J. Obstet. Gynaecol. Res. Vol. 36, No.

2: 248253, April 2010

doi:10.1111/j.1447-0756.2009.01125.x

Maternal mortality and cesarean delivery: An analytical

observational study
jog_1125

248..253

Gourisankar Kamilya, Subrata Lall Seal, Joydev Mukherji, Subir Kumar Bhattacharyya
and Avijit Hazra
Department of Obstetrics and Gynaecology, R. G. Kar Medical College, Kolkata, India

Abstract
Aim: Pregnant women and their doctors need to know the maternal risks associated with different methods
of delivery. There are few publications with ideal study design and adequate power to establish the relationship between maternal mortality and mode of delivery. The present retrospective cohort study was undertaken
to evaluate the intrinsic risk of maternal death, directly attributed to cesarean delivery (CD) compared to
vaginal delivery (VD), and to evaluate further the differential risk associated with antepartum and intrapartum
CD.
Methods: After exclusion of medical or obstetric comorbidities, all deliveries, either vaginal or cesarean, were
critically analyzed. The surviving mothers who had either method of delivery represented the two comparative
groups. In the same period, relevant clinical information of every maternal death was noted.
Results: Twenty seven mothers died among the 13 627 CD mothers and 19 died among 30 215 VD mothers.
CD was associated with a 3.01-fold increase in the risk of maternal mortality, compared with VD. The risk of
antepartum CD differed from intrapartum CD (OR 1.73 vs OR 4.86). There was a signicantly increased risk of
maternal death from complications of anesthesia, puerperal infection and venous thromboembolism. The risk
of death from postpartum hemorrhage did not differ signicantly (95% CI 0.73.95).
Conclusion: CD is increasingly perceived as a low-risk procedure. However, the present study clearly demonstrates that the risk of maternal death due to CD is signicantly high, particularly when performed in labor.
Therefore, CD should only be practiced when conditions clearly demand it.
Key words: cesarean delivery, intrapartum, maternal mortality, postpartum, vaginal delivery.

Introduction
Although cesarean delivery (CD), at one period of time,
was performed only subsequent to maternal death, the
indications for this procedure have dramatically liberalized over time, with approximately 32% of women
undergoing CD in our institution. There is a general
perception that CD substantially increases the risk of
maternal mortality and there are extensive data to
support that belief. The risk of maternal death has been
estimated to be several times that associated with
vaginal delivery (VD),13 but there are several problems
which make it difcult to interpret them.4,5 Many of the

publications are outdated as medical and obstetric

improvements have occurred and risks of CD have
declined.6 There are few publications with ideal study
design and adequate power to establish the relationship
between maternal mortality and mode of delivery. In
addition, only a minority of publications account for
medical and obstetric comorbidities that may have predisposed patients to CD. Thus, mortality attributed to
CD may be a result of preexisting disease rather than the
surgery itself. A recent literature review from developed countries concluded that the strongest publications suggest that there may not be an increased risk of
maternal mortality with elective CD compared to VD.7

Received: January 21 2008.

Accepted: April 20 2009.
Reprint request to: Dr Gourisankar Kamilya, BB 11/G, Salt Lake, Kolkata 700 064, India. Email: drgkamilya@gmail.com

248

2010 The Authors

Journal compilation 2010 Japan Society of Obstetrics and Gynecology

Maternal mortality and cesarean delivery

There is, however, a dearth of data from developing

countries which studies mortality due to inherent complications of CD. Proper anesthetic facilities, timely and
adequate supply of blood, and improper aseptic measures are the major factors that might increase the risk
of CD in less afuent countries. The present study was
undertaken to critically evaluate the intrinsic risks of
abdominal deliveries at a teaching hospital of a developing country. Our objectives were to provide a valid
estimate of the risk of maternal deaths directly attributable to CD compared with VD and to further
examine the differential risks associated with antepartum and intrapartum CD.

Methods
This retrospective cohort study was undertaken in the
Department of Obstetrics and Gynecology, R. G. Kar
Medical College, Kolkata, India, from 2003 to 2006.
During this period all deliveries, either vaginal or
abdominal, were critically analyzed regarding age,
parity, and the presence of preexisting medical conditions or coexistent obstetric complications. Indications
and type of CD, whether antepartum or intrapartum,
were also recorded. In the same period every maternal
death, either intrapartum or postpartum, was noted.
The relevant clinical information related to the
womans death was taken. Maternal deaths with a gestational period of more than 28 weeks or birth weight
>1000 g were only considered. The institutional ethical
committee approved the study.

Exclusion criteria
The exclusion criteria were as follows:
1 Deaths due to chronic conditions present before
pregnancy (e.g. diseases of the circulatory system,
hematological diseases, liver diseases, diseases of
the respiratory system, diseases of the digestive
system and chronic infections).
2 Deaths due to obstetric complications that developed during pregnancy, but before delivery, and
which may have had a direct inuence on maternal
death, like hypertensive disorders in pregnancy,
hemorrhage due to placenta previa or accreta,
abruptio placentae, multiple pregnancy and
chorioamnionitis.
Ultimately deaths due to conditions that occurred
during or after delivery were taken into account,
excluding deaths due to preexisting morbidities. In this
time period, all surviving mothers who delivered

Total cesarean delivery

n = 16 224

Total vaginal deliveries

n = 34 197

Exclusion: Medical or obstetrical complications during pregnancy

Total cesarean deliveries

without any medical or
obstetrical complications prior
to delivery
n = 13 628

Total maternal
deaths following
CD
n = 51

Total maternal deaths

following CD not
due to conditions
existing prior to
delivery
n = 28 (cases)

Total vaginal deliveries without

any medical or obstetrical
complications prior to
delivery
n = 30 215

Total maternal deaths

following vaginal
delivery not due to
conditions existing
prior to delivery
n = 19 (cases)

Total maternal
deaths following
vaginal delivery
n = 78

Figure 1 Process of selection of cases and controls in the

study of maternal mortality and cesarean delivery
(CD).

either by vaginal or CD with babies at a gestational age

of >28 weeks or a birth weight of >1000 g constituted
the two comparative groups. To make selection criteria
exact, the same exclusion criteria were used. Causes of
death both in the CD and VD groups were analyzed
(Fig. 1).
The primary predictive variable of interest was the
route of delivery. CD were further classied as antepartum and intrapartum. Differences in the distribution of
route of delivery between the two groups were examined, using c2 and Fishers exact tests. Crude odds
ratios (OR) of postpartum maternal death for CD were
calculated. Multivariable logistic regression analysis
was used to adjust for relevant covariates, like age and
parity, and adjusted OR were calculated. The level of
statistical signicance was P < 0.05. Minitab software
version 3.12 (Minitab, State College, PA, USA ) was
used for data analysis.

Results
Table 1 shows the total number of mothers who delivered by vaginal route and those who delivered by CD.
The total number of maternal deaths is also stated in the
table. After exclusion of preexisting medical or obstetric complications that may inuence the choice of route
of delivery or maternal death, it was found that 13 627
mothers had undergone CD. Similarly, after application of exclusion criteria, 30 215 mothers delivered
vaginally. These 46 maternal deaths (27 in the CD group
and 19 in VD group) represented the cases, and 43 796

2010 The Authors

Journal compilation 2010 Japan Society of Obstetrics and Gynecology

249

G. Kamilya et al.

Table 1 Mode of delivery and maternal death in the study of maternal mortality and cesarean delivery

Before exclusion criteria

After exclusion criteria

Vaginal
deliveries

Cesarean
deliveries

Maternal deaths
following vaginal
delivery

Maternal deaths
following cesarean
delivery

P-value

OR

95% CI

34 197
30 215

16 224
13 627

78
19

51
27

0.09
<0.001

1.38
3.16

0.971.96
1.755.68

Table 2 Distribution of characteristics of women among

the two groups (cesarean vs vaginal delivery)
Death

Other

P-value*

Age

All
<25
2529
3034
>34

46
22
14
8
2

43 796
29 796
10 212
3 148
640

<0.005

Parity

0
13
>3

19
24
3

26 576
17 025
195

<0.001

Discussion

*c2 test.

women who delivered either by vaginal or abdominal

route and survived represented the two comparative
groups. Characteristics of the mothers, both in the fatal
and surviving groups are shown in Table 2. Compared
to the surviving mothers, mothers who died were signicantly older and had high parity (P < 0.005 and
P < 0.001, respectively).
Table 3 shows the proportion of various indications
and number of maternal deaths in the CD group. Previous history of CD represents the most common indication (42.3%), followed by nonprogress of labor
(16.5%) and fetal distress (14.6%).
Table 4 shows that the proportion of cases among
CD group was signicantly higher (59.6% vs 40.4%),
although CD represents only 31.1% of the total deliveries. CD was associated with a signicantly increased
risk of maternal death (OR 3.15). After adjustment of
potential confounders, like age and parity, the risk of
CD became three times higher than VD (adjusted OR
3.01). When compared with VD, intrapartum CD was
associated with a signicantly increased risk of maternal death (P < 0.001). The risk of antepartum CD differed from the risk of intrapartum CD (adjusted OR
1.73 vs OR 4.86). The odds of death from intrapartum
CD were signicantly greater than the odds of death in
antepartum CD (OR 2.48, 95% CI 1.155.35).
To better understand the association between CD
and risk of maternal death, the specic cause of maternal death was further analyzed (Table 5) using Fishers
exact test. CD was associated with a signicantly

250

increased risk of postpartum maternal death from

complications of anesthesia, puerperal infection and
venous thromboembolism. The risk of death from postpartum hemorrhage did not differ signicantly
between the vaginal and CD groups (95% CI 0.703.95).
The specic cause of mortality could not be analyzed
separately for antepartum and intrapartum CD because
the number of deaths was too small.

Pregnant women and their doctors need to know the

maternal risks associated with different methods of
delivery. The present study focuses on the risks of
maternal mortality following CD. Previous studies,
mainly from developed countries, have reported heterogeneous, inconsistent and inconclusive results.
There is, therefore, a need to assess maternal mortality
following CD.
The authors of a recent French study8 reported the
risk of maternal death was 3.6 times higher after CD
than after VD. Our study shows that CD is associated
with a 3.01-fold increase in the risk of maternal mortality when compared with VD. In our study, maternal
deaths and their causes were assessed, based on
detailed information on the circumstances of death collected from admission and treatment records. This
allowed us to select the cases after excluding all maternal deaths attributed to preexisting complications. Like
the French study,8 the main strategy used here to control
for indication bias was to exclude deaths due to morbidities present before delivery so that risk related
exclusively to delivery route could be established. The
same analysis conducted without exclusion of these
deaths would have led to a different result, not representing the actual risk of CD (OR 1.38, 95% CI 0.971.96).
Without excluding preexisting comorbidities, our
results do not show an increased relative risk of mortality for CD. The reason for this could be the propensity to avoid CD in the serious and pre-terminally ill
gravid women, thus inating the maternal mortality
gures for VD cases. Maximum maternal death

2010 The Authors

Journal compilation 2010 Japan Society of Obstetrics and Gynecology

Maternal mortality and cesarean delivery

Table 3 Indications of cesarean delivery (CD) and maternal death

Indication

No. (%)

Maternal
deaths

Previous history of CD
Nonprogress of labor
Nonreassuring fetal heart rate pattern or fetal distress
Induction failure
Malpresentation
Intrauterine growth restriction
Elderly primigravida
Cephalopelvic disproportion
Bad obstetric history
Pregnancy following treatment of infertility

5767
2253
1987
988
704
534
556
420
261
160

12
4
4
1
3
0
2
1
0
0

(42.3)
(16.5)
(14.6)
(7.3)
(5.2)
(3.9)
(4.1)
(3.1)
(1.9)
(1.1)

Table 4 Risk of maternal death according to mode of delivery

Mode of delivery

Deaths
[n (%)]

Nonfatal
cases [n (%)]

OR

95% CI

Adjusted
OR

95% CI

P-value

Vaginal
Cesarean
Antepartum cesarean
Intrapartum cesarean

19 (40.4)
27 (59.6)
11 (25.5)
16 (34.1)

30 196
13 600
8 574
5 026

3.16
2.04
5.06

1.755.68
0.974.29
2.609.85

3.01
1.73
4.86

1.665.46
0.803.71
2.479.56

<0.001
=0.055
<0.001

(68.9)
(31.1)
(19.6)
(11.5)

Logistic model including maternal age and parity.

Table 5 Odds ratio for cause-specic maternal mortality associated with cesarean delivery
Causes of mortality

Cesarean delivery
no. (incidence/10 000)

Vaginal delivery
no. (incidence/10 000)

OR

95% CI

PPH
Venous thromboembolism
Amniotic uid embolism
Puerperal infection
Anesthetic complications
Uterine inversion
Mismatched blood Transfusion
All causes

9
6
3
5
3

1
27

12
1
3
2

19

1.66
13.31
2.22
5.54

3.16

0.703.95
1.60110.60
0.4510.99
1.0828.59

1.755.68

(6.6)
(4.4)
(2.2)
(3.7)
(2.2)
(0.73)

(3.97)
(0.33)
(0.99)
(0.66)
(0.33)

PPH, postpartum hemorrhage.

occurred when CD was performed for previous history

of CD. (Table 3). This is mainly because CD for these
indications was not always an antepartum operation.
Some of these patients were admitted during labor. The
risk of increased mortality was mainly due to adherent
placenta over a uterine scar, increasing the risk of postpartum hemorrhage and nonavailability of a large
amount of blood within a short time.
Though antepartum CD was not associated with a
signicant increase in maternal death (P = 0.055),
intrapartum CD was associated with a signicantly
increased risk (P < 0.001). The mortality risk was much
higher in intrapartum CD (OR 4.86) compared to prelabor CD (OR 1.73). The increased risk of maternal

mortality of CD was not uniformly distributed among

the various causes of death. CD was signicantly associated with an increased risk of maternal death due to
three causes that are classical perioperative or postoperative complications of any major surgery: venous
thromboembolism, infection, and complications of
anesthesia. Not surprisingly, recent studies have
reported that re-hospitalization for infectious morbidities and venous thromboembolism was more common
among mothers who had CD than among women who
had VD.8 An increased risk of puerperal sepsis-related
mortality has also been reported by others.9,10
Although anesthesia for CD has greatly improved,
our results suggest that CD is still associated with an

2010 The Authors

Journal compilation 2010 Japan Society of Obstetrics and Gynecology

251

G. Kamilya et al.

increased risk of maternal death from complications of

anesthesia. In the present study all three anesthesiarelated maternal deaths occurred after general anesthesia. There has been an increasing trend away from
general anesthesia for CD to regional techniques. Nevertheless, it is still necessary to monitor the anesthesiarelated mortality risk of CD.
Blood loss contributes a signicant portion of maternal mortality in both groups, probably due to nonavailability of adequate and emergency supply of blood,
and preexisting poor general condition of the patient.
Although blood loss during CD is much greater than
during VD, postpartum hemorrhage-related maternal
death was only slightly high in the CD group and was
not statistically signicant (95% CI 0.73.95). This result
is in agreement with previous studies in which the risk
of postpartum hemorrhage was not increased after CD
compared with VD.11 This is probably due to differential surveillance and care provided to parturient
women in the immediate postpartum period according
to their route of delivery. Clinicians may be more alert
about blood loss in a patient who has just undergone
surgery, leading to an earlier diagnosis and prompt
interventions.

Pitfalls of the study

The analysis of maternal mortality in the present study
is based on nal mode of delivery rather than as
planned using an intention to treat approach. In considering maternal mortality in primary elective CD, the
risk that needs to be determined is that associated with
an attempt at VD (which could result in VD, operative
VD or CD) as compared with elective CD. So an ideal
study design would compare maternal mortality of
planned primary elective CD to planned primary VD.12
Maternal death may occur during CD in a woman who
had originally planned VD. This death, although it
occurred following CD, ideally should be included in
the VD group for actual assessment of VD-related risk
of maternal death.
Again, it would take many years to follow women
throughout their reproductive careers and account for
long-term complications, such as uterine rupture, placental abruption or placenta percreta with subsequent
pregnancy that may impact maternal mortality.

Conclusion
There are some reports that suggest that there may not
be an increased risk of maternal death with elective CD
compared with VD. Although maternal mortality has

252

become rare in afuent countries, the picture is quite

gloomy for developing countries. CD is increasingly
perceived as a low-risk procedure; as a result there is
growing interest in primary elective CD. The present
study clearly demonstrates that the risk of maternal
death due to CD is signicantly high when compared
with VD, particularly when performed during labor.
Anesthetic complications, hemorrhage and sepsis,
which are mostly preventable in afuent countries,
may make major contributions towards maternal mortality in adverse conditions. This needs to be taken into
account by clinicians and women when considering
the riskbenet equation. CD is a major surgical procedure and should only be practiced when conditions
clearly demand it.

Acknowledgments
We must acknowledge the postgraduate students of
our institution for their tireless effort in recording
every detail of mothers who delivered babies during
this time period for the purpose of the study. We are
also thankful to the Principal and Head of the Department of Obstetrics and Gynaecology, R. G. Kar Medical
College, Kolkata, India for allowing us to conduct the
study and for giving necessary permission for publishing the present paper.

References
1. Rubin GL, Peterson HB, Rochat RW et al. Maternal death after
cesarean section in Georgia. Am J Obstet Gynecol 1981; 139:
681685.
2. Lilford RI, Vancoeverden HA, Bingham P. The relative risk of
caesarean section (intrapartum and elective) and vaginal
delivery: A detailed analysis to exclude the effects of medical
disorders and other acute pre existing physiological disturbances. Br J Obstet Gynaecol 1990; 97: 883892.
3. Schuitamaker N, Van Roosmalen J, Dekker G, Van Dongen P,
Van Geijn H, Gravenhorst JB. Maternal mortality after cesarean section in the Netherlands. Acta Obstet Gynecol Scand
1997; 76: 332334.
4. Minkoff H, Chervanak F. Elective primary cesarean delivery.
N Engl J Med 2003; 348: 946950.
5. Lydon-Rochlle M, Holt VL, Easterling TR, Martin DP. Caesarean delivery and postpartum mortality among primiparas in
Washington state, 19871996. Obstet Gynecol 2001; 97: 169174.
6. Burrows LJ, Meyn LA, Weber AM. Maternal morbidity associated with vaginal versus cesarean delivery. Obstet Gynecol
2004; 103: 907912.
7. Vadnais M, Sachs B. Maternal mortality with cesarean delivery: A literature review. Semin Perinatol 2006; 30: 242246.
8. Deneux-Tharaux C, Carmona E, Colle MB, Breart G. Postpartum maternal mortality and cesarean delivery. Obstet Gynecol
2006; 108: 541548.

2010 The Authors

Journal compilation 2010 Japan Society of Obstetrics and Gynecology

Maternal mortality and cesarean delivery

9. Liu S, Heaman M, Joseph KS et al. Risk of maternal postpartum readmission associated with mode of delivery. Obstet
Gynecol 2005; 105: 836842.
10. Ojo VA, Adetoro OO, Okwerekwu FE. Characteristics of
maternal deaths following caesarean section in a developing
country. Int J Obstet Gynecol 1988; 27: 171176.
11. Allen VM, OConnell CM, Liston RM, Baskett TF. Maternal
morbidity associated with cesarean delivery without labor

compared with spontaneous onset of labor at term. Obstet

Gynecol 2003; 102: 477482.
12. Wax RJ. Maternal request caesarean versus planned spontaneous vaginal delivery: Maternal morbidity and short term
outcomes. Semin Perinatol 2006; 30: 247252.

2010 The Authors

Journal compilation 2010 Japan Society of Obstetrics and Gynecology

253