William Osler.

Modern medicine : its theory and practice, in original contributions by American and foreign authors (Volume v. 5) online

. (page 44 of 126)
Online LibraryWilliam OslerModern medicine : its theory and practice, in original contributions by American and foreign authors (Volume v. 5) → online text (page 44 of 126)
Font size
QR-code for this ebook

(2) If so, is it from the digestive tract? (3) Is it gastric in origin, and if so,
does it come from near the cardiac end, from the duodenum, or from the
stomach proper? (4) What is the etiological and anatomical lesion?
(5) W'hat vessel is affected? (6) Is it amenable to medical treatment, or,
if not, is it accessible to operation?

HaMuatemesis is so often latent or unrecognized that one must make
careful observations both as regards the previous gastric history and the
present state of the abdomen and gastric contents or vomitus. Not infre-
quently there is a fulminating gastrorrhagia, in which no blood is vomited
and the diagnosis is made from the general sensations and symptoms of
hemorrliage and the previt)us history of gastric disease. Even when very
copious it may not be evacuated and the patient may die with merely the
signs of internal hemorrhage, or vomiting may be delayed for some time, or
the blood may only apjjear in the stools.

1. /.y Blood RcaUi) Present in the Vomitus? — It not infrequently happens,
both with patients and physicians, that a faulty diagnosis is made because
the food previously taken may resemble blood, e. g., red wine, cherries,
cranberries, red cabbage, red turnips, red sausages, coffee, and cinnamon.
Or pseudohcTmatemesis may be associated with various preparations of
iron or of bismuth; the bismuth grystals, moreover, under the microscope,
resemble to some extent those of htematin. In doubtful cases one should
always examine microscopically for blood cells or blood pigment, and where
doubt still exists, one should test with the spectroscope or by chemical
means, to prove the presence of these so-called occult hemorrhages.^

2. Is the Blood from the Digestive Tract? — The blood may not be from
the digestive tract at all, but from haemoptysis, epistaxis, or from a ruptured
aneurism. To differentiate from haemoptysis one must remember that in
hsematemesis the blood is dark red in color, usually resembling venous
blood; that it contains no air, has usually an acid reaction from contact with
the gastric juice; that it is mixed with food; that it is vomited, not coughed;
that it does not contain tubercle bacilli, for tuberculous ulcers of the stomach
are extremely rare; and lastly, local findings in the lungs are absent, while
those in the stomach are often obvious; or there may be signs of a heart
lesion, the possible cause of hemorrhagic infarction of the hmgs. Epistaxis
may give rise to error through swallowing of the blood and subsequent
vomiting. For this reason it is always well to examine the nasal cavity
carefully and to look for polypi and other source of nasal hemorrhage.
The same may occur in empyema and cardiac disease.

Aneurisms which leak gradually may simulate true gastric disease, as

' The most recent test for ocoilt l^lood, and an accurate one, is that devised by
Meyer, of Munich. Two grams of phenolphthalein, 20 srams of potassium hydrate,
and 10 grams of zinc dust are boiled in 100 cc. of distilled water until the mixture
is completely decolorized. It is then filtered and used as the reagent. Equal parts
of the reagent and of hydrogen peroxide are then mixed in a test tube, and to this
mixture the ftecal solution to be tested is added after having been boiled and cooled,
as in the benzine test (p. 194). A positive reaction is indicated by the production
of a carmine to pink color. It is said to be sensitive 1 to 1,000,000.


in a remarkable case recorded by Janowski/ whose patient entered the
hospital after vomiting one liter of blood and fainting. Several weeks
previously he had complained of a burning dry sensation in the oesophagus,
and for three years had had pain in the left hypochondrium. He had been
in hospital nine days when the tube was passed after the administration of
a test meal, and one minute later he died. An aneurism of the aorta had
perforated the oesophagus and been ruptured by the passage of the tube.

3. Having excluded other sources, one must decide whether the hemorrhage
be gastric, duodenal, or ossophageaL

(Esophageal Varices. — The oesophageal veins empty into the azygos veins,
which, in their turn, are tributaries of the vena cava superior and inferior
respectively. When portal obstruction occurs there is sometimes overflowing
of the vena cava inferior and therefore of the hemi-azygos and azygos veins,
and through these of the oesophageal veins. There is nothing distinctive
about the hemorrhage from the oesophagus. Q^lsophageal varices are most
commonly associated with early cirrhosis of the liver, and one should
therefore examine carefully for it. The evolution of the condition may
help the diagnosis. In two-thirds of such oesophageal hemorrhages the
patient improves, but recurrence takes place after intervals of days,
months, or years. Sometimes they are fulminating in character. The
etiology of cirrhosis, i. e., alcoholism, etc., will aid the diagnosis, as also
will the physical examination of the abdomen. (Esophageal cancers may
also simulate gastric ulcer, but bring with them other evidences of malignant
disease. (Esophageal ulcer causes pain after swallowing, increased secre-
tion and regurgitation of mucus and food, and progressive signs of stricture.

Duodenal hemorrhage from ulcer is sometimes accompanied by hsema-
temesis. Men are more commonly afflicted. As a rule, the stools show the
first evidence of this condition. The pain is usually situated more to the
right; it occurs later after meals than does that of ordinary gastric ulcer
(two to four hours), or, at all events, is increased at that time; the digestive
signs of pain, etc., coming on several hours after meals and the presence of
HCl help to establish the diagnosis.

4. The nature of the causative lesion is often most difficult to determine,
and is sometimes impossible. Neuropathic hoematemesis is a problematical
condition which is perhaps receiving less recognition each year; it is pre-
sumably more common in nervous people, in females than in males, comes
on after emotions with prodromata of heat, weight, or pain in the epigastrium
and sometimes vertigo. It co-exists or alternates with hemorrhages from other
organs which are presumed to have a similar cause. It is not usually fatal,
and seems to have no effect upon the system in general. As a rule, patients
thus afflicted pay but little heed to the hemorrhages when they occur. One
should look for the other stigmata of hysteria. In all such cases, however,
the exclusion of ulcer is not easy, and one will do well to diagnose the more
serious anatomical lesion until further proof of a mere neuropathic origin is

Vicarious hemorrhages usually occur in the absence of any gastric signs,
have more or less periodicity, and at best are quite rare. They occur either
at the time of the menses, or at times quite apart from that period. The
menses may be absent and replaced by the hsematemesis, or the latter may

1 Ztschr. f. klin. Med., 1902, xlvi, p. 43.


coincide with diniinisluHl menstrual flow, l.eoroiix had a patient in whom
monthly hpematemesis oeeurred dnrinji; the first seven months of ])reo;naney.

Varices in the stomach are very rare, and doubtless have the same causes
as those in the (esophagus. Letulle found only two cases in sixteen years.

Venous Stasis. — Although varices are rare, dilatation of veins is common,
and a rupture of the capillaries of the mucous membrane frequently occurs,
as in cirrhosis of the liver, inflammations or tumors of the ])ancreas, cancer
of the gall-bladder, wounds in the vicinity, etc. In the vomitus of peri-
tonitis and intestinal obstruction there may be a little showing of blood;
perhaps, however, these types are infectious in nature.

Hemorrhagic Erosions. — These show only bleeding and few or no other
symptoms. Often they are only foimd in the washings from the empty
stomach and usually then associated with chronic gastritis. The use of the
term to indicate a special disease as suggested by Einhorn is disputed by
Eisner and others. The bleeding may occur by the mouth or the rectum, as
in latent idcers; they are rarely fatal, and usually occur in men who are in poor
condition, with advanced tuberculosis, or alcoholics with cirrhosis, etc. C.
H. Miller' attributes many gastric hemorrhages to erosions from swollen or
ruptured lymphoid follicles in the stomach mucosa which has become thinned,
while the basement membrane is unusually thickened. Thus the follicles
are less covered, and under abnormal contlitions, many lymphoid follicles
become swollen, disintegrated, and softened. They in this way reach the
surface and may burst through the mucosa, thus leaving the basement mem-
brane temporarily exposed to the gastric juice, which digests some of the
numerous vessels of the submucosa. Hemorrhage then occurs.

Miliary aneurisms of the submucous arteries as a part of generalized
aneurisms cannot be distinguished clinically.

Simple Ulcer of the Stomach. — Hsematemesis coming on in healthy
individuals or in those who have suffered from stomach disorders is due
in the majority of cases to ulcer in some stage. The history with the symp-
tom complex, the oft-recurring hemorrhage in youthful subjects with good
appetites and good albumin digestion, the presence of HCl in the contents,
and the absence of cachexia aid in determining the presence of ulcer. The
important point is not so much the exact diagnosis of the lesion, as the
diagnosis of the value and possible success of intervention. The hemor-
rhage is often single and fulminating, as happens with the type of cases
known as "exulceratio simplex." The term should become obsolete in the
light of modern experience.

Cancer of the stomach is usually easy to diagnose from the history and the
findings. The patient is older, with ansemia, perhaps cachexia, anorexia, no
HCl in the contents, but lactic acid instead, etc. While sometimes hemor-
rhage in cancer is acute and fatal, as a rule, bleeding is small, recurrent,
and of a coffee-ground appearance. Occult blood in the stools is much more
constant. However, many exceptions occur, and all possible factors must
be considered together.

5. What Vessel is the Source? — This cannot usually be told from the
amount of blood present, the nature of the attack, or the evolution of the
disease. Experience teaches that the most copious bleedings may some-
times arise without any evidence of anatomical lesion, and, vice versa, small

* British Medical Journal, I9O6, p. 1547,

li /EM AT EM ESI ^ 325

continued hemorrhages may occur even when the aorta is the direct source
of the blood. Cases arc on record wliere a curdiogastric fistula occurred,
with hemorrhages lasting for a number of days prior to death. Similar
instances are published of fistula between the aorta and intestines. Perhaps
the large arterial trunk most commonly involved is the sf>lenic artery, and
yet here, also, where ulceration into it had occurred, sudden death followed
in only 3 out of 17 cases. When the bleeding occurs suddenly from an artery
of moderate size, the hemorrhage may be copious and emesis immediate;
in such a case the blood is usually bright in color, at all events much more
so than in slow oozings from small capillaries where one obtains merely
coffee-ground admixture of blood, mucus, food, etc. So few minutes are,
however, required to alter the color of the blood when in the stomach, that
the bright blood is much less frequently found than in haemoptysis.

Considering the possibility of copious hemorrhages from invisible sources,
the frequency of minute bleedings from large trunks, and the changes occur-
ring in the blood when in the stomach, one can appreciate the difficulties
of determining whether the source be capillary, venous, or arterial.

6. Are the Ulcer and Vessel Accessible for Operation? — Fresh ulcers usually
erode the superficial vessels in the wall and are readily reached; old ulcers
may be adherent to the pancreas and the liver and are sometimes hollowed out.
Adhesions form beyond the limits of the organ and are deeply situated, fixed
in immobile tissues, perhaps hard or friable the vessels are deeper, less acces-
sible, and often larger. In a word, they erode rather the extrinsic arteries,
such as the splenic, and make operation difficult, perhaps impossible.

Prognosis. — This depends chiefly upon the cause and, to a less degree,
upon the severity of the hemorrhage. In serious general conditions hsema-
temesis is one of the worst prognostic symptoms of the disease. When
blood flows from outside structures into the stomach the prognosis is bad.
In poisoning cases the amount of hemorrhage itself is of no prognostic im-
portance unless the contents be putrid, in which case a deeper ulceration
may be suspected and possibly death from sepsis or perforation. It is often
difficult to differentiate between rupture of the small and the large vessels,
and statistics vary very much because there is nothing in the evolution of
the condition or in the antecedent hemorrhage which can prognosticate a
recurrence or progressive gravity or origin ; one cannot say from the amount
of blood and the area of pain what vessel is affected or whether the hemor-
rhage will recur. The most copious hemorrhages may have no visible
anatomical lesions of the mucosa. Large hemorrhages, however, are rarely
fatal (3 to 4 per cent., Lebert).

Sometimes in gastric ulcer hemorrhage causes cessation of all the other
signs; in cirrhosis of the liver that organ often diminishes in size after a
hsematemesis. The cause is often more serious than the hemorrhage itself.
In ulceration the prognosis is difficult to determine. Hyperacidity after the
hemorrhage may act as an irritant and cause renewed ulceration and bleeding.

For one month after a severe hemorrhage the condition continues serious,
but from that time onward becomes less so, so far as life is concerned, but
there is always the same uncertainty as regards recurrence. After eight to
ten days the prognosis may be given as good, because of thrombus formation,
although one must always keep in mind the possibility of hemorrhage from
another part of the ulcer where a newly attached vessel has simultaneously
opened up. In cancer, hsematemesis may usually be considered as a fatal


sign because indicatinoi; advanced disease, although exceptions occur with
bleeding at an early stage.

Preble^ eontludes, from an analysis of GO fatal cases of hivmatemesis with
subsequent autopsies, that: (1) While not rare, fatal hemorrhage is at all
events not common in cirrhosis of the liver. (2) The first hemorrhage in
cirrhosis is fatal in only one-tiiirtl of all cases. (3) CKsophageal varices are
the source of the bleeding in 80 per cent, of cases; in half of these, ruptures
were found and are probably much more common than is generally realized.
(4) Fatal hemorrhages sometimes occur from invisible ruptures of numerous
capillaries of the gastric mucous membrane, and, as a rule, there are some
accompanying signs of early cirrhosis of the liver.

Treatment. — The indications are as follows: (1) Prophylaxis, when the
possibilities of liiTematemesis are recognized. (2) Discovery of the etiological
factor. (3) Control of the hemorrhage as quickly as ])ossible. (4) Pre-
vention of its recurrence. (5) Prevention of injurious sequehe.

Prophylaxis depends upon the discovery of the possible underlying con-
dition and its judicious treatment. Control of the hemorrhage requires rest
in general and of the stomach in particular. The patient should be put to
bed with the head low and kept as quiet as possible. All tight clothing should
be removed and the patient encouraged with the ex|)lanation that ha^ma-
temesis of itself is of no very serious import. Should faintness supervene,
the face should be sprinkled with cold water, ammonia placed beneath the
nostrils and if necessary camphor oil, ether, or caffeine, given hypodermically ;
a hot bottle may be placed to the feet. In the most serious cases an intra-
venous saline may be given and the arms and legs bandaged. An ice-bag
may be suspended from a cradle so that it will come into gentle contact with
the epigastrium. All food should be stopped for three days and the patient
should be allowed only some ice to suck, but not to swallow. After three
days, nutrient enemata may be given, three times daily. No food should be
given by the mouth until after the first week, and then, only gradually, boiled
milk, broth, or gruel.

Lavage of the stomach is best avoided, so, too, all local haemostatics.
General hsemostatics are perhaps beneficial; adrenalin, in doses of 25 drops
of the 1 to 1000 solution by mouth daily has apparently been of service.
One may use rectal injections of calcium chloride, 4 to 8 gm. (3j to 5ij)
daily, or gelatin is sometimes useful when given by the mouth. The reflex
effect of heat as administered by hot rectal injections (115°) has been
much recommended.

Accessory Treatment. — Cold water enemata are useful for retained blood
in the bowel. For tlie anaemia, which may persist, one should give good food,
and iron if the ])atient bears it with ease, but its use requires great care.

Surgical Interference. — Indications. — The blood count and estimation of
haemoglobin per se are of little value. Appreciation of the general condition
is more important than an estimate of the haematological state. If a low
blood count has supervened on a previous chlorosis the condition is more
grave. The pulse per se has doubtful value. The amount of hemorrhage
is also of little importance as compared with recurrence of hemorrhage,
which implies an end to medical treatment. For general surgical indications
see the section on the Treatment of Gastric Ulcer.

' American Journcil of the Medical Sciences, 1900, cxix.





Unfoktunately our knowledge of the physiology of the intestines has not
yet proven of much practical value in the study of intestinal diseases or
their treatment. The newer physiology of digestion (both gastric and intes-
tinal) gives promise, however, of some practical applicabiUty in the near
future, and for this reason requires to be kept prominently in mind.

Motor Functions of the Intestines. — The movements of the intestines
serve three important purposes: (1) The onward movement of the contents;
(2) thorough mixture of the chyme with the digestive secretions and close
contact of the contents with the absorptive mucous membrane; (3) the
propulsion of the venous blood and chyle away from the bowel. The old
idea that the main purpose of peristalsis is to cause the movements of the
intestinal contents toward the rectum has been disproven by recent investi-
gations, which show that certain conspicuous movements of the bowel serve
the purpose of thoroughly churning the contents and thus mixing the chyme
and the digestive fluids and at the same time bringing the contents of the
bowel into more complete contact with the mucous membrane.

Three forms of movement are recognized: (1) Peristalsis proper; (2)
rhythmic segmentation; (3) antiperistalsis.

Peristalsis Proper. — Peristalsis proper, or the movement that causes the
forward flow of the bowel contents, is a wave-like contraction that runs for
some distance, gradually diminishing toward the end of its course, where a
new wave starts up. The bowel contracts behind the contents and relaxes
in front of them (Starling). These movements. vary in rapidity and strength
under normal conditions, and may become excessive in pathological states.
Violent stimulation may cause a tonic contraction of a segment or a consider-
able portion of the bowel. The peristalsis of the large intestine causes an
alternate protrusion and retraction of the haustra.

Rhythmic Segmentation. — In skiagraphic studies of the movements of the
intestinal contents Cannon found that constrictions occur at various places,
and the portions of the bowel between the constrictions may subdivide by
secondary segmentation. The segments subsequently re-unite and divide
with renewed contractions. As many as thirty segmentations a minute were
observed. These segmentations involve the contents of the bowel and do not
indicate the kind of contractions that occur in the wall between the segments.
Possibly the swaying movements or waving movements described by Raiser
may be the occasion of the segmentation. These are backward and forward
contractions and relaxations affecting a limited part of the bowel and running
in a direction parallel to its length. There is no appreciable narrowing of



the lumen of the bowel. Such contractions mioht verv readily divide the
contents of the bowel into segments. Rhythmic segmentation continues
during sleep, but may be inhibited by excitement. These movements cause
a churning of the contents of the bowel and perhaps to some extent aid
peristalsis proj^er in their onward movement.

Antiperistalsis.— Cannon has found this to be the important motor phe-
nomenon of the large intestine. When the contents of the ileum are emptied
into the large intestine they are carried forward for some distance. Anti-
peristaltic waves then begin and run U})ward to the ileo-cir-cal valve. A new
si'niificance is thus made manifest for the valve. The antij)eristaltic waves
working toward the closed valve churn the fluid contents and promote
absorption by the mucosa. From time to time proper peristaltic contrac-
tions move the contents farther toward the rectum. Some investigators
like Nothnagel have not been able to demonstrate the antiperistaltic waves,
but the more modern methods of Cannon doubtless give more trustworthy

Mechanism of the Movements. — The movements of the bowel are caused
by the irritating effect of their contents: solid, liquid, or gaseous. When
empty, the intestines always are quiet. The irritation of the contents may
be mechanical or chemical. Indifferent substances may act by their tem-
perature or mere bulk. Of gases, hydrogen, oxygen, and nitrogen are indif-
ferent, while carbon dioxide, sulphuretted hydrogen, and the hydrocarbons
stimulate peristalsis.

Englemann believes the stimulus is transmitted through the bowel from
muscle cell to muscle cell, but Nothnagel insists that nervous influences are
always necessary, as do Gad and Liideritz. These nervous stimuli originate
in the plexuses of Auerbach and Meissner. Bayliss and Starling believe
the swaying movements are myogenic in origin.

External Nervous Mechanism. — Stimuli are carried to the intestine through
the vagus, and irritation of this nerve causes movements of the whole small
intestine and upper half of the large. Irritation of the splanchnic nerves
inhibits movements.

Cannon, however, has recently found that section of the splanchnics or
of the vagi and splanchnics disturbed normal peristalsis very little. Other
recent observers, like Bach and Ehrmann, found that one splanchnic nerve
is the motor nerve of the longitudinal fibers and the inhibitory nerve of the
circular fibers, while the vagus stimulates the circular and inhibits the longi-
tudinal. Bayliss and Starling deny any motor functions to the splanchnic.

The lower half of the colon and rectum are innervated through the inferior
mesenteric plexus and hypogastric plexus.

Pathological Peristalsis. — In pathological conditions various modifica-
tions of peristalsis may be met with. These may consist simply of excessive
contractions, but in some cases rolling movements (Nothnagel) cause a
contracted loop of the bowel to rotate like a wdieel and then relax, while in
other cases tonic contractions of a segment of bowel cause a stiffening and
narrow'ing of this portion, which may last sometime before it slowly relaxes.
Complete cessation of all movements may result from emotion, fright, fear,
etc., and is a common condition in certain diseases, such as peritonitis or
infarction from mesenteric embolism.

Varieties of Pathological Peristaltic Movements. — (1) Increased peristal-
sis; (2) tonic contractions; (3) antiperistalsis.


Increased 'peristalsis may occur: (1) When the bowel contents are
abnormal; (2) after the administration of purgatives; (3) after thermic or
chemical irritation; (4) after emotional shock; (5) in neurasthenia; (6) from
inflammations; (7) in cases of stricture of the bowel. Pathologically increased

Online LibraryWilliam OslerModern medicine : its theory and practice, in original contributions by American and foreign authors (Volume v. 5) → online text (page 44 of 126)