Hospital.) 
32. Hager. 
25 
1 (4 per cent.) 
Cent.ialbl. f. innere Mcdicin, 1894, No. 48. (Private practice.) 
33. Moeller. 
76 
27 35.5 per cent.) 
Ibid. (Hospital.) 
34. Kuntzen. 
25 
3 (12 per cent.) 
Deutsche Med. Wochenschrift, 1894. No. 49. (Hospital.) 
35. Schmidt. 
14 
3 (21.4 per cent.) 
Deutsche Med. Wochenschrift, 1894, No. 52. (Private practice.) 
36. Seiz. 
27 
1 (3.6 per cent.) 
30.6 per cent. 
Therapeut. Monatshefte, Dec., 1894. (Private practice.) 
Annales de la Soc. Royale des Sciences Medicaleset Naturellesde Bruxelles, 
T. Ill, p. 337 (1894). (Hospital.) 
37. Charon. 
13 
4 (30.8*per cent.) 
38. Washbourn, Goodall and Card. 
72 
14 (19.4 per cent.) 
36-41.8 per cent. 
Brit. Med. Jour., Dec. 22,1894. (Hospital.) 
39. Seitz. 
36 
2 (5.7 per cent.) 
Abst. in Miinchener Med. Wochenschrift, 1895, No. 12. (Hospital.) 40. Pavlik. 
13 
1 (7.7 per cent.) 
Wiener Med. Presse, 1895, Nos. 1 and 5. (Private practice.) 
41. Handler. 
32 
5 (15.6 per cent.) 
Ibid. No. 6. (Private practice.) 
42. Herringham. 
18 
3 (16.7 per cent) 
Brit. Med. Journ., Dec. 22,1894. (Hospital.) 
43. Caiger. 
30 
8 (26.6 per cent.) 
30 per cent. 
Ibid. Dec. 29,1894. (Hospital.) 
44. Hall. 
11 
3 (27.3 per cent.) 
31.25 per cent. 
Ibid. Jan. 19,1895. (Hospital ) 
45. Tirard and Willcocks. 
10 
1 (10 per cent.) 
38 per cent. 
The Lancet, Jan. 19,1895. (Hospital.) 
46. Buffer. 
274 
37 (13.5 per cent.) 
Cases treated in four London hospitals, British Med. Jour., Feb. 2, 1895. 
47. Epidemic in Trieste. Cases in 
private practice. 
72 
5 (6.5 per cent.) 
Das Oesterreichische Sanitatswesen, Jan. 3, 1895. 
48. Cases in hospital. 
180 
40 (22.2 per cent.) 
49. Blumenfeld. 
5.) 
2 (4 per cent.) 
Wiener Klin. Wochenschrift, 1895, No. 3. (Private practice.) 
50. Witthauer. 
36 
5 (14 per cent.) 
Therapeut. Monatshefte, Feb., 1895. 
51. Dreyfus. 
78 
15 (19.3 per cent.) 
50 per cent. 
Lyon Medical, Feb. 3, 1895. (Hospital.) 
52. Simon. 
16 
2 (i2.5 per cent.) 
La Medecine Moderne, Feb 6, 1895. (Private practice.) 
63. Malvoz. 
15 
1 (7 per cent.) 
Le Scalpel, Feb. 17, 1895. 
54. Gougenheim. 
125 
12 (9.5 per cent.) 
21.1 per cent. 
Annales des Maladies de l’Oreille, du Larynx et du Pharynx, 1895, No. 5. 
Abst. in the Medical News, June 15, 1895. (Hospital.) 
55. Rapmund. 
100 
7 (7 per cent.) 
20-30 per cent. 
Zeitschrift f. Medizinal-Beamte, Feb. 15, 1895. (Collective investigation of 
use of antitoxin in private practice in district of Minden.) 
56. Schiiller. 
32 
1 (3.1 per cent.) 
Allgem. Med. Central-Zeitung, 1895, No. 88. (Private practice.) 
57. Grtinfeld. 
12 
1 (8.3 per cent.) 
Abst. Schmidt’s Jahrbiieh. 1895, Bd. 246, p. 37. (Hospital.) 
58. Schaewen. 
15 
0 
Berliner Klin. Wochenschrift, 1895, No. 10. (Private practice.) 
59. Heidenhain. 
24 
3 12.5 per cent.) 
Ibid. (Private practice.) 
60. Risel. 
114 
9 (7.9 per cent.) 
Deutsche Med. Wochenschrift, 1895, No. 10. (Hospital and private practice.) 
61. Weiland. 
20 
0 
Ibid (From private practice of six physicians in Baden.) 
62. v. Muralt. 
58 
2 (3.4 per cent.) 
Correspondenzbl. f. Schweizer-Aerzte, 1895, No. 5. (Hospital.) 
63. Blattner. 
38 
9 (23.6 per cent.) 
Abst. in Miinchener Med. Wochenschrift, Mar, 5,1895. (Hospital.) 
64. Gerloczy. 
55 
15 (27 3 per cent.) 
Abst. Ibid. (Hospital.) 
65. D’Espine. 
60 
6 (10 per cent.) 
Rev. Med. de la Suisse ltomande, April 20,1895. (40 cases in hospital.) 
66. Mya. 
17 
2 (11.8 per cent.) 
Wiener Med. Blatter, 1895, p. 760. (Hospital.) 
67. v. Engel. 
39 
10 (25.5 per cent.) 
50 per cent. 
Prager Med. Wochenschrift, 1895. 
68. Fischer. 
225 
35 (15.5 per cent.) 
New Tork Medical Record, April 6, 1895. (Hospital, consultation and pri- 
vate practice.) 
69. Biggs—Cases in the city treated 
by sanitary inspectors. 
255 
40 (15.69 per cent.) 
25-45 per cent. 
Ibid. April 20,1895. 
70. Cases in Willard-Parker Hosp’l. 
164 
45 (27.4 per cent.) 
32 per cent. 
71. Lennox Browne. 
45 
2 (4.4 per cent.) 
34 per cent. 
Le Bulletin Medical, 1895, No. 21. (Hospital.) 
72. Codd and Whitehouse. 
11 
4 (36.4 per cent.) 
British Med. Jour., May 18,1895. (Hospital) 
73. Wink held. 
22 
4 (18.2 per cent.) 
36 per cent. 
Ibid. May 11,1895. (Hospital.) 
74. Horowitz. 
21 
1 (4.8 per cent.) 
10.7 per cent. 
Abst. in Miinchener Med. Wochenschrift, May 7,1895. (Hospital.) 
75. Sigel. 
100 
12 (12 per cent.) 
40-50 per cent. 
Ibid. May 21,1895 (abstract). (Hospital.) 
76. Howard. 
40 
3 (7.5 per cent.) 
The Medical News, June 1,1895. (Private practice.) 
77. Van Nes. 
52 
12 (23 per cent.) 
36-48 per cent. 
Deutsche Med. Wochenschrift, June 6,1895. (Hospital.) 
78. Leichtenstern and Wendelstadt 
123 
25 (20.3 per cent.) 
30.9 per cent. 
Miinchener Med. Wochenschrift, June 11, 1895. (Hospital.) 
79. Kurth. 
97 
10 (10.3 per cent.) 
Deutsche Med. Wochenschrift, July 11,1895. (35 cases—hospital.) 
80. Tiramer. 
31 
6 (19.4 per cent.) 
Abst. in Deutsche Mediziual-Zeiturig, June 10,1895. (Hospital.) 
81. Cases treated in Cartagena. 
156 
21 (13.5 per cent.) 
Abst. in British Med. Jour., July 6,1895. 
82. Mason. 
306 
81 (26.4 per cent.) 
45-52 per cent. 
Reported to Assoc, of American Physicians, May 31, 1895. Abst. in the 
Medical News, June 15,1895. (Hospital.) 
Total (82 Reports). 
7166 
1239 (17.3 per cent.) 
The fatality of 7166 cases of diphtheria treated with antitoxin was 17 ?> per cent. The previous or simultaneous fatality of cases not treated with 
antitoxin is stated in 46 reports. These contain 5406 cases treated with antitoxin with 1008 deaths, a fatality of 18.6 per cent. Estimating the 
number of deaths in these cases upon the basis of the previous or simultaneous fatality for each group (taking the lowest figures given), there would have 
been 2279 deaths or 42.1 per cent. There was, therefore, an apparent reduction of case mortality by the use of antitoxin of 55.8 per cent. 28 
There must be a much greater difference between the 
characters of the cases composing the two groups compared 
than appears from the statements of the writers and the 
details of the cases described, if this striking reduction in 
fatality is not due to the serum treatment. 
If we separate the hospital cases from those in private prac- 
tice we obtain from 61 reports of Table I 5777 cases of diph- 
theria treated with antitoxin in hospitals. These furnished 
1081 deaths, giving a percentage of 18.7. Although this is 
not an unheard of fatality of diphtheria in hospitals, it is most 
exceptional, and I am not aware that anything approaching 
it has been observed in hospitals receiving large numbers of 
cases of diphtheria in children. The fourth column in Table 
I gives the percentages observed in many such hospitals. 
There are 41 reports which give for the same hospital the 
previous percentage of deaths from diphtheria not treated 
with antitoxin. These furnish 4899 cases treated with anti- 
toxin, with 944 deaths or 19.3 per cent. If we calculate the 
number of deaths which would have occurred among these 
cases had the percentages of previous fatality obtained, select- 
ing the lowest percentages given, there would have been 2130 
deaths or 43.5 per cent. The apparent diminution in the 
number of deaths as the result of serum treatment is accord- 
ing to this estimate 55.6 per cent. If we had selected only 
the larger and most carefully analyzed and satisfactory 
statistics from the principal hospitals, in large part children’s 
hospitals, there would have been in over 3000 cases an apparent 
reduction in fatality of 60 per cent. 
There may occur considerable differences in the annual 
fatality from diphtheria in a hospital during a series of years, 
but such differences between the minimum and the maximum 
fatality as that just noted between the actual and the estimated 
fatality are most exceptional. In the Friedrichshain hospital 
in Berlin there has been observed a difference of 28 per cent, 
in the annual fatality from diphtheria. The largest differ- 
ence observed in the surgical clinic in Berlin during ten years 
was that between 43.2 per cent, in 1888 and 58.5 per cent, in 
1890.* In the report of the Metropolitan Asylums Board f in 
*V. Hirsch. Archiv f. klin. Chirurgie. Bd. 49, Hft. 4. 
f British Medical Journal, Dec. 22, 1894. 29 
London, where the case mortality from diphtheria in hospi- 
tals is generally much lower than on the Continent, the 
fatality in 1889 was 40.7 per cent.; in 1890, 33.5 per cent.; in 
1891, 30.6 per cent.; in 1892, 29.3 per cent.; in 1893, 30.4 per 
cent. This apparent reduction in fatality since 1889 in large 
part disappears if only patients under 15 years of age are 
considered, the corresponding percentages for these being 
respectively 40.7, 41.6, 36.9, 35.6, and 37. 
The natural interpretation of our statistics showing in over 
7000 cases, of which at least five-sixths are from hospital 
practice, treated with antitoxin an extraordinarily low per- 
centage of deaths for this class of cases, and showing an 
apparent reduction in fatality of from 50 to 60 per cent. By 
the use of antitoxin, is that antitoxin exerts a specific curative 
power over diphtheria. 
What are the objections which may be and have been urged 
against this natural interpretation of the statistical evidence ? 
In the first place it has been claimed that these observations 
have been made during the prevalence of unusually mild 
diphtheria. In some places the prevailing type of the dis- 
ease seems to have been mild, but the great majority of the 
observers quoted in the table consider that the prevailing 
diphtheria in their localities has been of average severity, and 
they cite in many instances the simultaneous fatality of cases 
not treated with antitoxin as proof that the disease is not of 
peculiarly mild type, indeed in several places it seems to have 
been of more than average severity. During the period in 
which Roux treated with antitoxin 300 cases in the Hopital 
des Enfants-Malades with a fatality of 26 per cent., the 
fatality in the Hopital Trousseau, also in Paris, and receiving 
a similar class of cases, was 60 per cent. 
But even if it be admitted for the sake of argument that 
the prevailing type of diphtheria during the past year has 
been mild, it is to be considered that the influence of this 
milder type upon the cases received in many hospitals appears 
chiefly in the reduction of their number, and far less in a 
change in the character of the cases admitted. This is the 
statement of von Ranke, of Bokai and of several other physi- 
cians in charge of diphtheria wards. They say that so far as 
their hospitals are concerned, as a rule, severe and advanced 30 
cases are sent there by physicians in the city, often for opera- 
tion to relieve laryngeal stenosis, and that when the epidemic 
is mild in character they receive fewer cases, but not many 
milder cases. Doubtless these conditions will not hold for 
all hospitals, particularly not for such as are intended for the 
compulsory isolation of all cases of diphtheria which cannot 
be properly isolated at their homes, but they are probably 
applicable in large part to most of the hospitals from which 
come the reports now under consideration. 
So far as I can judge, no proof has been brought forward in 
support of the opinion that the low percentage of fatality of 
diphtheria treated with antitoxin can be referred in any large 
measure to the prevalence of an unusually mild type of the 
disease, although in a few scattered groups of cases, particu- 
larly some of the smaller series in my table, this may be in 
part the explanation. 
I am inclined to attach decidedly more weight to a second 
criticism of antitoxin statistics which has been made, namely, 
that in hospitals where the serum treatment has been carried 
out, a proportionately larger number of cases are received 
now than formerly in the earlier stages of diphtheria. The 
advocates of the treatment have properly insisted upon the 
importance of early injection of the serum, and, especially 
during the time when the serum was not to any extent in the 
hands of general practitioners, it would be natural to suppose 
that physicians would send their patients and parents take 
their children to such hospitals as soon as possible after 
recognition of the disease. Inasmuch as with any approved 
method of treatment of diphtheria the results are better the 
earlier it is begun, it is evident that statistics based on the 
former experience with the treatment of diphtheria in hospi- 
tals would not be altogether comparable with the antitoxin 
statistics from the same hospitals. 
It is, however, very difficult to say how much allowance is 
to be made for this criticism. There has been wide-spread 
skepticism among physicians and the general public as to the 
value of the treatment. Thus Rapmund, in his efforts to 
establish in the district of Minden a collective investigation 
of serum therapy in diphtheria, found the physicians so 
skeptical that of 194 practitioners only 20 would use it at all, 31 
and only two employed it extensively. There are also state- 
ments as to the unwillingness of parents to have it tried on 
their children. Many of the reports state that during the 
period of serum treatment cases were not received in any 
earlier stages of the disease than formerly; in a few of the 
reports, as for example in that of Kuntzen from Oschers- 
leben, it is said that physicians were induced to send their 
patients early in the disease. In Berlin and some other cities 
there has been a marked increase in the number of patients 
with diphtheria admitted to hospitals since the introduction 
of the antitoxin treatment, and this has been without a corre- 
sponding increase in the total number of cases in the cities. 
Heubner in his recent address at the Congress of Internal 
Medicine in Munich admits that lighter cases of diphtheria 
go to the hospital now, but that this is not enough to explain 
the great difference in fatality. There are undoubtedly con- 
siderable differences in different hospitals as to the proportion 
of cases admitted in early stages of diphtheria, but in many 
of the hospitals where the benefits of antitoxin have been 
most apparent as contrasted with the previous results it is 
expressly stated that the number of mild cases admitted is no 
greater than formerly. 
If we make all due allowance for this possible increase in 
the proportion of early cases treated in hospitals, and certainly 
some allowance must be made, this factor is still altogether 
inadequate to explain the great reduction in fatality of diph- 
theria treated with antitoxin. This will also be apparent 
later when we consider the results of treatment according to 
the day of the disease on which it is begun. 
A third criticism, namely, that the bacteriological control 
of the diagnosis of diphtheria operates in favor of a low 
mortality in antitoxin statistics, has already been fully dis- 
cussed. 
It is manifestly improper to compare the average fatality 
of thousands of cases treated in hospitals with antitoxin with 
exceptionally favorable results at certain periods in a few 
hospitals in a comparatively small number of cases without 
serum treatment, and still more improper, as has even been 
done, to make such comparison with the most favorable per- 
centages which one can find reported from private practice or 32 
in municipal mortality statistics. Surely some consideration 
must be given to the previous and simultaneous results 
obtained from cases without serum treatment in the same 
hospitals from which the cases reported are derived. 
We have now considered the principal objections which 
have been made to the natural interpretation of statistics 
showing an apparently great reduction in the fatality of diph- 
theria by the use of antitoxin. I believe that it has been 
shown that even if all possible allowance be made for such 
assumptions as those considered, they are still wholly inade- 
quate to account for an apparent reduction in the deaths from 
diphtheria by antitoxic treatment of 50 to 60 per cent, in 
nearly 5000 cases collected from hospitals in Germany, France, 
Austria, Italy, England and America, and reported by forty 
different physicians, most of whom are of high reputation and 
large experience. These statistics seem to me to establish 
beyond all reasonable doubt the conclusion that antitoxin is a 
specific curative agent for diphtheria. 
It has been contended that the only absolutely convincing 
proof of the curative efficacy of antitoxin is the demonstra- 
tion of a marked reduction in the total number of deaths 
from diphtheria in a city or town in proportion to all of the 
cases. Municipal mortality and morbidity statistics are neces- 
sarily far less accurate than hospital statistics, and for reasons 
which have been stated, the prevalence of a mild type of 
diphtheria will have greater influence upon municipal mor- 
tality statistics for diphtheria than upon hospital statistics. 
It is to be expected that when sufficient time has elapsed and 
the employment of antitoxin in the treatment of diphtheria 
has become sufficiently general, the reduction in fatality by 
its use will be apparent in general fatality statistics. At 
present we have little information upon this point. The mere 
statement of the total number of deaths, without knowledge 
of the morbidity and of the prevailing type of disease, is of 
course not decisive for either side of the question, but so far 
as it goes it is interesting to learn that in Boston during the 
antitoxin period (January 1 to May 1, 1895) the total fatality 
from diphtheria was 14 per cent., as compared with a fatality 
of 31 per cent, during the corresponding period of previous 
years (Mason), and that in Cartagena, Spain, during four 33 
months of employment of antitoxin, the total number of 
deaths was only one quarter the average number for the same 
period of time during the preceding ten years. 
The only antitoxin statistics which T can find based upon 
such material as composes municipal fatality statistics are 
those of Risel and of Kurth. 
Risel reports the results in all of the cases treated by anti- 
toxin during two months in the city of Halle. They are 
derived from the practice of thirty physicians among the 
poor and the rich, in the houses of the patients and in hospi- 
tals, and include mild and severe cases as they presented 
themselves. Of the 89 patients treated in their homes, almost 
without exception children not over 7 years of age, 6 died, 
giving a fatality of 6.7 per cent. 19 of these had laryngeal 
diphtheria, of whom 4 died. Of the 25 patients treated in 
hospitals, 3 died, a fatality of 12 per cent. 15 of these had 
laryngeal involvement, of whom 3 died. The total fatality 
was 7.9 per cent. Ho data are given for comparison with 
the previous or simultaneous fatality of cases not treated 
with serum. In only a few cases, and these in hospitals, 
was the clinical diagnosis confirmed by bacteriological exam- 
ination. 
Kurth reports the results of serum treatment in the prac- 
tice of sixty physicians in Bremen and a few outlying villages 
from October 8, 1894, to January 31, 1895. A circular letter 
was sent from the Bacteriological Institute to every physician 
in the city, and apparently general co-operation on the part 
of the physicians and the public officials was secured. In 
97 cases treated with serum the diagnosis of diphtheria was 
established, in the great majority of cases by demonstration 
of the Loffler bacillus controlling the previous clinical diag- 
nosis. The total case mortality was 10.3 per cent. The 
fatality of the 64 cases treated in the city (hospital and pri- 
vate practice) was 7.8 per cent.; that of 33 cases derived from 
the surrounding country district was 15.2 per cent.; that 
of 35 cases treated in the city hospitals was 14.3 per cent. 
Laryngeal diphtheria occurred in 66 per cent, of the cases in 
country districts and only in 36 per cent, of the cases in the 
city. This and the generally less favorable results in country 
practice are attributed by Kurth not to greater severity of the 34 
epidemic in the former, but to the custom in the country of 
not calling the physician until the symptoms are urgent, and 
to the greater distance which the physicians have to travel. 
If the bacteriological control of the diagnosis be disregarded, 
that is, if all the cases diagnosed clinically as diphtheria and 
treated with serum be considered, the fatality was 9.4 per 
cent., another illustration that bacteriological control of the 
clinical diagnosis results in higher, not as some have claimed 
in lower, percentage of deaths. Of the 50 cases of clinical 
diphtheria, all of the cases being included which did not 
show Loffler bacilli, the fatality was only 6 per cent. During 
the serum period there occurred 25 cases of diphtheria not 
treated with antitoxin, with a fatality of 24 per cent. During 
the same period of the year in which the serum-treated cases 
occurred, there were during the preceding year 148 cases of 
diphtheria with a fatality of 32 per cent. It must be con- 
ceded that these interesting reports of Eisel and of Kurth speak 
strongly in favor of the possibility of bringing about a great 
reduction in the general fatality from diphtheria in cities by 
treatment with antitoxin. As a larger proportion of the 
cases in private practice can be treated in early stages of the 
disease than in the hospitals, this reduction should be greater 
than that already shown by hospital statistics. 
A most convincing demonstration of the healing power of 
antitoxin is furnished by the experience of Baginsky during 
an involuntary pause in the serum treatment caused by failure 
in the supply of serum. Between March 15, 1894, and March 
15, 1895, there were treated in Baginsky’s service by antitoxin 
525 children with a fatality of 15.6 per cent. During the 
period of forced interruption of the serum treatment, this 
period being chiefly the months of August and September, 
126 children were treated without antitoxin, with a fatality 
of 48.4 per cent. There was absolutely no selection of cases 
in either group. In his comments upon this experience Ba- 
ginsky says: “ It is all the more remarkable, as the ratio of 
mortality of those treated with the serum both before and 
after the period of interruption varied within very small per- 
centage figures. If one will permit figures to speak at all, 
there has scarcely been made on human beings a more demon- 
strative test of the curative power of a therapeutic agent. It 35 
was an experiment forced upon us, but it proved to us how 
terrible was the form of disease which we were treating, and 
how numerous would have been the victims without the use 
of the healing serum.” 
A similar experience has been reported by several other 
writers. Thus Korte noted a rise in fatality from 33.1 per 
cent, during the serum period to 53.8 during the period of 
failure in the supply of serum. G-anghofner, under similar 
conditions, a rise from 12.7 per cent, to 53.2 per cent.; Heim, 
from 22 per cent, to 65.6 per cent., and during the epidemic 
in Trieste the fatality rose from 18.7 per cent, to 50 per cent, 
when the serum failed. All of these highly significant obser- 
vations were made on cases occurring in the same epidemic, 
the period of enforced interruption of the serum treatment 
being preceded and followed by the periods of serum treat- 
ment. 
We have considered thus far mainly the hospital statistics. 
These are for manifest reasons more numerous, larger and 
more carefully analyzed than those from private practice. It 
is, however, in private practice, especially among those classes 
who are in the habit of calling the physician early in the 
disease, that the best results from serum treatment are to 
be expected, for here there is more frequent opportunity for 
timely treatment. A glance at Table I will show that in 
general the fatality of diphtheria treated with serum in pri- 
vate practice is much lower than in hospitals. 
If we summarize the 18 reports from private practice in 
Table I we have 663 cases of this class treated with antitoxin, 
and among these are 46 deaths, giving a fatality percentage 
of only 6.9. This would indicate that the serum treament 
may reduce the fatality from diphtheria in private practice to 
nearly one-third that under the same treatment in hospitals. 
Some of the reports of the results of serum treatment in pri- 
vate practice furnish, indeed, most remarkable evidence of 
the efficacy of this treatment. 
Most of the reports attempt some sort of classification of 
the cases treated with antitoxin. The simplest and most 
common, although not the most valuable, is the division 
according to degrees of apparent severity, expressed by such 
epithets as mild, moderate, severe, very severe. Such a classi- 36 
fication is, of course, only of limited value, as even the mildest 
case of diphtheria may unexpectedly assume a malignant 
character. If, as we believe to be proven, antitoxin injected 
in time arrests the local process and the constitutional dis- 
turbance, then many of the cases which appear under the 
head of mild cases in antitoxin statistics would under other 
methods of treatment have become severe cases and would be 
so recorded. Indeed, with the early administration of anti- 
toxin there should be comparatively few severe cases. 
The classification of diphtheria adopted by Roux into an- 
gina and croup, with and without microbic association, has 
been followed by some of his successors. So far as this clas- 
sification is actually based upon a correct separation of pure 
diphtheria from diphtheria with mixed infection, it is of the 
utmost importance in determining the relative value of anti- 
toxin in the treatment of these two divisions of diphtheria. 
I have already expressed the opinion, however, that the bacte- 
riological examination of the exudate and secretions in the 
throat is not decisive in determining the presence or absence 
of mixed infection. Still Roux’s analysis of his cases on this 
basis indicated clearly that the serum was far more efficacious 
in diphtheric anginas and croup which yielded pure cultures 
of the Loffler bacillus than in those which gave, in addition 
to the Loffler bacillus, cultures of other pathogenic bacteria. 
In presenting these results the epithet “pure” applied to 
angina or croup is to be understood to signify that only the 
Loffler bacillus was found in the cultures, and the epithet 
“associated” to signify that this bacillus was found in cul- 
tural association with the coccus of Brisou, staphylococci or 
streptococci, most commonly the last. The “corrected” per- 
centages are those obtained after subtracting deaths occurring 
within twenty-four hours after admission. 
Of the 300 cases treated with serum with a total fatality 
of 26 per cent, reported by Roux, Martin and Chaillou, there 
were 120 pure anginas, fatality 7.5 per cent, (corrected 1.7 per 
cent.); 49 associated anginas,fatality 24.2 per cent, (corrected 
17.7 per cent.); of not operated croup, 4 pure, fatality 0, 6 as- 
sociated, fatality 16.6 per cent.; of operated croup, 49 pure, 
fatality 30.9 per cent, (corrected 24.4 per cent.), and 72 associ- 
ated, fatality 56.9 per cent, (corrected 43.1 per cent.). Of Moizard and Perregaux’s 231 cases, total fatality 14.7 
per cent., there were 44 pure anginas, fatality 4.5 per cent.; 
42 associated, anginas, fatality 14.3 per cent.; 94 pure croup, 
fatality 18.5 per cent.; 51 associated croup, fatality 17.6 per 
cent. No correction is made in these percentages. 
Of Sevestre and Meslay’s 150 cases, total fatality 10 per 
cent., there were 29 pure anginas, fatality 3.4 per cent.; 24 
associated anginas, fatality 12.5 per cent, (corrected 8.3 per 
cent.); 67 pure croup, fatality 8.9 per cent, (corrected 7.5 per 
cent.); 30 associated croup, fatality 16.6 per cent, (corrected 
6.6 per cent.). 
It will be observed that in the last two reports the excess in 
fatality in the “associated” diphtherias is much less striking 
than in Roux’s statistics, and in some cases disappears 
altogether. This I am inclined to attribute to failure of the 
method employed to indicate properly the division into pure 
and mixed infections, for the testimony is unanimous that the 
serum is of far less benefit in mixed diphtheria than in uncom- 
plicated diphtheria, the most common and dangerous compli- 
cating micro-organism being the streptococcus pyogenes. 
Most noteworthy has been the improvement in the results 
of serum therapy of diphtheria in the Paris hospitals since 
Roux’s original communication to the Congress in Budapest 
in September, 1894. The fatality has descended from Roux’s 
original percentage of 26, in the later reports to 14.7, 12 and 
10 per cent., and according to a recent statement of Moizard 
and Bouchard (July, 1895), it at present oscillates between 8 
and 14 per cent. These are the best results which have 
hitherto been reported from any hospital for any large number 
of cases, and they are certainly most significant. As Moizard 
and Bouchard in their recent communication say, “ This 
result can no longer be attributed to fortunate series of cases, 
as was claimed at the beginning by adversaries of the method. 
Thousands of patients have been treated, and it can now be 
said that the controversy is closed.” This striking descent 
from Roux’s first figures is not, however, attributed by the 
writers wholly to improvements in the methods of serum 
therapy. As Roux pointed out in his first paper, the hygienic 
conditions in the two Paris hospitals from which these 
statistics come were very bad. These conditions have since 38 
then been greatly improved, and this reform has been espec- 
ially manifest in the reduction of the deaths from broncho- 
pneumonia. 
A most important classification of diphtheria for estimating 
the curative value of antitoxic serum is that into cases without 
and with laryngeal stenosis, and especially when such degrees 
of stenosis are considered as require operative interference by 
tracheotomy or intubation. I have therefore prepared the 
following table (Table II) which gives the results of antitoxin 
treatment in operated and not operated cases of diphtheria. 
This of course is not equivalent to a division into anginas and 
croup, as many cases of croup are included in the non-opera ted 
cases, but I have desired to submit the new method of treat- 
ment to the most severe test. No one can claim that laryn- 
geal diphtheria requiring intubation or tracheotomy is any- 
thing but a severe disease. If the benefits of antitoxin are 
unmistakably manifested in these operated cases of croup, 
then the test is an experimentum crucis and puts an end to the 
objections of those who assert that the apparently favorable 
results of serum therapy in diphtheria are attributable mainly 
to the large proportion of mild cases treated. 
The same reports with a few additional ones, for which 
references are given, have been used for Table II as for Table 
I, but many of the reports in Table I were not available for 
this table, as the writers did not always present their results 
in a form which fitted into the classification adopted. The 
table gives for each report the total number and fatality of 
cases treated, as in Table I, the number and fatality of cases 
not operated on (including cases of croup), the number and 
fatality of cases operated on, “ T ” signifying tracheotomy, 
“ I ” intubation, “ I and T ” signifying intubation followed 
by tracheotomy, and, so far as reported, the previous or simul- 
taneous percentage of fatality from operation in cases not 
treated with antitoxin. In the final column are pertinent 
statements concerning cases in the series. Some reports are 
inserted which do not give the number of cases under the 
different headings. These, of course, cannot be used in the 
summary giving the totals. TABLE II.—FATALITY IN OPERATED AND NON-OPERATED CASES OF DIPHTHERIA TREATED WITH ANTITOXIN. 
T Signifies Tracheotomy, I Intubation, I and T Intubation followed by Tracheotomy. 
Reporter. 
Total Cases. 
Deaths. 
Not Oper- 
ated Cases. 
Deaths. 
Operated 
Cases. 
Deaths. 
Previous 
Fatality from 
Remarks. 
Operation. 
Roux, Martin and 
Chaillou. 
300 
78 <26 per cent.) 
179 
22 (12.8 per cent.) 
121 T 
56 (46 per cent.) 
67 per cent. 
Deducting those which died in less than 24 hours after 
admission to the hospital, there remain 107 tracheoto- 
Kossel. 
117 
13 (11.1 per cent.) 
94 
1 (1.1 per cent.) 
23 T 
12 (52.2 per cent.) 
mies with 42 deaths (39.2 percent.). 
In all of the fatal operations it was necessary to perform 
KOrte. 
121 
40 (33.1 per cent.) 
79 
18 (22.8 per cent.) 
42 T 
22 (52.4 per cent.) 
77.5 per cent. 
tracheotomy within 12 hours after admission. 
Of 8 tracheotomized cases under 2 years of age, 5 died 
! 
(62.5 per cent.). The previous fatality for tracheotomy 
Sonnenburg. 
95 
16 (16.8 per cent.) 
61 
8 (13.1 per cent.) 
34 T 
8 (23.5 per cent.) 
38 per cent. 
under 2 years was 90.7 per cent. 
The fatality of 38 per cent, in the table is that of 47 
tracheotomies performed during the period in which 
the supply of antitoxin temporarily was exhausted and 
Baginsky. 
525 
83 (15.6 per cent.) 
430 
47 (10.9 per cent.) 
95 (53 T) 
36 (34 T) (37.8 p. c.) 
64.4 per cent. T 
could not be at once replaced. 
During two months in which the supply of serum failed 
(54 i ) 
(21) 
44.8 per cent. 1 
the general fatality rose to 48.4 per cent, and that from 
(In 12 cases 
(Ur the secondary 
tracheotomy and intubation to 62.2 per cent. The 
tracheotomy 
tracheotomies 12 
fatality at once fell upon re-introduction of the anti- 
followed in- 
died.) (37.9 per cent.) 
toxin treatment. 
v. Widerhofer. 
300 
tubation.) 
71 (23.7 per cent.) 
192 
20 (10.4 per cent.) 
108 T or I 
51 (47.2 per cent.) 
22 cases of laryngeal diphtheria (croup) recovered with- 
v. Ranke. 
Stintzing. 
96 
59 
19 (19.7 per cent.) 
12 (20.1 per cent.) 
54 
43 
6 (11.1 per cent.) 
5 (11.6 per cent.) 
42 / 
16 T 
13 (30.9 per cent.) 
7 (44 per cent.) 
61-75 per cent. 
49 per cent. 
out operation. 
21 cases of croup recovered without operation. 
SchrOder. 
63 
8 (12.7 per cent.) 
32 
5 (15.6 per cent.) 
31 T 
3 (9.67 per cent.) 
Vierordt. 
63 
16 (25 per cent.) 
48 
9 (18.75 per cent.) 
16 T 
7 (46 per cent.) 
Deducting 8 cases in a hopeless condition on admission, 
there was only one death among the non-trache- 
Rumpf. 
22 
2 (8 per cent.) 
15 
0 
7 T 
2 (28.6 per cent.) 
otomized cases, a death rate of only 2.7 per cent. 
Lebreton and Mag- 
delaine. 
258 
31 (12 per cent.) 
183 
8 (4.4 per cent.) 
75 (24 T) 
23 (9 T) (37.5 p. c.) 
73.2 per cent. T 
Of the intubated cases 7 died in less than 24 hours after 
(51 /) 
(14 /) (27.5 p. c.) 
admission. Deducting these, the fatality from intu- 
Legendre. , 
Moizard & Perregaux (_ 
398 ' 
52 (13.1 per cent.) 
316 
23 (7.6 per cent.) 
(51 T) 
82 (21 /) 
(20 T) (39.2 p. c.) 
29(5 /) (23.8 p. c.) 
86 per cent. T 
bation was only 18.1 per cent. 
Deducting 15 deaths in less than 24 hours after admission, 
the total fatality is reduced to 9.7 per cent. These 
(101 & T) 
(4 I & T) (40 p. c.) 
cases are those treated in the HSpital Trousseau in 
Ganghofner. 
110 
14 (12.7 per cent.) 
66 
8 (12 per cent.) 
44/or T 
(35.4 per cent.) 
6 (13.6 per cent.) 
59.8-78 per cent. 
Paris from the middle of Sept, to Dec. 25, 1894. 
12 laryngeal diphtherias recovered without operation. 
During period in which the supply of serum failed the 
general fatality rose to 53.2 per cent, and that of oper- 
Soltmann. 
89 
13 (14.6 per cent.) 
48 
2 (4.2 per cent.) 
41 / 
11 (27 per cent.) 
68 per cent. 
ated cases to 68.9 per cent. 
Bokai. 
120 
31 (25.5 per cent.) 
71 
10 (14 per cent.) 
49 / 
21 (43 per cent.) 
70 per cent. 
In 14 cases presenting symptoms of moderate laryngeal 
M 
stenosis upon admission, these symptoms disappeared 
Heim. 
27 
6 (22 per cent.) 
21 
4 (19 per cent ) 
6 (2 T) 
(1 T) (50 percent.) 
• • • • 
after injection of antitoxin without operation. 
During period when supply of serum failed the general 
f 
(4 I) 
(11) (25 per cent.) 
fatality rose to 65.6 per cent. Heim treated altogether 
t 
33.3 per cent. 
48 cases in two groups, but of his 2d group of 21 cases 
12 were still under treatment at date of report and 
BOrger. 
30 
2 (6.66 per cent.) 
25 
1 (4 per cent.) 
5 T 
1 (20 per cent.) 
these are not included in my table. 
313 cases of diphtheria treated in the same hospital 
(Greifswald) from Oct., 1893, to Sept., 1894, gave a fa- 
Moeller. 
76 
27 (35.5 per cent.) 
28 
8 (28.6 per cent.) 
48 T 
19 (39.6 per cent.) 
55.6 per cent. 
tality of only 14.5 per cent. 
Kuntzen. 
25 
3 (12 per cent.) 
20 
1 (5 per cent.) 
5 T 
(40 per cent.) 
Charon. 
13 
4 (30.8 per cent.) 
2 
0 
11 T 
4 (36.4 per cent.) 
Washbourn, Goodall 
and Card. 
Herringham. 
72 
14 (19.4 per cent.) 
63 
11 (17.5 per cent.) 
9 T 
3 (33.3 per cent.) 
91.7 per cent. 
In 13 previous series of tracheotomies in each group the 
18 
3 (16.7 per cent.) 
7 
0 
11 (10 T) 
(ID 
46 (2 T) 
3 (30 per cent. T) 
65 per cent. T 
average of deaths numbered 8K. 
The 7 unoperated cases were mild. 
Germonig— 
224 
42 (18.7 per cent.) 
178 
18 (10 per cent.) 
241 (54.5 per cent.) 
85.5 per cent. T 
During a period of exhaustion of the supply of serum 
the general fatality rose to 50 per cent. These data are 
(Civic of 
(44 /) 
75 per cent. / 
from abstracts in the British Med. Jour., Feb. 2,1895, 
and the Deutsche Med. Wochenschrift, 1894, No. 52. 
The general serum fatality is variously given as 18.7 
Witthauer. 
36 
5 (14 per cent.) 
16 
1 (6.25 per cent.) 
20 T 
4 (20 per cent.) 
25 per cent. 
per cent., 20.3 per cent, and 22 per cent. 
Only one adult in this series. 
Rapmund. 
100 
7 (7 per cent.) 
98 
7 (7.1 per cent.) 
2 T 
0 
Collection of cases treated in private practice by several 
Heidenhain. 
24 
3 (12.5 per cent.) 
22 
2 (9.1 per cent.) 
2 T 
1 (50 per cent.) 
physicians in Minden. 
Risel. 
114 
9 (7 9 per cent.) 
95 
5 (5.3 per cent.) 
19 T 
4 (21 per cent.) 
v. Muralt. 
58 
2 (3.4 per cent.) 
38 
0 
(14/) 
After intubation (14 cases) no death, 
20 (2 T) 
2 (10 per cent.) 
.... 
after intubation and tracheotomy 1 death, 
D’Espine. 
60 
6 (10 per cent.) 
49 
2 (8.2 per cent.) 
(4 I and T) 
11 T 
4 (36.4 per cent.) 
after tracheotomy 1 death. 
10 cases of croup recovered without operation. 
v. Engel. 
39 
10 (25.5 per cent.) 
30 
7 (23.3 per cent.) 
9 T 
3 (33.3 per cent.) 
85 per cent. 
The fatality of 42 unoperated cases not treated with 
antitoxin was 33.4 per cent., that of 20 tracheotomized 
Codd & Whitehouse. 
ii 
4 (36.4 per cent.) 
7 
2 (28.6 per cent.) 
4(3 T) 
0/ 
cases not so treated was 85 per cent. 
(11) 
2 T (66.6 per cent.) 
50 per cent, of oper- 
3 of the 4 deaths were from croup. 
Winkfield. 
ated cases. 
22 
4 (18.2 per cent.) 
19 
2 (10.5 per cent.) 
3 T 
2(66.6 per cent.) 
85.7 per cent. 
Sigel. 
100 
12 (12 per cent.) 
No. of cases 
No. of cases 
20.3 per cent. T 
63.3 per cent. 
The fatality of tracheotomized cases in 1894 up to the 
Galatti. 
No. of cases 
not given. 
No. of cases 
not given. 
15/ 
70 per cent. 
date of beginning treatment (Oct. 4) was 70 per cent., 
after treatment with antitoxin 20.3 per cent. 
6 (40 per cent.) 
52.6 per cent. / 
Quoted in Le Bulletin Medical, 1895, p. 364. 
Kraske. 
not given. 
not given. 
Not given. 
.... 
16 T 
5 (31.25 per cent.) 
28 per cent (in 
Says that he has tracheotomized 12 cases not treated with 
Bonain. 
Not given. 
5/ 
12 cases). 
antitoxin, with a fatality of only 28 per cent. Abst. in 
Miinchener Med. Wochenschrift, 1894, p. 1062. 
12 (23 per cent.) 
"30 
4 (13.3 per cent.) 
2 (40 per cent.) 
60.5 per cent. 
Le Bulletin Medical, April 17,1895. 
Van Nes. 
52 
22 (15 T) 
8 (36 per cent. / & T) 
48-73 per cent. 
During ten preceding yeais the lowest fatality of op- 
(7 /) 
erated cases was 52 per cent., of not operated 16 per 
cent. In 3 groups of 92 cases each preceding the intro- 
duction of antitoxin treatment, the lowest fatality of 
operated cases was 48 per cent., of not operated cases 
19 per cent. 
Of 17 tracheotomized cases less than 2 years old 9 died, 
Leichtenstern and 
123 
[25 (20.3 per cent.) 
86 
9 (10.4 per cent.) 
37 T 
16 (43.2 per cent.) 
64 per cent. 
Wendelstadt. 
between 2 and 4 years 4 out of 11 died ; of 13 unoperated 
cases under 2 years 2 died ; of 21 between 2 and 4 years 
2 died. 
Abst. Miinchener Med. Wochenschrift, June 11, 1895. 
Kobts. 
39 
Not given. 
Not given. 
9.1 per cent. 
Not given. 
30.77 per cent. T 
Timmer. 
31 
6 (19.4 per cent.) 
25 
5 (20 per cent.) 
6 Tor / 
1 
(16.6 per cent.) 
Mason. 
306 
81 (26.4 per cent.) 
275 
62 (22.5 per cent.) 
31 
19 
(61 per cent.) 
90 per cent.? 
The statement is that the mortality from tracheotomy 
Kurth. 
97 
10 (10.3 per cent.) 
82 
7 (8.5 per cent.) 
15 T 
was reduced by the antitoxin about one-third. 
3 
(20 per cent.) 
Of 32 cases of laryngeal diphtheria not operated upon, 30 
recovered. 
Total of 38 Reports. 
4294 
784 (18.3 per cent.) 
3127 
350 (11.2 percent.) 
1167 
434 
(37.2 per cent.) 
Of the total cases 27.2 per cent, were operated upon by 
tracheotomy or intubation. 
In 41 reports there are 648 tracheotomies with 258 deaths, a fatality ofZS.Sper cent.; 342 intubations with 99 deaths, a fatality of 28.9 per cent., and 26 intubations followed by 
tracheotomy with 14 deaths, a fatality of 53.8 per cent. 
There are 211 operated cases in which it is not stated how many are tracheotomy or intubation. These gave a fatality o/40.2 per cent. They were probably for the most part 
tracheotomies. 
The reports giving the previous or simultaneous fatality from tracheotomy contain 510 cases of tracheotomy with 217 deaths or 42.5 per cent. If the fatality of these cases be reck- 
oned on the basis of the preceding or simultaneous fatality, selecting the lowest figures given, there would have been 329 deaths or 64.5 per cent. There was, therefore, an apparent 
reduction in fatality of 34.1 by the serum treatment. 
Making a similar estimate on the basis of previous fatality from intubation, there were 250 intubations with 79 deaths or 31.6 per cent, instead of 156 deaths or 62.4 per cent. 
There was, therefore, an apparent reduction in the fatality of intubated cases of 49.5 per cent, as the result of serum treatment. 
The fatality of 3127 non-operated cases was only 11.4 per cent. 39 
Of the 4294 cases in Table II, 27.2 per cent, required 
tracheotomy or intubation. There were, however, many more 
cases of laryngeal diphtheria in this group than the ratio of 
operative cases would indicate, for it is the testimony of the 
great majority of the observers that the stenotic symptoms of 
laryngo-tracheal diphtheria are relieved without the necessity 
of operation in a much larger proportion of the cases treated 
with antitoxin than by any other method of treatment. As 
is well known, recovery without intubation or tracheotomy 
from descending laryngo-tracheal diphtheria, especially in 
children, is exceptional under all other methods of treatment, 
and the greater relative frequency with which such recovery 
occurs under serum treatment is a strong proof of the efficacy 
of antitoxin. 
Of Kossel’s 44 cases of laryngeal diphtheria treated with 
antitoxin, 21 (47.7 per cent.) recovered without operation; of 
von Widerhofer’s 130 stenotic cases treated with serum 22 
(16.9 per cent.) recovered without operation; of von Ranke’s 
63 cases, 21 (33.3 per cent.); of Vierordt’s 24 cases, 9 (37.5 
per cent.); of Ganghofner’s 56 cases, 12 (21.4 per cent.); of 
Bokai’s 63 cases, 14 (22.2 per cent.); of d’Espine’s 21 cases, 
10 (47.6 per cent.). Von Ranke says that before the use of 
serum at most 5 per cent, of his cases of laryngeal stenosis 
escaped operation, whereas now 33 per cent, escape. Of Gang- 
hofner’s stenotic cases formerly 12 per cent, escaped opera- 
tion, whereas now 21 per cent, escape. The experience of 
Heubner and many others is similar. 
In this respect, as in so many others, the results in the 
Paris hospitals have been most favorable. Of Moizard and 
Perregaux’s 145 cases of croup, 90 (62.1 per cent.) recovered 
without intubation or tracheotomy. Roux, Martin and 
Chaillou say, “ Of 169 children, admitted to the service for 
diphtheric angina, 56 presented laryngeal symptoms; 31 had 
hoarse voice, and in 25 the voice was so far extinguished and 
the dyspnoea (tirage) so marked that one might believe that 
the latter patients should be operated on. Under the influence 
of the serum (and in these cases one should not fear to make 
an injection every twelve hours), the dyspnoea diminished, 
then occurred only paroxysmally, the child coughed up false 
membranes, and at the end of two or three days the respira- 40 
tion became normal, to the great astonishment of the interns 
and personnel of the pavilion who, with their large experience 
of children affected with croup, indeed thought that operation 
could not be avoided. To-day in the presence of a child with 
dyspnoea it is not necessary to press for operation. One can 
inject the serum and wait as long as possible. Since the 
introduction of the serum the number of tracheotomies in the 
pavilion has diminished.” 
Out of his large experience Baginsky expresses himself in 
these vigorous words: “ Here again the observation of the 
individual cases of laryngeal stenosis, and more especially of 
those which do not come to the point of operation, speak to 
me more forcibly than the statistical figures. The surprising 
regression of the laryngo-stenotic respiratory phenomena, the 
freedom of breathing, the disappearance of the hoarse voice 
and the croupy cough, the euphoria of the children, the 
change in their general condition so that two days after the 
injection they are sitting up in bed, playing and contented 
and observant of their surroundings; all of these things 
produce in him who has had before his eyes for years the 
hopeless picture of continually progressing laryngeal stenosis, 
in very truth ineffaceable impressions.” 
Experience based upon such a large number of cases and 
careful clinical observation must be regarded as representing 
the norm. That there may be deviations from this norm, 
even in a fair number of cases, seems to be illustrated by the 
experience of Leichtenstern and Wendelstadt, who in 123 
cases of diphtheria, with 37 tracheotomies, were not able to 
note any material reduction in the proportion of cases requir- 
ing tracheotomy as compared with former series of cases. 
Their observations were uncontrolled by bacteriological diag- 
noses. 
Another point to be considered in this connection is of 
capital importance as an indication of the value of serum 
treatment. Cases which are free from symptoms of laryngeal 
involvement at the time of injection of the serum do not 
develop such symptoms later, or do so only very exceptionally, 
unless evidences of such involvement appear within twenty- 
four hours after the injection. 41 
Regarding' neither this nor any other point is there entire 
unanimity of opinion in the various reports, nor is such to be 
expected from observers of limited numbers of cases with 
unequal distribution in the various groups of mild cases, of 
early cases, of anginas, of croup, of pure diphtheria, of septic 
diphtheria, etc., to say nothing of the absence in some reports 
of any bacteriological control of the diagnosis and of treat- 
ment by insufficient doses or inferior quality of serum. I 
am only surprised that the conflicting statements are not 
more numerous.* But there are not many points concerning 
which there are so few differences of statement as concerning 
the efficacy of antitoxin in preventing descent of the diph- 
theritic process to the larynx and the trachea. Over and 
again one can read in the reports such statements as that in 
all of the patients who entered without laryngeal diphtheria, 
the larynx remained free, or that unless the symptoms of 
stenosis appeared within the first twenty-four hours after 
injection of the serum, they were not observed at all or only 
most exceptionally. Among the many vouchers for these 
statements may be cited Kossel, Roux, Baginsky, von Wider- 
hofer, Heubner, von Ranke, Yierordt, Ganghofner, Escherich, 
Bokai, Van Nes, Kurth. 
It is this power of antitoxin to check the spread of the 
diphtheritic process from the tonsils and pharynx into the 
larynx, and from the larynx into the bronchi, which has 
impressed many observers in favor of the new treatment more 
forcibly than any other feature of their experience with its 
action. Thus Vierordt observed that of 24 children with 
diphtheria who were admitted with unaffected larynx and 
treated with antitoxin, only one developed temporarily a 
hoarse cough on the third day. In all of the others the 
larynx remained free. Of 23 patients who were admitted 
with unaffected larynx not long before the introduction of 
the serum treatment, nine afterward developed croup. This 
is doubtless a somewhat unusual experience as regards the 
large proportion of cases of croup developing under previous 
methods of treatment. 
It follows from what has been said that the ratio of opera- 
tive cases in antitoxin statistics will in general be smaller 
than in statistics of cases of the same character treated by 42 
other methods. On the one hand there will be fewer laryn- 
geal stenoses developing after commencement of the treat- 
ment, and on the other hand a larger number of recoveries 
from laryngeal diphtheria without the necessity of operation. 
The following figures serve to illustrate this point. In the 
service from which the cases reported by Roux were derived, 
tracheotomy was performed before the serum period in 50 per 
cent, of the cases of diphtheria, after the introduction of serum 
in 40 per cent. The later Paris reports give a much greater 
reduction in the ratio of tracheotomies. Iii Baginsky’s ser- 
vice 43.9 per cent, of the cases were operated on before the use 
of serum, and 18.1 per cent, after its introduction; in von 
Ranke’s service the corresponding figures are 57 per cent, 
before and 43.5 per cent, after; in Bokai’s 65.6 per cent, before 
and 40.8 per cent, after. As already mentioned, Leichten- 
stern’s figures, 32 per cent, before and 30 per cent, after 
serum, are exceptional. 
It is furthermore to be considered that in view of the power 
of antitoxin to abate beginning and moderate symptoms of 
stenosis, operation will be delayed rather than hastened, and, 
when performed, the indications for it will generally be 
urgent. For manifest reasons, most of the operations will 
fall within a period not remote from the time of injection of 
the serum. Of the 121 tracheotomies in the report of Roux, 
Martin and Chaillou, 102 were performed either before the 
first injection of antitoxin or within 12 hours afterward; 14 
between the 12th and the 36th hour after inception of the 
serum treatment, and only 5 later than 36 hours after the 
injection of the serum. Of the 23 tracheotomies with 12 
deaths reported by Kossel, the operation was performed within 
the first twelve hours in all of the fatal cases, and of the 11 
successful cases it was performed in 9 on the day of admission 
to the hospital, in 1 on the second day and in 1 on the 
following day. Kossel refers the increase in the stenotic 
symptoms after injection of the serum in the two last cases 
to the separation of the false membranes, a point to which 
others have also called attention as an effect of antitoxin and 
which is to be borne in mind in cases of croup treated by 
antitoxin. 43 
Turning now to the results of tracheotomy and intubation 
in cases treated with antitoxin, we find in Table II that in 41 
reports there were 648 tracheotomies with 258 deaths, a 
fatality of 39.8 per cent., and 342 intubations with 99 deaths, 
a fatality of 28.9 per cent., and 26 intubations followed by 
tracheotomy with 14 deaths, a fatality of 53.8 per cent. These 
are not unheard of fatalities from these operations, but they 
are so low as to indicate decidedly remedial action of anti- 
toxin. 
The percentage of fatality from tracheotomy in diphtheria 
given by Monti from a total of 12,736 cases up to 1887 is 73.3. 
The percentage given by V. Hirsch in 1654 tracheotomies in 
diphtheria, in von Bergmann’s clinic in Berlin during the 
last ten years and seven months (up to July 31, 1894), is 68.7, 
the fatality during the first four years of this period being 
70.5 per cent., and during the last four years 63.8 per cent. 
The fatality during the first year of life was 98.8 per cent, 
and sank for each year to the ninth, when it was 41.7, and 
after the tenth year it rose again. 
More proper, however, than comparison with these latter 
percentages is comparison with the percentages of fatality in 
the same hospital or place from which the respective groups 
of cases are reported. It will be observed that with one 
exception in the table the percentage of deaths following 
operation in cases treated by antitoxin is lower, and generally 
very much lower, than the previous or simultaneous fatality. 
Kraske’s exceptional series is of so few cases (only 5 with and 
12 without serum) as to be without any significance. The 
lowest fatality thus far reported in a series is 3 deaths in 31 
tracheotomies with serum treatment, or a fatality of only 9.67 
per cent. This is reported by Schroeder from the hospital in 
Altona. 
If for each group of cases we estimate the number of deaths 
which would have occurred in the tracheotomized cases treated 
with serum on the assumption that the previous or simul- 
taneous fatality in cases not treated with serum had obtained, 
we obtain the following result: The actual percentage of 
deaths in 510 cases treated with serum was 
42.5. The percentage of fatality in these cases estimated on 
the basis of previous or simultaneous fatality in the same 44 
hospitals would be 64.5. There was therefore an apparent 
reduction in fatality by the serum treatment of 34.1 per cent. 
This difference between actual and estimated fatality is 
greater than is observed in any ordinary experience of varia- 
tions in fatality during a series of years in the same hospital 
from tracheotomy in diphtheria. 
I confess to some surprise that the analysis of the tracheoto- 
mized cases treated by serum should have yielded results so 
strikingly favorable to antitoxin treatment. When one con- 
siders that the benefits of serum treatment are most strikingly 
apparent when the treatment is begun early in the disease and 
become more and more doubtful after the third day, it would 
not have been a convincing argument against the treatment if 
these benefits were not conspicuously manifest in cases of 
diphtheria requiring tracheotomy, for, as has been explained, 
the great majority of these tracbeotomized cases are already 
the subject of advanced laryngeal stenosis when the antitoxin 
is first injected. There are, however, not a few cases which 
begin apparently as laryngeal diphtheria (croup d’emblee), or 
in which the involvement of the larynx occurs within twenty- 
four or forty-eight hours after the onset of the attack. That 
careful observation would reveal in many of these apparently 
primary or early laryngeal diphtherias a latent or slightly 
manifested diphtheric angina I believe to be true. 
It is interesting to note that in several reports the benefit of 
serum treatment has been much more evident in the operated 
cases than in those not operated on, although this is not the 
rule. Indeed Leichtenstern and Wendelstadt find in their 
series of 123 cases that the difference in favor of the serum in 
their non-operated cases was so small as to be without signifi- 
cance, whereas there was a difference in favor of the serum of 
20.8 per cent, in their tracheotomized cases with and without 
serum treatment. They attribute, therefore, the entire benefit 
of the serum in their experience to its action in tracheoto- 
mized cases. Their experience, however, is exceptional, 
although in a measure approached by that of Ganghofner and 
of Van Nes. On the other hand, in Vierordt’s experience the 
entire benefit of antitoxin seegied to be in the non-operated 
cases. As has been repeatedly explained, such diversities of 
experience with limited numbers of cases is to be expected, 45 
and the norm can be established only by observations of large 
numbers of cases in different places and at different times. 
This norm is that both operated and not operated cases are 
benefited by antitoxin, and that the difference in each class 
between serum fatality and fatality from other methods of 
treatment is a large one. 
The fatality of intubated cases in Table II, treated ’with 
antitoxin, is 28.9 per cent., which is 10.9 per cent, less than the 
fatality of tracheotomized cases. Before the introduction of 
the serum treatment a collective investigation was set on foot 
by the German Gesellschaft fur Kinderlieilkunde to determine 
the average fatality following intubation. In 1893 von Ranke 
reported to the Society that 1445 cases of diphtheria with 
laryngeal stenosis treated by intubation gave a fatality of 62.5 
per cent. This result was interpreted in favor of intubation 
as opposed to tracheotomy. There is a difference of 33.6 per 
cent, between this percentage and that obtained from our 342 
intubated cases treated with antitoxin. This difference is so 
great that, after making all possible allowance for differences 
in the series of cases entering into the two groups of statistics, 
it seems impossible to explain it otherwise than as a powerful 
additional support of the arguments already presented in sup- 
port of the claims of antitoxin. Here certainly the objection 
that the cases treated by antitoxin were light ones cannot be 
made. 
Table II enables us to compare the fatality of 250 intubated 
cases treated with antitoxin with the fatality estimated on the 
assumption that the previous or simultaneous fatality from 
intubation in the same hospital had obtained in the several 
groups. By this calculation we find the actual fatality to be 
31.6 per cent., and the estimated fatality 62.4 per cent. In 
other words, there was an apparent reduction in the fatality 
of intubated cases of 49.5 per cent, as the result of the serum 
treatment. 
However distrustful one may be of statistical evidence in 
therapeutics—and previous experience justifies much distrust 
—I fail to see on what credible assumption this striking- 
reduction of fatality can be explained otherwise than as 
demonstrative of the specific curative power of antitoxin in 
diphtheria. 46 
Lamentable for the victims but adapted to convince the 
skeptical were the experiences of Baginsky and Ganghofner 
during the periods of failure in the supply of serum. During 
the enforced two mouths’ interruption of the serum treatment 
(August and September) in Baginsky’s service there were 116 
cases of laryngo-stenosis with a fatality of 62.2 per cent., as 
opposed to a fatality of 37.8 per cent, in the serum periods 
which preceded and followed the pause. The percentage of 
operations rose to 55.2 as opposed to 18.1 per cent, during the 
periods of serum treatment, and this without any change in 
the .general character of the cases admitted. During the 
serum periods there were more intubations than tracheoto- 
mies, whereas during the pause there were 45 tracheotomies 
and 19 intubations, 13 of the latter requiring secondary 
tracheotomy. In Ganghofner’s service the fatality of the 
operated cases rose from 13.6 per cent, to 68.9 per cent, 
during the interruption in the supply of serum. 
There remain two points to be touched upon before dismiss- 
ing the laryngeal stenoses. These are the substitution of 
intubation for tracheotomy in a larger and larger proportion 
of the laryngeal diphtherias requiring operative interference 
and treated by the serum, and the shortening of the period 
during which the tube or the tracheal canula is required to 
be kept in the air passage. 
An agent which would arrest the progressive descent of 
the diphtheritic process from the larynx into the bronchi and 
hasten the disappearance of the obstructive exudate is just 
what was needed to make intubation the ideal operation for 
the relief of the great majority of cases of croup requiring 
operative interference. Such an agent we now possess in anti- 
toxin for a large group of cases, and we are not surprised, 
therefore, to find that the employment of intubation, as a 
substitute for tracheotomy, has been greatly extended by the 
introduction of serum therapy. 
Several writers give figures showing that serum therapy 
materially hastens the time when extubation or removal of 
the tracheal canula is permissible, but I have not attempted 
to collect these figures. 
Of the 3127 not operated cases, including as already stated 
many cases of croup, 350 died, giving a fatality of 11.2 per 47 
cent. In Y. Hirsch’s statistics of diphtheria from yon Berg- 
mann’s clinic for ten years the average fatality of not operated 
cases (1004) was 26 per cent., varying only from 25.9 per cent, 
during the first four years of the period to 27.3 per cent, during 
the last four years. There is, however, no general standard of 
fatality for cases of diphtheria not operated on. The varia- 
tions are within very wide limits, as might be expected. Only a 
comparatively small number of the reports give separately the 
previous or simultaneous fatality of non-operated cases not 
treated with serum. I find in the reports the following 
data on this point. In Roux, Martin and Chaillou’s report 
the previous fatality of non-operated cases averaged 33.9 per 
cent., the minimum being 32.1 per cent., and the maximum 
47.3 per cent., as opposed to 12.8 per cent, under the serum 
treatment; in Baginsky’s report the corresponding figures are 
31.6 per cent, versus 10.9 per cent.; in Bokai’s 34.5 per cent, 
versus 14 per cent.; in Ganghofner’s 15.8 per cent, (the lowest 
in a series of years) versus 12 per cent.; in Van Ues 33 per 
cent., the average of ten years, with a minimum of 16 per 
cent, and a maximum of 41 per cent, versus 13.3 per cent.; in 
Leichtenstern and Wendelstadt’s 15 per cent, versus 10.4 
per cent. 
Age is a factor of such prime importance in the prognosis 
of diphtheria that I have prepared the following table (Table 
III), in which the cases treated with serum collected from 
twenty-five reports are classified according to age. Unfortun- 
ately there is very little uniformity of system in tfie different 
reports in giving the results according to the ages of the 
patients, many of the reports simply stating the number of 
adults or the maximum age of the children or the number of 
cases under a certain age or the number between arbitrarily 
selected limits of age, etc., so that many of the reports were 
not used for the following table. In each space in the table 
the upper number is the total number of the cases belonging 
to the heading, and the lower number is the number of deaths 
among these cases. As the cases in Table III were not classified according to the ages by a uniform plan in the different reports they cannot 
all be summarized in a single table, but the chief results can be presented as follows: 
14 Reports. 
Total. 
0-2 yrs. 
2-4 yrs. 
4-6 yrs. 
6-8 yrs. 
8-10 yrs. 
10-12 yrs. 
12-15 yrs. 
Over 15 yrs. 
Undetermined. 
Cases 
1234 
187 
327 
297 
176 
114 
63 
32 
24 
(14) 
Deaths 
215 
60 
70 
48 
19 
8 
4 
5 
1 
(0) 
Percentages 
17.4 
32.1 
21.4 
16.2 
10.8 
7 
0.63 
1.6 
4.1 
(0) 
In the following table the cases under 4 years are from 20 reports containing 1630 cases (fatality 17.6$) and those over 4 
are from 17 reports containing 1451 cases (fatality 17.4$). 
0-2 years. 
2-4 years. 
4-10 years. 
10-15 years. 
Over 15 years. 
Cases 
263 
411 
681 
129 
40 
Deaths 
86 
84 
95 
11 
2 
Percentages 
32.7 
20.4 
13.9 
8.5 
5 
18 Reports. 
Total. 
Under 1 year. 
1-2 years. 
2-3 years. 
3-4 years. 
4-5 years. 
Over 5 years. 
Undetermined. 
Cases 
982 
34 
112 
118 
116 
140 
452 
(10) 
Deaths 
179 
10 
37 
36 
17 
31 
42 
(0) 
Percentages 
18 2 
47.1 
33 
30.5 
14.7 
22.1 
9.3 
(0) 
The table furthermore shows under one year 35 cases with 16 deaths or 45.7$/ under 2 years 291 cases with 97 deaths, or 33.3$/ under 3 years 
304 cases with 93 deaths or 30.6$, and under 4 years 692 cases with ]22 deaths or 17.6$ (each of these four groups of cases being from a total number 
of cases in the first group of 1080 cases, in the second group of 1914 cases, in the third group of 1140 cases and in the fourth group of 1665 cases, the 
average fatality for the whole number of cases being 17.3$.) 
The following table gives the results for each year np to 5 and over 5 years. TABLE III. 
AGES OF PATIENTS TREATED WITH ANTITOXIN. 
In each space the upper number is that of the cases and the lower that of the deaths, 
REPORTER. 
Total Cases. 
Under 
1 yr. 
1-2 
years. 
2-3 
years. 
3-4 
years. 
4-5 
years. 
5-6 
years. 
6-7 
years. 
7-8 
years. 
8-9 
years. 
9-10 
years. 
10-12 
years. 
12-15 
years. 
Over 
15 yrs. 
REMARKS. 
Kossel 
117 
3 
4 
18 
14 
20 
10 
n 
10 
7 
7 
12 
i 
13 (11 per ct.) 
1 
0 
2 
3 
3 
2 
0 
0 
1 
0 
0 
i 
KOrte 
121 
Of 15 infants under 2 years 8 recovered and 7 died. 
40 (33.1 per ct.) 
Baginsky 
Under 2 years. 
2-4 years. 
4-6 years. 
6-8 years. 
8-10 years. 
Of 4 cured cases the ages are not stated. 
525 
87 
146 
116 
79 
58 
20 
15 
83 (15.6 per ct.) 
22 
25 
20 
9 
3 
2 
2 
v. Widerhofer 
100 
8 
24 
20 
14 
16 
5 
7 
2 
1 
2 
1 
Deutsche Med. Wochenschrift, Jan. 10,1895. 
24 (24 per ct.) 
5 
9 
7 
0 
3 
0 
0 
(1 
0 
0 
0 
Schroeder 
Under 2 years. 
2-10 years. 
Over 10 years. 
63 
6 
52 
5 
8 (12.7 per ct) 
1 
Bokai 
120 
1 
22 
22 
24 
18 
13 
6 
7 
2 
1 
4 
69 cases were under 3 years of age. 
31 (26.5 per ct.) 
1 
10 
7 
4 
3 
1 
1 
3 
0 
0 
1 
Heim 
27 
6 
4 
2 
6 
3 
2 
2 
3 
6 (22 per ct.) 
1 
1 
0 
2 
0 
0 
2 
0 
Unterholzner 
8-13 years. 
Of 21 under 4 years of age 13 recovered. 
31 * 
2 
9 
7 
3 
2 
1 
2 
2 
3 
8 (25.8 perct.) 
1 
5 
2 
0 
0 
0 
0 
0 
0 
, 
BOrger 
30 
1 
2 
1 
1 
3 
2 
2 
2 
3 
2 
11 
16 cases were 10 years old or over. 
2 (6.6 per ct.) 
0 
0 
0 
1 
0 
0 
0 
0 
1 
0 
0 
Hager 
Over 6 years. 
25 
1 
1 
1 
3 
6 
4 
9 
1 (4 per ct.) 
0 
0 
0 
0 
1 
0 
Seiz 
27 
1 
4 
2 
6 
4 
4 
1 
2 
3 
Private practice. 
1 (3.6 per ct.) 
0 
0 
1 
0 
0 
0 
0 
0 
0 
Seitz 
Under 4 years. 
4-6 years. 
7-15 years. 
One fatal case was 13 months, the other 18 months 
35 
18 
10 
7 
old; both with laryngeal diphtheria. 
2 (5.7 per ct.) 
2 
Reported in British Medical 
5-10 years. 
10-15 years. 
16 
Journal, Feb. 2,1895. 
95 
4 
9 
6 
12 
10 
23 
15 
1 
22 (23.2 per ct.) 
2 
3 
1 
1 
3 
9 
2 
Goodall 
72 
4 
10 
7 
9 
10 
22 
10 
British Med. Jour., Jan. 12, 1895. These are the ages 
14(19.4 per ct.) 
1 
2 
1 
3 
5 
2 
0 
of the cases previously reported by Washbourn, 
Goodall and Card. 
Hall 
11 
1 
1 
1 
1 
2 
5 
3 (27.3 per ct.) 
0 
1 
1 
0 
i 
0 
Epidemic in Trieste. 
105 
9 
12 
12 
15 
16 
13 
7 
6 
4 
2 
6 
3 
Hospital Cases. 
27 (25.7 per ct.) 
5 
5 
6 
3 
5 
1 
0 
1 
0 
1 
0 
0 
Epidemic in Trieste. 
72 
2 
3 
8 
H 
12 
5 
1 
7 
2 
3 
7 
1 
10 cured cases are not accounted for in this series. 
Cases in private practice. 
5 (6.9 per ct.) 
0 
2 
1 
i 
0 
0 
0 
1 
0 
0 
0 
0 
Blumenfeld 
10-13 years. 
Private practice. 
50 
6 
5 
5 
5 
14 
5 
9 
1 
2 (40 per ct. 
1 
0 
0 
0 
0 
1 
0 
0 
Schaewen 
15 
1 
1 
2 
3 
3 
3 
2 
1 
Private practice. 
0 
0 
0 
0 
0 
0 
0 
0 
0 
Rapmund 
3-5 years. 
• 
5-10 years. 
10-20 yrs. Over 20 yrs. 
Private practice. 14 cases were under treatment at 
100 
1 
6 
12 
22 
43 
15 
1 
time of report. 
7 (7 per ct.) 
0 
1 
2 
1 
3 
0 
0 
v. Muralt 
1-3 years. 
Over 3 years. 
The latter fatal case was of a tracheotomized child 
58 
21 
3' 
4 years old admitted on the 6th day of the disease; 
2 (3.4 per ct.) 
1 
] 
the former case was of an infant two years old 
with syphilis, measles and pneumonia. 
Codd and Whitehouse 
11 
1 
1 
1 
1 
1 
2 
2 
1 
1 
The fatal case in an adult was of a woman 39 years 
4 (36.4 per ct) 
0 
1 
1 
0 
0 
0 
0 
1 
1 
old admitted in a state of collapse on the 8th day 
of the disease and dying in 24 hours. 
Winkfield 
22 
1 
2 
4 
2 
3 
2 
3 
2 
1 
2 
4(18.2 perct.) 
1 
1 
0 
1 
1 
0 
0 
0 
0 
0 
Van Nes 
52 
1 
7 
4 
9 
8 
7 
6 
3 
2 
5 
12 (23 per ct.) 
0 
4 
2 
4 
0 
0 
1 
1 
0 
0 
Leichtenstern and Wendel- 
0-2 years. 
2-4 years. 
Over 4 years. 
Of 17 tracheotomized infants less than 2 years old, 
stadt. 
123 
30 
32 
61 
9 died: of 13 unoperated patients less than 2 years 
25 (20.3 per ct.) 
11 
6 
old, 2 died; of 11 tracheotomized patients between 
2 and 4 years, 4 died; of 21 unoperated cases be- 
tween 2 and 4 years, 2 died. 49 
The most frequently quoted percentages of fatality in 
diphtheria according to the age are those of Herz, and are as 
follows: 
Under 1 year 80 per cent. 
I- years 45 “ 
3-5 “ 40 “ 
5-10 “ 17 “ 
Over 10 years 17 “ 
The percentages of fatality in Y. Hirsch’s statistics of 2658 
cases from the surgical clinic in Berlin for 10 years and 7 
months (ending July 31, 1894), according to age are: 
Under 1 year 88.3 per cent. 
1- years 82.5 “ 
3- “ 63.9 “ 
4- “ .56.6 “ 
5- “ 46.9 “ 
6- “ 43.7 “ 
7- “ 36.1 “ 
8- “ 28.1 “ 
9- “ 31.1 “ 
10-11 “ 21.2 “ 
II- “ 20.9 “ 
12- “ 18.5 “ 
13- “ 16.7 “ 
14- “ 15. “ 
15- “ 12.5 “ 
Adults (72 cases) 11.1 “ 
Baginsky gives the following percentages from his service 
in the Kaiser- und Kaiserin-Friedrich Children’s Hospital in 
Berlin as the mean of the four years 1890 to 1893 inclusive: 
Under 2 years 60.2 per cent. 
2- years 51.2 “ 
4-6 “ 38. “ 
6-8 “ 28.9 “ 
8-10 “ 24.5 “ 
10-12 “ 28.8 “ 
12-14 « 18.5 “ 50 
Baginsky’s results in Table III may be compared with this 
last list of percentages, otherwise I do not consider that these 
statistics of Herz, Hirsch and Baginsky furnish any certain 
standard of comparison for the percentages of fatality derived 
from Table III. I have cited them, however, in the absence 
of any such standard to show in a general way that these 
latter percentages indicate a low fatality according to age. 
The contrast between a fatality percentage of 33.3 for cases of 
diphtheria under two years of age treated with serum, and 
that of 60 to over 80 for cases of the same age not so treated 
is a striking one, even if a large allowance be made for differ- 
ences in the characters of the cases in the two groups. 
We come now to the consideration of the influence upon the 
fatality of the length of the interval between the onset of 
diphtheria and the first injection of antitoxin. In experiments 
upon animals this factor is decisive in determining the result. 
It is the factor which Behring from the first has put in the 
foreground. His claim is that no death will occur from 
diphtheria if antitoxin is injected in sufficient dose at the 
beginning of the disease, and that the fatality will fall under 
5 per cent, if the treatment in proper manner is begun before 
the third day of the disease. 
Of course the only significance of this great emphasis upon 
the importance of early treatment is as an expression of the fact 
that cure is rendered more difficult the larger the number of 
the diphtheria bacilli, the greater the amount and intensity of 
their toxins, the greater the damage already inflicted by the 
bacilli and their toxins, and the more serious the complica- 
tions and secondary infections. There is, however, no abso- 
lute parity between the length of time the disease has lasted 
before beginning treatment and the increase of these dangers. 
One case may become desperate within forty-eight hours after 
the onset, and another may present no grave symptoms after 
a week’s duration. The virulence, the number and the 
microbic associations of the infecting bacilli, and especially 
the local and general susceptibilities of the patient, are factors 
no less important than the single factor of time in influen- 
cing the issue. 
The individual peculiarities of each case must be consid- 
ered. If all is judged according to one simple uniform 51 
standard—antitoxin cures the case or it does not cure the 
case—audit must he confessed this is all which seems to he 
in the minds of many, then the practitioner will not come 
to any clear conception of the wonderful powers of the heal- 
ing serum. The sins of some observers in this matter seem 
incredible. They lump together indiscriminately all of their 
cases, including those complicated with measles, scarlet fever, 
tuberculosis and other diseases, the mixed infections, the 
anginas, the croups, the advanced and the early cases, the true 
and the false diphtherias, the infants and the adults, and throw 
them into the scale to be weighed for or against antitoxin. 
An unsuccessful case is put down to the discredit of anti- 
toxin without reference to its peculiarities. On the basis of 
experience in treating a dozen cases, the writer boldly attacks 
results established by the careful observation of hundreds of 
cases. It is true we need these brute figures for comparison 
with former fatality statistics of diphtheria, and they have 
served to demonstrate the curative efficacy of antitoxin, but 
reports of personal experience with the serum treatment 
should at least contain the data for an intelligent analysis of 
the cases treated. Such an analysis is requisite in order to 
reveal the full scope and capabilities of the new treatment. 
We have already seen that the study of the cases with refer- 
ence to laryngeal involvement has brought to light evidence 
in favor of the serum treatment more convincing than that 
derived from the gross statistics of all cases treated, and evi- 
dence of a kind which meets many of the objections which 
have been urged against the interpretation of the gross statis- 
tics as demonstrative of the efficacy of antitoxin. We shall 
now see that the analysis of the cases according to the day of 
the disease on which the serum treatment is begun almost, 
if not completely, substantiates Behring’s original claims, 
astounding as they seemed to be. 
There is, of course, in many cases considerable uncertainty 
as to the exact duration of the disease at the time when the 
patient is first seen by the physician. The statements of 
parents or of those in charge of the children are often the 
only evidence on this point which can be obtained. Satisfac- 
tory information will be particularly difficult to obtain in 
the class of patients in the diphtheria wards of hospitals, these 52 
patients being chiefly the children of laborers. We are also 
to consider that a diphtheric affection of the throat may exist 
without such manifest disturbance as to attract even intelli- 
gent observation, or it may be mistaken for a simple sore 
throat. A tabulation of cases of diphtheria according to the 
day of beginning treatment will be, therefore, only of relative 
value, but we can fairly assume that the duration of the 
disease will very rarely, if ever, be shorter, but often longer, 
than that stated. 
In the excellent reports on antitoxin treatment from the 
Paris hospitals, the cases are not analyzed according to the 
day of beginning treatment, as Eoux, whose scheme of classifi- 
cation has been followed by most other French writers, stated 
in his original article that it was practically impossible to 
obtain trustworthy statements on this point from parents of 
the children. Most of the reports, therefore, which enter into 
Table IV are from German and English sources. 
The statements as to the day of the disease are entirely from 
information obtained from parents and others, and are not 
estimates on the part of the physician, although in several 
instances the reporter says that the condition of the patient 
plainly indicated a longer duration of the disease than that 
assigned by the parents and put down in the report. It will 
be observed that not all of the reports in the table fit into 
any one system of classification, and therefore not all can be 
summarized in a single table. In each space the higher 
number is the total number of cases belonging to the heading, 
and the lower is the corresponding number of deaths. 
As is well known, the fatality from diphtheria by any ap- 
proved method of treatment is smaller the earlier in the 
disease the treatment is begun. This is clearly shown in the 
following table from the statistics of Y. Hirsch of the cases 
treated in the surgical clinic in Berlin for ten years preceding 
August, 1894, and of course before the employment of anti- 
toxin. The results are according to the day of the disease on 
which treatment was begun. 53 
First Day. 
Second Day. 
Third Day. 
Fourth Day. 
Oases, . . . 
241 
405 
323 
416 
Deaths, . . . 
44 
92 
124 
223 
Percentages, . 
18.3 
22.7 
38.1 
53.8 
Fifth Day. 
Sixth Day. 
Seventh 
Day. 
Eighth Day. 
Cases, . . . 
203 
525 
506 
239 
Deaths, . . . 
136 
219 
367 
191 
Percentages, . 
67 
67.4 
72.5 
81.6 
The preceding table is not intended to serve as a standard 
of comparison for my tables giving the results of cases treated 
by antitoxin, as the classes of cases in the two groups are not 
comparable. 
Philip* has reported from Baginsky’s service the results of 
treatment, before the use of antitoxin, begun in the earliest 
stages of diphtheria, the patients being brothers and sisters 
of children with diphtheria who were examined for Loftier 
bacilli, so that opportunity was given for recognition of the 
disease at its onset. The fatality was 10.5 per cent, lower 
in these cases recognized and treated early than in the others. 
The fatality of the cases treated by Baginsky with serum 
during the first three days of the disease was 32.2 per cent, 
lower than the preceding average fatality of cases not treated 
with serum. Plainly some more potent healing factor than 
merely that of early treatment was present. The only differ- 
ence in the methods of treatment of the two groups of cases 
was the use of antitoxin in the one and its absence in the 
other. 
Kohts, an opponent of the serum treatment, at the recent 
Congress for Internal Medicine in Munich, claims for his 
method of local treatment no deaths among cases treated on 
the first day of the disease. For later days his results are 
much higher than those in the serum statistics. The per- 
centages of deaths according to the day of beginning his treat- 
♦Philip: Arch. f. Kinderheilk., Bd. XVI. TABLE IV. 
FATALITY ACCORDING TO THE DAY OF DISEASE UPON WHICH ANTITOXIN IS INFECTED. 
(In each space the higher number is the total number of cases treated on the corresponding day, and the lower number is that of the deaths.) 
REPORTER. 
Total number 
of Cases 
Treated. 
1st 
Day. 
2nd 
Day. 
3rd 
Day. 
4th 
Day. 
5th 
Day. 
6th 
Day. 
After 
6th 
Day. 
Undeter- 
mined. 
REMARKS. 
Kossel 
117 
14 
30 
29 
9 
11 
6 
12 
6 
Fatality for first three days was 1.4 per cent. 
13 (11.1 per ct.) 
0 
1 
0 
1 
2 
3 
5 
1 
KOrte 
121 
Of 37 severe and moderately severe cases injected 
40(33 1 per ct.) 
during the first thi-eo days 8 died (21.6 per cent). The 
results following injection begun after the third day 
were less favorable. 
Baginsky 
525 
111 
134 
92 
52 
39 
13 
29 
55 
All of the three fatal cases of the first day were far 
83 (15.6 per ct.) 
3 
14 
13 
12 
14 
4 
12 
11 
advanced on admission, therefore the statements of 
the parents as to the date of beginning of the disease 
were probably erroneous (Baginsky). Fatality for 
first three days was 8.9 per cent. 
v. Mering 
74 
Treatment begun on 1st or 2nd day in nearly all cases. 
4 (5 per ct) 
v. Noorden 
81 
Treatment begun on the 3rd or later day in nearly all 
19 (23 per ct.) 
cases. 
Schroeder 
1st & 2nd Day. 
3rd & 4th Day. 
After 4th Day. 
63 
23 
27 
13 
8(12.7 per ct.) 
1 
3 
4 
Yierordt 
55 
3 
14 
17 
9 
7 
1 
4 
8 (14.6 per ct.) 
0 
2 
2 
0 
3 
0 
1 
Rumpf 
18 
3 
After 3rd Day. 
4 cases tvere still under treatment at date of the report. 
26 
1 
1 
5 
218 per ct.) 
0 
Ganghofner 
110 
3 
30 
35 
18 
9 
2 
13 
14 (12.7 per ct.) 
0 
2 
3 
4 
3 
0 
2 
Heim 
27 
9 
2 
7 
3 
1 
5 
6 (22 per ct.) 
0 
1 
1 
1 
0 
3 
Burger 
30 
3 
13 
9 
3 
1 
1 
2 (6.6 per ct.) 
0 
0 
0 
1 
0 
1 
Hager 
25 
14 
5 
4 
1 
1 
Private practice. The single fatal case died of compli- 
1 (4 per cM 
1 
0 
0 
0 
0 
cations after cessation of the diphtheria. 
Kuntzen 
25 
3 
6 
7 
2 
2 
1 
4 
3 (12 per ct.) 
0 
0 
1 
0 
0 
1 
1 
Schmidt 
14 
The three fatal cases were not treated until after the 
3 (21.4 per ct.) 
disease had lasted for 8 to 14 days. Private practice. Seitz 
35 
10 
12 
9 
After 4th Day. 
The two deaths were in infants IX years and 13 months 
2 (5.7 per ct.) 
old with advanced laryngeal diphtheria on day of 
admission, presumably therefore treated after 2nd 
or 3rd day, although this is not stated. 
Hail 
11 
2 
4 
4 
1 
3(27.3 per ct.) 
0 
0 
2 
1 
Epidemic in Trieste 
105 
0 
30 
29 
20 
11 
7 
2 
Hospital cases 
27 (25.7 per ct.) 
0 
5 
9 
5 
4 
3 
1 
Epidemic in Trieste 
72 
14 
27 
18 
8 
2 
2 
1 
Cases in private prartice 
5 (8.9 per ct.) 
0 
2 
1 
2 
0 
0 
0 
Witthauer 
30 
4 
8 
4 
0 
3 
11 
5 (14 per ct.) 
0 
2 
1 
0 
0 
2 
lllumenfeld 
1st & 2nd Day. 
3rd & 4th Day. 
Of the 2 fatal cases, in one the treatment began on the 
4th and in the other on the 5th day. Private practice. 
50 
40 
8 
2 
2 (4 per ct.) 
0 
1 
1 
Kapmuud 
100 
39 
34 
15 
The deaths are given for the 1st and 2nd, the 3rd and 
7 (7 per ct.) 
5 
4th and the 5th, 8th and after 8th day respectively. 
2 
3 
2 
Schaewen 
15 
0 
3 
1 
5 
Of the 5 undetermined cases the statement is that they 
0 
0 
0 
0 
0 
had been ill several days before the injection. 
Hisel 
114 
78 
21 
4 
4 
5 
2 
89 of these cases were treated at their homes, and 25 in 
9 (7.9 per ct.) 
4 
4 
0 
0 
1 
0 
hospital. They were mostly children. 
v. Muralt 
58 
11 
18 
17 
After 3rd Day 
2 (3.4 per ct.) 
0 
0 
0 
2 
2 
Codd and Wliitehouse 
11 
1 
2 
5 
2 
1 
4 (30.4 per ct.) 
0 
0 
2 
1 
1 
Winktield 
22 
8 
7 
4 
2 
1 
Of the 2 fatal cases in which it is stated the injection 
4(18.2 per ct.) 
2 
1 
1 
0 
0 
was made on the 2nd day, in one there were on admis- 
sion advanced laryngeal and nasal diphtheria with 
glandular swellings, and in the other thick mem- 
branes in the throat, nasal discharge and glandular 
swellings. 
Howard 
40 
24 
7 
2 
2 
3 
2 
Private practice. 
3(7.5 per ct.) 
0 
0 
0 
1 
0 
2 
Van Nes 
52 
2 
13 
10 
7 
8 
12 
No details given of fatal case said to be treated on 1st 
12 (23 per ct.) 
1 
1 
1 
2 
3 
4 
day of the disease, but it is noted that the determina- 
tion of the day was very uncertain in the class of 
~ 
patients (children of laborers) admitted. 
Kurth 
97 
12 
35 
10 
19 
5 
3 
7 
Of the 2 fatal cases in which it is stated the injection 
10 (10.3 per ct.) 
0 
2 
1 
3 
1 
1 
2 
began on the 2nd day, one was admitted with croup 
and died 18 hours afterward ; the other died a month 
after disappearance of the local diphtheria with 
albuminuria and broncho-pneumonia. The following table is the summary of the 19 reports of the preceding table in which the number of the cases with the results corresponding to the 
day of the disease on which antitoxin treatment was begun is given for each day up to and after the 6 th day. 
19 Reports. 
Total. 
1st Day. 
2nd Day. 
3rd Day. 
4th Day. 
5th Day. 
6th Day. 
After 6th 
Day. 
Undetermined. 
Cases 
1489 
222 
456 
311 
168 
116 
44 
104 
(68) 
Deaths 
212 
5 
37 
42 
32 
34 
15 
35 
(13) 
Percentages 
14.2 
2.2 
8.1 
13.5 
19 
29.3 
34.1 
33.7 
(17.6) 
Including with the preceding 19 reports those of Schroeder, Blumenfeld and Rapmund, we have the following table which gives the results of 
antitoxin treatment begun on the lsi and 2nd, on the 3rd and 4th, and after the 4th day. 
22 Reports. 
Total. 
1st and 2nd Day. 
3rd and 4th Day. 
After 4th Day. 
Undetermined. 
Cases 
1702 
811 
534 
286 
(68) 
Deaths 
229 
45 
81 
91 
(12) 
Percentages 
13.5 
5.5 
15.2 
31.8 
(17.6) 
It may also be computed from the table that of 1729 cases of diphtheria with a fatality of 14.9$, 1115 cases treated with antitoxin during the first 
three days of the disease yielded a fatality of 8.5$, whereas 546 cases in which antitoxin was first injected after the third day of the disease yielded a 
fatality of 27.8$. Of 232 cases in which treatment was begun on the first day 5 (2.15$) died; of 492 cases in which treatment was begun on the second 
day, 38 (7.7$) died; of 331 cases in which treatment was begun on the third day, 43 (13$) died. 57 
ment, as given by Kohts, without a statement of the number 
of cases treated, are as follows: 1st day, 0; 2nd day, 20 per 
cent.; 3rd day, 47 per cent.; 4th day, 55 per cent. 
Table IV shows that out of 232 cases in which it is alleged 
that antitoxin was injected on the first day of the disease, 5 
died. As a matter of fact, however, the assumed duration 
of the disease in each of these fatal cases is doubtful, as it 
rests solely on the statements of parents or those who cared 
for the children, and is apparently contradicted in at least 
the three cases concerning which any details are given by the 
condition of the patient on admission. Baginsky’s three 
cases (Nos. 311, 479 and 511 of his tables) when admitted 
were far advanced in the disease, with extensive membranous 
exudates, cyanosis and very bad general condition. Hager’s 
case may more readily be accepted, as it occurred in private 
practice, but here the patient died after disappearance of the 
diphtheria from complications, whether or not referable to 
the diphtheria is not stated. 
Of the fifth fatal case reported by Yan Nes no details are 
given in such form that the case can be identified from his 
description, but he himself places little reliance upon the 
alleged duration of the disease in the class of patients 
admitted to the hospital, these being the children of laborers. 
I am not aware of the report of any fatal case of diphtheria 
properly treated by antitoxin within the first 24 hours after 
the beginning of the disease, in which the duration was posi- 
tively determined, still as I have not read every article which 
has been published on the subject it is possible that such a 
case may have been reported. There are, however, many such 
cases of prompt recovery reported as that quoted from Hall in 
Table IV. 
It is noteworthy that the percentage of deaths in 814 cases 
in which treatment was begun before the third day of the 
disease is only 5.5. If the doubtful dehths attributed to the 
first day be excluded, the percentage actually falls a trifle 
short of 5. If we furthermore make allowance for the fact 
that the assigned duration of the disease can scarcely be 
shorter, but may readily be longer than the actual duration, 
then our tabulation of 1702 cases of diphtheria according to 
the day of beginning treatment verifies Behring’s original pre- 58 
diction. I do not, however, consider that it is justifiable from 
so small a number of cases and from material of the kind 
composing our table to draw any definite conclusions as to the 
exact percentages of deaths according to the date of beginning 
treatment. 
According to the table the percentage of deaths in cases in 
which the serum treatment is begun on the 3rd and 4th days 
of the disease is nearly three times greater than that in cases 
treated on the 1st and 2nd day, and the percentage after the 
3rd day is 34 times greater than that of cases treated within 
the first three days. 
We are of course not to infer from these results that anti- 
toxin may not he beneficial when administered after the 3rd 
or 4th day of the disease. There are cases which are still 
mild after this duration, but which subsequently become seri- 
ous, and even in desperate cases antitoxin holds out some hope 
of cure. 
It is apparent that the largest proportion of cures by anti- 
toxin are to be expected from private practice among those 
who call the physician in at an early stage of the disease. 
While a similar statement may be made concerning any other 
suitable method of treatment, it is not, I believe, true in the 
same measure as for the serum treatment. 
The main purpose of this article has been the study of the 
evidence thus far published concerning the curative power of 
antitoxin in diphtheria. I do not propose to consider the 
practical points relating to the employment of antitoxin, nor 
to consider in detail the specific effects of injection of the 
healing serum. There has been much diversity of opinion as 
to these effects, and I shall present briefly the principal points 
which seem to me to be established. 
Most writers approve of the continuance of such measures 
of local and general treatment as have hitherto been found to 
be useful, but recommend the avoidance of all irritating and 
caustic local applications. 
The injection of the serum may be followed in a few hours 
by local pain, swelling and redness, but there is no danger of 
abscess formation if the serum is uncontaminated and proper 
antiseptic precautions are taken. In over 3000 injections 
Martin observed the formation of an abscess only three times. 59 
In twenty-four to forty-eight hours after the injection the 
general condition of the patient is remarkably improved in 
the great majority of those patients who are in a condition to 
he benefited at all by antitoxin. This general improvement 
is accompanied by a fall of temperature, which may be a 
critical fall, especially if the disease is not far advanced; 
often it is a fall by lysis. Some hold that there may be a 
temporary rise of temperature as an immediate effect of the 
injection. Accompanying the fall of temperature is improve- 
ment of the pulse as to frequency and tension, but the heart’s 
action may for some time, even into the period of convales- 
cence, remain weak. 
In the favorable cases the local diphtheritic process is 
arrested, usually within the first twenty-four hours after the 
injection. Membrane may appear upon spots previously 
inflamed and invaded by the bacilli, but otherwise there is 
no extension of the membrane in the majority of the cases 
which are benefited. The area covered by membrane becomes 
sharply demarcated and the swelling of adjacent mucous 
membrane disappears. The membrane may disappear by 
rapid separation or by gradual softening. Sometimes it per- 
sists for several days after disappearance of all other local 
disturbance. Large membranous casts are coughed up from 
the larynx, trachea and bronchi under the serum treatment 
more frequently than under former methods. The rapid 
separation of the membrane in the lower air passages may 
cause sudden increase of stenotic symptoms. Nasal discharge 
is lessened. The swelling of the glands in the neck and the 
surrounding oedema disappear, so far as these are not referable 
to secondary infections. 
The most uncertainty prevails as to the influence of anti- 
toxin in preventing the three most important complications or 
sequelae of diphtheria, nephritis, heart failure and paralysis. 
The weight of evidence is that genuine nephritis is far less 
common in cases treated by antitoxin sufficiently early than 
under other methods of treatment, but it is questionable 
whether albuminuria is less common, although it is consid- 
ered to be by Kossel, Roux and others. If there is an albumi- 
nuria in any way directly referable to the injection of the 
serum, and this is by no means established, it is simple albu- 60 
minima with perhaps a few narrow hyaline casts hut without 
evidence of any serious damage to the kidney. Peptonuria, 
it is claimed by Hecker, is an effect of the serum, hut it 
is without clinical significance. Albuminuria is such an 
extremely common symptom of diphtheria that it must be 
very difficult to determine that it can be referred to the serum 
in any case. 
Many writers emphasize especially the favorable influence 
of antitoxin upon the heart, but there are some who have 
observed that with decided improvement in all other symp- 
toms the force of the heart may still remain weak and occa- 
sion anxiety. Baginsky’s experience is that the minor dis- 
turbances of the cardiac action are not less frequent in cases 
treated with serum, they appear to be even more frequent as a 
larger number of cases survive, but that actual death from 
heart failure is far less common in the serum cases than in 
others. 
Post-diphtheric paralyses may occur in cases treated with 
serum as early as the second or third day of the disease. 
Whether they occur in cases treated within the first twenty- 
four hours is not certain. According to some, paralysis is 
even more common in the serum cases than under former 
methods of treatment. This is doubtful, but if true, it may 
be attributed to the survival of a larger proportion of cases. 
It is apparent from what has been said that antitoxin is 
most strikingly beneficial in progressive fibrinous diphtheria 
and especially in the prevention and cure of laryngeal diph- 
theria. In septic diphtheria the serum treatment is of little 
avail. 
Antitoxic serum may produce unpleasant effects, but these 
do not involve danger to the patient. They are in all proba- 
bility referable to the serum as such and not to the healing, 
so-called antitoxic, substance contained in the serum. The 
most common undesired effect is some form of exanthem, 
usually erythema and urticaria, sometimes an eruption like 
measles or scarlatinal rash. The same ex anthems have been 
observed by Bertin after the injection of ordinary serum of 
the horse, and by Richardiere after injection of Marmorek’s 
anti-streptococcus serum. 61 
The serum from some horses is more likely to cause these 
exanthems than that from others, and there may be individual 
idiosyncrasies favoring their occurrence. Some writers report 
the occurrence of an exanthem in not more than five per cent, 
of their cases, others have observed them in over fifty per 
cent, of the cases treated with serum. They may be localized 
in the neighborhood of the seat of injection or extend from 
that over the greater part of the body, or make their first 
appearance at a distance from the point of injection. Often 
without noticeable fever they may be accompanied by consid- 
erable elevation of temperature and by pain and swelling in 
the joints. A rarer but more severe form of serum exanthem 
resembles erythema multiforme, and when this is accompa- 
nied, as it may be, by high fever, and severe pain in the bones 
and joints with swelling of the joints, the condition of the 
patient may really seem serious, but these patients recover. 
Some have attributed a petechial eruption to injection of the 
serum, hut this may occur in diphtheria without serum 
treatment. 
These occasional untoward effects of the healing serum are 
annoying, but, being unattended with danger to life and Avith- 
out serious consequences, they do not contraindicate the use 
of the serum. 
There have been a few cases reported in which the writers, 
without any satisfactory evidence whatever, have referred the 
death of the patient to the use of the serum. The essential 
harmlessness of the serum has been demonstrated by over a 
hundred thousand injections,* and if future investigations 
should show that through some idiosyncrasy on the part of 
the patient death ever is attributable to the injection of the 
serum, this would probably count for about as much as the 
rare deaths from the use of ether or chloroform. 
I shall leave untouched the question of the immunizing 
properties of antitoxin. 
The principal conclusion which I would draiv from this 
paper is that our study of the results of the treatment of over 
*This would seem to be at least a moderate estimate, as writing 
November 20, 1894, Behring says that there had been up to that 
date certainly over 40,000 injections (Das neue Diphtheriemittel, 
von Dr. Behring, Berlin, 1894, p. 25). 7000 cases of diphtheria by antitoxin demonstrates beyond all 
reasonable doubt that anti-diphtheric serum is a specific cura- 
tive agent for diphtheria, surpassing in its efficacy all other 
known methods of treatment for this disease. It is the duty 
of the physician to use it. 
The later reports show in general a decided improvement in 
the results of the treatment over the earlier ones, and there is 
every reason to believe that the results of the second year’s 
employment of the new treatment will make a much more 
favorable showing than those of the first year. We shall come 
to a clearer understanding of the mode of action of the heal- 
ing serum. Improvements in the methods of preparation and 
preservation of the serum, and possibly the separation of the 
healing substance, at least from other ingredients which pro- 
duce the undesired effects, may be expected. 
The discovery of the healing serum is entirely the result of 
laboratory work. It is an outcome of the studies of immunity. 
In no sense was the discovery an accidental one. Every step 
leading to it can be traced, and every step was taken with a 
definite purpose and to solve a definite problem. 
These studies and the resulting discoveries mark an epoch 
in the history of medicine. It should be forcibly brought 
home to those whose philozoic sentiments outweigh senti- 
ments of true philanthropy, that these discoveries which 
have led to the saving of untold thousands of human lives 
have been gained by the sacrifice of the lives of thousands of 
animals, and by no possibility could have been made without 
experimentation upon animals.