You are on page 1of 12

Research ajog.


Assessment of cesarean delivery
availability in 26 low- and middle-income
countries: a cross-sectional study
Rele Ologunde, BSc(Hons); Joshua P. Vogel, MBBS; Meena N. Cherian, MD;
Mariam Sbaiti, MBBS; Mario Merialdi, MD; James Yeats, BSc(Hons)

OBJECTIVE: We sought to assess the capacity to provide cesarean least 1 operating room. Of the facilities performing CD, 47.3% did not
delivery (CD) in health facilities in low- and middle-income countries. report the presence of any type of anesthesia provider and 17.9% did
not report the presence of any type of obstetric/gynecological or sur-
STUDY DESIGN: We conducted secondary analysis of 719 health fa-
gical care provider. In facilities reporting a lack of functioning equip-
cilities, in 26 countries in Africa, the Pacific, Asia, and the Mediter-
ment, 26.4% had no access to an oxygen supply, 60.8% had no access
ranean, using facility-based cross-sectional data from the World
to an anesthesia machine, and 65.9% had no access to a blood bank.
Health Organization Situational Analysis Tool to Assess Emergency and
Essential Surgical Care. CONCLUSION: Provision of CD in facilities in low- and middle-income
countries is hindered by a lack of an adequate anesthetic and surgical
RESULTS: A total of 531 (73.8%) facilities reported performing CD. In
workforce and availability of oxygen, anesthesia, and blood banks.
all, 126 (17.5%) facilities did not perform but referred CD; the most
common reasons for doing so were lack of skills (53.2%) and Key words: cesarean delivery, low- and middle-income countries,
nonfunctioning equipment (42.9%). All health facilities surveyed had at obstetric services, surgery

Cite this article as: Ologunde R, Vogel JP, Cherian MN, et al. Assessment of cesarean delivery availability in 26 low- and middle-income countries: a cross-sectional
study. Am J Obstet Gynecol 2014;211:504.e1-12.

F ollowing the adoption of the Mil-

lennium Declaration by the United
Nations in 2000, the Millennium notably MDG4 (reducing child mor-
malaria by the year 2026.3 However, ac-
cording to current estimates, the poorest
third of the worlds population receive
Development Goals (MDGs) were es- tality) and MDG5 (improving maternal only 3.5% of the 234 million surgical
tablished. These 8 international devel- health), has received relatively less procedures undertaken worldwide.4
opment goals, agreed on by all 189 attention.1 In 2010, the global maternal mortality
Member States, were aimed to be ach- Surgical disease has been estimated ratio was 210 deaths per 100,000 births,
ieved by the year 2015. Although much to account for 11% of the worlds which corresponds to a total of 287,000
attention has been focused on meeting disability-adjusted life years2 and is pre- maternal deaths worldwide.5 Major
the MDGs, the role of strengthening dicted to eclipse those of human im- complications due to obstetric condi-
surgical capacity to achieve these goals, munodeciency virus, tuberculosis, and tions such as antepartum hemorrhage,
obstructed labor, and eclampsia can be
prevented or managed with timely access
From the Faculty of Medicine, School of Public Health, Imperial College London, St Marys Hospital,
London, United Kingdom (Mr Ologunde, Dr Sbaiti, and Mr Yeats); Emergency and Essential Surgical
to cesarean delivery (CD).6 However, ac-
Care Program, Health Systems and Innovation (Mr Ologunde and Dr Cherian); Department of cess to and use of CD varies widely both
Reproductive Health and Research, United Nations Development Programme/United Nations Fund geographically and between low- and
for Population Activities/United Nations International Childrens Emergency Fund/World Health middle-income countries (LMICs)7,8
Organization (WHO)/World Bank Special Programme of Research, Development, and Research with rates <1% for many of the poorest
Training in Human Reproduction (Drs Vogel and Merialdi); and Faculty of Medicine, Dentistry and
Health Sciences, School of Population Health, University of Western Australia, Crawley, Australia
(Dr Vogel). To reduce the unacceptably high levels
Received Feb. 5, 2014; revised March 30, 2014; accepted May 15, 2014. of maternal mortality in LMICs,5 efforts
The authors report no conict of interest.
to improve the availability and accessi-
bility of emergency obstetric care must
Presented in poster format at the First World Congress of Surgery, Obstetrics, Trauma, and
Anesthesia, Port of Spain, Trinidad and Tobago, Oct. 16-17, 2013. be facilitated. This care encompasses all
Reprints: Rele Ologunde, Faculty of Medicine, School of Public Health, Imperial College London, St
care related to the treatment of peri-
Marys Hospital, Praed Street, London, W2 1NY, United Kingdom. partum complications, including the
0002-9378/free  2014 Elsevier Inc. All rights reserved. ability to perform safe CD.6 A checklist
of 9 signal functions denes the

504.e1 American Journal of Obstetrics & Gynecology NOVEMBER 2014 Obstetrics Research
minimum requirements of a facility to for referral of CD in those facilities that hospitals took over administration of
be considered a provider of compre- do not perform the procedure, avail- the SAT. Survey responses were kept
hensive emergency obstetric care ser- ability of essential surgical elements in anonymous. The data were entered into
vices.6 These signal functions are facilities performing and not performing and stored on the WHO DataCol SQL
indicators for a group of interventions CD, and availability of human personnel global database at WHO headquarters
that are used to manage the obstetric in facilities performing and not per- in Geneva, Switzerland, from December
complications that contribute to the forming CD. 2008 through the present. The hard
majority of maternal deaths world- The standardized WHO SAT to assess copies of the paper-based information
wide.10 The ability to perform surgery emergency and essential surgical care, were stored securely. In March 2013, a
(including CD) and to deliver a blood developed by the WHO Global Initiative database query was performed to extract
transfusion are the 2 indicators that for Essential and Emergency Surgical information on CD capacity.
distinguish a comprehensive emergency Care research group in November 2007,
obstetric care facility from a basic has been used to collect data from health Data analysis
emergency obstetric care facility. In the facilities in 44 LMICs from December Countries providing data on <5 health
absence of a CD, women with obstructed 2008 through the present. The SAT is a facilities were excluded from the ag-
labor are at increased risk of death or paper-based cross-sectional survey form gregated data to reduce potential bias of
developing a stula during childbirth, in used to quantify surgical capacity, in- including nationally unrepresentative
addition to risk of perinatal morbidity cluding trauma, obstetrics, and anes- data, in line with previous studies em-
and mortality.11 There is, however, evi- thesia, within participating facilities. The ploying the WHO tool.15,18 Health fa-
dence showing that basic surgical care analysis tool collects information on the cilities were included if they had 1
provision and investment in obstetric name, location, and type of participating operating rooms. A total of 719 health
capacity, particularly in LMICs, can be a facilities. The WHO SAT was pilot facilities met the inclusion criteria
cost-effective public health interven- tested in 8 facilities in The Gambia and (Table 1). Health facilities included
tion.12-14 United Republic of Tanzania15 and has health centers, district/rural/community
Effective intrapartum care is not been validated for assessing surgical ca- hospitals, provincial hospitals, general
limited to the capacity to perform CD; it pacity in health facilities in LMICs.16 hospitals, and private/nongovernmental
also requires trained skilled birth at- The WHO SAT has 108 data points organization (NGO)/mission hospitals.
tendants able to prevent, recognize, and divided into 4 sections: (1) 25 questions Results were grouped for aggregate
manage obstetric complications and on infrastructure and health facility de- analysis to avoid intercountry compari-
deliver a range of interventions. This mographics, including the availability of sons. Health centers are often present at
study is the largest cross-sectional survey essential surgical services such as oxygen, the subdistrict level where they provide
of availability of CD in LMICs to date an anesthesia machine, and a blood both preventive and curative services for
and provides the most comprehensive bank; (2) 8 questions on the availability their population19 and often represent
assessment of provision of this proce- of health care personnel (including the the lowest level of health facility. District
dure yet from a sample of a number of number of personnel for each relevant hospitals tend to represent the largest
facilities around the world rather than a category); (3) 34 questions assessing the level of health facility and are a rst
single geographic location. availability of surgical interventions; and referral point for patients who present
This study focuses on the critical (4) 41 questions on the availability of with conditions that require surgical
aspect of provision of CD as a lifesaving surgical equipment and supplies. Section intervention.20 District, rural, and com-
surgical intervention for women with 4 of the SAT is based on the WHO munity hospitals were grouped together
obstetric complications. We aimed to Essential and Emergency Equipment in keeping with previous studies using
quantify CD capacity in health facilities list.17 the WHO SAT.15,21-23 Provincial hospi-
in LMICs based on availability of the tals in many LMICs are typically tertiary
procedure, infrastructure and human Administration of the WHO SAT teaching hospitals. General hospitals
resources, and reasons for referral using Identication of health facilities for are similar to district hospitals and in
the World Health Organization (WHO) administration of the SAT was left to the some countries the terms are used
Situational Analysis Tool (SAT) to assess discretion of Ministry of Health, WHO interchangeably.23 Private/NGO/mission
emergency and essential surgical care. country ofce, and Global Initiative for hospitals are large well-resourced and
Essential and Emergency Surgical Care typically well-equipped institutions.
M ATERIALS AND M ETHODS representatives in individual countries. To minimize potential bias as a result
WHO SAT to assess emergency and As such, the data represent a sample of nonresponse, all reasonable attempts
essential surgical care of convenience. Representatives from were made to contact health facilities
The main outcome of this study was these organizations, during site visits to with missing data points. When health
to determine the proportion of health health facilities, performed data collec- facilities were contacted, verication of
facilities in LMICs performing CD. Sec- tion. Where this was not possible med- previously submitted data was also con-
ondary exposures of interest are reasons ical or surgical directors at respective ducted, to minimize potential bias from

NOVEMBER 2014 American Journal of Obstetrics & Gynecology 504.e2

Research Obstetrics

Characteristics of countries included in study and number of facilities contributed by each
MMR per Births by No. of facilities No. of facilities
No. Country LIC/MICa 100,000b CD, %c completing a survey included Data
1 Afghanistan LIC 460 3.6 26 22 3.06%
2 Argentina MIC 77 22.7 9 8 1.11%
3 China MIC 40 27.0 8 7 0.97%
4 Democratic Republic LIC 540 7.2 16 15 2.09%
of Congo
5 Ethiopia LIC 350 1.5 19 19 2.64%
6 The Gambia LIC 360 2.5 75 23 3.20%
7 Ghana LIC 350 6.9 22 22 3.06%
8 Haiti LIC 350 3.0 51 48 6.68%
9 India LIC 200 8.1 171 110 15.30%
10 Kenya LIC 360 6.2 54 53 7.37%
11 Liberia LIC 770 3.5 23 16 2.23%
12 Malawi LIC 460 4.6 16 15 2.09%
13 Mongolia LIC 63 21.0 43 29 4.03%
14 Myanmar LIC 200 32.1 20 20 2.78%
15 Niger LIC 590 1.0 21 21 2.92%
16 Nigeria LIC 630 1.8 121 115 15.99%
17 Pakistan LIC 260 7.3 9 8 1.11%
18 Papua New Guinea LIC 230 4.7 25 24 3.34%
19 Sao Tome and Principe LIC 70 5.3 5 1 0.14%
20 Sierra Leone LIC 890 4.5 12 11 1.53%
21 Solomon Islands MIC 93 6.2 9 7 0.97%
22 Somalia LIC 1000 1e2 14 14 1.95%
23 Sri Lanka MIC 35 23.8 39 18 2.50%
24 Uganda LIC 310 5.3 38 31 4.31%
25 United Republic LIC 460 4.5 49 43 5.98%
of Tanzania
26 Vietnam LIC 59 20.0 19 19 2.64%
Total 914 719 100.00%
CD, cesarean delivery; LIC, low-income country; MIC, middle-income country; MMR, maternal mortality ratio.
As defined by the World Bank Classification System based on 2011 Gross National Income per capita with LIC making $1025 and MIC making $1026-12,475; b Figures derived from trends in
maternal mortality: 1990 through 20105; c See Appendix for sources of CD rates.
Ologunde. Cesarean delivery availability in 26 LMICs. Am J Obstet Gynecol 2014.

possible response errors. Where a re- Redmond, WA). We used SPSS, version performing CD and those not per-
sponse for a data point was unobtain- 21.0 (IBM Corp, Armonk, NY) to forming CD with a P value of < .05 set
able, it was reported as missing and the perform c2 tests. We employed as statistically signicant.
health facility was excluded from the descriptive statistical analysis to Data used in this study did not require
subanalysis pertaining to that data point. compare individual elements of the ethics approval because no patient re-
Computerized spreadsheet tools survey between facilities performing cords or information was included. The
were used to generate descriptive sta- CD and those not performing CD. We data analyzed are for assessing the
tistics using Microsoft Excel for Mac performed bivariate analysis using c2 availability of surgical services for each
2011, version 13.3.4 (Microsoft, test to compare the results of facilities health facility.

504.e3 American Journal of Obstetrics & Gynecology NOVEMBER 2014 Obstetrics Research

Study flowchart

Ologunde. Cesarean delivery availability in 26 LMICs. Am J Obstet Gynecol 2014.

NOVEMBER 2014 American Journal of Obstetrics & Gynecology 504.e4

Research Obstetrics

Availability of CD in health facilities

Facilities excluded for not providing data on performance of CD (n 21) and not indicating whether or not they referred CD (n 36).
CD, cesarean delivery; NGO, nongovernmental organization.
Ologunde. Cesarean delivery availability in 26 LMICs. Am J Obstet Gynecol 2014.

Role of the funding source Health facility characteristics performing and referring CD were
The sponsors of the study had no role In all, 244 (33.9%) of the included fa- stratied by facility type and are illus-
in study design, data collection, data cilities were district/rural/community trated in Figure 2. Provision of CD was
interpretation, data analysis, or writing hospitals, 202 (28.1%) were private/ highest in private/NGO/mission hos-
of the report. The corresponding NGO/mission hospitals, 100 (13.9%) pitals, whereas referral was most com-
author had full access to all the data in were general hospitals, 78 (10.8%) were mon in health centers.
the study; all authors had nal re- health centers, and 78 (10.8%) were Of the facilities that did not perform
sponsibility for the decision to submit provincial hospitals. All health facilities but referred CD, the most common
for publication. surveyed had at least 1 operating room, reason for doing so was a lack of skills
with 233 (32.4%) reporting 2. (n 67, 53.2%). The next most com-
mon reasons were nonfunctioning
R ESULTS equipment (n 54, 42.9%) and lack of
A total of 18 countries were excluded CD provision and referral supplies/drugs (n 42, 33.3%). How-
from the aggregated data for providing In all, 531 (73.8%) of the 719 facilities ever, in general hospitals and private/
information on <5 health facilities surveyed reported performing CD, NGO/mission hospitals the most com-
(Figure 1). Of the remaining 914 facil- whereas 167 (23.2%) did not perform mon reason for referring CD was
ities, 195 did not have an operating room the procedure and 21 (2.9%) did not nonfunctioning equipment. Reasons for
and were excluded from the study. Of provide information on this. Of the 167 referring CD were stratied by facility
these, 29 were district/rural/community facilities that did not perform the pro- type among those not performing the
hospitals, 9 were general hospitals, 126 cedure, 36 did not state what they did procedure (Figure 3).
were health centers, 14 were private/ with regards to women requiring a CD
NGO/mission hospitals, 4 were provin- and were thus excluded from the anal- Essential surgical elements
cial hospitals, and 13 did not provide a ysis. Of the 719 facilities, 126 (17.5%) In facilities performing CD, there was
response for this data point. The 719 facilities reported that they referred the consistent availability of an oxygen supply
health facilities included in our analysis procedure to another facility. It is (cylinder or concentrator), an anesthesia
represent 14 African countries, 5 West- possible that a number of the facilities machine, and a blood bank at 417
ern Pacic countries, 3 Southeast Asian not providing information on referral of (78.7%), 350 (66.7%), and 199 (39.8%)
countries, 2 Eastern Mediterranean CD actually do not provide it because it facilities, respectively (Figure 4). In facil-
countries, and 2 North American coun- is not within the remit of procedures ities referring CD because of nonfunc-
tries. Demographic and study data for they perform (ie, a small rural health tioning equipment, only 21 (39.6%)
included countries are shown in Table 1. post) and thus pregnant women would facilities had a consistent availability of
Country classications were based on likely be aware not to seek such a facility oxygen (cylinder or concentrator), 18
WHO classication of world regions.24 in the event of labor. Establishments (35.3%) an anesthesia machine, and 4

504.e5 American Journal of Obstetrics & Gynecology NOVEMBER 2014 Obstetrics Research

Reasons for referring cesarean delivery

Facilities excluded for not providing data on reasons for referral (n 14).
NGO, nongovernmental organization.
Ologunde. Cesarean delivery availability in 26 LMICs. Am J Obstet Gynecol 2014.

(8.3%) a blood bank (Figure 4). A sta- to a lack of skills did not report the data demonstrate a lack of essential
tistically signicant difference was found presence of any type of anesthesia pro- equipment and skilled anesthesia, ob-
in the availability of essential surgical vider (Table 3). In addition, 251 (47.3%) stetric, and surgical care providers. This
elements between facilities performing of the facilities reporting that they per- raises concerns about patient safety and
and those not performing but referring formed CD did not report the presence the quality of care being delivered in
CD (Table 2). of any type of anesthesia provider. these facilities. To our knowledge this is
An anesthesia machine and a blood Surgeons were the most common the largest study to assess availability of
bank were the most common essential providers of obstetric or surgical care in CD, referrals, and barriers to provision
surgical elements that were lacking in facilities performing CD, whereas gen- of the procedure, using a standardized
facilities performing CD and in those eral doctors were the most common tool across a large number of LMICs.
referring due to nonfunctioning equip- surgical care providers in facilities that
ment (Figure 4). Of all the facilities sur- referred CD due to a lack of skills CD provision and referral
veyed, 67 (9.4%) reported not having an (Table 4). Additionally, of the facilities Kushner et al15 reported CD provision
oxygen supply, 243 (33.8%) reported not performing CD, 95 (17.9%) did not in 44% of facilities in a cohort of 132 fa-
having an anesthesia machine, and 345 report the presence of any type of ob- cilities, with at least 1 minor or major
(48.0%) reported not having a blood bank. stetric/gynecological or surgical care operating room, from 8 LMICs. A number
provider (Table 4). of single-country surveys have also been
Human resources conducted. A study by Contini et al21 in
The most common providers of anes- C OMMENT Afghanistan reported that 88% of facilities
thetic care in facilities performing CD, This study aimed to assess the surgical performed CD. In the United Republic
and in those referring due to lack of capacity of facilities providing CD in of Tanzania, 67% of surveyed health fa-
skills, were nurses or nonphysician LMICs, and the reasons for referral in cilities reported performing CD22 and in
medical practitioners (Table 3). In health those facilities unable to provide CD. The Gambia 58.8% of facilities, all con-
facilities referring CD due to a lack of We found that 73.8% of facilities an- taining at least 1 operating room, reported
skills at the facility, only 4 (6.0%) facil- alyzed performed CD, with 17.5% not performing the procedure.23 The differ-
ities had at least 1 anesthesiologist and performing but referring the proce- ence between our estimate and previous
only 6 (9.0%) facilities had at least 1 dure. Lack of skills and nonfunctioning ones may reect differences in sampling.
nurse or nonphysician medical practi- equipment were found to be major It may also be due to the limitations of
tioner providing anesthesia. In all, 57 barriers to provision of CD. Even in fa- the data used for these types of estimates,
(85.1%) of the facilities referring CD due cilities where CD was performed, our or may reect genuine variability in

NOVEMBER 2014 American Journal of Obstetrics & Gynecology 504.e6

Research Obstetrics

Availability of essential surgical elements in facilities performing CD, and those referring CD due to
nonfunctioning equipment

Facilities performing CD excluded for not providing data on oxygen supply (n 1), anesthesia machine (n 6), and a blood bank (n 31). Facilities
referring CD excluded for not providing data on oxygen supply (n 1), anesthesia machine (n 3), and a blood bank (n 6).
CD, cesarean delivery.
Ologunde. Cesarean delivery availability in 26 LMICs. Am J Obstet Gynecol 2014.

the provision of CD across countries. We that critical surgical services, such as which is much lower than previously
found that 17.5% of facilities surveyed oxygen, anesthetic machines, and blood published reports. While these 2 studies
referred CD; this is comparable with pre- banks, are often not available in LMICs.26 were methodologically similar to ours,
vious ndings from single-country sur- A survey of 590 facilities in 22 LMICs possible reasons for the differences in
veys of 30% by Abdullah et al20 and 18% found that 35% of facilities did not report ndings may be differences in sampled
by Choo et al.19 Although our ndings having access to any source of oxygen18 facilities and classication of responses,
suggest that over three quarters of facilities and a survey of 231 health facilities in or they may reect real differences in the
sampled provide CD, the availability of CD 12 African countries reported that 24.8% surgical capacity of facilities surveyed.
as a proportion of all health facilities in of facilities lacked access.26 We found Nevertheless, our study is the largest to
LMICs will be much lower given that we that 9.4% of all facilities surveyed re- date to quantify oxygen and anesthesia
only included health facilities that had an ported not having an oxygen supply, capacity and will be more representative
operating room and excluded those that
do not.

Essential surgical elements
Availability of infrastructure
Oxygen is crucial to the provision of safe
Facilities performing Facilities referring
surgical procedures and emergency
Item Total (n [ 719) CD (n [ 531) CDa (n [ 54) P value
resuscitation,25 yet our ndings suggest
that it is not universally available even in Oxygen 525 (73.3%) 417 (78.7%) 21 (39.6%) < .001
settings where CD is being performed. Anesthesia 415 (58.9%) 350 (66.7%) 18 (35.3%) < .001
We found that 21.3% of facilities per- machine
forming CD reported not having a reli- Blood bank 228 (33.8%) 199 (39.8%) 4 (8.3%) < .001
able supply of oxygen and 26.4% of those CD, cesarean delivery.
referring CD due to nonfunctioning a
Refer CD due to nonfunctioning equipment.
equipment did not have any supply. Ologunde. Cesarean delivery availability in 26 LMICs. Am J Obstet Gynecol 2014.
Previous analyses have demonstrated

504.e7 American Journal of Obstetrics & Gynecology NOVEMBER 2014 Obstetrics Research
meet the minimum threshold of health
TABLE 3 workers deemed necessary to deliver
Availability of anesthetic care providers maternal and child health services.30
Total Facilities performing Facilities referring We found that paraprofessionals and
Personnel (n [ 719) CD (n [ 531) CDa (n [ 67) P value nurses were the most commonly avail-
Anesthesiologist 172 (23.9%) 145 (27.3%) 4 (6.0%) < .001 able providers of anesthesia. Similar
General doctor 66 (9.2%) 62 (11.7%) 0 (0.0%) .003 ndings have been reported elsewhere
providing anesthesia in the literature with Iddriss et al23
documenting that anesthesia was de-
Nurse or NMP 229 (31.8%) 202 (38.0%) 6 (9.0%) < .001
providing anesthesia livered by nonphysicians in 72.2% of
facilities surveyed in The Gambia and
None present 389 (54.1%) 2 (47.3%) 57 (85.1%) < .001
Vo et al18 nding that nurses and clin-
CD, cesarean delivery; NMP, nonphysician medical practitioner.
ical assistants made up the majority of
Refer CD due to lack of skills.
anesthesia providers in the 590 facilities
Ologunde. Cesarean delivery availability in 26 LMICs. Am J Obstet Gynecol 2014.
they surveyed across 22 LMICs. Sur-
geons were the most common providers
of obstetric or surgical care in the fa-
of global capacity. It is also a signicant LMICs28 making the lack of a blood bank cilities we surveyed, however this varied
nding that <100% of all facilities at a health facility an urgent priority for between facilities performing CD and
providing CD lack an oxygen supply. action to improve obstetric outcomes. those referring with general doctors
Ability to provide anesthesia is typi- Previous studies describing the avail- being the most common provider in
cally subsumed with the ability to per- ability of essential surgical elements using the latter. Previous reports of surgical
form CD.6 Our ndings however, suggest the WHO tool have reported similar providers in the literature suggest that
that this should not be assumed as ndings to ours. In Afghanistan, a third paramedical professionals, including
over a quarter of facilities that reported of hospitals surveyed did not have a blood nonphysicians and nurses, make up
providing CD stated that an anesthesia bank.21 A study by Vo et al18 found that the bulk of the surgical workforce in
machine was not available at their facility. >25% of facilities surveyed did not have LMICs.20,23 The lack of specialists in
However, our study only assessed the an anesthesia machine. In the United obstetrics and gynecology in LMICs,
availability of an anesthesia machine. Republic of Tanzania, 23% of facilities both in the facilities we surveyed and in
Alternative forms of anesthesia such as surveyed did not have a blood bank and previous reports in literature, may have
spinal anesthesia and regional blocks, 33% lacked an anesthesia machine.22 signicant adverse effects in terms of the
which are commonly used in CD,27 may management of obstetric complications
have been available in participating Human resources in these facilities. Surprisingly, >50% of
health facilities. Availability of a blood The 2006 World Health Report called for all facilities included in our study did
bank was also low with almost half of all a rapid scaling up of the global health not report the presence of any of the
facilities reporting that this was unavai- workforce.29 Despite this, <10% of the types of anesthesia providers surveyed
lable. Postpartum hemorrhage is a lead- countries classied as low-income and almost a quarter of all facilities did
ing cause of maternal mortality in country by the World Bank actually not report the presence of a surgical
provider. Given that all facilities sur-
veyed reported having an operating
TABLE 4 room at the facility, the reasons hin-
Availability of obstetrical and surgical care providers dering surgical capacity are likely to be
Facilities Facilities due to a lack of personnel able to
Total performing referring perform CD at the facility. Where this is
Personnel (n [ 719) CD (n [ 531) CDa (n [ 67) P value the case there is the potential for task
Obstetrician/gynecologist 215 (29.9%) 187 (35.2%) 3 (4.5%) < .001 shifting, which may reduce inequalities
by extending care to underserved pop-
Surgeon 358 (49.8%) 307 (57.8%) 7 (10.4%) < .001
ulations. Task shifting in surgery has
General doctor 254 (35.3%) 195 (36.7%) 38 (56.7%) .002 already been shown to be effective in a
performing surgery
number of LMICs including Niger,31
NMP performing surgery 154 (21.4%) 86 (16.2%) 5 (7.5%) .061 where general practitioners are trained
None present 168 (23.4%) 95 (17.9%) 21 (31.3%) .009 in district surgery, and in Malawi,
CD, cesarean delivery; NMP, nonphysician medical practitioner. where nonphysician clinical ofcers
Refer CD due to lack of skills. have on-the-job training in surgery.32
Ologunde. Cesarean delivery availability in 26 LMICs. Am J Obstet Gynecol 2014. A recent review of task shifting in
maternal and reproductive health

NOVEMBER 2014 American Journal of Obstetrics & Gynecology 504.e8

Research Obstetrics

suggests that it is cost-effective and may safely. Efforts to increase surgical and indications is associated with an increased risk
increase access and availability of ser- obstetric capacity and thus availability of of adverse short-term maternal outcomes: the
2004-2008 WHO Global Survey on Maternal
vices without compromising on patient CD need to focus on addressing de- and Perinatal Health. BMC Med 2010;8:71.
outcomes.33 WHO guidelines on task ciencies in key infrastructural items 9. Ronsmans C, Holtz S, Stanton C. Socio-
shifting in maternal and newborn health and scaling up and meeting the training economic differentials in cesarean rates in
recommended that advanced-level as- needs of the health workforce. The developing countries: a retrospective analysis.
sociate clinicians could be used to per- limited availability of essential services Lancet 2006;368:1516-23.
10. World Health Organization. Maternal health
form CD in well-equipped facilities, in and infrastructure in many facilities that around the world. Geneva (Switzerland): WHO
the context of targeted monitoring and purport to provide CD raises questions 1997.
evaluation.34 about the quality and safety of the pro- 11. World Health Organization. Essential ele-
cedures being provided. - ments of obstetric care at rst referral level.
Limitations Geneva (Switzerland): WHO; 1991.
12. Laxminarayan R, Chow J, Shahid-Salles SA.
The Essential and Emergency Surgical ACKNOWLEDGMENTS Intervention of cost-effectiveness: overview of
Care database is a sample of convenience We are hugely grateful for the support of the main messages. In: Jamison DT, Breman JG,
and is susceptible to selection bias. The health facility visit teams in the various countries Measham AR, et al, eds. Disease control prior-
facilities in the data set are not demo- included in this survey and in particular would ities in developing countries, 2nd ed. New York:
graphically or geographically represen- like to thank Rev Dr Tomi Thomas (Catholic Oxford University Press; 2006:35-86.
Health Association of India), Mr Bakary Jargo 13. Spiegel DA, Abdullah F, Price RR,
tative of their country. The data were
(WHO country ofce, The Gambia), Dr Ananda Gosselin RA, Bickler SW. World Health Organi-
aggregated, and countries were not Gunasekera (Ministry of Health, Sri Lanka), Dr zation Global Initiative for Emergency and
weighted by their contribution. In ad- Hkon Angell Bolkan (CapaCare, Sierra Leone), Essential Surgical Care: 2011 and beyond.
dition, surveys completed without site Dr Tu Tran (University of California Haiti Initiative, World J Surg 2013;37:1462-9.
visits lack response integrity validation Haiti), Dr Samuel Likasi (United Republic of 14. McCord C, Chowdhury Q. A cost effective
Tanzania), Dr Opar Toliva (Uganda), and Dr small hospital in Bangladesh: what it can mean
and may be vulnerable to reporting bias.
Olayinka Ayankogbe (Nigeria). We are hugely for emergency obstetric care. Int J Gynecol
grateful for the support of Dr Graham Cooke, Obstet 2003;81:83-92.
Applications and future research Miss Florence Guida, Miss Fiona Constable, Dr 15. Kushner AL, Cherian MN, Noel L,
While the majority of facilities analyzed Luc Noel, Dr Laksmi Govindasamy, and Miss Spiegel DA, Groth S, Etienne C. Addressing the
performed CD, issues of the availability, Rikke Le Kirkegaard for their guidance and millennium development goals from a surgical
helpful discussion. perspective essential surgery and anesthesia in
accessibility, equity, quality, and safety
8 low- and middle-income countries. Arch Surg
of CD remain unanswered. Supporting 2010;145:154-9.
the training and continuing education REFERENCES 16. Osen H, Chang D, Choo S, et al. Validation
of health care personnel is paramount 1. PloS Medicine Editors. A crucial role for sur- of the World Health Organization tool for situa-
to improving safe surgical practices gery in reaching the UN millennium development tional analysis to assess emergency and
goals. PLoS Med 2008;5:e182. essential surgical care at district hospitals in
in LMICs. Future research should focus
2. Debas HT, Gosselin R, McCord C, Ghana. World J Surg 2011;35:500-4.
on developing the surgical capacity Thind A. Surgery. In: Jamison DT, 17. World Health Organization. WHO generic
of health facilities through frugal tech- Breman JG, Measham AR, et al, eds. Disease essential emergency equipment list 2006.
nologies that are appropriate for the control priorities in developing countries, 2nd Available at:
LMIC setting: for example, low main- ed. New York: Oxford University Press; 2006: EEEGenericListFormatted 06.pdf. Accessed
1245-59. May 7, 2013.
tenance, electricity-independent oxygen
3. Bae JY, Groen RS, Kushner AL. Surgery as a 18. Vo D, Cherian MN, Bianchi SL, et al. Anes-
concentrators that meet the needs of public health intervention: common mis- thesia capacity in 22 low and middle income
remote health facilities. Furthermore, conceptions versus the truth. Bull World Health countries. J Anesth Clin Res 2012;3:1-5.
to address the critical health worker Organ 2011;89:395. 19. Choo S, Perry H, Hesse AAJ, et al.
shortages in many LMICs, shifting of 4. Weiser TG, Regenbogen SE, Thompson KD, Assessment of capacity for surgery, obstetrics
et al. An estimation of the global volume of sur- and anesthesia in 17 Ghanaian hospitals using a
tasks, such as CD, may improve access to
gery: a modeling strategy based on available WHO assessment tool. Trop Med Int Health
the procedure.34 Further study in to data. Lancet 2008:372. 2010;15:1109-15.
the outcome of CD in women in LMICs 5. WHO, UNICEF, UNFPA, The World Bank. 20. Abdullah F, Choo S, Hesse AA, et al.
is also warranted to work towards im- Trends in maternal mortality: 1990 to 2010. Assessment of surgical and obstetrical care at
proving the safety prole and quality of Geneva (Switzerland): World Health Organiza- 10 district hospitals in Ghana using on-site in-
tion; 2012. terviews. J Surg Res 2011;171:461-6.
the procedure.
6. WHO, UNFPA, UNICEF, Mailman School of 21. Contini S, Taqdeer A, Cherian M, et al.
Public Health. Monitoring emergency obstetric Emergency and essential surgical services in
Conclusion care. Geneva (Switzerland): World Health Or- Afghanistan: still a missing challenge. World J
CD is a lifesaving obstetric intervention ganization; 2009. Surg 2010;34:473-9.
and is indicated in several complications 7. Betran AP, Merialdi M, Lauer JA, et al. Rates 22. Penoyar T, Cohen H, Kibatala P, et al.
of pregnancy and delivery. However, even of cesarean section: analysis of global, regional Emergency and surgery services of primary
and national estimates. Paediatr Perinat Epi- hospitals in the United Republic of Tanzania.
when CD is available, facilities in LMICs demiol 2007;21:98-113. BMJ Open 2012;2:e000369.
do not necessarily have the services, staff, 8. Souza JP, Gulmezoglu A, Lumbiganon P, 23. Iddriss A, Shivute N, Bickler S, et al. Emer-
and capacity to perform the procedure et al. Cesarean section without medical gency, anesthetic and essential surgical

504.e9 American Journal of Obstetrics & Gynecology NOVEMBER 2014 Obstetrics Research
capacity in the Gambia. Bull World Health Organ anesthesia for elective cesarean section at term: 31. Sani R, Nameoua B, Yahaya A, et al. The
2011;89:565-72. effect on the acid-base status of the mother impact of launching surgery at the district level in
24. WHO. World health statistics 2013, Geneva, and newborn. J Matern Fetal Neonatal Med Niger. World J Surg 2009;33:2063-8.
Switzerland. Available at: 2003;13:260-6. 32. Jiskoot P. On-the-job training of clinical of-
iris/bitstream/10665/81965/1/9789241564588 28. Khan KS, Wojdyla D, Say L, cers in Malawi. Malawi Med J 2008;20:74-7.
_eng.pdf. Accessed May 16, 2013. Gulmezoglu AM, Van Look PF. WHO analysis of 33. Dawson AJ, Buchan J, Dufeld C,
25. Brook MH, Brook J, Wyant GM. Emer- causes of maternal death: a systematic review. Homer CS, Wijewardena K. Task shifting and
gency resuscitation. Br Med J 1962;2: Lancet 2006;367:1066-74. sharing in maternal and reproductive health in
1564-6. 29. WHO. Rapid scaling up of health work- low-income countries: a narrative synthesis of
26. Belle J, Cohen H, Shindo N, et al. Inuenza force production 2006. Available at: http://www. current evidence. Health Policy Plan 2014;29:
preparedness in low-resource settings: a look at 396-408.
oxygen delivery in 12 African countries. J Infect Accessed May 19, 2013. 34. World Health Organization. Optimizing health
Dev Ctries 2010;4:419-24. 30. WHO. WHO global atlas of the health work- worker roles to improve access to key maternal
27. Petropoulos G, Siristatidis C, Salamalekis E, force 2010. Available at: and newborn health interventions through task
Creatsas G. Spinal and epidural versus general g_density.pdf. Accessed May 19, 2013. shifting. Geneva (Switzerland): WHO; 2012.

NOVEMBER 2014 American Journal of Obstetrics & Gynecology 504.e10

Research Obstetrics


Estimates and sources of CS rates
Country Year CS, %a Source
Afghanistan 2011 3.6
[Accessed May 11, 2013]
Argentina 2006 22.7
[Accessed May 11, 2013]
China 2008 27
[Accessed May 11, 2013]
Democratic Republic of Congo 2010 7.2
[Accessed May 11, 2013]
Ethiopia 2011 1.5
[Accessed May 11, 2013]
The Gambia 2010 2.5
[Accessed May 11, 2013]
Ghana 2008 6.9
[Accessed May 11, 2013]
Haiti 2006 3
[Accessed May 11, 2013]
India 2008 8.1
[Accessed May 11, 2013]
Kenya 2009 6.2
[Accessed May 11, 2013]
Liberia 2007 3.5
[Accessed May 11, 2013]
Malawi 2010 4.6
[Accessed May 11, 2013]
Mongolia 2010 21
[Accessed May 11, 2013]
Myanmar 2009 32.1
20Statistics%20Report%202009.pdf [Accessed May
14, 2013]
Niger 2006 1
[Accessed May 11, 2013]
Nigeria 2008 1.8
[Accessed May 11, 2013]
Pakistan 2007 7.3
[Accessed May 11, 2013]
Papua New Guinea 2002 4.7 Papua New Guinea Department of Health. Information
provided by Dr Nicholas Mann on 29 July 2003. Papua
New Guinea Department of Health, PO Box 807,
Waigani NCD, Papua New Guinea.
Sao Tome and Principe 2009 5.3
[Accessed May 11, 2013]
Sierra Leone 2010 4.5
[Accessed May 11, 2013]
Solomon Islands 2007 6.2
[Accessed May 11, 2013]
Ologunde. Cesarean delivery availability in 26 LMICs. Am J Obstet Gynecol 2014. (continued)

504.e11 American Journal of Obstetrics & Gynecology NOVEMBER 2014 Obstetrics Research

Estimates and sources of CS rates (continued)
Country Year CS, %a Source
Somalia 2012 1e2 Deyo NS. Cultural traditions and the reproductive
health of Somali refugees and immigrants
[Dissertation]. San Francisco, CA (US): University
of San Francisco; 2012.
Sri Lanka 2007 23.8
[Accessed May 11, 2013]
Uganda 2011 5.3
[Accessed May 11, 2013]
United Republic of Tanzania 2010 4.5
[Accessed May 11, 2013]
Vietnam 2011 20
[Accessed May 11, 2013]
CS, cesarean section.
Calculated as the total number of births by cesarean section over the total number of births within a given population over a given period of time, expressed as a percentage.
Ologunde. Cesarean delivery availability in 26 LMICs. Am J Obstet Gynecol 2014.

NOVEMBER 2014 American Journal of Obstetrics & Gynecology 504.e12