Geographical accessibility to functional emergency obstetric care facilities in urban Nigeria using closer-to-reality travel time estimates: a population-based spatial analysis

Abstract

Background
Better accessibility for emergency obstetric care facilities can substantially reduce maternal and perinatal deaths. However, pregnant women and girls living in urban settings face additional complex challenges travelling to facilities. We aimed to assess the geographical accessibility of the three nearest functional public and private comprehensive emergency obstetric care facilities in the 15 largest Nigerian cities via a novel approach that uses closer-to-reality travel time estimates than traditional model-based approaches.
Methods
In this population-based spatial analysis, we mapped city boundaries, verified and geocoded functional comprehensive emergency obstetric care facilities, and mapped the population distribution for girls and women aged 15–49 years (ie, of childbearing age). We used the Google Maps Platform's internal Directions Application Programming Interface to derive driving times to public and private facilities. Median travel time and the percentage of women aged 15–49 years able to reach care were summarised for eight traffic scenarios (peak and non-peak hours on weekdays and weekends) by city and within city under different travel time thresholds (≤15 min, ≤30 min, ≤60 min).
Findings
As of 2022, there were 11·5 million girls and women aged 15–49 years living in the 15 studied cities, and we identified the location and functionality of 2020 comprehensive emergency obstetric care facilities. City-level median travel time to the nearest comprehensive emergency obstetric care facility ranged from 18 min in Maiduguri to 46 min in Kaduna. Median travel time varied by location within a city. The between-ward IQR of median travel time to the nearest public comprehensive emergency obstetric care varied from the narrowest in Maiduguri (10 min) to the widest in Benin City (41 min). Informal settlements and peripheral areas tended to be worse off compared to the inner city. The percentages of girls and women aged 15–49 years within 60 min of their nearest public comprehensive emergency obstetric care ranged from 83% in Aba to 100% in Maiduguri, while the percentage within 30 min ranged from 33% in Aba to over 95% in Ilorin and Maiduguri. During peak traffic times, the median number of public comprehensive emergency obstetric care facilities reachable by women aged 15–49 years under 30 min was zero in eight (53%) of 15 cities.
Interpretation
Better access to comprehensive emergency obstetric care is needed in Nigerian cities and solutions need to be tailored to context. The innovative approach used in this study provides more context-specific, finer, and policy-relevant evidence to support targeted efforts aimed at improving comprehensive emergency obstetric care geographical accessibility in urban Africa.

Full text

Articles
www.thelancet.com/lancetgh Vol 12 May 2024
e848
Geographical accessibility to functional emergency obstetric
care facilities in urban Nigeria using closer-to-reality travel
time estimates: a population-based spatial analysis
Aduragbemi Banke-Thomas, Kerry L M Wong, Tope Olubodun, Peter M Macharia, Narayanan Sundararajan, Yash Shah, Gautam Prasad,
Mansi Kansal, Swapnil Vispute, Tomer Shekel, Olakunmi Ogunyemi, Uchenna Gwacham-Anisiobi, Jia Wang, Ibukun-Oluwa Omolade Abejirinde,
Prestige Tatenda Makanga, Ngozi Azodoh, Charles Nzelu, Bosede B Afolabi, Charlotte Stanton, Lenka Beňová
Summary
Background Better accessibility for emergency obstetric care facilities can substantially reduce maternal and perinatal
deaths. However, pregnant women and girls living in urban settings face additional complex challenges travelling to
facilities. We aimed to assess the geographical accessibility of the three nearest functional public and private
comprehensive emergency obstetric care facilities in the 15 largest Nigerian cities via a novel approach that uses
closer-to-reality travel time estimates than traditional model-based approaches.
Methods In this population-based spatial analysis, we mapped city boundaries, verified and geocoded functional
comprehensive emergency obstetric care facilities, and mapped the population distribution for girls and women aged
15–49 years (ie, of childbearing age). We used the Google Maps Platform’s internal Directions Application Programming
Interface to derive driving times to public and private facilities. Median travel time and the percentage of women aged
15–49 years able to reach care were summarised for eight trac scenarios (peak and non-peak hours on weekdays and
weekends) by city and within city under dierent travel time thresholds (≤15 min, ≤30 min, ≤60 min).
Findings As of 2022, there were 11·5 million girls and women aged 15–49 years living in the 15 studied cities, and we
identified the location and functionality of 2020 comprehensive emergency obstetric care facilities. City-level median
travel time to the nearest comprehensive emergency obstetric care facility ranged from 18 min in Maiduguri to 46 min
in Kaduna. Median travel time varied by location within a city. The between-ward IQR of median travel time to the
nearest public comprehensive emergency obstetric care varied from the narrowest in Maiduguri (10 min) to the
widest in Benin City (41 min). Informal settlements and peripheral areas tended to be worse o compared to
the inner city. The percentages of girls and women aged 15–49 years within 60 min of their nearest public
comprehensive emergency obstetric care ranged from 83% in Aba to 100% in Maiduguri, while the percentage within
30 min ranged from 33% in Aba to over 95% in Ilorin and Maiduguri. During peak trac times, the median number
of public comprehensive emergency obstetric care facilities reachable by women aged 15–49 years under 30 min was
zero in eight (53%) of 15 cities.
Interpretation Better access to comprehensive emergency obstetric care is needed in Nigerian cities and solutions
need to be tailored to context. The innovative approach used in this study provides more context-specific, finer, and
policy-relevant evidence to support targeted eorts aimed at improving comprehensive emergency obstetric care
geographical accessibility in urban Africa.
Funding Google.
Copyright © 2024 The Author(s). Published by Elsevier Ltd. This is an Open Access article under the CC BY 4.0
license.
Introduction
Despite a 43% reduction in global maternal mortality
since 2000, around 287 000 maternal deaths occurred in
2020 and about 70% of them were in Africa.1 Similarly,
44% of the estimated 2 million stillbirths worldwide
in 2019 occurred in Africa.2 Of these global figures,
Nigeria has the highest number of maternal deaths
(82 000 [29%]) and the third highest number of stillbirths
(171 428 [9%]).1,2 Evidence shows that half of maternal
deaths and three-quarters of intrapartum stillbirths are
preventable with timely access to high-quality emergency
obstetric care, meaning that delays increase the risk for
poor pregnancy outcomes.3
Current guidelines from WHO recommend that
comprehensive emergency obstetric care, including
caesarean section and blood transfusion, be available
within 2–3 h travel time.4 Pregnant women in urban areas
of Africa typically live closer to comprehensive emergency
obstetric care facilities compared with their rural
counterparts.5 However, this so-called urban advantage is
less apparent in some urban areas as travel can be long
due to poor road infrastructure, haphazardly built
Lancet Glob Health 2024;
12: e848–58
See Comment page e729
Faculty of Epidemiology and
Population Health, London
School of Hygiene & Tropical
Medicine, London, UK
(A Banke-Thomas PhD,
K L M Wong PhD); School of
Human Sciences
(A Banke-Thomas) and School
of Computing & Mathematical
Sciences (J Wang PhD),
University of Greenwich,
London, UK; Maternal and
Reproductive Health Research
Collective, Lagos, Nigeria
(A Banke-Thomas,
Prof B B Afolabi FRCOG);
Department of Community
Medicine and Primary Care,
Federal Medical Centre
Abeokuta, Abeokuta, Ogun,
Nigeria (T Olubodun FMCPH);
Department of Public Health,
Institute of Tropical Medicine,
Antwerp, Belgium
(P M Macharia PhD,
Prof L Beňová PhD); Population
& Health Impact Surveillance
Group, Kenya Medical Research
Institute–Wellcome Trust
Research Programme, Nairobi,
Kenya (P M Macharia); Centre
for Health Informatics,
Computing, and Statistics,
Lancaster Medical School,
Lancaster University, Lancaster,
UK (P M Macharia); Google,
Mountain View, CA, USA
(N Sundararajan PhD,
Y Shah BSc, G Prasad PhD,
M Kansal MBA, S Vispute MSc,
T Shekel MBA, C Stanton PhD);
Lagos State Ministry of Health,
Ikeja, Lagos, Nigeria
(O Ogunyemi MBBS); Nuffield
Department of Population
Health, University of Oxford,
Oxford, UK
(U Gwacham-Anisiobi MPH);
Dalla Lana School of Public
Health, University of Toronto,
Toronto, ON, Canada
(I-O O Abejirinde PhD); Women’s
College Hospital Institute for
Health System Solutions and

Articles
e849
www.thelancet.com/lancetgh Vol 12 May 2024
Virtual Care, Toronto, ON,
Canada (I-O O Abejirinde);
Surveying and Geomatics
Department, Midlands State
University Faculty of Science
and Technology, Gweru,
Zimbabwe (P T Makanga PhD);
Climate and Health Division,
Centre for Sexual Health and
HIV/AIDS Research, Zimbabwe
(P T Makanga); Department of
Health Planning, Research and
Statistics, Federal Ministry of
Health, Abuja, Nigeria
(N Azodoh MPH, C Nzelu PhD);
Department of Obstetrics and
Gynaecology, College of
Medicine of the University of
Lagos, Lagos, Nigeria
(Prof B B Afolabi)
Correspondence to:
Dr Aduragbemi Banke-Thomas,
Faculty of Epidemiology and
Population Health, London
School of Hygiene & Tropical
Medicine, London WC1E 7HT, UK
aduragbemi.banke-thomas@
lshtm.ac.uk
or
Dr Charlotte Stanton, Google,
Mountain View, CA 94043, USA
chstanton@google.com
environments, trac congestion, and expanding informal
settlements.6,7 By 2050, 70% of the global population will
be urban dwellers, and 40% of the projected additional
2·5 billion urban dwellers will be concentrated in Africa.8
To date, most explorations of geographical access
to emergency obstetric care have been based on
modelled travel time using the cost-friction approach and
open-source route mapping.9,10 However, such models
insuciently reflect the lived experience of travel to
care in urban settings.7,11 Ignoring the variability in
trac conditions results in as much as a three-time
overestimation of geographical coverage of services
compared with reality.12 Furthermore, almost all
geographical accessibility studies using modelled
approaches estimate travel time to the nearest health
facility,10 despite established evidence that pregnant
women in urban settings bypass the nearest health
facility even in emergencies.13,14
A comparative analysis of access to emergency obstetric
care in Lagos, Nigeria—Africa’s largest megacity—found
that navigation applications such as Google Maps provide
closer-to-reality travel time estimates to health facilities
Research in context
Evidence before this study
We systematically searched PubMed and Scopus on
Sept 16, 2023, without language and date restrictions, for
previously published articles that reported geographical
accessibility to emergency obstetric care using the search string
(“emergency obstetric care” OR “EmOC” OR “EmONC”) AND
(“distance” OR “travel time” OR “access*” OR “geography*”). We
also reviewed references from retrieved articles to identify any
additional studies. We found 12 studies that used models to
assess geographical accessibility to emergency obstetric care in
Africa. All but one of the 12 studies only estimated travel to the
nearest health facility in the public sector with no verification of
facility functionality, despite established evidence that pregnant
women might bypass the nearest health facility even in an
emergency, particularly due to issues of trust in quality of care. In
addition, all studies were conducted in rural areas or across an
entire country without urban-specific estimates. However,
concerns have been raised by experts that modelled approaches
do not reflect the reality of travel that many pregnant women
living in urban African settings undergo in accessing care.
Evidence from a 2021 study that compared estimates from
models (cost-friction surface approach and Open Source
Routing Machine) and the navigation application, Google Maps,
with actual replicated journeys that women would have taken to
reach comprehensive emergency obstetric care showed that
Google Maps offers closer-to-reality travel time estimates. In a
separate study, when applied retrospectively to map journeys
pregnant women took to care in Africa’s largest megacity, Lagos,
Nigeria, the method led to highly policy-relevant insights.
However, analysis prospectively accounting for choice of
different facilities and variable travel conditions at different
times of the day and days of the week had yet to be conducted.
Added value of this study
To the best of our knowledge, this study is the first to apply
Google Maps at scale using close-to-reality travel time estimates
to assess geographical accessibility to comprehensive
emergency obstetric care. We assessed travel times to these
facilities in the 15 most populated cities in Nigeria. We
integrated the element of choice by incorporating hospitals
from both public and private sectors, validated their
functionality, and assessed driving time to the three nearest
hospitals that provide comprehensive obstetric care. Estimates
also accounted for the day of the week and time of day. We
found that, even under peak travel conditions, the majority of
pregnant women in all cities were within 45 min of the nearest
public sector hospital. Geographical coverage was generally
poorest during weekdays, between 1800 h and 2000 h, and best
during the weekend, between 0100 h and 0300 h. This level of
precision adds to existing evidence that highlights where
inequities exist, to establishing when they are worse. In several
cities women cannot reach more than one hospital within 1 h,
especially if affordability limits them to the public sector. Within
cities, geographical inequity of accessibility to care existed to
various extents, with women in informal settlements and
peripheral suburbs particularly affected. Such a granular report
on geographical accessibility to emergency obstetric care has
only been possible because of our innovative approach of using
closer-to-reality travel time estimates combined with
verification of functionality status of health facilities.
Implications of all the available evidence
Travel time to reach obstetric care has important mortality and
morbidity implications for pregnant women and girls and their
babies. Navigation applications like Google Maps, which we
have used in our study, have the capacity to capture relevant and
context-specific data on travel time, which will be crucial for
service planning. The granularity of the approach used in this
study opens new opportunities in the field of within-city
accessibility to health care that extends beyond emergency
obstetric care. In addition, there is now a chance to layer other
datasets that allow more detailed and realistic characterisation
of the populations in greatest need. There is a clear need to
ensure that estimates are disaggregated by time of the day, day
of the week, and choice of health facilities to ensure more
policy-relevant evidence and action to bridge accessibility gaps.
Potential associations between travel time to health facilities
and care outcomes can now be studied, especially within the
context of urbanisation. These new capabilities constitute a step
change compared with widely used methods and will be crucial
towards universal health coverage. Future research needs to
explore possible integration of other dimensions of accessibility
including cost, quality of care, and availability of structures that
aid access to fully maximise the potential of this approach.

Articles
www.thelancet.com/lancetgh Vol 12 May 2024
e850
in urban areas compared with modelled approaches,7
because they use more realistic road and trac conditions
(appendix 1 pp 1–2). Evidence generated from such
analysis15 can inform priorities for health service
planning.
Nigeria consists of 36 states and a Federal Capital
Territory. The states are divided into 774 local government
areas, which are further divided into 8813 wards—the
lowest administration level. Of the country’s population
of 220 million, 53% live in urban areas, with a projected
increment to 70% expected by 2050.8 Evidence shows
significantly higher odds of maternal deaths and
stillbirths in urban areas compared with rural ones,
especially in the south of the country.16–18 The objective of
this study was to assess the geographical accessibility of
comprehensive emergency obstetric care using data
from a database of geocoded functional health facilities
in the 15 most populated cities in Nigeria and driving
time estimates to the three nearest facilities for eight
travel time scenarios spanning peak and non-peak hours,
including weekends and weekdays.
Methods
Study design
This population-based spatial analysis was conducted in
Nigerian cities with an estimated (2022) or projected
(2030) population of at least 1 million: Aba, Abuja, Benin
City, Ibadan, Ilorin, Jos, Kaduna, Kano, Lagos, Maiduguri,
Onitsha, Owerri, Port-Harcourt, Uyo, and Warri. These
15 cities accounted for 26% of the population in Nigeria
in 2022 (appendix 1 p 3).
This study involved assembly of data to define
extended city boundaries; to verify comprehensive
emergency obstetric care facility functionality, and
geographical location; and to map the population
distribution of girls and women aged 15–49 years (ie, of
childbearing age; herein referred to as women for
brevity) to estimate travel time to care. Details of the
methods used to collect and collate data for the study
and estimate travel time have been published elsewhere
(appendix 1 pp 1–2).19 Ethical approval was obtained
from the National Health Research and Ethics
Committee in Nigeria (NHREC/01/01/2007-11/04/2022)
and University of Greenwich Research and Ethics
Committee in the UK (UREC/21.4.7.8).
Data assembly
Because actual boundaries of the included cities were not
available, we delineated extended city boundaries
(including suburbs) of each city by its constituent local
government areas by spatially overlaying the vector file
of the local government area boundaries,20 WorldPop’s
gridded surface of population at 100 m² resolution (the
constrained 2020, UN adjusted version),21 Google Maps,
which was accessed as a basemap through ArcMap 10.5,
and Global Human Settlement (GHS) layers of gridded
surfaces (version GHS-SMOD R2023A).22 For each city,
all local government areas with parts of higher population
density than the surroundings within the area or marked
as urban, suburban, or peri-urban in the GHS layer
were selected for analysis.23 PMM led the city border
delineation process supported by coauthors familiar with
the context (AB-T, TO, OO, UG-A, and BBA).
A list of hospitals in the 15 selected cities and
information on facility name, ownership, location (local
government area and GPS coordinates), and operational
status (open or closed) was extracted from the 2018
Nigeria Health Facility Registry.24 These data were
complemented by state-specific lists in Lagos state and
from stakeholders familiar with health service provision
in other states. Data on emergency obstetric care service
availability were obtained through a facility functionality
assessment survey to identify facilities open 24 h a day
and able to conduct caesarean sections (used as a proxy
for comprehensive emergency obstetric care). Facility
ownership was also confirmed (ie, public [ federal or
state] or private [ for-profit, not-for-profit, faith-based,
military, or police-owned facilities]). The survey was led
by AB-T, TO, OO, UG-A, and BBA and involved in-person
facility visits conducted by research assistants using a
short questionnaire between March 1 and Aug 31, 2022.
The population distribution of women aged
15–49 years was obtained at 1 km² spatial resolution
from the WorldPop open spatial demographic data
portal.25 WorldPop uses dasymetric techniques to create
a constrained gridded surface by disaggregating 2006
census data from local government areas based on
weights derived from covariates such as land use, land
cover, and night-time lights.22 National estimates were
projected to match UN Population Division 2022
estimates while adjusting for rural–urban dierences.
Data for age and sex multipliers were derived from
census and household data and were applied to the
projected national estimates to derive the proportion of
women aged 15–49 years nationally. Geospatial layers of
various resolutions were resampled to 0·36 km² to
match the resolution used for travel time.26
Outcomes
The study outcome was comprehensive emergency
obstetric care geographical coverage, which was assessed
using three indicators: (1) the median travel time to
comprehensive emergency obstetric care facilities, (2) the
percentage of women aged 15–49 years who live within
three time thresholds (≤15 min, ≤30 min, and ≤60 min)
of a comprehensive emergency obstetric care facility, and
(3) the number of unique comprehensive emergency
obstetric care facilities reachable within the same three
time-thresholds under eight trac scenarios.
Statistical analysis
To obtain travel time to the identified comprehensive
emergency obstetric care facilities, we considered the
entire study region as being made up of level 14 S2 cells.
See Online for appendix 1

Articles
e851
www.thelancet.com/lancetgh Vol 12 May 2024
The S2 cell optimises the splitting of a spherical surface
into grid cells of approximately equal size. Specifically,
level 14 S2 cells are approximately 600 m by 600 m.26 This
resolution was selected to balance between accuracy and
computational feasibility needed for analysis. For each cell,
travel time by driving was extracted from the Google Maps
Platform’s internal Directions Application Programming
Interface (API), which oers closer-to-reality travel time
estimates compared with other approaches, such as the
cost-friction surface approach and Open Source Routing
Machine.7 This ability to generate closer-to-reality travel
time estimates is possible because of the API’s use of road
network data; crowdsourced data on road conditions; and
current and historic road trac patterns both in terms of
speed and human routing preference, within a machine
learning environment, to estimate travel time compared
with the assumed or user-defined speeds used by
alternative approaches (appendix 1 pp 1–2).7,27 For this
study, estimated travel times by driving were from each
cell centre, which represented the potential location of
women (ie, the origin of their journey to care) to the
three nearest comprehensive emergency obstetric care
facilities, by facility ownership (ie, public, private, and
both). For each ownership, travel time was derived during
eight trac scenarios covering weekdays and weekends
and a range of daily time periods spanning peak (0600 h to
0800 h and 1800 h to 2000 h) and non-peak trac (0100 h
to 0300 h and 1300 h to 1500 h).19 The extraction of travel
time from the API was done in January, 2023. To validate
the travel time estimates, journeys to care in each city
were randomly selected and API-generated travel time
for those journeys were compared with estimates from
the typical travel time function on the front end of
Google Maps.19
For each city, we estimated the three indicators of
geographical coverage of median travel time, percentage
of women aged 15–49 years who lived under three
dierent time-thresholds, and number of unique
facilities reachable within these three dierent time-
thresholds, using travel time to the three nearest
comprehensive emergency obstetric care facilities from
the centre of each S2 cell, at the level of the city and ward.
Ward-level median travel time (IQR) is reported for
assessment of within-city heterogeneity. Where possible,
findings were further contextualised against established
lists of informal settlements in the country.
In this paper, we present results for the cities of
Maiduguri and Kaduna, which had the median travel
time extremes (lowest and highest) and summary plots
for all cities.
Analysis and visualisation as static maps were done with
R (version 4.2.0) and ArcMap (version 10.8.1). Data used
are publicly available and described in detail elsewhere.19
Role of the funding source
The funder of the study had no role in study design, data
analysis, or data interpretation, but technical members of
the funder’s organisation contributed to data collection
of travel time estimates and contributed to the writing of
the report.
Results
The 15 cities consist of 104 774 local government areas with
a range of two in Maiduguri to 20 in Lagos. Across the
included cities there were 1440 wards. In 2022, the
estimated population of each city ranged from 1·1 million
in Maiduguri to 20·6 million in Lagos, and the total
number of women aged 15–49 years living in these cities
was 11·5 million (appendix 1 pp 3–4). The functionality
and location of 2020 comprehensive emergency obstetric
care facilities was identified, ranging from 26 in Maiduguri
to 796 in Lagos, with a median of 76 per city. Of all facilities,
1778 (88%) identified facilities were private for-profit,
121 (6%) were operated by the state or federal government,
and 121 (6%) were private not-for-profit. The total number
of functional comprehensive emergency obstetric care
facilities per 100 000 women aged 15–49 years ranged from
6·5 in Abuja to 33·3 in Owerri (appendix 1 p 4).
Results on travel time, the percentage of studied
women with comprehensive emergency obstetric care
coverage, and the number of comprehensive emergency
obstetric care facilities within reach under dierent
time thresholds are available for each city in appendix 1
(pp 5–14). Overall, the city-level median travel time to
the nearest comprehensive emergency obstetric care
facility during all eight travel scenarios was between
13 min and 18 min in Maiduguri (the city with the
lowest median travel time) and 42 min and 46 min in
Kaduna (the city with the highest median travel time;
table and figure 1).
Median travel time to the nearest (public or private)
comprehensive emergency obstetric care facility showed
more similarity across many of the 15 cities (range
8–30 min). Median travel time to the nearest (18 min),
second nearest (22 min), and third nearest (25 min) public
facility were similar in Maiduguri, but the increases in
travel time to reach the next nearest public comprehensive
emergency obstetric care facility were more substantial in
other cities, including Onitsha, Aba, and Owerri, due
mostly to the time to reach the third nearest facility
(figure 1A; appendix 1 pp 15–17). Accounting for both
public and private facilities, changes in median travel time
to the nearest, second nearest, and third nearest facilities
did not vary substantially across cities.
Also, at the city level, the percentage of women aged
15–49 years within 15 min of any comprehensive
emergency obstetric care facility ranged from 55% in
Abuja to 95% in Lagos, and, of comprehensive public
facilities, from 8% in Aba to 89% in Maiduguri (figure 1).
This coverage increased under the 30-min travel time
threshold but remained low in some cities (eg, 33% in Aba
for public facilities). The coverage of public comprehensive
emergency obstetric care facilities with 60 min travel time
ranged from 85% to over 90% in 13 cities.

Articles
www.thelancet.com/lancetgh Vol 12 May 2024
e852
Public Public or private
Weekday
0600 h to
0800 h
Weekday
1300 h to
1500 h
Weekday
1800 h to
2000 h
Weekday
0100 h to
0300 h
Weekend
0600 h to
0800 h
Weekend
1300 h to
1500 h
Weekend
1800 h to
2000 h
Weekend
0100 h to
0300 h
Weekday
0600 h to
0800 h
Weekday
1300 h to
1500 h
Weekday
1800 h to
2000 h
Weekday
0100 h to
0300 h
Weekend
0600 h to
0800 h
Weekend
1300 h to
1500 h
Weekend
1800 h to
2000 h
Weekend
0100 h to
0300 h
Maiduguri
Median travel time to a comprehensive emergency obstetric care facility (min)
Nearest 18
(11–30)
18
(12–30)
18
(12–30)
18
(12–30)
18
(11–30)
18
(12–30)
18
(12–30)
18
(11–30)
13
(8–26)
13
(8–26)
13
(8–26)
13
(8–26)
13
(8–26)
13
(8–26)
13
(8–26)
13
(8–26)
Second
nearest
22
(16–33)
22
(17–34)
22
(17–34)
22
(16–33)
21
(16–33)
22
(17–34)
22
(17–34)
22
(16–33)
16
(11–28)
16
(11–28)
16
(11–28)
16
(11–28)
16
(11–28)
16
(11–28)
16
(11–28)
16
(11–28)
Third nearest 24
(18–36)
25
(20–37)
25
(19–37)
24
(18–36)
23
(18–36)
24
(19–37)
24
(19–37)
24
(18–36)
18
(13–30)
18
(13–30)
18
(13–30)
18
(13–29)
17
(12–29)
18
(13–30)
18
(13–30)
17
(13–29)
Percentage of women aged 15–49 years within reach of a comprehensive emergency obstetric care facility
Within 15 min 236 509
(89%)
236 509
(89%)
236 509
(89%)
236 509
(89%)
239 166
(90%)
236 509
(89%)
236 509
(89%)
236 509
(89%)
249 796
(94%)
249 796
(94%)
249 796
(94%)
249 796
(94%)
249 796
(94%)
249 796
(94%)
249 796
(94%)
249 796
(94%)
Within 30 min 257 768
(97%)
257 768
(97%)
257 768
(97%)
257 768
(97%)
257 768
(97%)
257 768
(97%)
257 768
(97%)
257 768
(97%)
260 425
(98%)
260 425
(98%)
260 425
(98%)
260 425
(98%)
260 425
(98%)
260 425
(98%)
260 425
(98%)
260 425
(98%)
Within 60 min 265 740
(100%)
265 740
(100%)
265 740
(100%)
265 740
(100%)
265 740
(100%)
265740
(100%)
265 740
(100%)
265 740
(100%)
265 740
(100%)
265 740
(100%)
265 740
(100%)
265 740
(100%)
265 740
(100%)
265 740
(100%)
265 740
(100%)
265 740
(100%)
Number of comprehensive emergency obstetric care facilities within reach
Within 15 min 1 0 0 1 1 0 0 1 5 4 5 5 5 5 5 5
Within 30 min 3 3 3 3 3 3 3 3 17 16 16 17 17 16 16 17
Within 60 min 4 4 4 4 4 4 4 4 25 25 25 25 25 25 25 25
Kaduna
Median travel time to a comprehensive emergency obstetric care facility (min)
Nearest 42
(29–59)
45
(31–63)
46
(32–65)
42
(29–59)
40
(28–58)
44
(31–62)
46
(32–65)
42
(29–59)
30
(20–49)
30
(20–50)
30
(20–51)
29
(19–49)
29
(19–48)
30
(20–50)
30
(20–51)
29
(19–49)
Second
nearest
46
(35–63)
50
(37–68)
51
(38–70)
46
(34–63)
44
(33–61)
49
(37–67)
50
(38–69)
46
(34–63)
38
(27–56)
39
(28–58)
40
(29–59)
38
(27–56)
38
(27–55)
39
(28–58)
40
(29–59)
38
(27–56)
Third nearest 52
(40–70)
56
(43–74)
57
(44–77)
52
(40–70)
50
(39–68)
55
(43–73)
57
(44–76)
52
(40–70)
52
(30–58)
56
(32–61)
57
(33–64)
52
(30–58)
50
(29–57)
55
(31–61)
57
(32–63)
52
(30–58)
Percentage of women aged 15–49 years within reach of a comprehensive emergency obstetric care facility
Within 15 min 328 387
(64%)
292 470
(57%)
292 470
(57%)
323 256
(63%)
343 780
(67%)
302 732
(59%)
297 601
(58%)
328 387
(64%)
395 091
(77%)
389 960
(76%)
384 829
(75%)
395 091
(77%)
400 222
(78%)
384 829
(75%)
384 829
(75%)
395 091
(77%)
Within 30 min 451 532
(88%)
446 401
(87%)
446 401
(87%)
451 532
(88%)
456 663
(89%)
446 401
(87%)
446 401
(87%)
451 532
(88%)
472 057
(92%)
472 057
(92%)
472 057
(92%)
472 057
(92%)
472 057
(92%)
472 057
(92%)
472 057
(92%)
472 057
(92%)
Within 60 min 487 450
(95%)
487 450
(95%)
482 319
(94%)
487 450
(95%)
487 450
(95%)
487 450
(95%)
482 319
(94%)
487 450
(95%)
497 712
(97%)
497 712
(97%)
497 712
(97%)
497 712
(97%)
497 712
(97%)
497 712
(97%)
497 712
(97%)
497 712
(97%)
Number of comprehensive emergency obstetric care facilities within reach
Within 15 min 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0
Within 30 min 0 0 0 0 0 0 0 0 8 6 6 8 9 6 6 8
Within 60 min 2 2 2 2 2 2 2 2 30 26 25 30 31 27 26 30
There were 265 740 women aged 15–49 years in Maiduguri in 2022 and 513 105 in Kaduna.
Table: Median travel time, percentage of women aged 15–49 years, and number of facilities within reach under different time thresholds in Maiduguri and Kaduna

Articles
e853
www.thelancet.com/lancetgh Vol 12 May 2024
For the peak trac hours of 1800 h to 2000 h on
weekdays, the median number of public comprehensive
emergency obstetric care facilities reachable within
15 min for the study population was zero in all cities
(figure 1C). The median number of public comprehensive
facilities reachable within 30 min remained at zero in
eight (53%) of 15 cities (figure 1). Considering both public
and private comprehensive facilities, the city with the
highest number of facilities reachable within 15 min was
Onitsha (n=9), while the city with the highest number of
facilities within 30 min (n=42) and 60 min (n=182) was
Lagos. At the ward level (ie, within cities), travel time in
dierent wards showed substantial dierences (figure 2).
For the nearest public comprehensive emergency
Figure 1: Median travel time to comprehensive emergency obstetric care facilities (A), percentage of women aged 15–49 years within reach of these
facilities (B), and median number of comprehensive emergency obstetric care facilities within specific travel times (C) in the 15 cities
Comprehensive emergency obstetric care facility distance and type
Median travel time (min)
150
135
120
105
90
75
60
30
0
45
15
Nearest public or
private facility
Nearest public or
private facility
Second nearest
public or private
facility
Third nearest
public or private
facility
Nearest public or
private facility
Second nearest
public or private
facility
Third nearest
public or private
facility
Nearest public
facility
A
Travel time to comprehensive emergency obstetric care facility
100
80
60
40
20
0
≤15 min to the
nearest public or
private facility
≤30 min to the
nearest public or
private facility
≤60 min to the
nearest public or
private facility
≤15 min to the
nearest public
facility
≤30 min to the
nearest public
facility
≤60 min to the
nearest public
facility
≤15 min to the
nearest public or
private facility
≤15 min to the
nearest public
facility
B
Women aged 15–49 years within
reach of facilities (%)
Travel time thresholds
200
150
100
50
0
≤15 min to a
public or
private facility
≤30 min to a
public or
private facility
≤60 min to a
public or
private facility
≤15 min to a
public facility
≤30 min to a
public facility
≤60 min to a
public facility
≤15 min to a
public or
private facility
≤15 min to
a public
facility
C
Number of comprehensive emergency
obstetric care facilities within reach
Aba
Lagos
Abuja
Maiduguri
Benin City
Onitsha
Ibadan
Owerri
Ilorin
Port-Harcourt
Jos
Uyo
Kaduna
Warri
Kano

Articles
www.thelancet.com/lancetgh Vol 12 May 2024
e854
obstetric care facility, the IQR of median travel time
varied from the narrowest in Maiduguri (10 min) to the
widest in Benin City (41 min). Ward-level median travel
time values were less than 60 min in most wards (at
least 90%) but were over 2 h in some (<1%; appendix 1 p 18).
Ward-level travel times to either public or private facilities
were less variable based on IQR than median—4 min for
Lagos to over 20 min in Kaduna, Benin City, and Abuja.
The comparisons across wards were similar during
weekdays (1800 h to 2000 h) and weekends (0100 h to
0300 h), which were typically the longest and shortest
times to travel, respectively (figure 3). Median travel time
reduced substantially for the majority of wards in some
cities when considering public and private facilities
instead of public alone. For instance, in Lagos the median
travel time to the third nearest facility (public or private)
between 1800 h and 2000 h on a weekday was below
12·3 min in 283 (75%) of the 377 wards—a reduction
from 54·3 min if only public facilities were considered.
However, the median travel time remained greater
than 60 min in 11 (3%) wards regardless of inclusion of
private facilities.
The percentage of women aged 15–49 years within
30 min of a comprehensive emergency obstetric care
facility for the measured eight dierent times throughout
the week in a random sample of 40 (11%) of the 377 wards
in Lagos are shown in figure 3. Coverage of public
comprehensive care facilities varied substantially in
some wards by time of day (eg, 35% for weekday 1800 h
to 2000 h vs 100% for weekend 0100 h to 0300 h in
FESTAC I ward [AM002], for instance), while it remained
static in others (eg, in Ojo Town [OJO13] and Apapa II—
Liverpool road and environs [APA02] wards). Overall, for
the 40 randomly selected wards, there were more
fluctuations in the percentage of women aged 15–49 years
within 30 min of a public comprehensive emergency
obstetric care facility than in the percentage within
30 min of public and private combined (figure 4). Similar
patterns were also observed in other cities and wards,
although some cities, such as Maiduguri and Ilorin, had
a similar level of accessibility throughout the day and
week for public comprehensive care facilities in most
wards (appendix 1 p 18).
The percentage of women aged 15–49 years within
30 min travel time of public comprehensive emergency
obstetric care was typically lowest during weekdays
between 1800 h and 2000 h, and highest during the
weekend between 0100 h and 0300 h. In parts of Kano,
Lagos, Owerri, Port-Harcourt, and Uyo, coverage was
0–25% for weekdays at 1800 h to 2000 h and 75–100% for
Figure 2: Travel time to comprehensive emergency obstetric care facilities by ward in 15 cities, during weekdays 1800 h to 2000 h and weekends
0100 h to 0300 h
Edges of boxes represent quartiles, whiskers extend to 1·5 × interquartile range, and outlier values are plotted as individual points beyond the whiskers.
First nearest Second nearest Third nearest
Benin City
Port-Harcourt
Kaduna
Kano
Ibadan
Jos
Uyo
Abuja
Aba
Owerri
Onitsha
Lagos
Ilorin
Warri
Maiduguri
Benin City
Port-Harcourt
Kaduna
Kano
Ibadan
Jos
Uyo
Abuja
Aba
Owerri
Onitsha
Lagos
Ilorin
Warri
Maiduguri
Public only Public and private
Median travel time by ward (min)Median travel time by ward (min) Median travel time by ward (min)
150304560 75 90 105 120 150304560 75 90 105 120 150304560 75 90 105 120
Weekend 0100 h to 0300 h
Weekday 1800 h to 2000 h

Articles
e855
www.thelancet.com/lancetgh Vol 12 May 2024
weekends at 0100 h to 0300 h—dierences of at least
60 percentage points. In other cities, such as Aba,
Maiduguri, and Onitsha, coverage remained similar on
weekdays and weekends (appendix 1 p 18).
Ward-level dierences in coverage with and without
consideration of private facilities were largely similar in
Abuja, Maiduguri, and large parts of Warri (appendix 1
p 18) between 1800 h and 2000 h (when travel time is
typically the longest). In other cities, however, including
Aba, Jos, Lagos, and Uyo, ward-level coverage decreased
from 100% when considering public and private
comprehensive emergency obstetric care facilities to
0% when considering only public comprehensive care
facilities. In Lagos, wards with high coverage of public
and private comprehensive emergency obstetric care
facilities but low coverage of only public comprehensive
care facilities included areas characterised as informal
settlements (eg, Ajangbadi, Okokomaiko, Ilogbo-Elegba,
and Oto-Awori).
Discussion
This study assessed geographical accessibility to
comprehensive emergency obstetric care facilities in
15 large Nigerian cities using closer-to-reality travel time
estimates. To our knowledge, this study is the first on
emergency obstetric care accessibility to have considered
real-life trac dierentials at scale in a low-income and
middle-income country. We found that even under the
heaviest trac conditions, women aged 15–49 years in all
cities were within a median travel time of 45 min to the
nearest public comprehensive emergency obstetric care
facility. However, in five cities (Aba, Jos, Kaduna, Onitsha,
and Port-Harcourt), between 5–15% of women did not
have a comprehensive care facility within an hour’s travel,
especially within the public sector. During peak trac
times, the median number of public comprehensive
emergency obstetric care facilities available within 30 min
was zero in eight of the 15 cities. We identified within-city
geographical inequities of accessibility to comprehensive
care, notably for Benin City, Port-Harcourt, and Kaduna.
We found that most locations in the 15 cities were
within an hour’s drive to one or two public comprehensive
emergency obstetric care facilities. Our results suggest
reasonably acceptable geographical accessibility, based on
WHO’s 2-h benchmark to emergency obstetric care.4
Even in Kaduna, the city with the lowest coverage, 94% of
women were within 1 h travel time of a comprehensive
emergency obstetric care facility under peak trac
conditions. The contextualised spatial analysis of
comprehensive emergency obstetric care facilities and
accessibility to them suggests that the main drivers of
coverage to comprehensive care within these cities are
trac and distribution of functional comprehensive
care facilities across the city. For example, functional
comprehensive emergency obstetric care facilities in
Maiduguri appeared to be evenly spread across the city
centre and its suburbs, while in Kaduna, facilities were
mostly clustered in the centre. Also, cities with particularly
serious trac congestion (notably Abuja, Benin City, and
Port-Harcourt)28,29 were found to have the poorest coverage
when public and private facilities were combined. These
factors might explain why Lagos, despite having a
substantially higher number of functional facilities, did
not have the best availability (by number of facilities per
Figure 3: Percentage of women aged 15–49 years within 30 min travel time to the nearest comprehensive emergency obstetric care facility at eight different
times by ward in Lagos
Data are for a random sample of 40 wards in Lagos. Wards are listed on the right of lines. (A) Women aged 15–49 years within 30 min of the nearest public comprehensive
emergency obstetric care facility. (B) Women aged 15–49 years within 30 min of the nearest public or private comprehensive emergency obstetric care facility.
Weekday 0100 h to 0300 h
Weekday 0600 h to 0800 h
Weekday 1300 h to 1500 h
Weekend 0100 h to 0300 h
Weekend 0600 h to 0800 h
Weekend 1300 h to 1500 h
Weekend 1800 h to 2000 h
Weekday 1800 h to 2000 h
Departure time periods Departure time periods
Women aged 15–49 years within 30 min
of the nearest public comprehensive
emergency obstetric care facility (%)
Women aged 15–49 years within 30 min
of the nearest public or private comprehensive
emergency obstetric care facility (%)
AB
0
100
90
80
70
60
50
40
30
20
10
BDY07
BDY06
IUI05
EPE12
OJO13
AMO02
APA02
IKU14
EPE17
KSE19
IOE05
IKU13
Weekday 0100 h to 0300 h
Weekday 0600 h to 0800 h
Weekday 1300 h to 1500 h
Weekend 0100 h to 0300 h
Weekend 0600 h to 0800 h
Weekend 1300 h to 1500 h
Weekend 1800 h to 2000 h
Weekday 1800 h to 2000 h
0
100
90
80
70
60
50
40
30
20
10
BDY06
EPE12
APA02
IKU14
EPE17

Articles
www.thelancet.com/lancetgh Vol 12 May 2024
e856
100 000 women aged 15–49 years) or accessibility. In Aba,
geographical accessibility was particularly worse when
only considering public (and not private) comprehensive
emergency obstetric care facilities, because the only two
functional public care facilities are both in the northern
part of the city.
Although city-level estimates might be relevant for
comparative performance (ie, between city), granular
ward-level estimates can reveal within-city inequities
and help guide decision making in cities.30 We found
within-city inequity in the geographical accessibility
of comprehensive emergency obstetric care facilities.
Specifically, accessibility problems occur in places
without any local facilities and women therefore need to
travel across wards or neighbourhoods for care. Cities
such as Benin City, Port-Harcourt, and Kaduna had the
widest gaps across wards. In addition, layering geolocated
informal settlements with our data, we were able to
identify some informal settlements as areas with the
worst accessibility to public sector care.
Our analysis has several strengths. First, our study was
based on algorithmic outputs generated using Google
Maps’ internal Directions API, which has been shown
to oer closer-to-reality time estimates compared with
commonly used models such as the cost-friction surface
approach and Open Source Routing Machine.7 Second,
we included both public and private comprehensive
emergency obstetric care facilities and confirmed their
Figure 4: Ward-level differences in percentage points of women aged 15–49 years within 30 min reach of comprehensive emergency obstetric care facilities in
15 cities
AB
Abu Abuja Benin City
Ibadan Ilorin Jos
Kaduna
Maiduguri
Port-Harcourt UyoWarri
Onitsha Owerri
Kano Lagos
Abu Abuja Benin City
Ibadan Ilorin Jos
Kaduna
Maiduguri
Port-Harcourt Uy
oW
arri
Onitsha Owerri
Kano Lagos
Weekday 1800 h to 2000 h minus weekend 0100 h to 0300 h
0102030405060708090 100
Public and private facilities minus public facilities
Difference in percentage points

Articles
e857
www.thelancet.com/lancetgh Vol 12 May 2024
functionality, in contrast to other large-scale eorts
that have been limited to the public sector and not
confirmed functionality. Third, our approach integrated
the variability of trac throughout the day and week.
Fourth, we used several time thresholds for calculation of
coverage, in line with emerging evidence showing that
even travel time of 30 min increases odds of stillbirth and
maternal death during referral.17,18 Finally, our approach of
using a higher-resolution spatial scale of 0·36 km² made it
possible to granularly assess spatial accessibility at the level
of wards. This approach is required for higher accuracy
and robustness in model inputs to guide local policy and
decision making.30 For example, the approach allowed us
to identify informal settlements as having lower coverage,
which is relevant as 1 billion people globally and a third of
Africa’s urban population currently live in these settings.8
However, our results have limitations. First, we
acknowledge that bypassing the nearest facility is
common in care seeking. To account for this, our
accessibility metrics were based on the three nearest
functional comprehensive emergency obstetric care
facilities, thereby capturing some level of choice. Second,
the route mapped by the API is based on the routine
travel of commuters, which might not be aligned with
one a service user might take in an emergency. Third, we
assumed travel occurs by vehicle. Although it is true that
some pregnant women may access emergency obstetric
care by walking, motorised transport including taxis and
privately owned cars are used by a large majority in an
emergency.14 Linked to this is our assumption that a
vehicle was available to all women at the time of need.
Fourth, we made a pragmatic decision to use facilities
with caesarean section capacity as a proxy for all
comprehensive emergency obstetric care services.
However, we do not know availability of other services.
For example, assisted vaginal delivery or blood
transfusion might not be available in all facilities all the
time, irrespective of whether they are public or private.14
Finally, our urban delineation was based on qualitative
assessment as we included areas ranging from very
dense to peri-urban areas. Although this aligns with the
expansion of urban areas into peri-urban areas and
reflects the lived reality of rapid urbanisation, it also
means that some of the fringes of our selected areas
might include some less urban segments.
Our findings have substantial policy and planning
implications. Local governance which takes a multi-
sectoral approach involving health, urban and regional
planning, and transportation is needed to optimise
emergency obstetric care geographical accessibility.
Beyond suboptimal distribution of comprehensive
emergency obstetric care facilities, rapid urbanisation,
expansion of vehicle ownership, poor roads, and lack
of law enforcement, all contribute to the trac seen
in African cities. Innovative approaches aimed at
minimising trac and diversifying commuting options
for city populations are needed31 to ensure that roads are
more accessible to those who need them the most,
including pregnant women in an emergency. Urban
redesign initiatives focused on addressing comprehensive
emergency obstetric care geographical accessibility such
as ambulance-only lanes or roads and direct passes to
hospitals, targeting areas of inequities to situate available
ambulances, and community awareness campaigns
targeting reorientation of commuters to give way to
emergency vehicles are options to consider. To maximise
the usefulness of the evidence from this research, the
data that informed it was also used to develop a publicly
available digital dashboard allowing an interactive
visualisation of dierent travel scenarios to aid service
planning and decision making. Future developments
should include other emergency obstetric care services
and other dimensions of access including quality of care,
cost of care, and availability of structures to support
access such as ambulances,32 and reflect seasonal
variations in accessibility.
In conclusion, inequities in the geographical accessibility
of emergency obstetric care exist in Nigerian cities.
However, the location and magnitude of such inequities
dier across cities, with informal settlements being of
particular concern. Although our innovative approach can
be data intensive, requiring data from up-to-date facility
verification and closer-to-reality travel time estimates,15,19 it
oers an opportunity to generate more context-specific,
finer, and policy-relevant evidence to support improving
the geographical accessibility of emergency obstetric care
in urban Africa. Replication of this approach across urban
areas of the continent will aid the identification of targeted
cost-eective strategies that ensure that we leave no one
behind, including all pregnant women, in eorts to realise
universal health coverage.
Contributors
AB-T, KLMW, and PMM conceptualised the study and prepared the
analytical plan for the study with support from TO and LB. AB-T, TO,
OO, UG-A, and BBA led the facility functionality verification activity
conducted as part of the study. NS, YS, GP, MK, SV, TS, and CS were
involved in aggregating the travel time estimates from Google’s
internal Directions Application Programming Interface used for the
study. AB-T, PMM, PTM, LB, KLMW, UG-A, and I-OOA conducted the
literature review that informed the study. KLMW conducted the
analysis of the data with support from AB-T, PMM, and LB.
AB-T, KLMW, and PMM prepared the first draft of the manuscript.
AB-T, KLMW, TO, PMM, OO, UG-A, JW, I-OOA, PTM, NA, CN, BBA,
and LB contributed to the interpretation of data. All authors were
involved in editing of the article, commented critically on the
manuscript, and approved the final version of the manuscript. All
authors had full access to all the data in the study, and had final
responsibility for the decision to submit for publication.
Equitable partnership declaration
The authors of this paper have submitted an equitable partnership
declaration (appendix 2). This statement allows researchers to describe
how their work engages with researchers, communities, and
environments in the countries of study. This statement is part of
The Lancet Global Health’s broader goal to decolonise global health.
Declaration of interests
NS, YS, GP, MK, SV, TS, and CS are current or past employees of Google,
which developed the Google Maps Platform. AB-T and BBA are funded by
the Bill & Melinda Gates Foundation (investment identification
For the digital dashboard see
https://goo.gle/
emergencyobstetriccare
See Online for appendix 2

Articles
www.thelancet.com/lancetgh Vol 12 May 2024
e858
INV-032911). PMM was supported by Newton International Fellowship
(number NIF/R1/201418) of the Royal Society and acknowledges the
support of the Wellcome Trust to the Kenya Major Overseas Programme
(number 203077). UG-A is funded by a joint Clarendon, Balliol College,
and Nueld Department of Population Health DPhil scholarship. LB was
funded in part by the Research Foundation–Flanders as part of her Senior
Postdoctoral Fellowship. All other authors declare no competing interests.
Data sharing
The datasets used for this analysis are publicly available. These include a
database of health facilities with capacity for caesarean section in urban
Nigeria (https://doi.org/10.6084/m9.figshare.22689667) and a geospatial
database of close-to-reality travel times to obstetric emergency care in
15 Nigerian conurbations (https://figshare.com/s/8868db0bf3fd18a9585d).
Acknowledgments
We would like to express our most sincere appreciation to the Nigerian
Federal Ministry of Health and all state-level ministries of health involved
in the project. We are also indebted to the research assistants (fifth-year and
sixth-year medical students from the University of Ilorin, University of
Benin, University of Jos, University of Ibadan, Nnamdi Azikiwe University,
University of Uyo, University of Lagos, and Chukwuemeka Odumegwu
Ojukwu University; a nurse from Bingham University Teaching Hospital;
medical doctors from Ahmadu Bello Teaching Hospital and Lagos
University Teaching Hospital; and research assistants from the states of
Abia, Borno, Kano, Port-Harcourt, Imo, Delta, and the Federal Capital
Territory) who supported the health facilities validation exercise, from
May 10 to Aug 9, 2022. The study was funded by Google through a grant
awarded to AB-T, who is the Principal Investigator of the On Tackling In
transit delays for Mothers in Emergency (OnTIME) Consortium.
Editorial note: The Lancet Group takes a neutral position with respect to
territorial claims in published maps.
References
1 WHO, UNICEF, UN Population Fund, World Bank Group, UN
Department of Economic and Social Aairs/Population Division.
Trends in maternal mortality 2000 to 2020: estimates by WHO,
UNICEF, UNFPA, World Bank Group and UNDESA/Population
Division. Geneva: World Health Organization, 2023: 1–108.
2 UNICEF, WHO, World Bank, UN Procurement Division. A
neglected tragedy: the global burden of stillbirths. New York, NY:
United Nations Children’s Fund, 2020: 1–90.
3 Paxton A, Maine D, Freedman L, Fry D, Lobis S. The evidence for
emergency obstetric care. Int J Gynaecol Obstet 2005; 88: 181–93.
4 WHO, UN Population Fund, UNICEF, Averting Maternal Deaths
and Disabilities. Monitoring emergency obstetric care: a handbook.
Geneva: WHO Press, 2009.
5 Vearey J, Luginaah I, Magitta NF, Shilla DJ, Oni T. Urban health in
Africa: a critical global public health priority. BMC Public Health
2019; 19: 340.
6 Matthews Z, Channon A, Neal S, Osrin D, Madise N, Stones W.
Examining the “urban advantage” in maternal health care in
developing countries. PLoS Med 2010; 7: e1000327.
7 Banke-Thomas A, Wong KLM, Ayomoh FI, Giwa-Ayedun RO,
Benova L. “In cities, it’s not far, but it takes long”: comparing
estimated and replicated travel times to reach life-saving obstetric
care in Lagos, Nigeria. BMJ Glob Health 2021; 6: e004318.
8 UN. World urbanization prospects: the 2018 revision. New York,
NY: United Nations, 2019.
9 Sidze EM, Wekesah FM, Kisia L, Abajobir A. Inequalities in access
and utilization of maternal, newborn and child health services in
sub-Saharan Africa: a special focus on urban settings.
Matern Child Health J 2022; 26: 250–79.
10 Banke-Thomas A, Wright K, Collins L. Assessing geographical
distribution and accessibility of emergency obstetric care in sub-
Saharan Africa: a systematic review. J Glob Health 2019; 9: 010414.
11 Banke-Thomas A, Macharia PM, Makanga PT, et al. Leveraging big
data for improving the estimation of close to reality travel time to
obstetric emergency services in urban low- and middle-income
settings. Front Public Health 2022; 10: 931401.
12 Ahmed S, Adams AM, Islam R, Hasan SM, Panciera R. Impact of
trac variability on geographic accessibility to 24/7 emergency
healthcare for the urban poor: a GIS study in Dhaka, Bangladesh.
PLoS One 2019; 14: e0222488.
13 Dotse-Gborgbortsi W, Dwomoh D, Alegana V, Hill A, Tatem AJ,
Wright J. The influence of distance and quality on utilisation of
birthing services at health facilities in Eastern Region, Ghana.
BMJ Glob Health 2020; 4 (suppl 5): e002020.
14 Banke-Thomas A, Balogun M, Wright O, et al. Reaching health
facilities in situations of emergency: qualitative study capturing
experiences of pregnant women in Africa’s largest megacity.
Reprod Health 2020; 17: 145.
15 Banke-Thomas A, Wong KLM, Collins L, et al. An assessment of
geographical access and factors influencing travel time to
emergency obstetric care in the urban state of Lagos, Nigeria.
Health Policy Plan 2021; 36: 1384–96.
16 Meh C, Thind A, Ryan B, Terry A. Levels and determinants of
maternal mortality in northern and southern Nigeria.
BMC Pregnancy Childbirth 2019; 19: 417.
17 Banke-Thomas A, Avoka CK, Gwacham-Anisiobi U, et al. Travel of
pregnant women in emergency situations to hospital and maternal
mortality in Lagos, Nigeria: a retrospective cohort study.
BMJ Glob Health 2022; 7: e008604.
18 Banke-Thomas A, Avoka CK, Gwacham-Anisiobi U, Benova L.
Influence of travel time and distance to the hospital of care on
stillbirths: a retrospective facility-based cross-sectional study in
Lagos, Nigeria. BMJ Glob Health 2021; 6: e007052.
19 Macharia PM, Wong KLM, Olubodun T, et al. A geospatial database
of close-to-reality travel times to obstetric emergency care in
15 Nigerian conurbations. Sci Data 2023; 10: 736.
20 GRID3 Nigeria. National Local Government Administrative
Boundaries. 2022. https://grid3.gov.ng/dataset/national-local-
government-administrative-boundaries/resources (accessed
April 6, 2023).
21 Stevens FR, Gaughan AE, Linard C, Tatem AJ. Disaggregating
census data for population mapping using random forests with
remotely-sensed and ancillary data. PLoS One 2015; 10: e0107042.
22 Schiavina M, Melchiorri M, Pesaresi M. GHS-SMOD R2023A—
GHS settlement layers, application of the degree of urbanisation
methodology (stage I) to GHS-POP R2023A and GHS-BUILT-S
R2023A, multitemporal (1975-2030). 2023. https://data.jrc.ec.europa.
eu/dataset/a0df7a6f-49de-46ea-9bde-563437a6e2ba (accessed
Sept 16, 2023).
23 WorldPop, Center for International Earth Science Information
Network, Columbia University. Global high resolution population
denominators project. 2018. https://doi.org/10.5258/SOTON/
WP00660 (accessed Sept 2, 2023).
24 Federal Ministry of Health. Nigeria Health Facility Registry (HFR).
2019. https://hfr.health.gov.ng/download/facilities (accessed
June 26, 2023).
25 Bondarenko M, Tejedor-Garavito N, Priyatikanto R, Sorichetta A,
Tatem A. Interim: unconstrained and constrained estimates of
2021–2022 total number of people per grid square, adjusted to
match the corresponding UNPD 2022 estimates (1km resolution),
version 1.0. 2022. https://eprints.soton.ac.uk/472186/ (accessed
Sept 16, 2023).
26 The s2geometry.io. S2 cells geometry. 2022. http://s2geometry.io/
about/overview.html (accessed April 6, 2023).
27 Lau J. Google Maps 101: how AI helps predict trac and determine
routes. 2020. https://blog.google/products/maps/google-maps-101-
how-ai-helps-predict-trac-and-determine-routes (accessed
July 13, 2023).
28 TheNewspaperStand. 5 cities in Nigeria with the most trac. 2023.
https://www.linkedin.com/pulse/5-cities-nigeria-most-trac-
thenewspaperstand/ (accessed Dec 16, 2023).
29 Nwankwo W, Olayinka A, Ukhurebor K. The urban trac
congestion problem in Benin City and the search for an ICT—
improved solution. Int J Sci Tech Res 2019; 8: 65–72.
30 Hierink F, Boo G, Macharia PM, et al. Dierences between
gridded population data impact measures of geographic access to
healthcare in sub-Saharan Africa. Commun Med (Lond) 2022;
2: 117.
31 Rwakarehe E. Review of strategies for curbing trac congestion in
sub-Saharan Africa cities: technical and policy perspectives.
Tanzania J Eng Tech 2022; 40: 24–32.
32 Penchansky R, Thomas JW. The concept of access: definition and
relationship to consumer satisfaction. Med Care 1981; 19: 127–40.