- Email: [email protected]
A critical analysis of cesarean section in a large municipal hospital
RFI”ERFKi(XS 1 1
(1) Dodds, E. C’., Goldberg, L., l,nwso~r, T7’., lou7 Robinson, K.: Sature, Lontluu 141: 247, 1938. (2) Dodds, E. C., tccu~son, TV., md Noblr. T:. L.: T,ewet 1: 13SH, 1935. (3) Pdmtr, A., nwl Zuckrrmntz, S. : Lnncet 1: R%? 1939 (4) Bishop, I’. IV. F., Boycott, ;V., und Zzcckermvn, S.: hncet 1: 5, 1939. ’ (5)’ iTintrrton, m7. K., und MacGregor, J. N. : Brit. M. J. 1: IO, 1939. (0) Geist, A’;. H., and Salmon, 1:. d.: New York J. Med. 39: 1759, 1939. (7) I;‘arnngot, .7.: Bull. SW. d ‘obst. vt de gyn&. 28: 160, 1939. (8) Z’uscher, H.: Klin. \Vchnschr. 10: X08, 1959. i!)) Knrwky. K. ,I.: Fhuth. M. J. 32: 813, 1939. (10, P~reisseckr, E.: Zentrslbl. f. Qynlik. 63: 1904, 1939. (11) Rnhon, 1”. J., and Franl;, E. T.: Prw. Sot. Exper. Hiol. B A.\lt:d. 33: 612, 1936. (12) Frank, R. .l’.. ant/ Xdmon, I:. ,J.: Ibicl. 33: 311, 19%. (131 Albright, P. : Endocrinology 20: “4: 193H. (14) .Joilrs. .M. S., trnd MacGw,qor, .I. x. : Lancet 2: 974, 1936. (15) (1~ pn5~s.j ( I fi) .lIncHryr7c. I’. N.. Ftwdmn~~, II.. and Loeffel, E.: .T. A. M. A. 113: 2320, 1939. (17) Shorr, II‘., Bob%nzon. F. B.. trw/ Papnicolaou, G. A’. : .J. A. M. A. 113: “31$ 1939. (18) Zh~.rion, (‘. I,., W,K$ I?q/lt, B. T.: J. A. 11. A. 113: 2318, 1939. (lY! I,orsrr. .I. 3.: firit. hf. J. 1: 13, 1939. (20) Geist, N. Ii., cfnd Ralmo~~, r. ./.: AM. .J. ORS’I~. B C+vsw. 38: 39t’. I!hW.
A CRITICAL ABR~IHALI
B.
ANALYSIS OF (:ESAREAN SECTION LARGE MViW ‘IPAL HOSPITA I,” TAMIS,
M.D.,
F.A.(1.8., NEW
lFrotn
the Obstetrid
AND
YORK,
Sr,rcicr
Iv.
I~frr,To~
1).
IN a4
KLEIS,
iI’Ll->..
I-.
C’ifg
, 3lorri,scmin
Hospifot,
A
PERUSAL of the current literature ~n~c~ys the impression that the maternal risk of c.vsa.t.ran section is less thali 3 per cent.‘, 3, 8, 11,12, 1.1 Occasionally a maternal mortnlibp xltr of’ 5 per cent and 6 per cent is acknowlec1getl.” The maternal mortality studies of large communities.“* i’ia ‘; which more truly reflect the results for obstetrics as it is genc~rally practiced, yield much higher rates anal substantiate the contention oi Watson1 that when all cases are included, the caesarean mort.ality fog the United Stat,ps is probably f’ron~ 10 to 15 per crnt. If any advance is to be matlr in the reduction of cesarean mortality, it is important, that, every institx~tion shoultl, from time to time, review its records and present the facts for discussion. whether they be good or bad. Before judgment is passed, however. the critic should bear in mind that lower cesarean mortality rates are to be expected where elective sections predomi~tai t’. and that the mortality percentage *Read Academy
at a meeting of Medicine,
of the February
Section
27. 1940.
of
Obstetrics
and
c:ynecology,
The
New
Pork
TAMIS
AND
KLEIN:
CESAREAN
251
SECTION
is likely to rise when emergency and other nonelective sections eonstitute the greater proportion of cases. This report is based on a review of all cesarean operations at Morrisania City Hospital, one of several municipal hospitals in the City of New York. The Obstetrical Service was organized in October, 1929. Up to Dec. 31, 1937 there were 14,489 deliveries. More than one-half of the obstetric admissions are referred from the Out-patient Antepartum Clinic. The remainder either apply directly or are transferred from other municipal hospitals. About 20 t,o 30 per cent of the in-patients are colored. The service is supervised by t,he attending staff, resident, and assistant resident. STATISTICBL
RiSUMf
One hundred and twelve cesarean sections were performed, an incidence of 0.8 per cent, or 1 section in 129 deliveries. The average cesarean incidence in New York City is two to four times higher, or about 1 section in 25 to 50 deliveries. It would appear from this that we have either been very conservative in t.he management of the various obstetric difficulties for which cesarean section might be indicated, or that we have been rather fortunate in having a minimum of patients requiring this procedure. We believe that the explanation of our low incidence is due to a conservative policy. In fact from a consideration of the data which follow, we may have carried this policy too far in some instances. Our uncorrected maternal morality rate for cesarean section is 10.7 per cent. However, this high death rate is not reflected in our uncorrected general maternal mortality rate for the same period, which is 5.3 per thousand living births. This resuIt compares favorably with that of other simiIar institutions. The indications for which ccsarean sections were done at our institution are enumerated in Table I. TABLE
I.
INDICATIONS
FOR
CESAREAN
SECTION NUMBER CASES
Cephalopelvic disproportion Contracted inlet 42 Contracted outlet 2 Normal pelvis 3 Previous cesarean section Placenta previa Central 15 Partial 4 Premature separation of placenta Complete 9 Partial 1 Obstructing tumor Cardiac disease Eclampsia Rupture of uterus Previous attempt at delivery Vaginal stenosis Cervical stenosis Total *One
case
included
Cephalopelvic per cent). is too large for (41
in
group
of
“previous
47
6
19 19
1 2
10
2
6 4 1 2” 1 1 3 112 cesarean
Disproportion.-This constituted The problem for the obstetrician the pelvis or the pelvis too small
MATERNAL DEATHS
1
12
section.”
the largest of the groups is identical whether the baby for the baby. Of the 47 pa-
252
AMERICAN
JOURNAL
OF
OBSTETRICS
AXD
GYNECOLOGY
tients operated upon for this indication, 6 women died, a maternal mortality rate of 12.7 per cent for the group. In 11 instances the disproportion was absolute, and elective sections were done immediately on admission with no maternal or fetal deaths. In the 36 patients with relative disproportion a test of labor was permitted before eesarean section was performed. Rome hospitals have a hard and fast rule that, unless rapid progress is made section is to be perin six to eight hours, with intact membranes, a cesarean formed. With this yardstick the patient is operated upon under ideal conditions with a prognosis for recovery not less than 97 per cent. By the same token it is also quite obvious that many women so readily sectioned might conceivably have delivered spontaneously or with the assistance of forceps if given a longer test Our practice has been to individualize each case. The test lasted six hours, or even twenty-four hours, depending on the amount of progress made. Vaginal examinations were prohibited. Rectal examinations were done as often as required. During the past few years our judgment has been somewhat influenced by x-ray pelvimetry. The method used is that described by Weitzner,? which gives a rapid and accurate measurement of the conjugata vera as well as a comprehensive view of the presenting part in its relation to the pelvic inlet. By following this plan, most of our patients with relative cephalopelvic disproportion were successfully delivered by the vaginal route. Tn 1038, for example, of 47 full-term primiparas who began labor with an unengaged presenting part, only 2 required a cesarean section. The rest were delivered spontaneously or with forceps, with a loss of only- 2 babies and no maternal mortality. If insufficient progress was made during a test of labor, the Imby was deor by craniotomy (if dead). It is siglivered by cesarean section (if alive), nificant and interesting to note that during the period covered 1ly this report craniotomy was resorted to onlp once. Of the 36 patients with relative disproportion delivered by cesarean section, 6 died. Four of these deaths were due to peritonitis. The type of cesarean operation chosen was an important factor in causing the high mortality. Approximately one-half of the 47 patients with cephalopelvic disproportion were operated upon by the classical and the other half by the low segment, met,hod (Table II). These operations were performed 1j.v the same mcmhers of the obTABLE
II.
I
~EPHALOpELvlC
DISPROPORTION,
CLASSICAL
cases / Primiparas __Hours in labor Number of cases ITiFlFternal deaths 1 - ! t’”
LOW
TYPE
OF OPERATION PERITONEAL ESCLUSTON
SEGMENT
Number of ______~--
A
Peritonitis Paralytic Pneumonia *Died.
ileus
1
3’
1
! 1* / -
L
-,-
_____-_--
/ 1”
1
lx ! lx 1 -
1
-
1 1
,
;., 1% I”
1 -
-
stetric staff in the same hospital for the same indication, and are therefore comparable for evaluating the superiority of one operation over the other. Those treated by the classical cesarean section had the advantage of intact membranes and shorter durat,ion of labor. In spite of this, 3 patients (13.6 per
TAMIS
AND
KLEIN:
CESAREAN
253
SECTIOS
cent) died of peritonitis. Of the low segment group, only died of the .same cause. The security against peritoneal segment operation was three times greater than that for section.
1 patient (4 per cent) infection for the low the classical cesarean
Previous Cesarean, Sectiolz.-Of the total number of patients admitted to the hospital with a history of a previous section, 19 women required a repeated operation. Eighteen were registered in our Prenatal Clinic and were instructed Thus it happens that most of the to enter the hospita1 at the first sign of labor. patients were admitted under ideal conditions, with membranes intact and in labor less than six hours (Table III). One patient died, a mortality of 5.3 per cent for this group.
TABLE III. Hours
CESAREAN
CLASSICAL 0 O-12 F--g---I-Y ] -
labor
Maternal deaths Hours membranes tured
PREVIOUS
rup-
Int.
O-6
-G--T----Maternal Stillbirths Neonatal Peritonitis *Ruptured
deaths deaths
I
0
SECTION,
12-24
0
-
-
(i-12
Int.
1
2 /-
TYPE
OF OPERATIOK-
TWO FLAP O-l? 13-24 --i--- 1
O-6 _ (I
U
6-12 Int. - -yY-
PORRO O-12 J*
12-24 -
O-6
6-l 2
1
0
1 (died‘) uterus.
Our attitude toward patients who have been previously sectioned is more elastic than is permitted by the adage “once If the a eesarean, always a eesarean. ” previous section was performed for indications other than obstructed labor, and if the post-partum course was not complicated bv morbidity, the patient was allowed a test of labor, provided she remained under observation from the very onset. If rapid and progressive advancement appeared within a relatively short time, delivery was allowed to be completed normally or assisted with forceps. If labor slowed up far any cause, a cesarean section was done. For example, during the year 1934 alone, 7 patients were admitted with a history of having had a previous section, Three patients delivered spontaneously and one was assisted by forceps. The remaining three patients were subjected to section immediately on admission because of a contracted pelvis.6 One patient, a negress, disobeyed instructions and remained at home while in labor with the hope of delivering spontaneously. A few hours later the uterine scar gave way and she was brought to the hospital in shock due to a ruptured uterus and an intra-abdominal hemorrhage. A Porro section was done. The fetus was stillborn. The patient had a stormy convalescence but fortunately recovered. The patient who died had a large incisional hernia which was repaired immediately after the section was completed. Autopsy examination indicated an extension of a Staphylococcw C~UTPZCRinfection of the abdominal wound into the general peritoneal cavity. The uterine incision was uninvolved. This death was preventable. It taught us to restrict our surgery during cesarean section to actual necessities, and postpone to some later date other procedures which increase the risk to the patient. Placenta Previa.-(Table IV.) Nineteen patients with placenta previa were treated by cesarean section. Of these, 15 were central and 4 were partial. Two mothers died, one of embolism and one of peritonitis. The maternal mortality for this group was 10.5 per cent. The patient who died of peritonitis had a classical section within three hours of the initial hemorrhage and immediately after only one vaginal examination. Six infants died. All were premature. Only 3 could he considered viable.
-7----
Number Multipams Months’ Number Stillbirth Neonatal Maternal
ot cases
-Embolrsm
--..
-
-
-
I..
-.-
gest.ation of casea
-... ^. ‘y-y;
L.L--
death death -----
,j
iI
/
j.
j
:: -1:
~-.?
_
../_. --.
1:
.~_ ..~ I:!.__ c
:
g
..I.
.~;;._... -.7.
:
1,; .
- / .~ ___.
.-
.-,/-
-
The superior results claimed hy the advocates of the Dublin routine for the t,reatment of complete separation of the placenta have impressed us considerably. The techniqur consists of rupturing the membranes, packing cervix and vagina, repeated injections of pituitrin, application of a Spanish windlass, replacement of blood loss by lwge transfusions, and prompt delivery when full dilatation is obtained. TVe Our applied this routine in tire last 3 cases of ablatio placentae seen on t,he wards. courage faltered after observing 1 ho tirhf cir.se for twelve hours 2s the patient seemed to be going downhill. She was delivered by a classical cesarean section. When the next 2 patients were admitted we determined to carry out the method to completion, and did. Both patients made an uneventful recovery.
TAMIS
AND
It was a trying experience, cooperation on the part of the method should be used only and where a trained personnel interest of the patient is best Obstructing Tumors (Table variety of pelvic masses found of cervical fibromyomas, and cyst of the ovary, pelvic kidney, TABT,E VI.
KLEIN:
CESAREAN
255
SECTION
for it meant constant surveillance and concerted nursing and interne staffs. We feel that the Dublin where blood transfusions can be readily obtained is always available. Lacking such facilities, the served by a cesarean section. VI).-This group is mainly of interest because of the There were 2 cases obstructing vaginal delivery. one each of embryonal tumor of the ovary, dermoid and retroperitoneal sarcoma. OBSTRUCTISG
Cervical fibromyoma Embryonal tumor of ovary Dermoid cyst of ovary Pelvic kidney Retroueritoneal sarcoma
PELVIC -.
TUUORS 2 1 I 1 1
cases case case case case
The one neonatal death resulted from preThere were no maternal deaths. maturity. Cardiac Disease.-Cesarean section as means of terminating labor in patients with heart disease was resorted to in four instances. One ease was a primiparous cardiac who had been hospitalized for twenty-four days because of decompensation. The other three were multiparas with histories of previous deeompensations in whom sterilization was advocated and a cesarean section done, therefore, at term to effect this procedure. All these patients were functionally classified as Class 2b. In spite of a poor myocardium they were able to withstand the strain of the operation. One patient eviseerat.ed on the twenty-first day postoperative, then developed a bronchopneumonia and pleural effusion, but survived. The case illustrates the difficulty of judging the cardiac reserve until the heart is actually under stress. The classical section was employed in each instance because of the rapidity with which it can be performed and because of the access to the Fallopian tubes. There were no maternal or fetal deaths in this small group. A point worth emphasizing is the choice of anesthesia and the manner of its administration. Patients with heart disease are very prone to devefop pulmonary For this reason the anesthesia should be given by an expert complications. anesthetist. We prefer the use of gas-oxygen inhalation, EcZampsia.-We have not found it necessary to resort to cesarean section for pre-eclampsia. Our experience with a modified Stroganoff regime followed bj induction of labor has proved most satisfactory. Perhaps the same conservative attitude may also be taken with the convulsive states. We had an unfortunate experience with the only patient with eclampsia who was treated surgically. This woman, a nonclinic primipara, was admitted to the hospital in coma, with frequent convulsions and was practicall> moribund. Examination revealed her to be at term and not in labor. A classical cesarean section was done under ethylene anesthesia but the patient died twenty-four hours later. Rupture of the Uterus.-Included in this series are two cases of rupture of the uterus. One occurred through a previous eesarean sear. The other had a spontaneous rupture of the uterus. This patient, a thirty-six-year-old multipara, was admitted to the hospital in shock after having had abdominal pains for eight hours. The cervix was I finger dilated, membranes were intact, presenting part unengaged, and there wa,s no external evidence of bleeding. She was given a transfusion but failed to rally. The diagnosis of rupture of the uterus was then suspected and a laparotomy performed. Free blood was present in the peritoneal cavity and found to originate from a tear in the anterior wall of the uterus. A stillborn baby was delivered through this tear and the uterus closed in layers. Convalescence was uneventful.
256
AMERICAN
JOURNAL
OF
OBSTETRICS
AND
GYNECOLOGY
Unswccessfzll Attempt at Delivery.-The danger of performing a cesarean section, after previous attempts at delivery have been unsuccessful, is appreciable. Kerr4 quotes maternal mortality rates under such conditions as high as 15 to 20 per cent. He advocates craniotomy as the method of choice, because the fetus is dead as a rule, and the risk is less unless the pelvis is markedly contracted. Our experience at Morrisania City Hospital with this type of patient has been limited to 2 patients, both of whom were admitted in shock. One was treated by craniotomy. The other was subjected to a Hirst peritoneal exclusion operation because of extensive bladder injury which followed an attempted version in another hospital. The trauma to the bladder precluded any possibility of doing Both these patients recovered after a a Latzko or a low segment operation. stormy convalescence. Stenosis of the Peginrc and of the Cervix.-A douche of undiluted lysol in an attempt to produce an abortion resulted in an extensive cicatrization of the vagina in one case which made cesarean section mandatory. stenosis due to a previous trachelectomy were deThree patients with cervical livered by vaginal section in one instance and by abdominal seetion in two.
There were no maternal Mater& tion are
Mortality.-The listed in Table
or fetal deaths in this group. VII.
cause of onr Peritonitis VII.
TABLE .___ I
7 (58.3
per
cent)
of
resort
since
death
,
2_”
the
12 deaths.
appeared
seefor
OF DEATH FOR OPERATION
ZZ I -i'YPE
,
OPERATION-
-
At
least
two
of
these
deaths
sidered as preventable, because autopsy revealed the infection troduced at the time of operation. The two patients who died of toxemia were critically ill at mission aud died in less than twenty-four hours. The eclamptic a suitable risk for abdominal surgery. The operation was used last
cesarean accounting
1 -
: I 1
Toxemia
CAUSES
INDICATION
I
Peritonitis Pulmonary embolus Paralytic ileus Bronchopneumonia
maternal deaths following took the heaviest toll,
may
be con-
to have been inthe time of adpatient was not as a measure of
inevitable. DISCUSSION
The prevention of peritonitis deserves intensive investigation as it causes from 40 to 50 per cent of deaths following cesarean section.5r 8l I5 The relation of ruptured membranes, duration of labor, and vaginal examination to the spread of infection during labor requires no discussion. Their importance has been adequately stressed by previous studies and needs no further emphasis here. Two other factors, however, have played a prominent role in our incidence of peritonitis, namely : introduction of infection at the operating table ; and type of operation selected. The mostly
incidence of our wound infections, which fortunately consisted of mild stitch abscesses,was close to 20 per cent in spite of the
TAMIS
AND
KLEIN:
CESAREAN
SECTIOiV
257
most rigorous precautions in preparing the abdomen for operation. Autopsies on two of the seven deaths from peritonitis clearly indicated an extension of the infection from the abdominal wound to the general peritoneal cavity. Tn checking over the possible sources for this break in technique, the operating room itself came under suspicion. The hospital has two major operating rooms for the use of all surgical cases. As less than 50 per cent of our cesarean sections were elective, not infrequently the operation followed shortly aft,er a potentially or actually infected surgical case. The nursing 1,ersonnel remained the same for both operations. It is not difficult, therefore, to imagine cross-infection under such circumstances. We believe, with DeLee and others, that the obvious remedy is a complete operating room unit for the exclusive use of cesarean section, with nursing personnel devoted only to this division. Arrangements to incorporate such a unit on our obstet,ric floor have already been made. The second factor of importance affecting the incidence of peritonitis is related to the type of cesarean operation selected. The classical cesarean section was performed on 70 patients. Six died of peritonitis, a maternal death rate of 8.5 per cent. The low segment operation was performed 37 times with one death from peritonitis, a maternal death rate of 2.7 per cent. disproportion, ” the condiAs was described under ‘ ’ cephalopelvic tions under which the low segment operation was performed were less favorable than those of the classical group. Nevertheless, notwithstanding this disadvantage, the low segment operation gave three times more security against the occurrence of peritonitis. The classical cesarean section should be utilized only where speed is necessary or where it is essential to perform other operative procedures. It is safer for the pat,ient, however, to resort to the low segment operation, t,he Latzko extraperitoneal operation, or the approach to the lower uterine segment recently described by Waters.g The final criterion of obstetric management does not. rest on the end results of any one operative procedure but on the outcome of all deliveries, normal or otherwise. This information may be obtained from general maternal mortalit,y rates which are uncorrected. In New York City, this rate is said to be 6.1 per thousand living births.5 In the very best hands in English hospitals, the rate is 5.8.1o ,4t Morrisania City Hospital, the rate is 5.3, notwithstanding a relatively high maternal mortality for cesarean operations. Therefore, we feel justified in maintaining our present policy toward the obstetric patient, except for the few changes already mentioned. SUMMARY
AKD
CONCLUSIONS
1. An analysis of cesarean sections at Morrisania City Hospital is presented. The incidence of this operation is 0.8 per cent, or one in 129 deliveries. The uncorrected cesarean maternal mortality rate is
32IXEJ)
Al)ENOCARClNO~~JA ANI) SQI-AklOUS CARCINO31 h ( ) F’ THE ITTERI’S”
IR,Z (“. &INNER,
M.I).,
C~~:JIJA
,\NL) vJ01-i~ R. ~%J)oNAI,D, M.i)., &IIXK. fhi Muyo Clitric)
RO(WES.I’ER,
(From
M
ALIGNANT neoplasms are ovcusionally found in which there is differentiation into a type of cell entirely foreign to t.he organ in which it is primary. Home of these neoplasms van be explained as arising in teratomas while in ot,ht>rs no such explanation is possible. Primary osteogenic sarcomas have hvcn described in the thyroid gland (Broders and Pembert.on’) and in the breast.. Synnmons cell carcinomas are of frequent occurrence in thv bronchus and yet. squamous epithelium is not normally found here. Synamous epithelium has been described in the gall bladder ( Broders2 ) , prostak (Burrows and Kennaway:’ ) , duct,s of the pancreas, and gastrointMina1 tract (Yinson and &ode&). *Abridgment of thesis submittwl b> IIt-. Skinner to the faculty of the Graduate School of the IlniversitY of Minnesota. in partial fulfillment of the requirements for the degree of M. S. in surgery. Submitted for publication April 26, 1940.
(1) Dodds, E. C’., Goldberg, L., l,nwso~r, T7’., lou7 Robinson, K.: Sature, Lontluu 141: 247, 1938. (2) Dodds, E. C., tccu~son, TV., md Noblr. T:. L.: T,ewet 1: 13SH, 1935. (3) Pdmtr, A., nwl Zuckrrmntz, S. : Lnncet 1: R%? 1939 (4) Bishop, I’. IV. F., Boycott, ;V., und Zzcckermvn, S.: hncet 1: 5, 1939. ’ (5)’ iTintrrton, m7. K., und MacGregor, J. N. : Brit. M. J. 1: IO, 1939. (0) Geist, A’;. H., and Salmon, 1:. d.: New York J. Med. 39: 1759, 1939. (7) I;‘arnngot, .7.: Bull. SW. d ‘obst. vt de gyn&. 28: 160, 1939. (8) Z’uscher, H.: Klin. \Vchnschr. 10: X08, 1959. i!)) Knrwky. K. ,I.: Fhuth. M. J. 32: 813, 1939. (10, P~reisseckr, E.: Zentrslbl. f. Qynlik. 63: 1904, 1939. (11) Rnhon, 1”. J., and Franl;, E. T.: Prw. Sot. Exper. Hiol. B A.\lt:d. 33: 612, 1936. (12) Frank, R. .l’.. ant/ Xdmon, I:. ,J.: Ibicl. 33: 311, 19%. (131 Albright, P. : Endocrinology 20: “4: 193H. (14) .Joilrs. .M. S., trnd MacGw,qor, .I. x. : Lancet 2: 974, 1936. (15) (1~ pn5~s.j ( I fi) .lIncHryr7c. I’. N.. Ftwdmn~~, II.. and Loeffel, E.: .T. A. M. A. 113: 2320, 1939. (17) Shorr, II‘., Bob%nzon. F. B.. trw/ Papnicolaou, G. A’. : .J. A. M. A. 113: “31$ 1939. (18) Zh~.rion, (‘. I,., W,K$ I?q/lt, B. T.: J. A. 11. A. 113: 2318, 1939. (lY! I,orsrr. .I. 3.: firit. hf. J. 1: 13, 1939. (20) Geist, N. Ii., cfnd Ralmo~~, r. ./.: AM. .J. ORS’I~. B C+vsw. 38: 39t’. I!hW.
A CRITICAL ABR~IHALI
B.
ANALYSIS OF (:ESAREAN SECTION LARGE MViW ‘IPAL HOSPITA I,” TAMIS,
M.D.,
F.A.(1.8., NEW
lFrotn
the Obstetrid
AND
YORK,
Sr,rcicr
Iv.
I~frr,To~
1).
IN a4
KLEIS,
iI’Ll->..
I-.
C’ifg
, 3lorri,scmin
Hospifot,
A
PERUSAL of the current literature ~n~c~ys the impression that the maternal risk of c.vsa.t.ran section is less thali 3 per cent.‘, 3, 8, 11,12, 1.1 Occasionally a maternal mortnlibp xltr of’ 5 per cent and 6 per cent is acknowlec1getl.” The maternal mortality studies of large communities.“* i’ia ‘; which more truly reflect the results for obstetrics as it is genc~rally practiced, yield much higher rates anal substantiate the contention oi Watson1 that when all cases are included, the caesarean mort.ality fog the United Stat,ps is probably f’ron~ 10 to 15 per crnt. If any advance is to be matlr in the reduction of cesarean mortality, it is important, that, every institx~tion shoultl, from time to time, review its records and present the facts for discussion. whether they be good or bad. Before judgment is passed, however. the critic should bear in mind that lower cesarean mortality rates are to be expected where elective sections predomi~tai t’. and that the mortality percentage *Read Academy
at a meeting of Medicine,
of the February
Section
27. 1940.
of
Obstetrics
and
c:ynecology,
The
New
Pork
TAMIS
AND
KLEIN:
CESAREAN
251
SECTION
is likely to rise when emergency and other nonelective sections eonstitute the greater proportion of cases. This report is based on a review of all cesarean operations at Morrisania City Hospital, one of several municipal hospitals in the City of New York. The Obstetrical Service was organized in October, 1929. Up to Dec. 31, 1937 there were 14,489 deliveries. More than one-half of the obstetric admissions are referred from the Out-patient Antepartum Clinic. The remainder either apply directly or are transferred from other municipal hospitals. About 20 t,o 30 per cent of the in-patients are colored. The service is supervised by t,he attending staff, resident, and assistant resident. STATISTICBL
RiSUMf
One hundred and twelve cesarean sections were performed, an incidence of 0.8 per cent, or 1 section in 129 deliveries. The average cesarean incidence in New York City is two to four times higher, or about 1 section in 25 to 50 deliveries. It would appear from this that we have either been very conservative in t.he management of the various obstetric difficulties for which cesarean section might be indicated, or that we have been rather fortunate in having a minimum of patients requiring this procedure. We believe that the explanation of our low incidence is due to a conservative policy. In fact from a consideration of the data which follow, we may have carried this policy too far in some instances. Our uncorrected maternal morality rate for cesarean section is 10.7 per cent. However, this high death rate is not reflected in our uncorrected general maternal mortality rate for the same period, which is 5.3 per thousand living births. This resuIt compares favorably with that of other simiIar institutions. The indications for which ccsarean sections were done at our institution are enumerated in Table I. TABLE
I.
INDICATIONS
FOR
CESAREAN
SECTION NUMBER CASES
Cephalopelvic disproportion Contracted inlet 42 Contracted outlet 2 Normal pelvis 3 Previous cesarean section Placenta previa Central 15 Partial 4 Premature separation of placenta Complete 9 Partial 1 Obstructing tumor Cardiac disease Eclampsia Rupture of uterus Previous attempt at delivery Vaginal stenosis Cervical stenosis Total *One
case
included
Cephalopelvic per cent). is too large for (41
in
group
of
“previous
47
6
19 19
1 2
10
2
6 4 1 2” 1 1 3 112 cesarean
Disproportion.-This constituted The problem for the obstetrician the pelvis or the pelvis too small
MATERNAL DEATHS
1
12
section.”
the largest of the groups is identical whether the baby for the baby. Of the 47 pa-
252
AMERICAN
JOURNAL
OF
OBSTETRICS
AXD
GYNECOLOGY
tients operated upon for this indication, 6 women died, a maternal mortality rate of 12.7 per cent for the group. In 11 instances the disproportion was absolute, and elective sections were done immediately on admission with no maternal or fetal deaths. In the 36 patients with relative disproportion a test of labor was permitted before eesarean section was performed. Rome hospitals have a hard and fast rule that, unless rapid progress is made section is to be perin six to eight hours, with intact membranes, a cesarean formed. With this yardstick the patient is operated upon under ideal conditions with a prognosis for recovery not less than 97 per cent. By the same token it is also quite obvious that many women so readily sectioned might conceivably have delivered spontaneously or with the assistance of forceps if given a longer test Our practice has been to individualize each case. The test lasted six hours, or even twenty-four hours, depending on the amount of progress made. Vaginal examinations were prohibited. Rectal examinations were done as often as required. During the past few years our judgment has been somewhat influenced by x-ray pelvimetry. The method used is that described by Weitzner,? which gives a rapid and accurate measurement of the conjugata vera as well as a comprehensive view of the presenting part in its relation to the pelvic inlet. By following this plan, most of our patients with relative cephalopelvic disproportion were successfully delivered by the vaginal route. Tn 1038, for example, of 47 full-term primiparas who began labor with an unengaged presenting part, only 2 required a cesarean section. The rest were delivered spontaneously or with forceps, with a loss of only- 2 babies and no maternal mortality. If insufficient progress was made during a test of labor, the Imby was deor by craniotomy (if dead). It is siglivered by cesarean section (if alive), nificant and interesting to note that during the period covered 1ly this report craniotomy was resorted to onlp once. Of the 36 patients with relative disproportion delivered by cesarean section, 6 died. Four of these deaths were due to peritonitis. The type of cesarean operation chosen was an important factor in causing the high mortality. Approximately one-half of the 47 patients with cephalopelvic disproportion were operated upon by the classical and the other half by the low segment, met,hod (Table II). These operations were performed 1j.v the same mcmhers of the obTABLE
II.
I
~EPHALOpELvlC
DISPROPORTION,
CLASSICAL
cases / Primiparas __Hours in labor Number of cases ITiFlFternal deaths 1 - ! t’”
LOW
TYPE
OF OPERATION PERITONEAL ESCLUSTON
SEGMENT
Number of ______~--
A
Peritonitis Paralytic Pneumonia *Died.
ileus
1
3’
1
! 1* / -
L
-,-
_____-_--
/ 1”
1
lx ! lx 1 -
1
-
1 1
,
;., 1% I”
1 -
-
stetric staff in the same hospital for the same indication, and are therefore comparable for evaluating the superiority of one operation over the other. Those treated by the classical cesarean section had the advantage of intact membranes and shorter durat,ion of labor. In spite of this, 3 patients (13.6 per
TAMIS
AND
KLEIN:
CESAREAN
253
SECTIOS
cent) died of peritonitis. Of the low segment group, only died of the .same cause. The security against peritoneal segment operation was three times greater than that for section.
1 patient (4 per cent) infection for the low the classical cesarean
Previous Cesarean, Sectiolz.-Of the total number of patients admitted to the hospital with a history of a previous section, 19 women required a repeated operation. Eighteen were registered in our Prenatal Clinic and were instructed Thus it happens that most of the to enter the hospita1 at the first sign of labor. patients were admitted under ideal conditions, with membranes intact and in labor less than six hours (Table III). One patient died, a mortality of 5.3 per cent for this group.
TABLE III. Hours
CESAREAN
CLASSICAL 0 O-12 F--g---I-Y ] -
labor
Maternal deaths Hours membranes tured
PREVIOUS
rup-
Int.
O-6
-G--T----Maternal Stillbirths Neonatal Peritonitis *Ruptured
deaths deaths
I
0
SECTION,
12-24
0
-
-
(i-12
Int.
1
2 /-
TYPE
OF OPERATIOK-
TWO FLAP O-l? 13-24 --i--- 1
O-6 _ (I
U
6-12 Int. - -yY-
PORRO O-12 J*
12-24 -
O-6
6-l 2
1
0
1 (died‘) uterus.
Our attitude toward patients who have been previously sectioned is more elastic than is permitted by the adage “once If the a eesarean, always a eesarean. ” previous section was performed for indications other than obstructed labor, and if the post-partum course was not complicated bv morbidity, the patient was allowed a test of labor, provided she remained under observation from the very onset. If rapid and progressive advancement appeared within a relatively short time, delivery was allowed to be completed normally or assisted with forceps. If labor slowed up far any cause, a cesarean section was done. For example, during the year 1934 alone, 7 patients were admitted with a history of having had a previous section, Three patients delivered spontaneously and one was assisted by forceps. The remaining three patients were subjected to section immediately on admission because of a contracted pelvis.6 One patient, a negress, disobeyed instructions and remained at home while in labor with the hope of delivering spontaneously. A few hours later the uterine scar gave way and she was brought to the hospital in shock due to a ruptured uterus and an intra-abdominal hemorrhage. A Porro section was done. The fetus was stillborn. The patient had a stormy convalescence but fortunately recovered. The patient who died had a large incisional hernia which was repaired immediately after the section was completed. Autopsy examination indicated an extension of a Staphylococcw C~UTPZCRinfection of the abdominal wound into the general peritoneal cavity. The uterine incision was uninvolved. This death was preventable. It taught us to restrict our surgery during cesarean section to actual necessities, and postpone to some later date other procedures which increase the risk to the patient. Placenta Previa.-(Table IV.) Nineteen patients with placenta previa were treated by cesarean section. Of these, 15 were central and 4 were partial. Two mothers died, one of embolism and one of peritonitis. The maternal mortality for this group was 10.5 per cent. The patient who died of peritonitis had a classical section within three hours of the initial hemorrhage and immediately after only one vaginal examination. Six infants died. All were premature. Only 3 could he considered viable.
-7----
Number Multipams Months’ Number Stillbirth Neonatal Maternal
ot cases
-Embolrsm
--..
-
-
-
I..
-.-
gest.ation of casea
-... ^. ‘y-y;
L.L--
death death -----
,j
iI
/
j.
j
:: -1:
~-.?
_
../_. --.
1:
.~_ ..~ I:!.__ c
:
g
..I.
.~;;._... -.7.
:
1,; .
- / .~ ___.
.-
.-,/-
-
The superior results claimed hy the advocates of the Dublin routine for the t,reatment of complete separation of the placenta have impressed us considerably. The techniqur consists of rupturing the membranes, packing cervix and vagina, repeated injections of pituitrin, application of a Spanish windlass, replacement of blood loss by lwge transfusions, and prompt delivery when full dilatation is obtained. TVe Our applied this routine in tire last 3 cases of ablatio placentae seen on t,he wards. courage faltered after observing 1 ho tirhf cir.se for twelve hours 2s the patient seemed to be going downhill. She was delivered by a classical cesarean section. When the next 2 patients were admitted we determined to carry out the method to completion, and did. Both patients made an uneventful recovery.
TAMIS
AND
It was a trying experience, cooperation on the part of the method should be used only and where a trained personnel interest of the patient is best Obstructing Tumors (Table variety of pelvic masses found of cervical fibromyomas, and cyst of the ovary, pelvic kidney, TABT,E VI.
KLEIN:
CESAREAN
255
SECTION
for it meant constant surveillance and concerted nursing and interne staffs. We feel that the Dublin where blood transfusions can be readily obtained is always available. Lacking such facilities, the served by a cesarean section. VI).-This group is mainly of interest because of the There were 2 cases obstructing vaginal delivery. one each of embryonal tumor of the ovary, dermoid and retroperitoneal sarcoma. OBSTRUCTISG
Cervical fibromyoma Embryonal tumor of ovary Dermoid cyst of ovary Pelvic kidney Retroueritoneal sarcoma
PELVIC -.
TUUORS 2 1 I 1 1
cases case case case case
The one neonatal death resulted from preThere were no maternal deaths. maturity. Cardiac Disease.-Cesarean section as means of terminating labor in patients with heart disease was resorted to in four instances. One ease was a primiparous cardiac who had been hospitalized for twenty-four days because of decompensation. The other three were multiparas with histories of previous deeompensations in whom sterilization was advocated and a cesarean section done, therefore, at term to effect this procedure. All these patients were functionally classified as Class 2b. In spite of a poor myocardium they were able to withstand the strain of the operation. One patient eviseerat.ed on the twenty-first day postoperative, then developed a bronchopneumonia and pleural effusion, but survived. The case illustrates the difficulty of judging the cardiac reserve until the heart is actually under stress. The classical section was employed in each instance because of the rapidity with which it can be performed and because of the access to the Fallopian tubes. There were no maternal or fetal deaths in this small group. A point worth emphasizing is the choice of anesthesia and the manner of its administration. Patients with heart disease are very prone to devefop pulmonary For this reason the anesthesia should be given by an expert complications. anesthetist. We prefer the use of gas-oxygen inhalation, EcZampsia.-We have not found it necessary to resort to cesarean section for pre-eclampsia. Our experience with a modified Stroganoff regime followed bj induction of labor has proved most satisfactory. Perhaps the same conservative attitude may also be taken with the convulsive states. We had an unfortunate experience with the only patient with eclampsia who was treated surgically. This woman, a nonclinic primipara, was admitted to the hospital in coma, with frequent convulsions and was practicall> moribund. Examination revealed her to be at term and not in labor. A classical cesarean section was done under ethylene anesthesia but the patient died twenty-four hours later. Rupture of the Uterus.-Included in this series are two cases of rupture of the uterus. One occurred through a previous eesarean sear. The other had a spontaneous rupture of the uterus. This patient, a thirty-six-year-old multipara, was admitted to the hospital in shock after having had abdominal pains for eight hours. The cervix was I finger dilated, membranes were intact, presenting part unengaged, and there wa,s no external evidence of bleeding. She was given a transfusion but failed to rally. The diagnosis of rupture of the uterus was then suspected and a laparotomy performed. Free blood was present in the peritoneal cavity and found to originate from a tear in the anterior wall of the uterus. A stillborn baby was delivered through this tear and the uterus closed in layers. Convalescence was uneventful.
256
AMERICAN
JOURNAL
OF
OBSTETRICS
AND
GYNECOLOGY
Unswccessfzll Attempt at Delivery.-The danger of performing a cesarean section, after previous attempts at delivery have been unsuccessful, is appreciable. Kerr4 quotes maternal mortality rates under such conditions as high as 15 to 20 per cent. He advocates craniotomy as the method of choice, because the fetus is dead as a rule, and the risk is less unless the pelvis is markedly contracted. Our experience at Morrisania City Hospital with this type of patient has been limited to 2 patients, both of whom were admitted in shock. One was treated by craniotomy. The other was subjected to a Hirst peritoneal exclusion operation because of extensive bladder injury which followed an attempted version in another hospital. The trauma to the bladder precluded any possibility of doing Both these patients recovered after a a Latzko or a low segment operation. stormy convalescence. Stenosis of the Peginrc and of the Cervix.-A douche of undiluted lysol in an attempt to produce an abortion resulted in an extensive cicatrization of the vagina in one case which made cesarean section mandatory. stenosis due to a previous trachelectomy were deThree patients with cervical livered by vaginal section in one instance and by abdominal seetion in two.
There were no maternal Mater& tion are
Mortality.-The listed in Table
or fetal deaths in this group. VII.
cause of onr Peritonitis VII.
TABLE .___ I
7 (58.3
per
cent)
of
resort
since
death
,
2_”
the
12 deaths.
appeared
seefor
OF DEATH FOR OPERATION
ZZ I -i'YPE
,
OPERATION-
-
At
least
two
of
these
deaths
sidered as preventable, because autopsy revealed the infection troduced at the time of operation. The two patients who died of toxemia were critically ill at mission aud died in less than twenty-four hours. The eclamptic a suitable risk for abdominal surgery. The operation was used last
cesarean accounting
1 -
: I 1
Toxemia
CAUSES
INDICATION
I
Peritonitis Pulmonary embolus Paralytic ileus Bronchopneumonia
maternal deaths following took the heaviest toll,
may
be con-
to have been inthe time of adpatient was not as a measure of
inevitable. DISCUSSION
The prevention of peritonitis deserves intensive investigation as it causes from 40 to 50 per cent of deaths following cesarean section.5r 8l I5 The relation of ruptured membranes, duration of labor, and vaginal examination to the spread of infection during labor requires no discussion. Their importance has been adequately stressed by previous studies and needs no further emphasis here. Two other factors, however, have played a prominent role in our incidence of peritonitis, namely : introduction of infection at the operating table ; and type of operation selected. The mostly
incidence of our wound infections, which fortunately consisted of mild stitch abscesses,was close to 20 per cent in spite of the
TAMIS
AND
KLEIN:
CESAREAN
SECTIOiV
257
most rigorous precautions in preparing the abdomen for operation. Autopsies on two of the seven deaths from peritonitis clearly indicated an extension of the infection from the abdominal wound to the general peritoneal cavity. Tn checking over the possible sources for this break in technique, the operating room itself came under suspicion. The hospital has two major operating rooms for the use of all surgical cases. As less than 50 per cent of our cesarean sections were elective, not infrequently the operation followed shortly aft,er a potentially or actually infected surgical case. The nursing 1,ersonnel remained the same for both operations. It is not difficult, therefore, to imagine cross-infection under such circumstances. We believe, with DeLee and others, that the obvious remedy is a complete operating room unit for the exclusive use of cesarean section, with nursing personnel devoted only to this division. Arrangements to incorporate such a unit on our obstet,ric floor have already been made. The second factor of importance affecting the incidence of peritonitis is related to the type of cesarean operation selected. The classical cesarean section was performed on 70 patients. Six died of peritonitis, a maternal death rate of 8.5 per cent. The low segment operation was performed 37 times with one death from peritonitis, a maternal death rate of 2.7 per cent. disproportion, ” the condiAs was described under ‘ ’ cephalopelvic tions under which the low segment operation was performed were less favorable than those of the classical group. Nevertheless, notwithstanding this disadvantage, the low segment operation gave three times more security against the occurrence of peritonitis. The classical cesarean section should be utilized only where speed is necessary or where it is essential to perform other operative procedures. It is safer for the pat,ient, however, to resort to the low segment operation, t,he Latzko extraperitoneal operation, or the approach to the lower uterine segment recently described by Waters.g The final criterion of obstetric management does not. rest on the end results of any one operative procedure but on the outcome of all deliveries, normal or otherwise. This information may be obtained from general maternal mortalit,y rates which are uncorrected. In New York City, this rate is said to be 6.1 per thousand living births.5 In the very best hands in English hospitals, the rate is 5.8.1o ,4t Morrisania City Hospital, the rate is 5.3, notwithstanding a relatively high maternal mortality for cesarean operations. Therefore, we feel justified in maintaining our present policy toward the obstetric patient, except for the few changes already mentioned. SUMMARY
AKD
CONCLUSIONS
1. An analysis of cesarean sections at Morrisania City Hospital is presented. The incidence of this operation is 0.8 per cent, or one in 129 deliveries. The uncorrected cesarean maternal mortality rate is
32IXEJ)
Al)ENOCARClNO~~JA ANI) SQI-AklOUS CARCINO31 h ( ) F’ THE ITTERI’S”
IR,Z (“. &INNER,
M.I).,
C~~:JIJA
,\NL) vJ01-i~ R. ~%J)oNAI,D, M.i)., &IIXK. fhi Muyo Clitric)
RO(WES.I’ER,
(From
M
ALIGNANT neoplasms are ovcusionally found in which there is differentiation into a type of cell entirely foreign to t.he organ in which it is primary. Home of these neoplasms van be explained as arising in teratomas while in ot,ht>rs no such explanation is possible. Primary osteogenic sarcomas have hvcn described in the thyroid gland (Broders and Pembert.on’) and in the breast.. Synnmons cell carcinomas are of frequent occurrence in thv bronchus and yet. squamous epithelium is not normally found here. Synamous epithelium has been described in the gall bladder ( Broders2 ) , prostak (Burrows and Kennaway:’ ) , duct,s of the pancreas, and gastrointMina1 tract (Yinson and &ode&). *Abridgment of thesis submittwl b> IIt-. Skinner to the faculty of the Graduate School of the IlniversitY of Minnesota. in partial fulfillment of the requirements for the degree of M. S. in surgery. Submitted for publication April 26, 1940.