December, 1951 425
Primary Postpartum Hemorrhage
GEORGE F. MELODY, M.D., San Francisco
generally known, are here considered in the light of
SUMMARY present knowledge:
Postpartum hemorrhage is the outstanding Uterine Atony
cause of maternal mortality, and redoubt-
a Myometrial exhaustion, or what is more com-
able contributor to puerperal death from monly called uterine atony, is undoubtedly the most
other causes, notably infection and renal fail- frequent cause of postpartum hemorrhage. The usual
ure. The clinical situations in which hemor- predisposing factors are prolonged or precipitate
rhage is liable to occur must be better labor, overdistention of the uterus, antepartum hem-
known, so that anticipatory and preventive orrhage, deep anesthesia, and mismanagement of the
measures can be taken. Recent knowledge third stage. Prolonged labor is usually owing to
about defibrinated blood in women with de- cephalopelvic disproportion, to uterine inertia, or to
generative changes at the placental site must the premature and injudicious use of analgesics or
be incorporated in the thinking and practice conduction anesthesia. Cephalopelvic disproportion
of physicians dealing with obstetrical cases. should be recognized in good time, and never per-
The indications, limitations, and hazards of mitted to eventuate in desultory labor. Primary iner-
the various anesthetic methods available for tia should be treated by support and stimulation, in
parturient women should be carefully con- contrast to secondary inertia which should be man-
sidered in the circumstances of each case. aged by rest and sedation. Labor can be prolonged
many hours by starting continuous caudal analgesia
prematurely; labor can be brought to a complete
PRIMARY postpartum hemorrhage is excessive standstill for hours by the untimely administration
bleeding from the genital tract during the first of "terminal" spinal anesthesia. Postpartum bleed-
24 hours after birth. A blood loss in excess of ing is almost inevitable after the relaxation induced
600 cc. is usually regarded as hemorrhage. The by deep anesthesia, particularly ether. Uterine atony
bleeding may occur before or after the third stage of follows precipitate labor just about as often as pro-
labor is completed. Although the blood loss is usu- longed labor; and one should be cautious about the
ally external, in certain instances it may be more or induction of labor, as induced labor is often pro-
less concealed. The bleeding may be sudden and furi- longed or precipitate.
ous, erratically repeated, or a slowly exsanguinating Overdistention of the uterus from multiple preg-
trickle. It may follow spontaneous or operative vag- nancy, hydramnion, or excessively large fetus, is
inal delivery, cesarean section, or cesarean-hysterec- notoriously liable to be followed by postpartum hem-
tomy. Postpartum hemorrhage occurs in approxi- orrhage. In such circumstances, the uterus should be
mately 5 per cent of all deliveries. Puerperal hemor- decompressed slowly and general anesthesia used
rhage after the first 24 hours is known as late post- cautiously. Cross-matched blood should be in readi-
partum hemorrhage and has been considered else- ness.
where.6 Trauma of Delivery
There are both immediate and delayed hazards in Injury to the uterovaginal tract continues to be a
postpartum hemorrhage. The immediate threat is major cause of postpartum hemorrhage. The ana-
death from loss of blood and irreversible shock. De- tomical lesions are cervical and vaginal lacerations,
layed complications are puerperal infection, trans- including rupture of the vagina, paravaginal hema-
fusion reactions, renal failure from lower nephron tomas, and rupture of the uterus. With most of these
nephrosis or cortical necrosis, thromboembolic dis- accidents the bleeding is external, in contrast to the
ease, and finally, Simmonds' disease, or pituitary more or less concealed hemorrhage which occurs
cachexia. with ruptured uterus and paravaginal hematomas.
Most cases of postpartum hemorrhage can be an- The external bleeding is typically a slow, steady
ticipated, more can be prevented, and all can be trickle of bright (arterial) blood, in distinction to
treated. No woman should die of hemorrhage, the the darker blood which escapes from the placental
number one cause of maternal mortality, and a for- site. The possibility of these complications should be
midable contributor to puerperal death from other thought of after precipitate delivery, after difficult
causes. The best treatment for postpartum hemor- obstetrical maneuvers, and with the parturition of
rhage is its anticipation and prevention. The several elderly primiparae.
major causes of postpartum hemorrhage, although Laceration of the cervix is inevitable if delivery
From the Department of Obstetrics and Gynecology,
is attempted through an incompletely dilated os.
University of California School of Medicine, San Francisco. Manual or digital completion of dilatation has no
Presented before the Section on Obstetrics and Gyne-
cology at the 80th Annual Session of the California Med-
place in modern obstetrics; if conditions are propi-
ical Association, Los Angeles, May 13 to 15, 1951. tious for vaginal delivery except for a persistent
426 CALIFORNIA MEDICINE Vol. 75, No. 6
remaining ring of cervix, Diihrssen's incision undertaken here, but this much should be said: The
should be made. The first clue to a contracted mid- conduct of the third stage should begin during the
pelvis may be brisk bleeding from puncture wounds second stage by the deliberately slow delivery of the
over pronminent ischial spines. Lacerations of the fetus to allowv time for the uterus to accommodate
vaginal vault are not uncommon after breech ex- itself to its diminished volume and thus peel off the
traction. forceps rotations, and the inept use of placenta in a physiological manner. The Crede ma-
Kielland forceps. A rotary or torsion force may neuver is dangerous and obsolete. No attempts to
actually rupture the vagina (colporrhexis) by creat- deliver the placenta should be made until it has
ing a circular tear around the cervix, usually in the separated, and the fundus is firmly retracted. Squeez-
posterior fornix, which communicates with the peri- ing or kneading the uterus traumatizes the myo-
toneal cavity. The bleeding with colporrhexis is usu- metrium, dislodges thrombi from the placental site,
ally furious because of laceration of the internal and invites inversion. If and when deep anesthesia
pudendal, middle hemorrhoidal, or vaginal artery. is used, it should be stopped the moment the fetus
No woman should be returned to bed if there is a is delivered. Mismanagement of the third stage of
steady trickle of blood. The cervix and entire vaginal labor was responsible for four out of seven cases of
vault should be inspected if there is suspicion of acute or subacute inversion recently reported from
trauma. Sometimes a physician has to rely on tac- the Sloane Hospital for Women.4 Vigorous bleeding
tile sense to locate a laceration which is hidden by at any time calls for immediate extraction of the
the invagination of its edges. Lacerations should be )lacenta. In addition to massage, bimanual com-
visualized (or palpated) and promptly repaired. A pression, and use of oxytocics, manual exploration
gently placed vaginal pack is often adequate for should always be done if bleeding persists after the
superficial abrasions, but if the vagina is packed too placenta is out, and it should be done before a
tightly bleeding may be aggravated by forcing the tampon is placed. Retained placental tissue, a sub-
ends of the severed vessel apart; moreover, over- mucous myoma, rupture or inversion may be noted.
packing the vagina may cause neurogenic shock. An emergency unit of type IV (0), Rh-negative
In rupture of the vagina, if hemostasis cannot be ob- blood should be kept refrigerated on every lying-in
tained per vaginam, ligation of the hypogastric ar- suite. The equipment necessary for laparotomy
tery may be necessary. should be immediately available in or adjacent to
Vulvovaginal and paravaginal hematomas are not the delivery pavilion. It is well to transfer post-
uncommon and are usually not preventable. Back- partum women to a recovery room for the first 24
bleeding from a retracted submucosal vessel in the hours, so that they may be observed constantly by
episiotomy or a laceration is the usual mechanism. competent personnel. This would do much to prevent
By starting the repair two centimeters above the serious secondary relaxations, and the "flooditn-s of
apex of the episiotomy or laceration, the likelihood the newly laid women."
of "missing" a submucosal vessel is lessened. Hema- 1)efibrinated Blood
tomas should be recognized early, evacuated, and
bleeding points secured. If the process is left to Thanks mainly to the Boston group of obstetri-
nature, the extravasation of blood may extend past cians, more and more is being learned about the
the broad ligament to the lumbar plate. (levelopment during pregnancy of conditions mak-
Traumatic rupture of the uterus occurs much ing the blood incoagulable. Many physicians have
more often than vital statistics would indicate.
observed cases in which the patient bled to death
Deaths attributed to "obstetrical shock" are often because the blood had lost the ability to clot. Hem-
due to unrecognized or disavowed rupture. Rupture orlhage not only from the placental site, but from
should be thought of after every difficult forceps de- surgical incisions, venipunctures. and even from the
livery. version and extraction, and breech extrac- nasopharynx has been observed; and in such in-
stances death may have occurred despite heroic
tion.2 If there is any doubt about the integrity of the measures, including ample blood replacement and
uterus, manual exploration should be done. If rup- lhysterectomy. Since the report of Maloney anid his
ture is noted, the uterus should be packed, and a associates5 in 1949 a significant literature has accu-
continuous transfusion given until arrangements are inulated on the subject of acquired afibrinogenemia
completed for laparotomy. Hysterectomy is neces- of pregnancy. It is now known that whenever there
sai-v in about half of such cases, although in many is necrosis or degeneration at the placental site,
instances the uterus can be saved by repairing the there is a chance that particulate matter. thromubo-
Errors of the Third Stage
p)lastin, toxins, and anticoagulants can be absorbed
into the maternal circulation. In extreme cases the
More women die during the third stage of labor woman's blood may become totally defibrinated an(d
than during the first two stages combined. Proper quite incoagulable. The hematologic findings include
management of the third stage is perhaps the obste- critically low values for fibrinogen and prothrombin.
trician's greatest responsibility. The third stage is and often a circulating fibrinolysin is present. A sam-
all too often the fatal climax of a neglected first stage ple of such venous blood when incubated will either
and a traumatic second stage worsened by the ef- not clot at all, or else the clot will be so fragile that
fects of injudicious anesthesia. Full description of the slightest vibration will cause it to disintegrate.
the proper management of the third stage cannot be When] the plasma fibrinogen level is reduced from
December, 1951 PRIMARY POSTPARTUM HEMORRHAGE 427
Figure Decidua basalis of same speciien as Figurie
1 Note extensive necosis of the dlecidual cells and round
cell infilti ation. The dea(l fetus hcad been carried in uter o
foi foui %v eeks pi'ioi'to delivery. X 43.
Figur e 1.-Placental site six hours postpartum. Hyste- decidual junction (Figure 2), the source of the anti-
rectomiiy done for intractable hem-elorrhage due to afibr ino- coagulants. Even though the placenta had been im-
genenmia. Note sinuses filled withi liquid blood. X 17. planted high in the corpus the patient was admitted
the normal concentration of 0.5 gm. per 100 ml. to in early labor with prolapse of the placenta, no doubt
0.2 gm. per 100 ml. the circulating blood is inco- due to the combined effects of necrosis of the pla-
agulable. cenital site plus Braxton-Hicks contractions. In
microscopic examination of the placenta (Figure 3)
It is now known that abruptio placentae, fetal endarteritic obliteration of the villous vessels and
death in utero, toxemia of pregnancy, and sometimes atrophy of the chorionic epithelium were noted.
placenta praevia may be associated with defibrina- Venous blood drawn at the time the patient was ad-
tion of the maternal blood, depending upon the mitted to hospital did not clot even after five days.
absorption of anticoagulants into the circulation.
Women with these complications should be studied Antepartum Hemorrhage
hematologically as well as obstetrically. In addition Antepartum bleeding predisposes to postpartum
to replacement of whole blood, specific deficiencies hemorrhage. The value of prophylactic blood trans-
of fibrinogen and prothrombin must be eliminated fusions in cases of placenta praevia and abruptio
if uncontrollable postpartum hemorrhage is to be placentae was emphasized as long ago as 1919 by
averted. Fibrinogen in the form of Fraction I, from Bill2 of Cleveland. The vicious circle of bleeding
which the virus of homologous serum jaundice has followed by uterine relaxation, followed by more
been removed by nitrogen mustard, is packaged in
flasks of 2 gm. (2,000 mg.), and may be dissolved
in 300 to 500 cc. of 5 per cent glucose or normal
saline solution for intravenous administration.3
It is important that a uterine bag, or metreuryn-
ter, not be used in such cases. A uterine bag makes
a closed system, so that when contractions begin,
particulate matter, break-down products and antico-
agulants are swept into the maternal sinuses, thence
into the circulation, with risk of sudden death from
amniotic fluid embolism, or aggravation of the in-
coagulable blood status. Simple rupture of the mem-
branes, and the draining off of amniotic fluid with
its noxious compounds, is a rational precautionary
measure, and should be done when feasible.
In a recent case in which antepartum fetal death
occurred at six months due to the Rh-factor, afibri-
nogenemia developed in the patient during the four
weeks that the dead fetus was carried in utero. In
examination of the specimen obtained at hysterec-
tomy, which was six hours postpartum for intrac-
table hemorrhage, unclotted blood in the placental Figure 3.-Choiionic X illi from samne case as Figures 1
sinuses was observed (Figure 1), and there were and 2. Placenta gr adually became detached and eventu-
ally pr olapsed. Note disappearance of chor ionic vessels
extensive degenerative changes in the trophoblastic- and epithelium. X 100.
428 CALIFORNIA MEDICINE Vol. 75, No.6
bleeding until irreversible shock sets in, should be blood loss frequently amounts to between 500 and
anticipated, and never permitted to occur. Blood lost 1,000 cc. The sources of bleeding are the uterine
antepartum should be replaced antepartum or intra- incision, including transection of the uterine vessels
partum to prevent postpartum hemorrhage. on one side, unrecognized uterine rupture, and the
The interests of mother and fetus will best be placental site. If the placenta is attached to the de-
served in cases of placenta praevia and abruptio by fective scar from a previous cesarean section, the
ample blood replacement, and simple rupture of the bleeding may be furious and intractable; SlemonsO
membranes, or cesarean section, depending upon the advised hysterectomy in such instances.
particular case. The author regards the use of dilat- Hemorrhage should be anticipated in every cesar-
ing uterine bags as dangerous in either complica- ean section. There should be a functioning venocly-
tion. In both conditions there is the risk of amniotic sis with a No. 18 gauge needle started before the
fluid embolism and of provoking the defibrinated anesthetic is begun. In addition to preoperative
blood syndrome. In placenta praevia there is the blood replacement there should be a unit of cross-
added hazard of furious postpartum hemorrhage matched blood available in the operating room for
from a laceration of the friable cervix or bleeding even the "routine" case. If the operation is done for
from a sinus in the placental site due to the trauma antepartum hemorrhage or toxemia, spinal anesthe-
of a dilating bag. A bleeding sinus in the lower seg- sia should be avoided because of the inherent risk
ment encountered at cesarean section can be visual- of hypotension. Loss of blood will be reduced if
ized and transfixed. In both placenta praevia and the surgeon uses local infiltration. The placental site
abruptio placentae, hysterectomy is occasionally should be spared the trauma of rough sponging and
tiecessary as a life-saving measuire. the suction tip. Blood will be saved and morbidity
Toxemias of Pregnancy reduced if myomectomy is avoided. In event of in-
tractable bleeding from profound atony, placenta
Postpartum hemorrhage should be anticipated in praevia, abruptio with uteroplacental apoplexy, rup-
all cases of "toxic" pregnancy, for the following rea- ture or perforation (from "failed forceps"), hyste-
sons: (1) Vasomotor collapse with consequent ute- rectomy may be necessary.
rine atony and hemorrhage frequently follows deliv- The author has observed two cases in which the
ery in such cases, for the vasomotor response to patient underwent cesarean-hysterectomy and sev-
emptying the uterus is unpredictable. Therefore, eral hours after operation lapsed into severe shock
spinal and general anesthesia should be avoided; from internal hemorrhage. The increased vascularity
local infiltration is preferable. (2) Often in "toxic" of the pelvic structures in the pregnant state makes
pregnancy there are degenerative changes at the cesarean-hysterectomy a formidable undertaking.
trophoblastic-decidual junction, and the risk of de- All vascular pedicles should be doubly ligated, as
fibrinated blood is considerable. (3) As was first the risk of a major vessel's retracting is consider-
pointed out by Slemons8 in 1933, in toxemia of able. A progressively falling volume of erythrocytes
pregnancy the myometrium may be diffusely infil- as determined by hematocrit is the best laboratory
trated with intermuscular hemorrhages identical with indication of internal hemorrhage. If the physician
those observed in cases of abruptio placentae. A is convinced there is continuing bleeding it is man-
uterus thus affected may not retract postpartum, and datory that he reopen the abdomen. Total hysterec-
hysterectomy may have to be done. In the case re- tomy is fraught with special hazards in the pregnant
ported by Slemons, hysterectomy was necessary nine state, and, like all surgical operations during gravid-
hours after cesarean section, because of repeated ity, should be avoided unless really necessary. Bleed-
severe hemorrhages due to a blood-infiltrated myo- ing from the vaginal cuff and internal hemorrhage
metrium. from vascular pedicles are calculated risks from this
Uterine Tumors operation.
Uterine neoplasms are only rarely a cause of post- 490 Post Street.
partum hemorrhage. Multiple intramural myomas or REFERENCES
extensive adenomyosis9 may interfere with proper 1. Bill, A. H., Barney, W. R., and Melody, G. F.: Rupture
retraction. More important, however, is a submucous of the uterus, Am. J. Obst. & Gynec., 47:712-717, May 1944.
myoma which has undergone red degeneration. Such 2. Bill, A. H.: Observations on the problem of hemor-
a myoma can become agglutinated to the placenta, rhage in obstetrical cases, Am. J. of Obst. & Dis. of Women
and prevent its complete separation in the third & Child., 80:1-4, 1919.
stage. If, upon inspection of the delivered placenta, 3. Diamond, L. K.: Personal communication, March 1951.
a clean, punched-out defect is noted, a submucous 4. Fenton, A. H., and Singh, B. P.: Acute puerperal in-
myoma should be suspected and manual exploration version of the uterus, Obst. & Gynec. Survey, 5:781-795,
carried out. Exceedingly rare causes of postpartum Dec. 1950.
bleeding are cervical carcinoma, cervical polyps, ec- 5. Maloney, W. C., Egan, W. J., and Gorman, A. J.: Ac-
quired afibrinogenemia in pregnancy, New Eng. Jour. Med.,
topic decidua7 which may involve the cervix, and 240:596-598, April 14, 1949.
hemangioma of the uterine wall. 6. Melody, G. F.: An inquiry into late postpartum hemor-
Cesarean Section and Cesarean-Hysterectomy rhage, Am. Jour. Surg., 78:821-842, Dec. 1949.
7. MIelody, G. F.: Deciduation and massive hemorrhage
Postpartum hemorrhage is a major cause of death of the omentum in the final month of pregnancy, West. Jour.
following cesarean section. It is probable that the Surg. Obst. & Gynec., 58:460-462, Sept. 1950.
December, 1951 PRIMARY POSTPARTUM HEMORRHAGE 429
8. Slemons, J. M.: Hemorrhage following cesarean sec- partum hemorrhage. The practice of placing the patient on
tion, Am. Jour. Obst. & Gynec., 26:656-661, Nov. 1933. iron therapy without such a routine hemoglobin evaluation
9. Szenes, A.: Internal adenomyosis causing severe hem- only creates a false sense of security. It should be severely
orrhage in the third stage of labor and in the puerperium, condemned.
Arch. f. Gynak., 134:546, 1928. For the most part, I agree wholeheartedly with Dr.
Melody. However, in the matter of handling the third stage
of labor, I still feel that the Cred6 maneuver, when properly
Discussion by DONALD W. DECARLE, M.D., San Francisco carried out, is a valuable procedure. If one's hand is on the
fundus during the third stage, certainly there is far less
To me, there is but one obstetrical complication in the chance of the uterus' filling with blood concealed behind an
presence of which I must, at times, admit to actual panic. undelivered placenta. The same is true of the Crede maneu-
That complication is postpartum hemorrhage. This is par- ver when an oxytocic agent is employed at the onset of
ticularly true in the presence of bleeding which one can the third stage. Otherwise, the placenta may become incar-
not only see but also hear. I have known men far more cerated.
experienced than I to be so completely unnerved in its pres- Dr. Melody has very definitely changed our way of think-
ence that clear, sound judgment is lost. It is the complica- ing on at least two points. In the first place, employment of
tion which must always be feared as long as it is potentially the Voorhees bag has been largely replaced by more modern
present at every delivery. procedures. The danger of introduction of toxic material
As a subject for discussion, it has, therefore, been wisely into the uterine sinuses by the use of the uterine bag should
chosen. It can never be exhausted as long as it leads the lead to complete and universal discarding of this procedure.
list of fatal complications in the field of obstetrics. As Dr. Secondly, the same holds true,of our attitude toward the
Melody has pointed out, in order to combat postpartum fetus which has died in utero. It is apparently much safer
hemorrhage, one must be conversant with every phase of the to empty the uterus as soon as a diagnosis of intrauterine
conditions underlying it. This necessitates routine and fre- death has been made.
quent rechecking of the blood throughout every pregnancy. Through the efforts of the essayist, a supply of Fraction I
Potential or actual anemia as distinguished from the physio- is now available at Children's Hospital in San Francisco. It
logical hydremia of pregnancy can be recognized and prop- was my recent misfortune to watch as a patient died of
erly treated before the onset of labor only by such a method. bleeding which could not be controlled even by hysterec-
I deem this most important, since it can mean the margin tomy. Had this material been available at the time, such
of difference between life and death in the event of post- an outcome might have been prevented.