THE 
DEATH-RATE OF LYING-IN HOSPITALS 
IN THE 
UNITED STATES. 
BY 
OBSTETRICIAN TO THE MATERNITY HOSPITAL OF PHILADELPHIA, AND OBSTETRICAL 
REGISTRAR TO THE PHILADELPHIA HOSPITAL. 
BARTON COOKE HIRST, M.D., 
FROM 
THE MEDICAL NEWS, 
March 5, 1887. [From THE MEDICAL NEWS, March 5, 1887.] 
THE DEATH-RATE OF LYING-IN HOSPITALS 
IN THE UNITED STATES.1 
BARTON COOKE HIRST, M.D., 
OBSTETRICIAN TO THE MATERNITY HOSPITAL OF PHILADELPHIA, AND 
OBSTETRICAL REGISTRAR TO THE PHILADELPHIA HOSPITAL. 
Statistics showing the death-rate in general 
hospitals, while interesting and instructive in them- 
selves, are of little value for purposes of comparison, 
for the class of cases treated in one institution can 
differ so much from that in another as to make any 
comparison between the two impossible. This is 
true to a much less degree of hospitals devoted to 
the reception of women about to be confined; here 
the conditions are constant, the affection to be 
treated is always the same. 
The collection and comparative analysis of the 
statistics of such institutions has been brought to a 
condition of great perfection in Europe, especially 
in Germany, with the result of exciting a spirit of 
emulation among the authorities of the obstetrical 
hospitals as to which one of them shall possess the 
best record, and the wide publication of these re- 
ports has had, moreover, the very valuable effect of 
1 Read before the Philadelphia Obstetrical Society, March 3, 
1887. 2 
HIRST, 
making it impossible that a hospital should have a 
conspicuously bad record without giving some ex- 
planation for it. Such a collection of statistics in 
regard to the lying-in hospitals of this country does 
not exist; to supply this deficiency, therefore, the 
writer has devoted considerable time and attention, 
with the result of obtaining the records of 19,902 
confinements during the six years, 1880-1885, 
divided among 34 hospitals. 
Of the importance of possessing such knowledge 
it is almost unnecessary to speak. By placing the 
statistics in a tabulated form, the records of those 
hospitals which are unusually good may be com- 
pared with those that are conspicuously bad, and if 
the divergence in results between them is great some 
explanation for this fact must be sought. Then, 
again, the total result, the average mortality, may 
be compared with the average hospital mortality in 
other countries, and, most important of all, it will 
be necessary to contrast the mortality of obstetrics 
in hospital practice with that of general practice, as 
on the result of this comparison should depend the 
very existence of these institutions; for it must be 
conceded that no hospital, whether supported by 
the community or by private charity, has the right 
to subject its inmates to a danger of death much 
greater than if they had remained in their own 
homes, no matter how squalid they might be. 
The study of the following statistical table, there- 
fore, will be divided under three heads: First, a 
study of the table itself, with a comparison of the 
individual records, and an explanation, if possible, 
of their wide divergence; second, a comparison of 
the average hospital mortality in this country with 
that of some other countries; and, third, a study of 
the relation that hospital mortality bears to the mor- 
tality of general practice. DEATH-RATE OF LYING-IN HOSPITALS 
3 
Total 
Per cent, of 
Institution. 
confine- 
ments, 
1880-85. 
deaths. 
deaths 
from 
sepsis. 
Baltimore, Maternity 
694 
2-S9 
0.86 
Boston, Lying-in Hospital 
1599 
3.81 
3.10 
New England, Moral Reform Society . 
191 
0.00 
0.00 
Hospital for Women and Children 
552 
0.72 
0.36 
Brooklyn, St. Mary’s Hospital . . . 
298 
2.01 
0.33 
Maternity 
552 
1.63 
1.44 
King’s County Hospital 
Chicago, Women’s and Children’s Hos- 
834 
2.63 
1.19 
pital 
SIS 
i-3S 
0.58 
Women’s Refuge 
Cincinnati, St. Joseph’s Lying-in Hos- 
60 
6.66 
3 33 
pital 
27s 
0.72 
0.72 
Cleveland, The Retreat 
Detroit, Women’s Hospital and Found- 
192 
I.O4 
0.52 
ling’s Home 
295 
4-74 
3 05 
Hartford Hospital ........ 
139 
2.87 
2 87 
Indianapolis City Hospital 
323 
5.26 
4-33 
Louisville City Hospital 
428 
2-33 
0.70 
Minneapolis, Northwestern Hospital . 
451 
4-44 
2.22 
Newark City Hospital 
772 
3-89 
2 59 
New Orleans, Charity Hospital . . . 
808 
S-32 
2 97 
New York, Blackwell's Island .... 
2842 
2.92 
1,82 
Nursery and Child’s Hospital . . 
1239 
290 
1 S3 
Infant Asylum 
1057 
1.79 
0.85 
Emigrant’s Hospital, Ward’s Island 
Syr3 
3-SS 
2 64 
Foundling Asylum 
847* 
1,41 
Infirmary for Women and Children 
613 
0.48 
0.16 
Asylum for Lying-in Women . 
464 
1.29 
0.32 
Philadelphia, Almshouse Hospital . . 
IOQQ4 
4-34 
Maternity 
503 
i-39 
0.79 
Women’s College Hospital . . . 
458 
2.83 
1 52 
Lying-in Chanty 
174 
0-57 
0-57 
Pittsburg, City Hospital (Almshouse) . 
206° 
3-53 
2 65 
Rochester City Hospital 
139 
0.71 
0.71 
Monroe Co. Hospital 
93 
1.07 
1.07 
St. Louis, Lying-in Hospital .... 
623 
1,92 
1.28 
Washington, Columbia Hospital . . 
797 
1.25 
0 50 
Total number of women delivered. . . 
19,902 
“ “ “ “ died 
516 
1 1882-85. 2 1882-85 
3 Records prior to 1881 unreliable. 
4 Causes of death not given. 
6 Only reckoned for last three years. 4 
HIRST, 
It cannot be claimed that the preceding list is a 
complete one of all the lying-in hospitals in this 
country; there were some that ignored repeated 
requests for their statistics, there were others that 
refused to give them, and still others whose records 
when obtained were found valueless for the purposes 
of this paper. In addition to these hospitals una- 
voidably omitted from the list there may be others 
of whose existence the writer is unfortunately ignor- 
ant. It is improbable, however, that the addition 
of any records that have been omitted would seri- 
ously affect the final result. The number of cases 
dealt with is so large as to insure the approximate 
accuracy of the conclusions arrived at. 
A death-rate of 2.59 per cent, of the women con- 
fined. It must be surprising to one who has not 
devoted much attention to the subject to find that, 
of the women of this country who perform the 
perfectly physiological function of childbearing in 
hospitals designed for their reception at such a time, 
1 out of every 38 must die. One looks involun- 
tarily to the table to learn if this represents the true 
state of affairs in all lying-in hospitals. If they all 
have a death-rate closely approximating this, then 
the only conclusion possible is that such a mortality 
is inseparable from childbirth in hospitals. A 
mere glance, however, suffices to show that the onus 
of producing a mortality as high as 2.59 percent, 
of the women confined, rests not upon all, but 
conspicuously upon a few. Ten hospitals may be 
selected whose average mortality is 4.48 per cent, 
of the women confined ; or, of every 22 women who 
are compelled to seek shelter in these institutions, 
1 will die. On the other hand, ten hospitals may 
be selected, of which the average mortality is only 
o 87 per cent., in which 115 women must be deliv- 
ered before 1 dies. Can there be a satisfactory DEATH-RATE OF LYING-IN HOSPITALS. 
5 
explanation for a divergence in results so wide as 
this ? 
It has often been asserted that unmarried primi- 
parae form the most unfavorable class of cases possi- 
ble ; that, in addition to the primiparity which is 
acknowledged everywhere as an unfavorable factor 
in obstetrics, the shame connected with the illegiti- 
mate birth exerts an almost murderous influence in 
raising the mortality among these women.1 
It might, therefore, be expected that only hospitals 
devoted to this class of cases would be found among 
those contributing to raise the general mortality, 
while in those where the mortality was so low there 
would be found a large proportion of multiparse. 
Again, a reference to the table of statistics will show 
the truth or the fallacy of this surmise. We shall 
find that the record of the Maternity Hospital of 
Philadelphia, an institution devoted entirely to the 
care of unmarried primiparse, will show a death- 
rate of 1.39 per cent., while the Boston Lying-in 
Hospital, in which more than half the patients are 
married women and presumably raultiparae, although 
the annual reports give no information on this 
point, has a mortality of 3 81 per cent. 
It may also be claimed that a high mortality is 
inseparable from a lying-in hospital that deals with 
a large number of women yearly; but if this is the 
case, it must be due to over-crowding, which can 
occur just as well in a small as in a large hospital, 
but is unjustifiable in either. And, in fact, if we 
look again to the table, we shall find the highest 
mortality in a hospital which reports only 60 cases 
for the six years. It is useless to look further for a 
satisfactory explanation for the high death-rate that 
prevails in many of our lying-in hospitals, for no 
1 Tail; Hospital Mortality. 
* 6 
good reason can be found, and the conviction is 
forced upon us that such excessive mortality is 
utterly unjustifiable. That this is true and that, 
even in institutions whose previous record had been 
simply shocking, an earnest effort toward improve- 
ment will accomplish much, is shown by the follow- 
ing table of the death-rate by years in two large 
hospitals. 
HIRST, 
Death-rate per cent, of Women Confined. 
Institution. 
1880. 
l88l. 
l882. 
1883, 
1884 
1885. 
Boston, Lying-in 
Hospital . . . 
5-58 
3,18 
6.00 
S-88 
1.94 
I-3I 
Philadelphia, Block- 
ley Almshouse 
6.01 
4.08 
3-92 
7-3i 
3-19 
1-52 
The sudden decrease in the mortality of the 
Boston Lying-in Hospital, in 1884, was the result 
of a letter calling the attention of the managers to 
the really murderous death-rate of former years in 
that institution. The improvement begun in 1885 
in the Philadelphia Almshouse still continues, the 
death-rate for the past year having been only 0.87 
per cent. It is also gratifying to note the gradual 
improvement for the whole country shown in the 
following table: 
Years. 
Women confined. 
Women died. 
Death-rate per ct. 
1880 
2800 
85 
3-03 
1881 
3320 
78 
2-34 
1882 
3272 
90 
2.74 
1883 
3207 
12 C 
3-77 
1884 
3595 
67 
1.86 
1885 
3704 
75 
2.02 DEATH-RATE OF LYING-IN HOSPITALS. 
This improvement is not steady, but, as may be 
seen, the death-rate for the last two years is lower 
than for any two years previous. 
To conclude the study and analysis of the table 
of statistics, it would be well to fix a maximum mor- 
tality for lying-in hospitals, beyond which none 
should go without giving a satisfactory explanation. 
We have seen that there are 10 hospitals on the list 
with an average mortality of only 0.87 per cent., 
and among these hospitals are some which deal with 
the worst class of cases imaginable. It would seem, 
therefore, reasonable and liberal to fix this limit at 
1.5 per cent, of the women confined, and I am 
happy to say that in future the State Board of Char- 
ities in Pennsylvania will probably demand annual 
reports from the lying in hospitals throughout the 
State, in order that if any one of them shows an ex- 
cessive death-rate, that fact may be pointed out to 
the managers of the hospital with a request for an 
effort to secure better results. 
Now we may pass to the second division of the 
subject and compare the records of lying-in hospitals 
Period. 
Delivered 
Death-rate 
per cent. 
Germany, 47 hospitals* . . 
1874-83 
104.328 
1-37 
England, 10 hospitals2 . . 
1870-75 
18,369 
I 05 
Vienna, I. Geburts. u. Gyn. 
klinik des Allgemein. kran- 
kenhaus.3 
1880-85 
15.070 
O./O 
Vienna, II. Geburts. u. Gyn. 
klinik des Algemein. kran- 
kenhaus 4 
1882-85 
8.355 
0.52 
Paris, Maternite5 .... 
1876-82 
1,223 
O.49 
1 Dohrn: Zeitschr. f. Geburts. u. Gyn., B. xii. H. i. 
2 Lawson Tait: Hospital Mortality. 
3 Carl Braun; Beilage zu Nr. 35, der Wiener Mediz. Woch., 1886. 
4 Ehrendorfer; Archiv f. Gyn., B. xxvii. H. 2. 
5 Tarnier: Bulletin et Mcmoires de la Soci6t6 Medicate des 
Hcpitaux de Paris, 1882. 8 
HIRST, 
in this country with those of some hospitals in the 
most highly civilized states of Europe. 
It would be impossible, were it necessary or desir- 
able, to present a complete list of the lying-in 
hospitals, with their statistics, of the foremost coun- 
tries of Europe. The brief little table on page 7, 
however, will supply us with sufficient material to 
enable us to contrast the mortality of child-birth in 
hospitals here with the hospital mortality of some 
other countries. 
We shall have little cause for congratulation if we 
compare the records of our lying-in hospitals with 
those of the 47 German hospitals collected and pub- 
lished by Professor Dohrn. These hospitals, although 
dealing with an immense number of women and 
making use of them, as. a rule, for purposes of 
instruction, present a record twice as good as our 
own. The English hospitals, whose records are 
given by Tait, even ten years ago presented a very 
much better showing than ours do to-day; and 
when we see the really astonishing results achieved 
in the obstetrical clinic of the General Hospital in 
Vienna, where under one roof more women are 
delivered in a year than in all the hospitals of the 
United States, we have still greater reason for dis- 
satisfaction with what we have done in this direction, 
especially as there can be no good reason why we 
should occupy such a disadvantageous position in 
this international comparison, for with our greater 
wealth and the superior physical and mental condi- 
tion of our populace, as compared with the peasantry 
of Europe, we should in every lying-in hospital 
strive, at least, to equal the results obtained by 
Tarnier in 1884, when, in the Maternite, he had 
1000 successive accouchements with but a single 
death.1 
1 Annates de Gynecologic, 1884. DEATH-RATE OF LYING-IN HOSPITALS. 
9 
Finally, in order to determine the relation that 
hospital mortality bears to the mortality of general 
obstetrical practice, we are confronted by the neces- 
sity of fixing a normal death-rate after childbirth, 
of knowing how many women die during or after 
labor in a large city; for to make the comparison 
with hospital results a just one, we must have in 
both cases the same environment. 
The almost insuperable difficulty of determining 
the exact mortality of obstetrical practice in a large 
community will be at once apparent. If we consult 
the city’s official record of deaths, we may be sure 
that many of those which are attributed to pyaemia 
and septicaemia really occurred during the puerpe- 
rium, and that often a death directly connected 
with childbirth will, from ignorance, carelessness 
or design of the attendant, be ascribed to some 
other cause. The experience of the Registrar-Gen- 
eral1 of England in 1882 will be interesting in this 
connection. In that year there were reported, with- 
out further specification, 413 deaths from septicaemia 
and pyaemia, and 430 deaths from peritonitis, of 
women of childbearing age. A careful inquiry 
developed the fact that of the first group (413 
deaths) 123 were due to childbirth, and that of the 
second group (430 deaths) 179 followed childbirth; 
in both groups of cases more than half the inquiries 
remained unanswered. This addition to the number 
of deaths after childbirth raised the mortality among 
the women who had been delivered in 1882 by 0.1 
per cent., making the death-rate for the year through- 
out England 0.5 per cent. We may realize, there- 
fore, how little reliance is to be placed upon official 
records if we would learn the true death-rate in 
obstetrical practice in a large city, where many of 
1 Annual Report, 1883. HIRST, 
the women in labor fall into the hands of ignorant 
midwives or of physicians not much more skilful. 
I may as well at once assert my belief that this 
death-rate will closely approximate, if, indeed, it 
does not exceed, i per cent, of the women confined ; 
but as there will be many, from the days of Le Fort 
to the present time, who will dispute the truth of 
this statement, it becomes necessary to support it 
by proofs that, to me at least, are convincing. 
If we are unable to look for these proofs in our 
own official records, we have, nevertheless, several 
methods1 at our command by which we may obtain 
the information desired. We might, as Tarnier 
once did, ascertain the number of women of child- 
bearing age who had died in a certain district, and 
if by the record of births it was shown that any one 
of these women had given birth to a child within a 
month of her death, the death might be ascribed to 
childbirth. This method, if the record of births 
and deaths be reliable, can be depended upon 
to give a fairly accurate idea of the number of 
deaths from all causes after childbirth. We might, 
on the other hand, consult the records of other 
cities or countries which we have reason to believe 
more accurate than our own, or we may learn the 
results of the private practice of obstetricians who 
have attended a large number of women and have 
kept a careful record of their cases. Again, we 
might adopt McClintock’s2 plan in order to obtain 
the number of deaths from all causes after child- 
birth, including those from non-puerperal causes, as 
such deaths must sometimes contribute to raise the 
mortality of hospitals. Collecting the results of the 
house practice of nine reliable physicians in 16,774 
cases, McClintock found that more than a quarter 
1 Garrigues : Transact, of the American Gyn. Soc., 1877. 
2 Dublin Quarterly Journal, 1869. DEATH-RATE OF LYING-IN HOSPITALS. 
of the deaths after childbirth were due to non- 
puerperal causes. If, therefore, 25 per cent, is 
added to the number of deaths appearing in a city’s 
records as due to childbirth, a fair idea of the total 
mortality among puerperal women will be obtained. 
Employing Tarnier’s method, in Glasgow and 
Edinburgh, Duncan1 found the mortality among 
women after childbirth to be 0.93 per cent. Ap- 
plying McClintock’s method to the records of Paris, 
St. Petersburg, and Dublin, he found the mortality 
in those cities to be respectively 0.83, 0.89 and 1.16 
per cent, of women confined. By a careful analysis 
of the records preserved in the Health Department 
of New York City, Lusk2 found that 1.17 per cent, 
of the women confined in New York, during the 
years 1868-75, died as a consequence of childbirth. 
In the Duchy of Baden there were delivered, in the 
year 1884, 54,996 women.3 Of these, 0.4 per cent, 
died of puerperal fever; 0.7 per cent, died after the 
performance of the different obstetrical operations, 
from accidents during labor, or from eclampsia; but 
the cause of death in some of these cases was un- 
doubtedly septic poisoning, and they would conse- 
quently be included in the first percentage, so that 
the death-rate cannot be estimated by adding these 
two percentages together, which would make 1.1 per 
cent., but the death-rate is probably not less than 
0.9 per cent, of the women confined, and this in an 
agricultural district containing no very large towns 
and with a comparatively healthy and prosperous 
population. ' 
A. H. McClintock,4 speaking before the British 
Medical Society, gave the results achieved in private 
1 Matthews Duncan; Mortality of Childbed and Maternity 
Hospitals. Edin., 1871. 
2 The Science and Art of Midwifery, p. 645. New York, 1886. 
3 Aerztliche Mittheilungen aus Baden, Nr. 5, 1886. 
* British Medical Journal, Aug. 10, 1878. 12 
DEATH-RATE OF LYING-IN HOSPITALS. 
obstetrical practice by nine reputable physicians, 
men of no mean skill and intelligence, as the follow- 
ing list will show: 
Campbell (Paris), of 1500 cases lost 13 
Simpson “ 180 “ “ 4 
Matthews Duncan,u 736 “ “ 8 
McClintock, “ 1000 “ “ 12 
Thos. Hamilton, “ 402 “ “ 7 
Chas. Egan, “ 400 “ “ 8 
W. T. Greene, “ 1500 “ “ 12 
George Jones, “ 2000 “ “ 16 
Uvedale West, “ 3100 “ “ 23 
10,818 103 
These figures certainly seem to justify the asser- 
tion that, at least, 1 per cent, of the women deliv- 
ered in a large city will die. If one hesitates to 
admit that such is the mortality associated with what 
is usually regarded as a simple physiological process, 
what shall be said of the death-rate attending con- 
finement in the lying-in hospitals of this country, 
which is more than twice as great as that of general 
practice, and in 'individual instances four and five- 
fold greater? 
The title, charitable, can be conferred upon many 
of these institutions only by the followers of Malthus; 
but as their existence has become a necessity in all 
populous communities, a necessity which will become 
more pressing as the population of this country be- 
comes more densely concentrated at certain points 
and the poverty of the lower classes increases, surely 
common humanity demands that they.be controlled 
by State or municipal supervision, in order that they 
may fulfil the ends for which they were intended— 
that they may truly afford aid to a woman at a time 
when it is most urgently needed—and that they may 
not, as many of them do to-day, make childbearing 
as dangerous to the woman as an attack of typhoid 
fever or the performance of laparotomy. 
A death-rate of 0.95 per cent.