Speizer et al.

BMC Pregnancy and Childbirth 2014, 14:398

http://www.biomedcentral.com/1471-2393/14/398

RESEARCH ARTICLE Open Access

Factors associated with institutional delivery in

Ghana: the role of decision-making autonomy
and community norms
Ilene S Speizer1,2*, William T Story1 and Kavita Singh1,2

Abstract
Background: In Ghana, the site of this study, the maternal mortality ratio and under-five mortality rate remain high
indicating the need to focus on maternal and child health programming. Ghana has high use of antenatal care (95%)
but sub-optimum levels of institutional delivery (about 57%). Numerous barriers to institutional delivery exist including
financial, physical, cognitive, organizational, and psychological and social. This study examines the psychological and
social barriers to institutional delivery, namely women’s decision-making autonomy and their perceptions about social
support for institutional delivery in their community.
Methods: This study uses cross-sectional data collected for the evaluation of the Maternal and Newborn Referrals
Project of Project Fives Alive in Northern and Central districts of Ghana. In 2012 and 2013, a total of 2,527 women aged
15 to 49 were surveyed at baseline and midterm (half in 2012 and half in 2013). The analysis sample of 1,606 includes
all women who had a birth three years prior to the survey date and who had no missing data. To determine the
relationship between institutional delivery and the two key social barriers—women’s decision-making autonomy and
community perceptions of institutional delivery—we used multi-level logistic regression models, including cross-level
interactions between community-level attitudes and individual-level autonomy. All analyses control for the clustered
survey design by including robust standard errors in Stata 13 statistical software.
Results: The findings show that women who are more autonomous and who perceive positive attitudes toward facility
delivery (among women, men and mothers-in-law) were more likely to deliver in a facility. Moreover, the interactions
between autonomy and community-level perceptions of institutional delivery among men and mothers-in-law were
significant, such that the effect of decision-making autonomy is more important for women who live in communities
that are less supportive of institutional delivery compared to communities that are more supportive.
Conclusions: This study builds upon prior work by using indicators that provide a more direct assessment of perceived
community norms and women’s decision-making autonomy. The findings lead to programmatic recommendations that
go beyond individuals and engaging the broader network of people (husbands and mothers-in-law) that influence
delivery behaviors.
Keywords: Institutional delivery, Ghana, Maternal health, Autonomy, Social support

* Correspondence: Ilene_speizer@unc.edu
1
Carolina Population Center, University of North Carolina at Chapel Hill,
Chapel Hill, NC, USA
2
Department of Maternal and Child Health, Gillings School of Global Public
Health, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA

© 2014 Speizer et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative
Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain
Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article,
unless otherwise stated.
Speizer et al. BMC Pregnancy and Childbirth 2014, 14:398 Page 2 of 13
http://www.biomedcentral.com/1471-2393/14/398

Background Often, the financial and physical barriers are examined

As we approach 2015 and the deadline for attainment of together to capture issues around transportation and
the Millennium Development Goals (MDG), increased distance to a facility as well as costs to reach or use the
attention and effort are being given to reaching target facility [13,15-17]. Ghana has implemented the National
populations with specific programmatic strategies, espe- Health Insurance Scheme in an effort to reduce these
cially in countries that are not on track to attain the types of financial barriers. Recent studies from Ghana
MDGs. In Ghana, the site of this study, where the ma- have found that women with health insurance were
ternal mortality ratio (MMR) remains high at 350 mater- more likely to have an institutional delivery [18-20] and
nal deaths per 100,000 live births [1] and under-five insurance was associated with better maternal and child
mortality is estimated at 82 deaths per 1000 live births health outcomes [20]. Cognitive barriers are related to
[2], there has been increased attention to initiatives to misconceptions about services offered and concerns
improve maternal, infant, and child health services. To about quality of services [17]. Organizational barriers are
help in the attainment of improved infant and child focused on the role of the providers in terms of atti-
health (MDG 4) and improved maternal health (MDG tudes, availability, and services offered [17,21]; these
5), programs in Ghana and elsewhere promote skilled at- have been found to be particularly important from quali-
tendance at delivery; in many low income countries, this tative studies in Ghana [14,21]. Finally, psychological
is equated with institutional (also called facility) delivery and social barriers are related to community norms and
[3]. Skilled attendance at delivery means having an attitudes toward facility delivery and toward the staff at
accredited health professional, including a midwife, doc- the facility [3,17]. These barriers to health care have
tor, or nurse, who has been trained in the skills needed been demonstrated in various cultural contexts including
to manage a normal or uncomplicated pregnancy and Bangladesh [22], Cambodia [17], Nepal [23], rural Kenya
childbirth and to support the woman in the immediate [13], and Northern Ghana [10] mostly using qualitative
postpartum period. This person should also be able to data or nationally representative data from Demographic
identify, manage and refer complications experienced by and Health Surveys.
the woman or the newborn [4]. Social barriers to institutional delivery, such as com-
Numerous studies have demonstrated that in sub- munity attitudes towards institutional delivery and levels
Saharan Africa, including Ghana, there is often high use of decision-making autonomy among women, have re-
of antenatal care services but lower use of institutional ceived much less attention in the literature [24]. Com-
delivery [2,5,6]. For example, the 2008 Ghana Demo- munity beliefs and attitudes about maternal health
graphic and Health Survey (GDHS) showed that more behaviors have been shown to influence a woman’s indi-
than 95% of women who had a birth in the last five years vidual decision to seek care. For example, in a study of
received antenatal care from a skilled provider prior to six countries in sub-Saharan Africa, Stephenson and col-
birth [6]. As compared to high antenatal care use, only leagues [25] found that community norms about facility-
57% of women had an institutional delivery and only based delivery and women’s decision-making autonomy
59% delivered with a skilled attendant present [6]; simi- were potential pathways that influenced the decision to
lar distinctions are found in the 2011 Multiple Indicator deliver a child in a health facility. In rural Tanzania, com-
Cluster Survey [2]. Wide variability in institutional deliv- munity beliefs that facility delivery is important for the
ery and skilled attendance at delivery was observed by health of the mother and baby were associated with use of
region of residence with the Northern region having the facility-based delivery [26]. In a separate study among the
lowest percentage of women delivering in a facility and same population in Tanzania, male partners’ opinions
the Central region falling in the middle on percentage of about institutional delivery were associated with actual in-
births in a facility [6]; these are the two regions covered stitutional delivery, such that spouses who disagreed about
in this study. the importance of institutional delivery were less likely to
A number of studies using the GDHS demonstrated have one compared to spouses who agreed that delivering
important demographic- and policy-level factors associ- in a health facility was important [27]. In rural Mali,
ated with institutional delivery [7-9]. Recent global stud- mothers-in-law’s beliefs and attitudes were demonstrated
ies have examined common barriers to antenatal care to have an influence on their daughters-in-law’s maternal
and institutional delivery [3,10-17]. Much of this re- health care-seeking behaviors [28].
search has focused on transportation, distance, and cost Previous studies have also shown that women’s
[13,15,16]. A recent qualitative study by Matsuoka and decision-making autonomy is associated with the use
colleagues in Cambodia (2010) demonstrated five types of health facilities for delivery. In Nigeria, women with
of barriers to utilization of government maternal health greater decision-making autonomy were more likely to de-
services [17]. These barriers were: financial; physical; liver in a health facility, which may indicate that these
cognitive; organizational; and psychological and social. women were better able to advocate for and access a
Speizer et al. BMC Pregnancy and Childbirth 2014, 14:398 Page 3 of 13
http://www.biomedcentral.com/1471-2393/14/398

health facility for childbirth [29]. In Bangladesh, house- Multiple survey instruments were used at baseline and
holds in which husbands made decisions alone were asso- midline, including a household survey, a community leader
ciated with less use of antenatal care and skilled delivery survey, and facility surveys (with clients, providers, trad-
care compared to households that practiced joint decision- itional birth attendants, and chemical sellers). This analysis
making [30]. The relationship between women’s decision- focuses on the household survey data. The purpose of the
making autonomy and use of maternal health services may household survey was to obtain information on knowledge,
be due to women’s power to realize their preferences, attitudes and practices regarding maternal and child health
which includes a stronger preference for ensuring their services.
own health [31]. Although some studies have demon- The household survey used the 30-by-N cluster sam-
strated the importance of community norms and women’s ple design; this method is commonly used in child sur-
decision-making autonomy on the decision to deliver in a vival programs [25,32]. Cluster sampling is an efficient
health facility, there have been few studies that have looked sampling method because it provides a means to obtain
at the two pathways together and examined the ways in a representative sample from the region without under-
which community norms and household decision-making taking a census of households in the community. How-
autonomy interact with one another. ever, cluster sampling leads to biased standard errors
This study contributes to our understanding of auton- due to the correlation between observations from the
omy and social barriers to institutional delivery in Ghana same cluster. We explain our approach for accounting
using recently collected quantitative data from two re- for the biased standard errors in the analysis section.
gions of Ghana. Because the focus of the study was to The overall sampling strategy was designed to meet the
obtain information on barriers to institutional delivery evaluation objectives for the Maternal and Newborn Re-
and women’s referral experiences, this study includes a ferrals Project [18]. At baseline, the goal was to include
large sample of women who recently delivered a child, a large sample of recently pregnant women (pregnant in
providing rich information on barriers to institutional the last 12 months) to identify their experiences with
delivery in these regions. The objectives of this study are pregnancy, childbirth, and newborn health. Thus, we
to examine whether women’s decision-making roles and used a 30-by-7 sampling approach to identify thirty clus-
their perceptions about social support for facility deliv- ters per region (thirty from the three districts in the
ery, measured at the individual and community levels, Northern region and thirty from the three districts in
are associated with women’s actual place of delivery in the Central region), and seven recently pregnant women
Ghana. Furthermore, we examine how the relationship in each cluster were randomly selected for interview.
between community-level attitudes and institutional de- Random selection of clusters was undertaken from an
livery differs for households in which women have a say exhaustive list of communities in the six study districts.
in their own health care and those that do not. The recently pregnant women were randomly sampled
from a list of all recently pregnant women in the com-
Methods munity (determined through interviews with community
Data collection leaders and health workers in the community). To sup-
The cross-sectional data for this study come from base- plement the sample of 210 recently pregnant women per
line and midline surveys that were used during the region, we also included 14 nearby neighbor women
evaluation of the Maternal and Newborn Referrals Pro- (ages 15–49) who may or may not have been recently
ject of Project Fives Alive in the Northern and Central pregnant to permit an examination of maternal and
Regions of Ghana. The Maternal and Newborn Referrals newborn health knowledge, attitudes, and behaviors of
project is being implemented by the Institute for Healthcare women in the community. At midline, the same 30-by-
Improvement (IHI), the National Catholic Health Service N cluster design was employed, however, a new sample
(NCHS) and the Ghana Health Service (GHS). The evalu- of communities was drawn from the same districts. As
ation is being led by the University of North Carolina at with the baseline survey, in all selected clusters, seven
Chapel Hill and the University of Ghana. Baseline data were recently pregnant women were surveyed as well as 14
collected between May and June 2012 to help design the nearby neighbors.
project and midline data were collected between October For the purpose of this study, which uses baseline and
and November 2013 to help strengthen project implemen- midline data, and accounting for plausible design effect,
tation. Since the focus of this study is not significantly our sample size is adequate to obtain precise estimates
affected by the interventions implemented as part of the of our key outcome (institutional delivery). A total of
Maternal and Newborn Referrals Project and there 2,527 women were interviewed in the two rounds of data
were delays in project initiation until July, 2013, the collection (1,267 women were interviewed at baseline
baseline and midline data were merged to provide a lar- and a new sample of 1,260 women were interviewed at
ger cross-sectional sample. midline). This analysis of institutional delivery focuses
Speizer et al. BMC Pregnancy and Childbirth 2014, 14:398 Page 4 of 13
http://www.biomedcentral.com/1471-2393/14/398

on a sub-sample of the 2,527 women interviewed, which These community-level responses were calculated by cre-
excludes women who did not have a birth in the last ating an average value of all women living in the cluster,
three years, were not in union, or had missing informa- removing each individual woman from the calculation. A
tion on the key variables of interest. Thus the final ana- similar approach was undertaken for the community-level
lysis sample is 1,606 women. men’s attitude and the community-level mother-in-law’s
Ethics review approval for the study was obtained by attitude.
the University of North Carolina at Chapel Hill and the All models control for demographic factors previously
Ghana Health Service. Informed consent was obtained found to be associated with facility delivery including
from all study participants. age, education, ethnicity, employment status, religion,
parity, wealth, region, and time period (baseline or mid-
Variables line) [24]. See Table 1 for a description of these variables.
The key dependent variable for this analysis is the place of The wealth variable was created based on three house-
delivery of the last birth in the last three years. Women hold characteristics: type of toilet, type of fuel used in
who delivered in a health facility are coded one, whereas the household, and location of the kitchen. Households
all women who delivered at home or in the home of some- with a non-improved toilet facility (as defined in the
one else (e.g., a relative or a health worker) are coded zero Ghana DHS), that use wood for their source of fuel, and
(i.e., non-institutional delivery). that have a kitchen outside their household were coded as
The main independent variables for this analysis focus being the poorest households. Households with two out of
on decision-making autonomy and attitudes toward in- three of these lower quality scenarios were considered to
stitutional delivery. First, all women were asked: “Who be medium, and households with none or just one of these
usually makes decisions about health care for you?” lower quality scenarios were considered to be the richest.
Women who reported that they make the decisions This is the same approach that was used in an earlier ana-
alone or make the decisions jointly with their partner lysis of health insurance effects on facility delivery using
were the reference group (high decision-making auton- these same data [18]. In the full sample, based on this clas-
omy), and were compared to women who reported that sification, about 40% of the women were in the poorest
their partner makes the decision alone (low decision- category, 40% in the medium category, and only 19% were
making autonomy). A third category was also created for in the richest category (see Table 1). It is worth noting that
the small number of women who reported that someone use of antenatal care (ANC) during the pregnancy was not
else makes the decision. The other independent variables included as an independent variable in the reduced form
of interest are attitudes toward institutional delivery, models presented. Previous research has demonstrated
represented by three separate questions. First, all women that use of ANC is endogenous and would introduce bias
were asked: “How many women do you think in your into the models presented [33].
community deliver their baby in a health facility?” Re-
sponse options were: none, some, most, and all (coded Analysis
1–4); the small number of women who reported “don’t We use bivariate analyses to examine the association be-
know” (n = 114) were dropped from the analysis. Second, tween the key independent variables and institutional de-
women were asked: “In your opinion, what percentage livery, controlling for key demographic variables described
of men in your community is supportive of facility deliv- earlier and adjusting for the clustered survey design. Be-
ery?” Response options were: no men, few men, some cause the dependent variable of interest (institutional deliv-
men, most men, and all men (coded 1–5); the 149 ery) is binary, and we are interested in both individual and
women who reported “don’t know” were dropped from community-level attitudes, we use multi-level logistic
the analysis. Finally women were asked: “In your opin- regression models. To examine if the relationship between
ion, what percentage of mothers-in-law in your commu- community-level attitudes and institutional delivery differs
nity is supportive of facility delivery?” Response options by women’s decision-making autonomy, we use models
were: no mothers-in-law, some mothers-in-law, most with cross-level interactions between community-level atti-
mothers-in-law, and all mothers-in-law (coded 1–4); the tudes and individual-level autonomy. Since the interactions
172 women who reported “don’t know” were dropped cannot be evaluated by looking at the sign, magnitude, or
from the analysis. statistical significance of the odds ratio for nonlinear
To examine community-level attitudes toward facility models [34], we plot the interaction using the margins
delivery, we also created comparable variables at the command in Stata 13. All regression analyses adjust for the
community-level for each of the three attitude questions. clustered survey design by including robust standard errors
In particular, for each woman, we calculated the average in Stata 13 statistical software. Regression results are pre-
response on how many women in the community she sented by showing the odds ratios and the 95% confidence
perceived had delivered their baby in a health facility. intervals.
Speizer et al. BMC Pregnancy and Childbirth 2014, 14:398 Page 5 of 13
http://www.biomedcentral.com/1471-2393/14/398

Table 1 Characteristics of total sample (baseline and midline), recent birth sample, and analysis sample from Ghana
evaluation of Maternal and Newborn Referrals Project, 2012, 2013
Characteristics Full sample Recent birth sample (birth in the Analysis sample (in union, birth in the last
last 3 years) 3 years, no missing information)
Percent Number (n = 2527*) Percent Number (n = 1840*) Percent Number (n = 1606)
Age:
<19 7.8 196 7.5 138 4.7 76
20-24 23.7 597 25.4 468 23.4 375
25-34 44.7 1,127 48.7 896 51.9 833
35-49 23.9 604 18.4 338 20.1 322
Education:
None 47.4 1,199 45.3 835 50.0 803
Primary 20.1 508 20.5 378 18.4 295
Secondary or higher 32.4 819 34.2 629 31.6 508
Ethnicity:
Akan 47.9 1,210 50.3 926 45.6 732
Mole-Dagbani 32.1 811 30.8 568 34.5 554
Konkomba 9.6 242 8.9 164 9.3 150
Other 10.4 264 10.0 184 10.6 170
Work status:
Unpaid/unemployed 58.3 1,474 59.3 1,092 57.7 927
Self employed 38.1 964 37.1 684 38.7 622
Paid work 3.5 89 3.6 66 3.6 57
Religion:
Christian 56.0 1,414 58.1 1,071 54.4 874
Muslim 29.8 752 28.1 517 30.7 493
None/traditional/other 14.3 361 13.8 254 14.9 239
Marital status:
Not currently in union 15.1 381 12.7 234 0.0 0
Currently in union 84.9 2,146 87.3 1,608 100.0 1606
Parity:
0 4.9 124 0.0 0 0 0
1 20.8 526 23.2 427 18.4 296
2 15.7 397 17.9 329 17.3 277
3 14.0 352 14.9 275 16.2 260
4 12.7 322 13.1 242 14.2 228
5 9.6 242 9.9 183 10.9 175
6+ 22.3 564 21.0 386 23.0 370
Region:
Central 50.1 1,267 52.6 968 48.0 770
Northern 49.9 1,260 47.5 874 52.1 836
Wealth category
Poorest 40.4 1,022 40.5 745 41.5 667
Medium 40.3 1,019 40.3 742 39.3 631
Richest 19.2 486 19.3 355 19.2 308
Speizer et al. BMC Pregnancy and Childbirth 2014, 14:398 Page 6 of 13
http://www.biomedcentral.com/1471-2393/14/398

Table 1 Characteristics of total sample (baseline and midline), recent birth sample, and analysis sample from Ghana
evaluation of Maternal and Newborn Referrals Project, 2012, 2013 (Continued)
Time
Baseline 50.1 1,267 45.7 841 45.8 735
Midline 49.9 1,260 54.3 1011 54.2 871
*The sample size is slightly smaller for some variables that had missing data.

Results of men’s opinions of facility delivery indicate that the

Table 1 presents the descriptive characteristics of three majority of women (about 70%) believe that most men
sample populations: the full sample (n = 2,527), the sam- or all men are supportive of facility delivery. About 30%
ple of women who had a recent birth (n = 1,840), and of women report that no men, few men and some men
the final analysis sample of women in union who had a are supportive of facility delivery. Third, women’s per-
recent birth and had non-missing information on all ceptions of mothers-in-law’s opinions are similar to
study variables (n = 1,606). Women who were not in women’s perceptions of men’s opinions of facility deliv-
union were dropped from the analysis since one of the ery. Fourth, about half of the women report that they
key independent variables on autonomy was specifically alone, or with their partner, make decisions about their
about decision-making in union. The only difference ob- own health care. Forty-seven percent of women report
served between the full sample and the final analysis that their husband alone makes these health care deci-
sample is that the analysis sample is made up of a sions and only 4% report that someone else makes these
greater percentage of women from the midline sample. decisions. When comparing the attitude and autonomy
At midline, the study recruited a greater number of re- variables by whether or not the woman had a facility de-
cently pregnant women in the neighbor sample, increas- livery, a significant difference is found in the expected
ing the proportion of the sample from midline in the direction, such that those women who had a facility de-
recently pregnant sample. More than two-fifths of the livery have attitudes that are more supportive of facility
women surveyed for this study are uneducated, and delivery than those women who did not have a facility
more than two-fifths are of Akan and one-third is of delivery.
Mole-Dagbani ethnicity. More than half of the women Table 3 presents the odds ratios and 95% confidence in-
are unemployed or doing unpaid work. More than half tervals for the analysis of having a facility delivery com-
of the sample is Christian, and a little less than a third is pared to not having a facility delivery among women in
Muslim. In the analysis sample, all women have had at union who had a birth in the last three years. The key vari-
least one birth (since the outcome is place of delivery of ables of interest to this analysis are at the bottom of the
the last birth) with about equal proportions having one, table. Note that because of high correlation between the
two, or three births. Twenty-three percent of the women attitude variables (e.g., attitudes about facility delivery for
have had six or more births. Finally, the full sample is all women, attitudes about men’s support for facility deliv-
evenly divided between the Central and Northern re- ery and attitudes about mother-in-law’s support for facility
gions, whereas the analysis sample includes a greater delivery), models were run separately for each of these in-
percentage of women from the Northern region; this may dependent variables.
be suggestive of more pregnancies among women in Model 1 demonstrates that women who perceive that
union in the Northern region. To assess if differences be- a greater number of women in their community deliver
tween the baseline and midline samples are influencing in a facility are significantly more likely to have delivered
the results, we re-ran the regression analyses with only the their last birth in a facility (OR: 1.83; 95% CI: 1.53-2.19)
baseline sample; the results were similar for the key inde- compared to women who think fewer women deliver in
pendent variables of interest (results not shown). a facility. Furthermore, women who report that someone
Table 2 presents the distribution of the key independ- other than herself or her spouse makes decisions about
ent variables (perceptions about facility delivery and her own health care were more likely to deliver in a fa-
decision-making autonomy) in the analysis sample and cility (OR: 2.22; 95% CI: 1.09-4.53) compared to women
then by whether or not the woman had a facility deliv- who are involved in the decision-making process. This
ery. First, women’s attitudes toward how many women variable is not significant in the other models and may
deliver in a health facility indicate that more than half of merit consideration in future analyses. Community-level
women report that most women deliver in a health facil- attitude toward facility delivery is also significant in
ity (52%), and another 13% report that all women deliver Model 1 indicating that, controlling for the woman’s own
in a facility. About 35% of women report none or some attitude toward the number of women who deliver in a fa-
women deliver in a facility. Second, women’s perceptions cility, women who live in communities where more women
Speizer et al. BMC Pregnancy and Childbirth 2014, 14:398 Page 7 of 13
http://www.biomedcentral.com/1471-2393/14/398

Table 2 Attitudes toward facility delivery and who makes decisions in the household and distribution by whether the
woman had a facility delivery in Ghana 2012, 2013
Characteristics Analysis sample Distribution by whether
(in union with had a facility delivery for
recent birth)* last birth
Percent Number Non-facility delivery Facility delivery
(47.0%) (53.0%)
How many women do you think in your
community deliver their baby in a health facility?
None 5.7 87 10.7 1.3
Some 29.1 446 40.1 19.7
Most 52.4 803 43.4 60.1
All 12.9 197 5.8 18.9***
In your opinion, what percentage of men in your
community are supportive of facility delivery?
No men 1.9 29 4.0 0.1
Few men 11.5 174 17.7 6.2
Some men 17.0 257 20.8 13.7
Most men 52.8 797 46.9 57.9
All men 16.8 253 10.6 22.1***
In your opinion, what percentage of mother-in-laws in your community are
supportive of facility delivery?
No mothers-in-law 3.9 59 7.6 0.8
Some mothers-in-law 29.8 446 37.4 23.2
Most mothers-in-law 51.2 767 44.0 57.4
All mothers-in-law 15.1 227 11.1 18.7***
Who usually makes decisions about health care for you:
Woman alone/both partners 49.3 791 39.9 57.5
Husband only 46.6 748 57.7 36.7
Other 4.2 67 2.4 5.8***
*Note, those who reported “don’t know” to the attitude questions were dropped, this was 114 for the question on community attitudes; 149 for the question on men’s
attitudes; and 172 for the question on mother-in-law attitudes. Significance testing compares facility delivery to non-facility delivery. ***F-test p value ≤0.001.

perceive higher use of facility delivery are more likely to de- more likely to have a facility delivery than all others
liver in a facility than women who live in communities (OR: 3.70; 95% CI: 2.27-6.05). Model 4, which adds an
where women perceive fewer facility deliveries (OR: 4.13; interaction term between women’s decision-making au-
95% CI: 2.49-6.85). The patterns for the key variables of tonomy and men’s community-level attitudes, shows that
interest remained the same in Model 2 and there was no the effect of the husband making the decision alone be-
significant interaction between decision-making autonomy comes more important (and remains negative), particu-
and community-level attitudes. larly for women with less male support for facility
Models 3 and 4 tell a slightly different story. In Model delivery at the community-level (OR: 0.04; 95% CI: 0.01-
3, women who perceive that more men are supportive of 0.32). The interaction, depicted in Figure 1, shows that
facility delivery are significantly more likely to deliver in when husbands make decisions alone and community
a facility (OR: 1.36; 95% CI: 1.18-1.55). In addition, attitudes toward men’s support are low, the predicated
women who report that the husband alone makes deci- probability of facility delivery is low. Conversely, when
sions about her health care are significantly less likely to community attitudes toward men’s support are high,
have a facility delivery than women who report that she whether the husbands make decisions alone or not, the
is involved in decision-making (OR: 0.74; 95% CI: 0.58- probability of facility delivery is higher and more similar
0.96). Model 3 also shows that controlling for women’s to when the wife is involved in health care decisions.
own attitudes toward men’s support for facility delivery, Finally, Models 5 and 6 provide similar findings. Spe-
women who live in communities where men are per- cifically, Model 5 shows that women who perceive that
ceived to have more positive attitudes are significantly more mothers-in-law are supportive of facility delivery
Speizer et al. BMC Pregnancy and Childbirth 2014, 14:398 Page 8 of 13
http://www.biomedcentral.com/1471-2393/14/398

Table 3 Logistic regression odds ratios and 95% confidence intervals of association between attitudes and
decision-making autonomy and whether a woman delivered a recent birth in a facility, Ghana, 2012, 2013
Characteristics Facility delivery Facility delivery Facility delivery Facility delivery Facility delivery Facility delivery
vs. non-facility vs. non-facility vs. non-facility vs. non-facility vs. non-facility vs. non-facility
delivery delivery delivery delivery delivery delivery
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
Time
Baseline 1.0 1.0 1.0 1.0 1.0 1.0
Midline 1.20 (0.80-1.81) 1.21 (0.81-1.81) 0.83 (0.56-1.25) 0.85 (0.57-1.27) 0.84 (0.55-1.28) 0.86 (0.56-1.32)
Age:
<25 1.0 1.0 1.0 1.0 1.0 1.0
25-34 1.00 (0.71-1.39) 1.00 (0.71-1.40) 1.06 (0.77-1.46) 1.07 (0.77-1.47) 1.00 (0.73-1.37) 1.00 (0.73-1.38)
35-49 1.17 (0.74-1.85) 1.18 (0.74-1.86) 1.24 (0.79-1.93) 1.26 (0.81-1.98) 1.13 (0.75-1.71) 1.15 (0.76-1.74)
Education:
None 1.0 1.0 1.0 1.0 1.0 1.0
Primary 1.40 (0.94-2.09)+ 1.41 (0.94-2.10)+ 1.31 (0.92-1.88) 1.32 (0.92-1.90) 1.43 (1.01-2.03)* 1.45 (1.02-2.07)*
Secondary or higher 1.92 (1.27-2.92)** 1.93 (1.27-2.92)** 2.04 (1.40-2.96)*** 2.06 (1.42-2.99)*** 2.21 (1.52-3.23)*** 2.26 (1.55-3.30)***
Ethnicity:
Akan 1.0 1.0 1.0 1.0 1.0 1.0
Mole-Dagbani 1.43 (0.43-4.84) 1.40 (0.41-4.72) 1.25 (0.34-4.61) 1.15 (0.32-4.20) 1.51 (0.45-5.09) 1.47 (0.44-4.93)
Konkomba 5.31 (1.69-16.69)** 5.43 (1.74-16.96)** 3.25 (0.97-10.88)+ 3.33 (1.01-10.95)* 2.34 (0.74-7.36) 2.40 (0.77-7.46)
Other 1.45 (0.60-3.51) 1.45 (0.60-3.48) 1.35 (0.56-3.27) 1.30 (0.55-3.12) 1.49 (0.67-3.31) 1.46 (0.66-3.24)
Work status:
Unpaid/unemployed 0.75 (0.57-1.00)+ 0.76 (0.57-1.01)+ 0.80 (0.61-1.06) 0.81 (0.62-1.07) 0.72 (0.55-0.95)* 0.73 (0.56-0.96)*
Self-employed/paid work 1.0 1.0 1.0 1.0 1.0 1.0
Religion:
Christian 1.0 1.0 1.0 1.0 1.0 1.0
Muslim 1.34 (0.67-2.66) 1.33 (0.67-2.63) 1.25 (0.57-2.77) 1.23 (0.57-2.70) 1.25 (0.58-2.72) 1.23 (0.57-2.64)
None/traditional/other 0.53 (0.33-0.86)** 0.53 (0.33-0.86)** 0.50 (0.30-0.83)** 0.51 (0.31-0.84)** 0.47 (0.30-0.75)** 0.47 (0.29-0.75)***
Parity (continuous): 0.95 (0.92-0.98)** 0.95 (0.92-0.98)** 0.95 (0.92-0.98)** 0.95 (0.92-0.98)** 0.96 (0.93-0.99)** 0.96 (0.93-0.99)**
Region:
Central 2.83 (0.96-8.33)+ 2.73 (0.92-8.09)+ 1.98 (0.62-6.39) 1.84 (0.57-5.93) 2.15 (0.72-6.44) 2.06 (0.68-6.23)
Northern 1.0 1.0 1.0 1.0 1.0 1.0
Wealth category
Poorest 1.0 1.0 1.0 1.0 1.0 1.0
Medium 1.30 (0.99-1.72)+ 1.30 (0.98-1.72)+ 1.24 (0.93-1.63) 1.21 (0.91-1.61) 1.28 (0.97-1.68)+ 1.26 (0.96-1.67)
Richest 1.93 (1.32-2.82)*** 1.93 (1.32-2.83)*** 1.87 (1.25-2.81)** 1.87 (1.24-2.81)** 1.94 (1.29-2.90)*** 1.95 (1.30-2.92)***
Decision-making
about woman’s healthcare
Woman/joint 1.0 1.0 1.0 1.0 1.0 1.0
decision-making
Husband alone 0.81 (0.63-1.05) 0.37 (0.07-1.99) 0.74 (0.58-0.96)* 0.04 (0.01-0.32)** 0.73 (0.57-0.94)* 0.08 (0.01-0.50)**
Other 2.22 (1.09-4.53)* 5.73 (0.14-227.93) 1.65 (0.82-3.31) 0.41 (0.01-23.09) 1.77 (0.83-3.76) 1.11 (0.03-42.74)
Perceived number of 1.83 (1.53-2.19)*** 1.84 (1.53-2.20)*** Na Na Na Na
women that delivery
in facility
Speizer et al. BMC Pregnancy and Childbirth 2014, 14:398 Page 9 of 13
http://www.biomedcentral.com/1471-2393/14/398

Table 3 Logistic regression odds ratios and 95% confidence intervals of association between attitudes and
decision-making autonomy and whether a woman delivered a recent birth in a facility, Ghana, 2012, 2013 (Continued)
Perception of men’s Na Na 1.36 (1.18-1.55)*** 1.36 (1.19-1.56)*** Na Na
attitude toward
facility delivery
Perception of MIL Na Na Na Na 1.40 (1.18-1.65)*** 1.41 (1.19-1.66)***
attitude toward
facility delivery
Community attitudes
Attitudes toward 4.13 (2.49-6.85)*** 3.62 (2.02-6.51)*** Na Na Na Na
number that delivery
in facility
Men’s attitudes toward Na Na 3.70 (2.27-6.05)*** 2.48 (1.47-4.19)*** Na Na
facility delivery
MIL attitudes toward Na Na Na Na 3.46 (1.85-6.45)*** 2.20 (1.11-4.34)*
facility delivery
Interactions
Community attitude Na 1.33 (0.72-2.45) Na 2.14 (1.26-3.62)** Na 2.25 (1.15-4.42)*
*Husband alone decides
Community attitude Na 0.70 (0.19-2.59) Na 1.45 (0.50-4.22) Na 1.18 (0.30-4.62)
*Other decide
Na – not applicable for that model. Sample size slightly smaller than full sample since small number of missing observations were dropped for perception
questions by model.
+p ≤ 0.10; *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001.

are more likely to deliver in a facility and women with low law’s community-level attitudes. The interaction, depic-ted
decision-making autonomy are less likely to deliver in a fa- in Figure 2, demonstrates that women who have low
cility. Furthermore, controlling for women’s own attitudes decision-making autonomy (husbands make decisions
toward mothers-in-law’s support for facility delivery, alone) have a lower predicted probability of a facility deliv-
women living in communities where there is greater per- ery when community attitudes toward mothers-in-law’s
ceived support for facility delivery by mothers-in-law are support for facility delivery are low. When community
more likely to deliver in a facility than all others (OR: 3.46; attitudes toward mothers-in-law’s support are higher, the
95% CI: 1.85-6.45). Model 6 includes the interaction be- effect of lower decision-making autonomy is more similar
tween women’s decision-making autonomy and mother-in- to when women have higher decision-making autonomy.
1 .8
Probability of facility delivery
.2 .4 0 .6

1 1.5 2 2.5 3 3.5 4 4.5 5

Community perceptions about men's attitudes on women delivering in facility

Woman/joint Husband only

Figure 1 Adjusted predictions at representative values with 95% confidence intervals for Model 4. Community perceptions of husbands’
attitudes toward facility delivery by decision-making autonomy.
Speizer et al. BMC Pregnancy and Childbirth 2014, 14:398 Page 10 of 13
http://www.biomedcentral.com/1471-2393/14/398

1 .8
Probability of facility delivery
.2 .4 0 .6

1 1.5 2 2.5 3 3.5 4

Community perceptions about MIL's attitudes on women delivering in facility

Woman/joint Husband only

Figure 2 Adjusted predictions at representative values with 95% confidence intervals for Model 6. Community perceptions of mother-in-law’s
attitudes toward facility delivery by decision-making autonomy.

Discussion levels (including perceived attitudes of husband’s and

Using these unique data, we find that supportive mothers-in-law).
community-level attitudes are associated with greater odds Our findings corroborated findings from two previous
of facility delivery, even after controlling for a woman’s studies that have shown that support for facility deliver-
own perceptions about community attitudes. In addition, ies by mothers-in-law and husbands is associated with
women with lower decision-making autonomy regarding institutional delivery [27,28]. White and colleagues [28]
their own health care have lower odds of facility delivery found similar results in that women in rural Mali whose
compared to women who are involved in health care deci- mothers-in-law agreed with traditional and cultural
sions. In models that include community-level attitudes practices were less likely to deliver in a health facility.
about men’s and mothers-in-law’s support for facility de- The traditional views of mothers-in-law may contradict
livery, the interaction between autonomy and community more contemporary views of pregnancy and childbirth.
attitudes is significant indicating that women with low Contrary to our results, rural Malian women’s own per-
decision-making autonomy who also live in communities ceptions of delivery practices of women in the commu-
that are less supportive of facility delivery are the least nity, and the perceptions of their husbands, were not
likely to have a facility delivery. Conversely, women who associated with place of delivery. Furthermore, Danforth
have low decision-making autonomy who live in commu- and colleagues [27] found that spousal agreement about
nities that are more supportive of facility delivery are more the place of delivery was associated with facility delivery
similar to women with high decision-making autonomy. in rural Tanzania. This evidence supports our notion
Therefore, perceived men’s and mothers-in-law’s commu- that the opinions of both partners are important when
nity attitudes are influential, even for the least empowered deciding where to deliver their child.
women. Previous studies that have explored community-level
Previous studies have also demonstrated that there are attitudes about facility delivery did not have the same
psychological and social barriers to maternal and child amount of detail that our study provides. Kruk and col-
health services [3,10,17]. These studies have generally fo- leagues [26] showed that women’s perceptions of the im-
cused on qualitative data whereas our study uses quantita- portance of facility delivery at the village level were
tive data to demonstrate similar influences. Our findings associated with facility delivery, controlling for women’s
are consistent with other studies from Ghana that used individual perceptions. However, they did not provide in-
Demographic and Health Survey data in terms of demo- formation on the perceptions of other influential people in
graphic characteristics associated with facility delivery the woman’s social network, including her husband and
[7,8]. However, our study is unique because of its focus on her mother-in-law. Stephenson and colleagues [25] used
maternal and child health care utilization with specific community-level variables to predict facility delivery in six
questions on attitudes toward facility delivery at multiple African countries, but the variables they used were proxies
Speizer et al. BMC Pregnancy and Childbirth 2014, 14:398 Page 11 of 13
http://www.biomedcentral.com/1471-2393/14/398

for community norms and gender dynamics. For example, This study is not without limitations. First, this is a
they used female educational attainment and level of facil- cross-sectional study and therefore it is not possible to
ity delivery to approximate social norms related to the determine the direction of causality. Possibly, women
power of women and facility delivery, respectively. Our who deliver in a health facility become more favorable
study builds upon these results by using indicators that toward facility delivery after the fact rather than their at-
provide a more direct assessment of perceived community titudes toward facility delivery influencing their delivery
norms and women’s decision-making autonomy. behaviors. Therefore, we are only able to show associa-
There have been mixed results related to the association tions with the available data. To better understand if at-
between women’s decision-making autonomy and the use titudes influence behaviors, it would be necessary to
of maternal and child health services. Studies on women’s have longitudinal data on women’s attitudes toward fa-
decision-making power have shown associations between cility delivery prior to any pregnancy experience. The
higher decision-making autonomy and increased antenatal second limitation with these data is that we are using
care use [23,30], having a normal body mass index [35], perceived attitudes of men and mothers-in-law. Unfortu-
having a birth preparedness plan [36], childhood immu- nately, data from these influential individuals were not
nization [23], and sick child care [35]. However, the link- collected in this study. That said, there is prior research
age between women’s decision-making autonomy and that shows that perceptions of attitudes and norms are
facility delivery in sub-Saharan Africa is less commonly important to understanding health behaviors [37] and
studied and has shown mixed results [29,30]. In a study might matter more than actual attitudes in influencing
using the 2008 Ghana Demographic and Health Survey, health behaviors [38]. Third, responses of women’s atti-
Moyer and colleagues [3] found that women who did not tudes and their perceptions of their husbands and
participate in decision-making regarding their own health mothers-in-law were correlated which meant that we
care were less likely to deliver in a health facility. However, were not able to determine if one is more influential
when other factors were considered, such as maternal lit- than the other. Finally, given our use of quantitative
eracy, health insurance coverage, and wealth, the associ- data, it is not possible to answer the questions of “why”
ation between women’s decision-making autonomy and and “how” autonomy and community-level attitudes in-
facility delivery was no longer significant. When control- fluence behaviors. This can be explored with qualitative
ling for perceptions of mothers-in-law and husbands data that goes into depth on decision-making autonomy
(Table 3, Models 3–6), we find that women who have and community attitudes simultaneously.
lower decision-making autonomy about their own health Future studies of barriers to women’s use of facility de-
care have lower odds of facility delivery compared to livery in Ghana and elsewhere in sub-Saharan Africa
women who are involved in health care decisions. This should consider the role of decision-making autonomy
suggests that there is something unique about household and community norms. To do this correctly will require
dynamics and spousal communication, beyond maternal collecting data from multiple players (e.g., women, part-
education and household wealth, in these two districts in ners, mothers-in-law, and health care providers) as well
Ghana. However, when controlling for perceptions of as collecting data longitudinally to better understand the
other women, decision-making by someone other than the multiple influences. In particular, over time, women’s at-
woman or her spouse was significantly related to facility titudes may change as well as community attitudes. Fol-
delivery (Table 3, Model 1). Women in this decision- lowing communities and women longitudinally will
making scenario need to be examined more closely as this permit a determination of which of these are the most
was a rare response with a large confidence interval. important for future program development.
The relationship between women’s decision-making au-
tonomy and facility delivery is even more important when Conclusions
the attitudes and beliefs in the community in which The findings from this study are useful for the Maternal
women live are considered (Figures 1 and 2). The associ- and Newborn Referrals Project of Project Fives Alive, and
ation between community norms and facility delivery is other similar maternal and child health programs, to in-
greater (i.e., the slope is steeper) among women whose form future strategies to increase the use of institutional
husbands make their health care decisions compared to (or skilled) delivery. First, programs should continue to in-
women who are involved in health care decisions about volve influential players, including husbands and mothers-
their own health. This has important implications for the in-law, and not just pregnant women. Notably, since the
importance of community norms and how they interact time these data were collected, the Maternal and Newborn
with household gender dynamics. Communities that sup- Referrals Project of Project Fives Alive has tested strategies
port contemporary delivery practices may also be more to engage husbands and mothers-in-law. This has been
supportive of gender equity with regards to household done through community engagement and mobilization
decision-making. led by traditional birth attendants, health staff, and quality
Speizer et al. BMC Pregnancy and Childbirth 2014, 14:398 Page 12 of 13
http://www.biomedcentral.com/1471-2393/14/398

improvement project staff. Community conversations in- 4. WHO: Making Pregnancy Safer: The Critical Role of the Skilled Attendant.
volve influential individuals and encourage discussion on Geneva: WHO, ICM, FIGO; 2004. http://whqlibdoc.who.int/publications/2004/
9241591692.pdf. Accessed July 23, 2014.
the importance of men attending antenatal and delivery 5. WHO: Health services coverage statistics. http://www.who.int/whosis/
care and the importance of delivering in a health facility. whostat/EN_WHS09_Table4.pdf. Accessed July 23, 2014
The final evaluation will permit an assessment of whether 6. Ghana Statistical Service (GSS): Ghana Health Service (GHS), and ICF Macro.
2009. Ghana Demographic and Health Survey 2008. Accra, Ghana: GSS, GHS,
women (and their network members) who were exposed and ICF Macro; 2009. http://www.dhsprogram.com/pubs/pdf/SR172/SR172.
to these types of interventions were more likely to deliver pdf. Accessed July 23, 2014.
in a health facility compared to women who were not 7. Doku D, Neupane S, Doku PN: Factors associated with reproductive health
care utilization among Ghanaian women. BMC Int Health Hum Rights 2012,
exposed to these interventions. Second, programs that 12:29. http://www.biomedcentral.com/content/pdf/1472-698X-12-29.pdf.
undertake community outreach activities to change com- 8. Johnson FA, Padmadas SS, Brown JJ: On the spatial inequalities of
munity norms are likely to be more effective than pro- institutional versus home births in Ghana: a multilevel analysis.
J Community Health 2009, 34:64–72.
grams that simply target individuals. By addressing 9. Penfold S, Harrison E, Bell J, Fitzmaurice A: Evaluation of the delivery fee
community norms related to institutional delivery, espe- exemption policy in Ghana: population estimates of changes in delivery
cially birth location preferences among women and men, service utilization in two regions. Ghana Med J 2007, 41:100–109.
10. Finlayson K, Downe S: Why do women not use antenatal services in
programs have the potential to help families become better low- and middle-income countries? A meta-synthesis of qualitative
prepared for obstetric emergencies should complications studies. PLoS Med 2013, 10(1):e1001373.
arise [39]. With these types of interventions, community 11. Pell C, Meñaca A, Were F, Afrah N, Chatio S, Manda-Taylor L, Hamel MJ,
Hodgson A, Tagbor H, Kalilani L, Ouma P, Pool R: Factors affecting
norms should become more favorable toward facility deliv- antenatal care attendance results from qualitative studies in Ghana,
ery and women will be more likely to deliver in safer Kenya and Malawi. PLoS One 2013, 8(1):e53747.
settings, even if they do not have the decision-making 12. Arthur E: Wealth and antenatal care use: Implications for maternal health
care utilization in Ghana. Heal Econ Rev 2012, 2(14):14. doi:10.1186/2191-
authority. Overall, maternal and child health programs that
1991-2-14.
involve influential individuals and communities are likely 13. Mwangome FK, Holding PA, Songola KM, Bomu GK: Barriers to hospital
to be the most successful at helping Ghana and other sub- delivery in a rural setting in coast province. Kenya: community attitude
Saharan African countries to attain their Millennium and behaviours. Rural Remote Health 2012, 12:1852.
14. Ykong VN, Rush KL, Bassett-Smith J, Bottorff JL, Robinson C: Women’s
Development Goals for women’s and children’s health. experiences of seeking reproductive health care in rural Ghana:
challenges for maternal health service utilization. J Adv Nurs 2010,
Competing interests 66(11):2431–2441.
The authors declare that they have no competing interests. 15. Gething PW, Johnson FA, Frempong-Ainguah F, Nyarko P, Baschieri A,
Aboagye P, Falkingham J, Matthews Z, Atkinson PM: Geographical access
to care at birth in Ghana: a barrier to safe motherhood. BMC Public Health
Authors’ contributions
2012, 12:991.
ISS, WTS, and KS conceived of the topic, ISS performed the majority of the
16. Kitui J, Lewis S, Davey G: Factors influencing place of delivery for women
analyses, ISS and WTS wrote the first draft of the paper. All authors
in Kenya: an analysis of the Kenya demographic and health survey,
contributed to the intellectual content and writing process of the paper. All
2008/2009. BMC Pregnancy & Childbirth 2013, 13:40.
authors read and approved the final manuscript.
17. Matsuoka S, Aiga H, Rasmey LC, Rathavy T, Okitsu A: Perceived barriers to
utilization of maternal health services in rural Cambodia. Health Policy
Acknowledgments 2010, 95:255–263.
We thank Sodzi Sodzi-Tettey and Pierre Barker for their review and feedback 18. Singh K, Osei-Akoto I, Otchere F, Sodzi-Tettey S, Barrington C, Huang C,
into an earlier draft of this paper. Funding for the data collection and analysis Fordham C, Speizer I: Ghana’s national health insurance scheme and
of this paper came from a subcontract to the University of North Carolina maternal and child health: a mixed methods study. 2014, Under review.
from the Institute for Healthcare Improvement (IHI). The data reported in this 19. Dzakpasu S, Soremekun S, Manu A, ten Asbroek G, Tawiah C, Hurt L, Fenty J,
paper are part of baseline and midline data collected as part of the Owusu-Agyei S, Hill Z, Campbell OMR, Kirkwood BR: Impact of free delivery
evaluation of the IHI Maternal and Newborn Referrals Project of Project Fives care on health facility delivery and insurance coverage in Ghana’s Brong
Alive! funded by the Bill & Melinda Gates Foundation. We are also grateful to Ahafo Region. PLoS One 2012, 7(11):e49430.
the Carolina Population Center for training support (T32 HD007168) and for 20. Mensah J, Oppong J, Schmidt C: Ghana’s national health insurance
general support (R24 HD050924) during the development of this manuscript. scheme in the context of the health MDGs: an empirical evaluation
using propensity score matching. Health Econ 2010, 19:95–106.
Received: 30 July 2014 Accepted: 19 November 2014 21. D’Ambruoso L, Abbey M, Hussein J: Please understand when I cry out in
pain: Women’s accounts of maternity services during labour and delivery
in Ghana. BMC Public Health 2005, 5:1–40.
References 22. Choudhury N, Ahmed SM: Maternal care practices among the ultra poor
1. WHO and UNICEF: Accountability for Maternal, Newborn and Child Survival: households in rural Bangladesh: a qualitative exploratory study.
The 2013 Update. Geneva: WHO; 2013. http://www.who.int/woman_child BMC Pregnancy & Childbirth 2011, 11:15.
_accountability/ierg/reports/Countdown_Accountability_2013Report.pdf. 23. Allendorf K: Couples’ reports of women’s autonomy and health-care use
Accessed July 23, 2014. in Nepal. Stud Fam Plann 2007, 38(1):35–46.
2. Ghana Statistical Service: Ghana Multiple Indicator Cluster Survey with an 24. Moyer CA, Mustafa A: Drivers and deterrents of facility delivery in
Enhanced Malaria Module and Biomarker, 2011, Final Report. Accra, Ghana: sub-Saharan Africa: a systematic review. Reprod Health 2013, 10:40.
2011. http://www.unicef.org/ghana/Ghana_MICS_Final.pdf. Accessed July 23, 25. Stephenson R, Baschieri A, Clements S, Hennink M, Madise N: Contextual
2014. influences on the use of health facilities for childbirth in Africa. Am J
3. Moyer CA, Adongo PB, Aborigo RA, Hodgson A, Engmann CM, DeVries R: Public Health 2006, 96(1):84–93.
“It’s up to the woman’s people”: How social factors influence facility- 26. Kruk ME, Rockers PC, Mbaruku G, Paczkowski MM, Galea S: Community and
based delivery in rural northern Ghana. Matern Child Health J 2013. health system factors associated with facility delivery in rural Tanzania:
doi:10.1007/s10995-013-1240-y a multilevel analysis. Health Policy 2010, 97(2–3):209–216.
Speizer et al. BMC Pregnancy and Childbirth 2014, 14:398 Page 13 of 13
http://www.biomedcentral.com/1471-2393/14/398

27. Danforth EJ, Kruk ME, Rockers PC, Mbaruku G, Galea S: Household decision-
making about delivery in health facilities: evidence from Tanzania.
J Health Popul Nutr 2009, 27(5):696–703.
28. White D, Dynes M, Rubardt M, Sissoko K, Stephenson R: The influence of
intra familial power on maternal health care in Mali: perspectives of
women, men and mothers-in-law. Int Perspect Sex Reprod Health 2013,
39(2):58–68.
29. Singh K, Bloom S, Haney E, Olorunsaiye C, Brodish P: Gender equality and
childbirth in a health facility: Nigeria and MDG5. Afr J Reprod Health 2012,
16(3):123–129.
30. Story WT, Burgard SA: Couples’ reports of household decision-making
and the utilization of maternal health services in Bangladesh. Soc Sci Med
2012, 75:2403–2411.
31. Allendorf K: The quality of family relationships and use of maternal
healthcare services in India. Stud Fam Plann 2010, 41(4):263–276.
32. Henderson R, Sundaresan T: Clustering sampling to assess immunization
coverage: a review of experience with a simplified sampling method.
Bull World Health Organ 1982, 60(2):253–260.
33. Adjiwanou V, LeGrand T: Does antenatal care matter in the use of skilled
birth attendance in rural Africa: a multi-country analysis. Soc Sci Med
2013, 86:26–34.
34. Ai C, Norton EC: Interaction terms in logit and probit models. Econ Lett
2003, 80(1):123–129.
35. Singh K, Bloom S, Brodish P: Gender equality as a means to improve
maternal and child health in Africa. Health Care for Women International.
2013. doi:10.1080/07399332.2013.824971.
36. Becker S, Fonseca-Becker F, Schenck-Yglesias C: Husbands’ and wives’
reports of women’s decision-making power in Western Guatemala and
their effects on preventive health behaviors. Soc Sci Med 2006,
62(9):2313–2326.
37. Fishbein M: A reasoned action approach to health promotion. Med Decis
Making 2008, 28(6):834–844.
38. Speizer IS: Are husbands a barrier to women’s family planning use? The
case of Morocco. Soc Biol 1999, 46(1–2):1–16.
39. Cofie LE, Barrington C, Singh K, Sodzi-Tetty S, Barker P, Akaligaung AJ: Birth
location preferences of mothers and fathers in rural Ghana. Working paper.
2014.

doi:10.1186/s12884-014-0398-7
Cite this article as: Speizer et al.: Factors associated with institutional
delivery in Ghana: the role of decision-making autonomy and
community norms. BMC Pregnancy and Childbirth 2014 14:398.

Submit your next manuscript to BioMed Central

and take full advantage of:

• Convenient online submission

• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution

Submit your manuscript at

www.biomedcentral.com/submit