A bacteriologic study of 500 consecutive abortions, with treatment and results

A BAC1'ERIOLOGIC STUDY OF 500 CONSECUTIVE ABORTIONS, WITH TREATMEKT AND RESULTS* T. K. BRowN, B.S., M.S., M.D., AND GEoRGE A. HuNT, B.S., M.S., Pn.D., ST. Lours, Mo. (From the Department of Obstetrics and Gynecology, Washington Uni1Jersity School of Medicine and the St. lA"mis City Hospital)

T

HE purpose of this study is to emphasize the importance of the group of organisms, known as amwrobes, in 500 consecutive cases of abortion. 'rreatmen t and results vdll also be discussed.

Kronigt recognized and describefl anaerobi1• streptoeoeei in 1895. In 1905, Little2 reported a case of puerperal infection due to anaerobie streptococci oecurring in the service of Dr. Whitridge Williams. The importance of this grouiJ of organisms was not recognizeported, in 1910, a series of twenty-five eases with a 50 per cent mortality. He stated that the anaerobic streptococcms is a virulent pathogenic organism and cannot be regarded as a parasite, because once having invaded the tissues, thl'ombi or blood stream, it has pathologic properties. Schwarz and Dieekmann4 reported forty-two uterine cultures in 1925 and corrob· orated the findings of Schottmiiller. In an attempt to show the probable source of the anaerobk organisms, Soule and Brown;; (19:J2) studied the vaginal flora of normal clinic patients during preg>mnc;y am! found anaerobic growth in 60 per cent of the cases. These studies were continued and in 1934, Schwarz and BrownB reported an incidence of 83.3 per cent anaerobic organisms in 228 eases of puerperal infection treated in the previous ten yearR on the Obstetrical Service of the Washington University School of Medicine.

It seemed desirable to study another group of cases and compare results. Such a study has been accomplished on the Gynecological Service of the St. lJonis City Hospital, No. 1 (for white patients), in a period of eighteen months (500 cases). The treatment of the patients has been carried out under the direction of the Resident Gynecologist, and the bacteriology has been determined by the Bacteriology Laboratory of the hospital. Cases with positive uterine cultures were divided into three groups: (1) Aerobic, showing only aerobic growth. (2) Mixed,t showing both aerobic and anaerobic growth. (3) Anaerobic, showing only anaerobic growth. Table I shows the incidence of positive cultures and then stresses the predominance of the anaerobic group of organisms. Table II gives some idea as to the types of abortion encountered. If the doubtful group is considerecl as probably imluced (although not admitted as such), we find that m·er half of the abortions in this series are likely induced. *Read, by invitation, at the Sixty-First Annual Meeting of tho American Gynecological Society, Absecon, N. J., ::\.fay 25 to 27. 1936. 'tlncluded in the "mixed" cultures are the ca:oes in which the facultative anaerobes were demonstrated. This was done intentionally, in order to simplify the bacterip!oglcal study so that it could easily be applied 1n a clinical way.

SO,J,

805

STUDY OF CONSECUTIVE ABORTIONS

BROWN -HUNT :

Table III eombines the findings in Tables I and II and indicates the relatively high percentage of negative cultures in spontaneom abortions. The very high percentage of anaerobic growth in the eases with positive cultures in all types of a hortions is illustrated. Table IV shows tlu~ very low incidence of positive blood cultures and analyze~ them a<'cording to the types of uterine eultmes. It will be noted that one cm;e with an anaerobic uterine culture was found to han~ a blood culture wl1ich showed the presence of B. typhomw anrl tlw patit>nt wa~ imme(liately transferred to the lRolation Hospital. TABLE

I.

UTERI~~ CULTURES

An.alysis of 300 Positive Uterine Cult1weg COMBI:t-.TED 'rOTALS

%

CASES

24 Aerobic y· d 1Aerobic . 1136 u:e {AnaerobiC J 140 Anaerobic

CASES

8.0

53.:1

276

92.0

45.3 46.6

TABLE

II.

TYPE OF ABOttTIO!'-r

36.6% 19.0%

Induced Doubtful Spontaneous Stem Total of all eases TABLE TYPE OF· ABORTION

Pessaries (stem) 'l'otal

44.4%

III.

0

200

ANALYSIS 0~' UTEBINE CULTURES

MIXED

AEROBIC

NEGATIVE

Induced 56 (30.6%) Doubtful 29 (30.5%) Spontaneous 115 (52.5%)

%

I nO

10 (5.4%) 6 (6.3%) 8 (3.6%) 0

Uterine Cultures Negative Aerobic

Mixed

Anaerobie

(37.7%) (30.5%) (16.8%) (33.3%)

136

24 TABLE

69 29 37 1

IV.

ANAEROBIC

(26.2%) (32.6%) (26.9%) (66.6%)

48 31 59 2

'rOTAL TOTAL CASES ANAEROBES

183 95 219 3

92.1% 90.9% 92.3% 100%

500

140

BLOOD CULTURES

0

1 ()

fever)

Table V gives the analysis of aerobic organisms obtaine(l in the aerobic anti ruixed groups of uterine cultures. Considerable emphasis is given to the role played by the various types of staphylococci as frequent uterine contaminants.

806

AMERICAN JOURNAL OF OBSTETRICS AND GY~ECOLOGY

TABLE V. ANALYSIS

0~'

AEHOBIC

CULTURES

AEROBIC GROUP

B. coli Staphylococcus (unidentified) Staphylooocoos albu.s Staphylococ008 aureus Sporefor:ming bacillus Diphtheroids N onhemolytic streptococcus Hemolytic streptococcus Streptooocoos viridans Diplococcus (unidentified) Gonococcus Yeast ~ ot classified Total TABLE VI.

30

0

12

29

5 5 7

15 10 14

1

0

0 0

24

54

1

0

2

}

11 7 (2 died) 0 2 2

2 (2 died)

3 136

-·····-·--·-···--·---

HOSPITALIZA'l'ION AFTER TREATMENT

rTERINE CULTURES

CASES

Negative Aerobic Anaerobic Mixed Total

200 24 140 136 500

TABLE VII.

MIXED GROUP

2 0

DIAGNOSIS

_\1'\fl TRf~ATMJ>;N1' 011'

AVERAGE DAYS

5.8

5.5

6.6

7.4

6.3 PUERPERAT, JNF')iC1'10N

1. Differential diagnosis. 2. Preparation of patient: do not catheterize because of danger of urinary in· fection. 3. Intrauterine culture and examination of patient. 4. Emptying of the uterine cavity of debris with patient in seminarcosis, but with· out anesthesia by : a. Foerster's sponge-holding forceps. b. Uterine wiper. 5. Intrauterine douche with KMn04 (1-1,000) at 110• to 115• F., under 15 em. of water pressure or less (irrigating can resting on symphysis), using a large size Bozeman intrauterine douche cannula. 6. Administer ergot preparation, but avoid pituitary preparations. I. Follow-up treatment: Semi-Fowler position. Ice-bag to lower abdomen. Ergot preparations. Rapidly increasing diet (if no peritonitis). Low pressure vaginal douches, KMn04 (1-1,000) beginning the thinl day. Start getting up on fourth or fifth day. Transfuse for anemia. In cases with signs of peritoneal irritation: Nothing by mouth. Intravenous glucose. Subcutaneous saline. Transfusions, large and frequent. Wangenstein apparatus, No laxatives. Daily tap water enema. 8. If temperature remains elevated reexamine every second or third day to deter· mine presence of pelvic thrombophlebitis or beginning pelvic abscess formation. 9. Patient discharged at end of one week under conservative instructions. 10. Reexamine in two or three weeks.

BROWN-HUNT:

STUDY OF CONSECUTIVE ABORTIONS

807

Next in frequency is the Bacillus coli. The incidence of the Btreptooocous hemolyticus is very low, but wben it is the offender, the mortality rate i~ high. ln two cases the gonococcus was cultured as the offending organism. 'l'ahle VI illustrates the very short period of hospitalization requirPd after trPat, ment. This point is of great interf'st hoth to patient~ :mtl the ho~pital admini!
The treatment of puerperal infection has been divided into (1) conservative and (2) active. It appears that the conservative method would be an active method of treatment as has been used in this series of cases. In other words, the inactive treatment is not always conservative, because it may permit the spread of a local infection. When a patient in the puerperal state develops signs of infection, which cannot be accounted for after the usual diagnostic methods have been used to rule out all other possible infections, then it seems justifiable to investigat~ the uterine cavity. The surgical principles involved in the treatment of this series of cases are ( 1) drainage of an infected wound site and (2) debridement of a potentially infected cavity. This is done with a minimal amount of manipulation in order that further trauma may be avoided. We have demonstrated that this procedure can be safely performed by a junior interne under proper supervision. A general anesthetic is not to be used, because this would remove the inhibiting influence of the patient 'g response to pain which is usually inoicatire of too active manipulation. The patient should he under the influence of some sNlative so as not to he too apprehensive. Morphia (% gr.) with hyoscine hydro bromide ( 0.0005 gm.) followed l1y hyoscine (0.0005) in forty-five minutes are given so that the second do~11 iR given at least thirty minutes before the patient is prepared for examination. The perineum is prepared with 5 per cent neutral acriflavin in 10 per cent acetone and 50 per cent ethyl alcohol. The bladder is not catheterized beeauHe of thf' danger of contamination which may frequently superimpoRe a urinary infection. Drapes are placed. A Graves' vaginal speculum is used to obtain expoS\l!'P of the cervLx. If this is not satisfactory, large vaginal retractors will be necessary. The vagina and cervix are then prepared with the ahove solution oJ' 5 per cent neutral ucrifiavin. Treat the cervical canal with this solution, then dry carefully with sterile gauze. A culture is then obtained from the uterine cavity with a modified Little tube. If it is not possible to have careful bacteriologic in'lestigation of the material obtained, at least several smears of the material ~an be made and stained by the Gram method. This has been done routinely and found to check very well with the bacteriologic findings. Both aerobic and anaerobic blood agar slants should be made. The Wright7 anaerobie technic is used in search for anaerobic organisms. After the culture has been obtained, a gentle bimanual examination is done to determine evidence of any spread of the uterine infection, pelvic abscess or thrombophlebitis. This is also done without an anesthetic in order that the patient's reaction to pr.in will limit the extent of the examination. Too much pressure will not be used, which might cause rupture of an abscess internally. The uterine cavity is very carefully lDvestigated with a Foerster's sponge-holding forcep~, plain jaw. Dilatation of the cervix is usually sufficient to permit this. The sponge forceps is inserted closed to the depth desired, opened, closed, and removed to see

808

AMERICAN JOURNAL OF OBSTETRICS AND GYNECOLOGY

The forceps should not be advanced with the jaws open, beeaus11 of the danger of gra~ping tht> uterine muscle. The uterine wall is next explored Yery sy~tematir~tlly awl with gn'at g·entlenellR hy means of a "uterine wiper'' (formt>rly eallerl a \·aginal deprosAor) mueh as a windshield •.viper wipes the surface of the windshiel.J, without marring the surface. (NO curettage is done.) The uterus is usually found to be relaxed in suclt cases. After mechanical removal of any retained debris, the patient is given an intrauterine douche using a Bozeman's extra large intrauterine douche nozzle. This instrument permits the free exit of the solution from the uterine cavity without any pressure being established. Two liters of a 1:1000 solution of potassium permanganate in sterile water at 105° to 110° F. are used. This solution is usually acidulated with 50 c.c. N/1 sulphmic acid. The bottom of the douche can is held at the level of the symphysis so that the water pressure of the solution as it enters the douche nozzle is 15 em. or less. Such a douche results in ( 1) removal of small bits of tissue remaining after mechanical emptying of the cavity, (2) firm contraction of the uterus, including the cervix, so that bleeding is controlled and the sinuses are closed, (3) elimination of the putrid discharge which is characteristic of the anaerobic infections. Table VII outlines the treatment and postoperative care. if any tissue has fallen within the jaws.

We agree with Schottmiiller that the manner of spread of such contamination is by way of the endometrium; the organisms invade the uterine wall, parametrium, and thrombosed veins. Therefore, the earlier the local endometritis is treated, the more prompt is the cure. Less complications develop and the period of convalescence is much shortened. The possibility of future sterility is greatly lessened as evidenced by patients returning shortly with another abortion and others presenting themselves for delivery at term. The thrombosed veins in contradistinction to the circulating blood offer a fine culture medium. The proliferating organisms desire hemoglobin as their food and low oxygen tension in their environment. These organisms usually have a definite proteolytic tendency and are able to disintegrate thrombi so that small particles break off and gain access to the blood stream and otl1er organs, particularly the lungs. TABLE

VIII.

MORTALITY

PERCENTAGE TOTAL DEATHS CASES 1.4 7 500 and the admission upon complications inflammatory serious had Six of these cases seventh was almost exsanguinated. ~'YPES

OF t:TERJNE INFECTION

Aerobic Mixed Anaerobic

4

1

BACTERIOLOGY OF FATAL

Staphylococcus aureus Anaerobic streptococcus Hemolytic streptocoecus

1

2 4

14.2% 28.5% 57.1%

~7

The Streptococcus hemolyticus was cultured from the uterus in 9 instances among this series of 500 cases-an incidence of 1.8 per cent. Only 4 of those eases terminated fatally.-a mortality of 44.4 per cent.

BROWN-HUNT:

STUDY OF CONSECUTIVE ABORTIONS

80!1

The proteolytic properties of anaerobic streptococci probably are responsible for the marked loss of hemoglobin in patients with this type of infection. Therefore, one is able to see the rationale of large and frequent blood transfusions iu the treatment iu such cases. Since treatment has been instituted at an earlier time in the course of the infection, the demand for transfusions has greatly diminished. 'l'he

101

100

qq

10

II Fig. 1.

'"'

13

disease seldom runs the prolonged or chronic course formerly expected. Leucopenia is a common finding in cases of puerperal infection which are uncomplicated. RESULTS

Fig. 1 illustrates a typical temperature chart, with the prompt return to normal nfter treatment. Table VIII outlines the mortality figures and we observe the hopelessness at the timA of admission of the patients who died. The cases are analyzed as to type

Peritonitis Septicemia Peritonitis Perforation of uterus

Mixed (Hemolytic Strep.) Mixed (Anaerobic Strep.)

2 4

., "

Peritonitis septicemia Peritonitis Septicen1ia

Pos. Pos. Pos.

Self-induced (Medicine dropper)

Criminal (Midwife)

Self-induced (Catheter)

D. W. A-4031 3 days

S. G. A-10239 13 days

N. S. A-17168 8 days

Septicemia Early pneumonia

Peritonitis Multiple lung abscesses Aerobic (Hemolytic Strep.)

2

Exsanguinated

Keg.

Self-induced (Catheter)

N.J. 19808 12 days

Aerobic (Hemolytic Strep.)

Septicemia Peritonitis Lung abscess, Lt.

Mixed (Hemolytic Strep.)

4

Peritonitis septicemia

Pos.

Criminal (Midwife)

E. 8. 17104 24 days

> t<

Septicemia

Pos.

Self-induced (Catheter 30 da.)

P. C. 14344 1 day

Septicemia

Mixed (Staph. aureus)

0

§ Peritonitis Multiple lung abscesses

Anaerobic (Anaerobic Strep.)

8

Pelvic abscess

>1

0 0

t<

0

a

zi?:i

>1

0

zt:J

>

0 m

:;>j

>-3

t'j

>-3

m

!::;!

0

0 ":!

:>;! ~

0...

z

TYPE OF INFECTION

TRANSFUSIONS

Neg.

COMPLICATION

>

r')

~

t'j

i!:::

> CAUSE OF DEATH (CHECKED BY AUTOPGY)

Criminal (Midwife)

TYPE OE' ABORTION

IX.

F. H. 11653 32 days

HISTORY

BLOOD CULTURE

TABLE

DEATHS

I-'

m 0

BROWN-HUNT:

STUDY OF CONSECUTIVE ABORTIONS

81]

of uterine infection and also as to the predominant organism. 'l'he relative un· importance of the StreptocociYUS hemolyticus as to morbidity is stressed, and then it9 great importance as to mortality is emphasized. The situation in this particular type of infection is by no means hopeless and trentmf'nt should be similar in all types of cases, irrespective of bacteriologic :findings. Table IX analyzes the fatal eases, showing the hospital days and other im" portant :findings. Autopsy was obtained in every instanre to cheek the cause of death. CONCLUSIONS

1. Anaerobic organisms play a predominant role in the bacterial contamination of the uterine cavity following abortion as they are shown to be present in 92 per cent of the patiPnts with positive cultures in this series of 500 consecutive cases. 2. This finding has been obtained by a different group of clinical assistants aided by another laboratory corps, but checks very closely with the findings of the Department of Obstetrics and Gynecology of Washington University School of Medicine. 3. Sixty per cent of uterine cultures are positive. 4. The incidence of StreptococctfS hernoZyHcus as the infecting organism was very low (1.8 per cent), but its importance as a factor in mortality (57.1 per cent) must not be overlooked. 5. The small number of positive blood cultures should indicate its minor importance as a diagnostic sign. It means much more as to prognosis. 6. The period of hospitalization is very short, average 6.3 days after treatment. 7. Treatment should be early and active-culture, gentle evacuation of the uterine cavity, followed by an intrauterine douche. 8. Cases coming to fatal termination were moribund upon admission. (I wish to express my thanks to the Resident Staff of St. Louis City Hospital, No. 1, for theh· excellent cooperation.) REFERENCES

(1) Kronig: Zentralbl. £. Gynik. 19: 409, 1895. (2) Little, H. M.: Bull. Johns Hopkins Hosp. 16: 136, 1905. (3) Sohottm1i,ller, H.: Mitt. a. d. Grenzgeb. d. Med. u. Chlr. 21: 450, 1910. (4) Sohwa.rz, 0. H., and Dieckmann, W. J.: South. M. J. 19: D., ama Brown, T. K.: AM. J. OBST. & GYNEC. 23: 532, 470, 1926. (5) Soule, 1932. (6) Sohwa.rz, 0. H., and Brown, T. K.: AM. J. OBST. & GYNEC. 31: 379, 1936. (7) Wright, J. H.: J. Boston Soc. Med. Sc. 5: 114, 1900.

s.

630 SOUTH KINGSHIGHWAY BOULEVARD DISCUSSION

DR. WILLIAM J. DIECKMANN, CHICAGO, !LL.-ln evaluating the management of abortion eases we have always considered only the mortality. In his closing remarks I hope Dr. Brown will say something about the follow-up of these eases with reference to subsequent sterility, incapacity, subsequent abortions, and ectopic pregnancy as a result of the infection, and something abont fibrotic uteri that may develop due to the subinvolution.

812

AMERICAN JOURNAL OF OBSTETRICS AND GYNECOLOGY

The importance of anaerobic streptococei is shown by his report. Whenever uterine cultures are grown both aerobically and anaerobically, it is noted, usually with surprise, that the majority of the organisms are obligate anaerobes. At the Queen Charlotte's Isolation Hospital, where some of the best research work on puer· peral infection is being done, about one-third of the infecting organisms are anaerobic streptococci. At the Chicago Lying-In Hospital, our incidence of abortions is small, but it is interesting that in a group of 25 cases of postabortal and puerperal in· fections, ten or 40 per cent were due to anaerobic streptococci. In 70 uterine cui· tures, 40 per cent were anaerobic streptococci, 14 per ('ent mixed and, a very interest· ing thing, only 6 per cent hemolytic streptococci. In the subsequent discussion, I hope that the question of radical versus conserva· tive treatment will not be introduced. \Vhut Dr. Brown has done is not radical. Adair and Davis in their work on the ergot preparations have introduced bags into the uterus in more than 300 patiE'nts with no patient having a temperature of 38° for longer than twenty·four hours. Koff at .Tohns Hopkins introduced a bag in 70 cases with only one patient developing subsequent puerperal infection. The uterus can be invaded without danger, if proper technic is used. I think it is of importance to know what type of organism is present, and particularly with the anaerobic streptococci to remove the fragments of placenta that are left. It is not a curettage that Dr. Brown has advised. The purpose of the douche is to stimulate the uterus to contral't. There iR no hope of washing out the bacteria. DR. OTTO H. SCHWARZ, ST. Lours, Mo.-Various maternal mortality reports have stated that the number of deaths from puerperal infections was entirely too great and that the cause of these deaths was entirely due to neglect. I have main· tained that such a sweeping statement was a mistake, because anerobic infection would occur in many cases whether or not the patient was contaminated. Puerperal infections are due to two types of organisms, those harbored by the patient and those that are introduced. The anerobic are the chief cause of the endog· enous infections. They can develop under certain conditions and can cause all the conditions of pathogenicity, and our problem today as obstetricians is not so much in preventing the exogenous infections by good technic as in preventing these endogenous infections. Nevertheless our problem is still twofold, the endogenous and the exogenous infections. Of course, with bad technic the exogenous prevail, but with good technic it is the endogenous infections that we must bear in mind. WhP.n we do as Dr. Brown has done we ~hall see fewer cases of puerperal infection in well-regulated maternity hospitals. DR. BROWN (closing) .-That the infection is favored by inadequate drainage is demonstrated very thoroughly in the many cases that we have had. If we can establish drainage, get the blood clots out of the cervical canal and empty the uterus, the patient responds very quickly with a normal temperature and no complications develop. We have had no serious complications in these 500 patients treated by this method. We have had patients come back after several months with subsequent abortions, and quite a number of patients return within a year for delivery. We do try to educate the patient against the use of abortion, not by preaching but by trying to show the dangers. Many of them come back for subsequent delivery, which I think is of some aid in the lowering of maternal mortality. Subsequent examination shows the pelvis to be absolutely negative without masses, tenderness, or induration, and the menstrual cycle returill! to normal. Since the early

DICKINSON-HARTMAN:

CERVIX OF RHESUS ~t:ONKEY AND WOMAN

813

treatment of these patients, thrombophlebitis is almost unheard of. The patient either comes in with it already developed or she does not get it. Transfusions have been diminished in number tremendously since active treat· ment. The patient immediately improves, and she regenerates her blood very quickly as soon as the infection and bleeding are arrested. We do not do a curettage. We empty the uterus and wipe very, very gently. I do not believe in wiping with the fingers because in doing so it is necessary to push the uterus down in order to reach it and get the finger in. In doing this you will squeeze the uterus as a sponge and disseminate the infection.

SIMILARITY IN CERVIX OF RHESVS :VIONKEY AND W01\IAN* RoBERT

L.

J;'.A.C.S., NEw YoRK, N. Y., AND G. HARTMAN, M.A .. PH.D., BALTIMORE, MD.

DICKINSON, M.D.,

CARJ,

(Prom National Co·mmittee on Maternat Health, Z.nc., and Ca1·ne.gi.e Institution of Washington)

O .JUDGE by thirty adults in the colony of the Carnegie Institution in Baltimore, examined March 6, 1935, the difference in form and pathologic behavior between the vaginal portion of the cPrvix of the macaque and that of woman is almost altogether one of size, a matter of four-tenths in diameter. Hence, in tbis field biology and gynecology may find teamwork profitable. To permit easy visual comparison, drawings to scale were made by one of us (D.). These have been set over against examples in women, each opposite a human cervix which, by employing the above average, is scaled down to that of the monkey. The human types were selected from more than a thousand sketches in the office records of the senior author of this report. These cervix drawings had been made from life, usually life-size, duly checked by measurement, and often in color. This habit developed because of the experience of fifty years that much time could be saved by making life-size sketches as compared with writing entries. By visualized comparison, the similarity is here shown to be astonishingly exact, so mueh so that, in the accompanying figures, one will not be able to tell which is whieh save in four instances. The human and the simian por·tio vaginal£s will be seen to agree in conformation and invagination, laceration and inflammation, eversion and erosion, secretion and vascularity, ·1:njantilism and asymmetry. Cystic degeneration and polyp, cancer and certain infections alone were missing in this short series. Three differences appear. These are: 1. Facile dilatability of the untorn lower cervical canal and of each external os, a condition almost unknown in women.

T

*Read at the Sixty-Fir•st Annual Meeting: of the Amet·iean Gynecological Soeiet)·, Absecon, N. J .. May 25 to 27, 1936.