REPRINTED FROM 
©fiMmiSorl 
GYN/tC6LoGYi 
•TffiD-PKDinTKYi 
A Monthly Krricw of Gymncology, Obstetrics, 
Abdota!n*l Surgery, and Ole Diacuacs ot Children. 1 
ElttCST W. CUSHING. UO, C 
RICHARD C. hCR'S. MX, 
a* irtrru >»*. ' * T a -■ - - I 
o.ax >1 
Eubscripticn Price, $3.00 a year in advance, - A 
mz.a> »0 t>T. ..TO XMUX. U1L HBUta luua OZUAlt. r 
PHILADELPHIA M 
VMVKRsrrv oy Pennsylvania press 
PUBLISH \ 
»Mafc2«S£B»ESi3C!S*»«:-X<Sf3®ES.»B? 
JUNE, 1893 
UTERINE THROMBOSIS FOLLOWING POST-PARTUM 
HAEMORRHAGE, AND ITS RELATION TO 
PUERPERAL INFECTION. 
BY W. REYNOLDS WILSON, M.D., 
PHILADELPHIA. Uterine Thrombosis Following Post-Partum Haemorrhage, 
and Its Relation to Puerperal Infection.1 
BY W. REYNOLDS WILSON, M.D., 
PHILADELPHIA. 
Post-partum haemorrhage is fol- 
lowed by a series of consequences de- 
pendent upon thrombosis. The most 
conspicuous of these are phlegmasia 
alba dolens, and pyaemic infection, 
although a general infection depend- 
ent upon the same cause is to be ob- 
served, as demonstrated by the histo- 
logical study of the subject. In nor- 
mal involution the contraction and 
retraction of the uterine muscle is 
sufficient to prevent bleeding from the 
sinuses, by causing an approximation 
of the vessel walls. In the absence 
of normal uterine -contractions de- 
pendent upon want of muscular de- 
velopment, or upon loss of blood, as 
in placenta praevia, or upon over-dis- 
tention from twins, or hydramnios, 
haemorrhage is prevented by the for- 
mation of thrombi. On the part of 
the blood itself the increase of fibrin, 
consequent to the loss of blood, is an 
important factor in thrombosis. This 
natural means for controlling haemor- 
rhage approaches a pathological con- 
dition, in that it admits of an exten- 
sion of the thrombi into the veins 
surrounding the uterus, namely, those 
of the parametrium and broad liga-' 
ments. In this way a direct commu- 
nication between the endometrium 
and pelvic veins is set up. 
In active involution the blood cur- 
rent is diminished, and the absorptive 
power of the veins and lymphatics is 
decreased, whereas in defective invo- 
lution the amount of blood in the 
uterus is increased, and the lymphatic 
circulation called more actively into 
play. According to Winckel (i) the 
outcome of physiological thrombosis 
is described as a conversion of the 
thrombi, by the immigration of wan- 
dering cells, probably with the aid of 
the endothelium and vasa-vasorum, 
into a firm connective tissue cord, 
which at times becomes canaliculated, 
possibly by the passage of red blood 
corpuscles, so that the blood current 
is restored. Thus, under the condi- 
tions in which uterine inertia exists,, 
we have haemorrhage giving rise to 
increased tendency to inertia, and to> 
the formation of thrombi, which serve 
as a dangerous means of communica- 
1 Read before the Philadelphia Obstetrical Society, 
May, 1893. 2 
W. REYNOLDS WILSON. 
tion with the central venous circula- 
tion, and as a stimulation of the lym- 
phatics surrounding them. Where 
elements of infection are absent the 
thrombi shut off the uterine cavity 
from the circulation, but where septic 
material is present they offer, when 
once affected by the putrefactive 
changes about them, a means of en- 
trance into the system for the micro- 
organisms which attack them. 
Bacteriological research has shown 
that the normal uterine lochia con- 
tain no germs, and may be injected 
into the body of any animal in any 
amount without injury. Doederlein 
(2) found that after a normal labor 
with a temperature not exceeding 
98.4° there were no germs, but when 
fever was present, bacilli and cocci 
were found until the temperature fell, 
being eliminated by the very abund- 
ant secretion, especially when this 
was purulent. Micro-organisms may 
find entrance into the uterine cavity 
by various means, and when in con- 
tact with the endometrium give rise 
to infection. It has been positively 
shown that the endometrium is the 
usual source of infection, for in puer- 
peral ulcers of the vagina we have 
only a mild form of infection accom- 
panying the local signs, and although 
the same micro-organisms are present 
as those which are found within the 
uterus in puerperal endometritis they 
occur only at the seat of infection, 
and are not found penetrating into 
the neighboring tissues (3). 
Having, therefore, a case of haem- 
orrhage with the occurrence of dila- 
tation thrombosis, and the presence 
of septic material, we have the liabil- 
ity of infection, the process attacking 
the endometrium, and spreading by 
means of the disorganization of the 
thrombi along the course of the veins, 
especially at the placental site, and 
invading the general circulation. 
It will be of interest to study the 
means by which infecting germs find 
entrance into the uterus. These may 
be present before the occurrence of 
labor in cases in which haemorrhage 
is likely to occur, their presence and 
the liability to haemorrhage being 
dependent upon the same cause. 
Namely, in cases of endometritis we 
have, as has been so forcibly main- 
tained by Pozzi, the presence of path- 
ogenic organisms, the prevailing spe- 
cies being staphylococci (pyogenes 
aureus, albus and citreus), and various 
kinds of streptococci. As to the part 
played by the uterus in cases of haem- 
orrhage due to metritis and endome- 
tritis with the presence of the usual 
pathogenic organisms, Winckel states 
that a limited metritis, or premature 
fatty degeneration of the muscle of 
the pregnant uterus is likely to inter- 
fere with the contractile power of the 
affected area. Endometritis having 
existed during pregnancy, and pres- 
ent at the time when uterine contrac- 
tion and retraction are essential to 
the arrest of haemorrhage, predisposes 
to bleeding; first, on account of the 
hyperaemia; secondly, by reason of 
erosion of already occluded vessels 
from the presence of mycotic ele- 
ments ; and thirdly, by interference 
with involution. The question arises, 
what determines the presence of mi- 
cro organisms within the uterus in 
endometritis, and why, if in any such 
case their existence is proven, should 
puerperal endometritis and its conse- 
quences be the exception rather than 
the rule ? The answer to this lies in 
the fact that the tissues of the gen- 
ital tract possess, under normal con- UTERINE THROMBOSIS. 
3 
ditions, a power of resistance to the 
pathogenic action of the germs which 
may be present. The vitality of 
these germs becomes more and more 
attenuated as they are acted upon by 
the normal secretions and cellular ele- 
ments of the tissues. This antago- 
nism of the tissues against the inva- 
sion of pathogenic germs continues 
as pregnancy advances, up to the time 
of the beginning-of labor. The com- 
pletion of labor, marked by the expul- 
sion of the placenta and discharge of 
liquor amnii, affords the natural means 
by which the genital tract is flushed 
out, and the possibility of the lodg- 
ment of germs is prevented. In 
pathological conditions, on the other 
hand, that is, in simple endometritis, 
in contra-distinction to puerperal en- 
dometritis, the mucous membrane 
becomes infected by the invasion 
of germs which are indigenous to 
the genital tract. According to 
Pozzi (4) there exists in the genital 
tract of the female a zone rich in 
micro-organisms, situated at the level 
of the internal os. The activity of 
this zone is increased by the general 
debility of all the tissues, which re- 
duces cellular vitality, or by trauma- 
tism. In endometritis the mucous 
membrane becomes infected from 
this source, and the ordinary lesions 
and symptoms follow. In puerperal 
endometritis, in cases where hetero- 
infection can be excluded, we may 
ascribe the condition to an ante-par- 
tum infection dependent upon an 
earlier endometritis, the earlier patho- 
logical changes in the uterine mucosa 
and connective tissue predisposing 
to haemorrhage, by interfering with 
contractions, with infection of the 
resulting thrombi by the germs which 
are already present. In cases of atony 
from other causes (want of muscular 
development, over-distention) the 
treatment which is used to avert 
the haemorrhage, and the necessary 
manipulations, may be responsible 
for the infection. Frequent exami- 
nations during the course of labor, 
hasty and careless manipulations at 
the time when the patient is bleed- 
ing, and carelessness of the prin- 
ciples of asepsis owing to the loss of 
self-control on the part of the attend- 
ant, and the introduction of infected 
instruments, all contribute to the risk 
of infection. There exist, therefore, 
under these circumstances, ample 
opportunities for the invasion of bac- 
teria. 
What, on the other hand, are the 
natural means of resisting these bac- 
teria in cases of non-infection, and 
what are the local changes in septic 
cases, resulting from the action of 
micro-organisms ? Immediately after 
the expulsion of the placenta the 
uterus contracts and obliterates the 
cavity recently occupied by the ovum. 
This contraction is influenced largely 
by the nervous condition of the wo- 
man, and may be considered as an 
active process. Under normal cir- 
cumstances the innervation of the 
organ produces the active power of 
contraction irrespective either of the 
elasticity of the fibres or of the dimi- 
nution due to shortening of the fibres 
by retraction. The blood supply is 
lessened by this contraction, and the 
vessels at the placental site are com- 
pressed by the uterine fibres and 
emptied of blood. Both the free con- 
tents of the uterus, namely, blood and 
the remaining amniotic liquid, and 
the adherent shreds of decidua are 
expelled. As soon as the tonic con- 
tractile power of the uterus is estab- 4 
W. REYNOLDS WILSON. 
lished, retraction of the muscle (fatty 
degeneration of the muscular fibres) 
and regeneration of the mucous mem- 
brane take place. Together with the 
lessening in size of the uterus by re- 
traction, there is an increase in the 
development of intraglandular tissue 
and a reconstruction of the mucosa 
from the epithelium springing from 
the remaining glands. The exudation 
which accompanies this process, to- 
gether with the migration of white 
corpuscles and the secretion from the 
cervical canal, constitute the lochial 
discharge. As to the local changes 
occurring in the course of infection, 
we have these normal processes modi- 
fied as follows : First, as a predispos- 
ing cause of infection we have the 
absence of uterine contraction. As 
a result, the haemorrhage from the 
sinuses is controlled not by pressure, 
but by thrombosis; secondly, the 
uterus contains, also incident to the 
absence of contraction, remnants of 
decidua or placental debris, these, to- 
gether with the thrombi projecting 
from the placental site, act as foreign 
bodies, and are prone to putrefactive 
changes; thirdly, the reconstruction 
of the mucous membrane is replaced 
by necrosis of the epithelium and base- 
ment membrane; fourthly, the normal 
constituents of the lochial discharge 
are replaced by the putrefactive debris 
of disorganized thrombi, the remnants 
of decidua and necrotic mucous mem- 
brane, mixed with the various micro- 
organisms which accompany these 
putrefactive changes. In order to ap- 
preciate the relation of such changes 
to the development of infection, it will 
be necessary at this point to study the 
histology of puerperal endometritis. 
According to Bumm we have com- 
monly to deal with the following 
forms : 
Putrid Endometritis.-In this form 
putrefaction occurs from the presence 
of saprophytic organisms. The bac- 
teriology of this condition is still un- 
developed ; as to the histology, we 
find that the necrotic decidua is cut 
off by a zone of cellular infiltration, by 
which the various micro-organisms 
present are prevented from penetrat- 
ing into the underlying tissues. In- 
vasion of the thrombi, however, at the 
placental site, is not prevented by any 
such zone of reaction on account of 
the want of organization of the throm- 
botic tissue. 
Septic Endometritis, occurring in 
two forms.-First, a localized septic 
process in which a granulation zone 
occurs (contrary to-what in 
the form mentioned abovb) shutting 
off the necrotic endometrium and pre- 
venting the invasion of germs. The 
uterine lymphatics are not actively 
involved. The placental site, as in 
the putrid form, is most markedly 
affected. Secondly, a septic endo- 
metritis, accompanied by a general 
infection. Bumm has studied five 
cases belonging to this class, and 
has found in three instances that in- 
fection has occurred by invasion 
through the lymphatic system, and in 
two instances along the course of the 
veins. In the first set of cases the 
placental site is free from micro-or- 
ganisms and thrombi, so that it is not 
likely that this pathological condition 
bears upon that form of infection re- 
sulting from haemorrhage in which 
thrombi, especially at the placental 
site, occur. In the second set of 
cases the smaller lymphatic branches 
surrounding the sinuses are marked 
by colonies of cocci, which extend 
into larger lymphatics underlying 
the peritoneal covering of the uterus. 
The decidua is disorganized and in- UTERINE THROMBOSIS. 
5 
filtrated with a fibrinous exudate, 
presenting a diphtheritic appear- 
ance. In this class of cases, as well 
as in that about to be described, the 
granulation zone is absent. This fact 
has evidently an important bearing 
the function of such a zone of 
demarkation, in combating the pro- 
gress of micro-organisms into the un- 
derlying tissue. 
Thirdly, a thrombotic form of infec- 
tion; and this is the form which con- 
cerns us principally in the discussion 
of post-haemorrhagic infection. This 
is characterized by both a putrid and 
septic endometritis. It is described 
by Bumm as follows : 
" The decidual layer of the uterine 
cavity, in a state of necrosis, is beset 
with micro-organisms. In the neigh- 
borhood of the colonies of strepto- 
cocci, outlined by the staining process, 
are scattered innumerable colonies of 
putrefactive germs. The histologi- 
cal relation of the tissues, that is, the 
decidual, glandular and muscular tis- 
sues, in the necrotic area, is unrecog- 
nizable." 
The zone of reaction is marked. 
The placental site presents no remains 
of the placenta, but is marked by the 
projection of thrombi. The latter are 
found to be infected by various path- 
ogenic germs, are disorganized, and 
offer, by reason of their disorganiza- 
tion, a direct means of entrance for 
the septic products into the current 
of the blood. The disorganization 
occurs first in the axis of the thrombi. 
The endothelium and the vessel wall 
become rapidly affected and break 
down into a mass of necrotic tissue 
mixed with white corpuscles and in- 
fected with cocci and bacilli. 
In conclusion, we may summarize 
the development of infection as a 
result of thrombosis by noting the 
following events : First, a predisposi- 
tion to infection arising in cases of 
haemorrhage the result of atony of the 
placental site; secondly, the forma- 
tion of thrombi which offer, on ac- 
count of their want of vital organiza- 
tion, an improper means of resistance 
to infecting germs; and thirdly, the 
presence of infecting material either 
from the pre-existing endometritis or 
from contamination at the time of 
delivery by careless or frequent ex- 
aminations. When these factors are 
present we have a resulting infection 
occurring in accordance with the his- 
tological changes described above. 
Clinically, we are apt to consider pyae- 
mia as the type of infection occurring 
as a result of thrombosis. Such a view 
is based upon-first, the frequency of 
the occurrence of phlegmasia follow- 
ing phlebitis, either by extension from 
the veins of the broad ligament or by 
the lodgment of coagula washed from 
the placental site and carried into the 
hypogastric veins and obstructing the 
flow of blood through the crural veins; 
and secondly, upon the occurrence 
of embolism from the detachment of 
thrombi from the placental site or the 
parametrium. But it is more likely, 
from the histology of endometritis in 
the puerperal state, that the thrombi 
act more as a channel by which path- 
ogenic germs find entrance into the 
organism than as a direct means of 
conveyance by their detachment and 
circulation in the blood current. We 
have observed in the thrombotic form 
of endometritis that the disorganiza- 
tion of the thrombi is a pathological 
change dependent upon the action of 
bacteria, and that the natural barrier 
to the entrance of infecting elements 
is removed by this process of disor- 6 
W. REYNOLDS WILSON. 
ganization. According to this, the 
blood current is likely to be contami- 
nated, not by the remnants of uterine 
coagula, but by the presence of path- 
ogenic bacteria and their chemical 
products. These, carried along in the 
blood current, may be reasonably sup- 
posed to set up inflammatory changes, 
causing phlebitis, and, especially, to 
produce the development of a general 
septicaemia. There is no doubt that 
the formation of emboli is a common 
result of the detachment of thrombi 
from the placental site, but in the 
study of the subject from a histologi- 
cal point we are not warranted in ac- 
cepting the occurrence of pyaemia as 
the universal clinical associate of 
thrombotic infection the result of 
haemorrhage, and, on the other hand, 
we warranted in assuming the pos- 
sibility of a marked state of septic 
endometritis occurring after haemor- 
rhage without the early association of 
pyaemic symptoms. 
REFERENCES. 
(1) Winckel: Lehrbuch der Geburt- 
shiilfe, 1889. 
(2) Doederlein: Centr. f. Gynaek., 
1888, Nos. 23 and 28. 
(3) Bumm: Archiv. f. Gyn., Bd. 
xxiii, p. 237. 
(4) Pozzi: Medical and Surgical 
Gynaecology, translated by Brooks H. 
Wells, M.D., 1891. 
DISCUSSION. 
Dr. Edward P. Davis : 
As a clinical illustration of some of the 
points made by Dr. Wilson may be mentioned 
the not infrequent occurrence of haemorrhage 
in syphilitic women after labor. When such 
haemorrhage occurs septic processes are apt 
to follow unless special care is taken. We 
recall such a case recently. A woman with 
syphilis had haemorrhage after labor. Fol- 
lowing this a septic process developed, with 
dissemination of micro-organisms through 
the body. I feared a tubal abscess and 
opened the abdomen, but found nothing war- 
ranting removal of the tubes. They were 
allowed to remain and the abdomen closed. 
There then occurred multiple thrombosis of 
one lower extremity, thrombosis of the pop- 
liteal vein and some of the smaller veins. 
The patient recovered under careful nursing, 
and illustrated well the clinical course of a 
case complicated as described by Dr. Wilson. 
In regard to treatment of haemorrhage, the 
use of the intra-uterine iodoform gauze tam- 
pon is based upon the fact that after haemor- 
rhage the thrombi are apt to become infected; 
the use of the antiseptic or aseptic gauze is 
an attempt to maintain an aseptic condition 
of the thrombi at the placental site for 
twenty-four or thirty-six hours after delivery. 
Dr. T. Ridgway Barker : 
I hoped that Dr. Wilson, in treating of this 
subject, would add something to our knowl- 
edge of the causes of thrombosis, as this dis- 
ease is one in which I am much interested. 
I do not think that there is any doubt that 
thrombosis of the uterine veins is due in a 
large measure to a hyperinotic condition of 
the blood incident to pregnancy. 
While the exact cause is often doubtful, the 
predisposing are unquestionably, first, early 
rupture of the membranes leading to prema- 
ture separation of the placenta, with the 
formation of soft coagula in the sinuses; 
second, sudden fall of blood pressure, usu- 
ally the result of haemorrhage; and third, 
uterine inertia. 
The question arises, how shall we combat 
by prophylactic measures these predisposing 
causes? We always have associated with 
gestation an increase of the fibrin in the 
blood, and it becomes anaemic in character. 
By sunlight, fresh air, and exercise we may 
counteract this tendency to deterioration. 
With reference to premature separation of 
the placenta we have only to delay rupture of 
the membranes, while in the case of uterine 
inertia a timely application of the forceps will 
usually prevent its occurrence. 
I would say in conclusion with regard to 
the avoidance of puerperal infection of the 
thrombosis, that I believe ergot administered 
after labor is a very important measure, given 
either hypodermatically or in the form of 
tablet triturate by the mouth. UTERINE THROMBOSIS. 
7 
It tends to produce firm tonic uterine con- 
traction and retraction, and insures obliter- 
ation of the uterine sinuses, and, moreover, 
causes complete expulsion of all debris and 
clots, which otherwise would remain within 
the cavity. 
value-1300 deliveries without a death from 
any cause. 
Dr. Richard C. Norris: 
Dr. Wilson's remarks are of great scien- 
tific interest and value, but the clinical aspect 
of this question is of especial interest to me. 
It is an interesting problem to practical ob- 
stetricians to arrive at any conclusion as to 
when infection of the uterine thrombus takes 
place. We know that it is the custom with 
many when elevation of temperature occurs 
to begin local disinfection, and some would 
resort to the use of the curette. There have 
been a few cases in which after this curetting 
prompt elevation of temperature has oc- 
curred. When I find this elevation of tem- 
perature I fear the possibility of infection of 
the thrombi, and that the curette may have 
done harm. It breaks up the clots, and, per- 
haps, aids in the absorption of more septic 
material. In one case this symptom of sec- 
ondary rise of temperature occurred, and not 
long afterward septic pneumonia appeared, 
probably from embolism. I believe that in 
the absence of local signs of disease of the 
tubes, ovaries, or cellular tissue, and where 
we curette the uterus and find debris, and 
this is followed by a prompt rise in tempera- 
ture, we should look upon the case as one of 
infection of the thrombi in the uterus, and 
desist from further use of the curette. 
Dr. W. Reynolds Wilson: 
I think that perhaps the most important 
bearing of this subject is as to the existence 
of pathogenic germs within the uterus-the 
existence of these germs even before the 
chance of distribution and infection following 
labor. I think, as I said in the paper, that 
the researches mentioned by Pozzi as to the 
cause of endometritis and the general infect- 
ing inflammatory troubles of the uterus are 
of especial interest at this time. I think that 
in obstetrical and gynaecological work we are 
coming to a time when this subject should be 
looked into. 
The remarks of Dr. Norris are very practi- 
cal. We find general rapid infection occur- 
ring principally through the lymphatics. In 
the case mentioned by Dr. Norris, it is prob- 
able that there was beginning a general in- 
fection through the lymphatic system, and 
that the disturbance of the thrombi added a 
local septic process to the general infection 
already developed. 
Dr. J. Price : 
I may say a word on the practical side of 
this subject. Some of you may have read 
the interesting researches in regard to the 
presence of micro-organisms in the uterus, 
of Dr. La Place, whose investigations have 
been republished in the British Gynecologi- 
cal Journal. Speaking purely from the clin- 
ical side of the question, I would state that 
my experience has been rather peculiar, and 
differs somewhat from that of others. The 
Germans have condemned the expression 
method of delivering the placenta, and I be- 
lieve that it has been condemned by a few 
teachers in this country. In my own expe- 
rience I value the expression method above 
all others. With me it begins with the deliv- 
ery of the child. When the head is delivered 
I anticipate delivery of the shoulders, and 
place my left hand on the uterus, and follow 
the extremities. My practice is that of early 
delivery of the placenta, and securing early 
firm contraction and retraction. This I hold 
on to for some time, not permitting the relax- 
ation that so commonly follows slow or de- 
layed delivery of the placenta. I have no 
recollection in the Retreat work of a single 
bubble of air or clot that you find in late 
delivery of the placenta. You are called to 
deliver the placenta because some one has 
failed. You find the uterus large and relaxed, 
and it takes some time to stimulate a contrac- 
tion. If you succeed, you succeed in deliv- 
ering the placenta without introducing the 
hand, but with it you have a great quantity 
of clot, and a great blubber of air. 
I am satisfied that tardy delivery of the 
placenta, and the slow contraction and retrac- 
tion of the uterus, is at the bottom of much 
mischief that the reader has dwelt upon. 
At the Retreat I have no knowledge of a post- 
partum haemorrhage in 1300 cases. I have 
not had a post-partum haemorrhage for ten 
years in my own work. I find the practice of 
early delivery of the placenta, and following 
the uterus down, very valuable. It makes 
me very comfortable to leave a woman after 
such a practice. Following this practice the 
statistics of the Retreat strongly fortify its