Background

Globally, the number of Caesarean sections (CS) has been on the rise over the last decade [1, 2, 3, 4]. While CS are potentially life-saving, the adverse maternal and perinatal outcomes when a CS is not medically necessary have become a major public health concern as the associated expenses decrease resources available for other maternal and child health interventions [5, 6]. According to the World Health Organization (WHO), as a population’s CS rates approach 10%, maternal and newborn deaths decrease [7]. A medically necessary CS can prevent maternal and infant mortality; however, there is no evidence that CS benefits women who do not require the procedure [7]. The WHO has estimated—based on rates of fistula—that in 15.5% of pregnancies in Nigeria, a CS is medically necessary [1, 8, 9]. Underutilization of CS is of particular concern in most of Africa where 7.4% of all births occurred by CS in 2014 [3].

Although most African countries have regional hospitals with surgical services available to perform CS, multiple individual and health system characteristics impede access and contribute to the delay in women seeking services during pregnancy and delivery. Thaddeus and Maine (1994) developed the “three-delay” model that has been widely accepted as a framework to explain the obstacles in obtaining adequate healthcare during pregnancy and delivery [10]. This three-tiered framework includes: delay in decisions to seek care, delays in reaching a healthcare facility, and delays in receiving adequate treatment for obstetric complications.

In sub-Saharan Africa (SSA), delayed access to healthcare services during pregnancy and delivery can be influenced by multiple factors. Lack of knowledge of the importance of perinatal care and an inability to pay for healthcare services are common reasons for delaying healthcare utilization [10]. Women also delay seeking treatment during pregnancy and delivery because of poverty, gender inequalities in household decision making, cultural barriers, and geographical and transport barriers [10]. Increased age, education and wealth are all positively associated with deciding to have a doctor present at delivery in SSA [11, 12]. When life-threatening complications occur during labor, delays in seeking adequate care can increase maternal mortality even when lifesaving CS is performed.

In 2013, Nigeria had the second highest number of maternal deaths and had the 11th highest crude birth rate, making it an important country in which to study the barriers to obtaining adequate obstetric care [13, 14]. In 2014, 2% of births in Nigeria utilized CS [3]. More than 75% of all CS in Nigeria are linked to obstetric emergencies that could have been prevented by earlier medical care [15]. Even with birth plans in place, many Nigerian women opt to deliver with an unskilled birth attendant in a setting other than a hospital because of barriers in seeking treatment, including cost and geographical/transportation difficulties [10, 15, 16]. Cultural factors such as gender inequalities and the acceptance of home deliveries, compared to hospital deliveries, also influence a woman’s decision to deliver at a healthcare facility [17]. Delays in seeking treatment results in women attempting to access care at healthcare facilities only after life-threatening complications develop. Increasing access to obstetric care, such as CS, decreases maternal and infant morbidity and mortality [18, 19]. However, fears associated with having a CS may further delay a woman’s decision to seek treatment. Common fears associated with having a CS in Nigeria include: cultural beliefs that vaginal birth is a confirmation of womanhood, stigma of being mocked by other women, death, violation of religious beliefs, post-operative pain, future infertility, expense, and medical incompetence [20–22]. A woman’s socioeconomic status also influences access to CS, with the richest women, as measured by wealth index, having better access to CS compared to the poorest women [9]. Education and age are also strong predictors of a woman’s willingness to have a CS in Nigeria [17, 18]. Women who are either younger or less educated are more likely to refuse a CS even when it is medically necessary, and this usually stems from concerns about the expense [12, 23, 24].

Even women who do attempt to access healthcare facilities during delivery, often encounter poorly trained personnel and a lack of proper equipment and supplies [12]. Most studies evaluating pregnancy outcomes in Nigeria equate having a doctor present at delivery to having access to quality healthcare. However, this metric may not be an accurate predictor of the facility’s ability to perform a CS. Many healthcare facilities within Nigeria cannot offer a CS, and ambulance services are virtually non-existent [25]. In fact, one study demonstrated that only 1 in 21 health facilities in Nigeria is equipped to perform CS [26]. This complicates the ability of pregnant women to obtain adequate healthcare during pregnancy and delivery.

In order to meet the Sustainable Development Goal target of reducing the global maternal mortality ratio to less than 70 per 100,000 live births by 2030 [27], increased in access to lifesaving obstetric measures such as CS is needed [18, 19, 28]. Examining factors associated with having a CS will help provide insight into ways to increase access to healthcare during pregnancy and delivery. In addition, socioeconomic and comorbid conditions are often not examined together when exploring factors associated with CS in Nigeria. Therefore, the aims of this paper were two-fold: 1) to establish the rates of CS in Enugu, southeast Nigeria; and 2) to determine socioeconomic or medical risk factors associated with having a CS.

Methods

Survey

Data for this study were derived from the Healthy Beginning Initiative study (HBI), which has been described in detail elsewhere [30]. Briefly, the parent study was a two-arm randomized cluster trial aimed to assess rates of HIV testing. HBI used congregation-based sampling to recruit pregnant women and their partners in 40 churches from four dioceses (the Anglican Diocese of Enugu, the Catholic Diocese of Enugu, the Anglican Diocese of Oji-River, and the Catholic Diocese of Agwu). Women who were self-identified as pregnant were included in the study. Recruitment occurred at the level of the churches and participants (in that order), while randomization occurred only at the church level—that is, churches were randomly assigned to either the intervention or control groups. The intervention group participated in educational games about healthy pregnancy habits in addition to HIV acquisition modes, and effective prevention of mother-to-child transmission. They were also offered free prenatal care in the form of blood samples taken on-site to test for HIV, hemoglobin, malaria, hepatitis B, sickle cell gene, and syphilis. Women who tested positive for HIV were linked to local HIV care. Women in the control group were encouraged to attend prenatal care through nearby health facilities and were also referred to the health facility for testing. The research team maintained direct contact with health facilities to confirm HIV testing and prevention of mother-to-child transmission completion. Women in both the control and intervention groups completed a post-delivery questionnaire, which was available every 2 to 3 months at church. Maternal mortality was not ascertained in the parent study.

In Nigeria, approximately 35% of pregnant women deliver at a healthcare facility; therefore, a community-based sampling technique was employed to obtain a more representative sample of pregnant women [31]. Also, because pregnant women were recruited from churches in Enugu State, Nigeria, a more representative sample of pregnant women was expected as the population is more than 95% Christian and church attendance approaches 90% [30]. Pregnant women interested in the study were asked to read and sign a consent form in either English or the local language, Ibo. If the participant was illiterate, the consent form was read aloud to her in the local language; then the participant affixed her thumb print as an indication of her consent to participate in the study [30]. Demographic characteristics of participants were obtained at baseline via a semi-structured questionnaire [30]. Trained research staff and church-based health advisors administered this questionnaire written at a 6th grade reading level. Participants had the option of reading the survey themselves or having study personnel read to them. Because of inherent risks associated with having multiples (i.e. twins, triplets etc.), only women between the ages of 17–45 at baseline, who had singleton deliveries, were retained for this analysis. The length of pregnancy at baseline was not ascertained. Participants remained in the study until post-delivery. Post-delivery questionnaires were used to ascertain the mode of delivery, i.e., CS or vaginal birth, and singleton or multiple deliveries. Gravidity was dichotomized as primigravida and multigravida. Overall 76.6% (n = 2300) of participants who gave informed consent answered questions regarding their mode of delivery (Control n = 1042; Intervention n = 1258). Only women who answered the question regarding mode of delivery on the post-delivery questionnaire were retained for the analysis described in Table 1. However, not all women had complete data on socioeconomic and comorbid conditions; therefore, only 1,680 women were retained for the analysis described in Table 2. The study took place from January 2013 to August 2014.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Comparison of participant baseline characteristics and infant gender with mode of delivery.

https://doi.org/10.1371/journal.pone.0174369.t001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Crude and logistic regression models of the odds of C-Section vs vaginal birth.

https://doi.org/10.1371/journal.pone.0174369.t002

Results

Results from univariate analyses are shown in Table 1; 166 (7.22%) women had CS and 2134 (92.78%) had vaginal deliveries. A woman’s age was statistically associated with having a CS (p<0.01), with a greater percentage of women aged 35–45 having had a CS (11.22%) than women aged 25–34 (7.05%) or 17–24 (3.27%). Education was statistically associated with having a CS (p<0.01), with a greater percentage of women with tertiary education having had a CS (15.06%) than those with a secondary education (5.95%) or a primary education or less (5.03%). Employment status was also statistically associated with having a CS (p = 0.03); women with full-time employment had higher percentages of CS than women who worked part time or who did not indicate they were currently employed (8.72%, 5.01%, and 6.89%, respectively). Area of residence (i.e., rural vs urban), was significantly related to having a CS (p<0.01) with more women in urban settings having a CS (12.00%) compared to women in rural settings (5.55%). A mother’s baseline malaria parasitemia was significantly associated with having a CS (p = 0.03); more women with high malaria parasitemia at baseline had a CS than women who displayed low levels of malaria parasitemia (8.80% vs 6.08%, respectively). No significant relationship was observed between CS and number of people in the household, gravidity, distance to nearest healthcare facility, marital status, HIV, sickle cell disease, or anemia. When analyses were restricted to women who did not have a prior pregnancy (n = 333), 7.81% (n = 26) had CS and 92.19% (n = 307) had vaginal deliveries (Table 1). Among these primigravida women, living in an urban environment was significantly related to having a CS (p<0.01) with 14.29% of urban women having a CS compared to only 5.11% of rural women. No other participant characteristics were significant predictors of having a CS among primigravida women.

Table 2 presents the crude and adjusted odds ratios (95% CIs) for having had a CS or a vaginal birth by participant characteristics. The adjusted models showed that, compared to women aged 17–24, the odds of having a CS were higher when the mother was aged 25–34 years (adjusted OR (aOR): 2.00; 95% CI: 1.03–3.87) and when the mother was aged 35–45 years (aOR: 2.69; 95% CI: 1.24–5.82; p-trend <0.01). Compared to those with a primary school education or less, the odds of having a CS were higher if the mother had at least a tertiary education (aOR: 2.85; CI: 1.50–5.40), but not if she had a secondary education (aOR: 1.26; CI: 0.71–2.23; p-trend <0.01). The odds of having a CS were significantly lower if participants were employed part-time compared to full-time (aOR: 0.55; CI: 0.32–0.96). Compared to women who lived in an urban setting, those who lived in a rural setting had a significant reduction in the odds of having a CS (aOR: 0.58; CI: 0.38–0.89). Significantly higher odds of having a CS were seen among those with high peripheral malaria parasitemia compared to those with low parasitemia (aOR: 1.54; CI: 1.04–2.28). After adjustment for confounders, no relationship was found between CS and number of people in the household, gravidity, marital status and not being employed. Among primigravida women, adjusted logistic regression models showed a significant relationship between woman living in an urban vs rural environment and having a CS (aOR: 0.27; CI: 0.09–0.81). No other significant relationships were found among primigravida women between having a CS and a woman’s baseline characteristics.

Discussion

Women in SSA continually struggle to obtain adequate obstetric care. Increased access to emergency obstetric care, such as CS, decreases maternal and infant morbidity and mortality [18, 19]. Nigeria has one of the fastest growing populations in the world, making it a key location to study access to healthcare in pregnancy and delivery. Overall results indicate that 7.22% of women in Enugu, southeast Nigeria had a CS while 92.78% had a vaginal delivery. Percentages of CS increased as maternal age and/or education increased. Compared to women who had full-time employment, women who worked part-time had 45% lower odds of having a CS after adjusting for potential confounders. Likewise, significantly lower odds of having a CS were observed among women who live in a rural setting compared to those who reside in an urban setting in both the full sample and among primigravida women. After adjustment for confounding, this study demonstrated 54% higher odds of having a CS if participants had high peripheral malaria parasitemia compared to those with lower peripheral malaria parasitemia.

The present work demonstrated higher percentages of CS in older women and those with more education. In SSA, education has been shown to be a strong predictor of using professionally-assisted delivery services [11, 36]. Older and more educated women in SSA are thought to be more confident and influential in their household decision-making, including the use of healthcare services [11, 37]. Likewise, women with more education and/or women who are employed often have greater control over family resources and play a larger part in reproductive decision-making [17, 37, 38]. These variables may be a proxy of a woman’s ability to access healthcare, thereby increasing her chances of having had a CS.

Although the results did not demonstrate an association between having had a CS and distance to nearest health facility, there was a statistically significant relationship between having had a CS and living in rural vs urban environments. Distance has consistently been an important barrier to seeking healthcare in rural settings [10, 12, 39]. It is possible that in this self-report study, area of residence (i.e., rural vs urban) was an indirect assessment of the ease of reaching a healthcare facility for childbirth. Women living in rural parts of Enugu State, Nigeria—like women in other rural areas—may have had increased difficulty accessing healthcare facilities that can perform a CS because of limited transportation options, poor road conditions, and poverty [10, 12, 17]. Women in rural areas also refrain from using healthcare facilities during pregnancy because of financial constants, cultural norms that discourage delivery in a hospital setting, or gender inequalities in household decision-making [10, 12, 17]. Therefore, it remains unclear if distance to the nearest healthcare facility was a measure of difficulty accessing a healthcare facility or willingness to use a healthcare facility and it does not indicate if the nearest facility had CS access.

To the authors’ knowledge, this study is the first epidemiological investigation to report that high malaria parasitemia is associated with higher odds of CS. The literature and current guidelines are based on case studies [40–42]. It is unknown if a biological pathway exists underlying this relationship, or if women with higher malaria parasitemia lack adequate health care overall, which inherently makes their pregnancies higher risk. The relationship between malaria parasitemia and CS warrants further attention.

In SSA, some evidence has suggested that women with SCD are more likely to have a CS [43, 44]. However, it is difficult to determine whether those with SCD receive any benefits from having a CS, because both SCD and CS are related to high risk of adverse maternal and neonatal outcomes in SSA [44–46]. It has been established that malaria and SCD are associated with anemia during pregnancy [47–50]. There has been much debate in the literature on whether anemia is related to an increase in maternal morbidity and mortality in the context of developing countries [51]. Having a CS would complicate this relationship and also warrants further attention.

Our study is consistent with a meta-analysis that found HIV-infected women were no more likely to have a CS than those uninfected [52]. Evidence from resource unconstrained areas suggests that having a CS is beneficial if a woman’s HIV-RNA level is above 1000 copies/ml near delivery [53]. Because women in resource-constrained areas are often unaware of their viral load before delivery, having a CS could be beneficial for HIV-infected women [54]. However, in resource-constrained areas, CS are often unavailable and unsafe; therefore, the WHO guidelines do not currently recommend that HIV positive women in resource constrained regions have an elective CS [55]. Instead, the WHO recommends that HIV positive women take 3 or more antiretroviral medications in order to decrease mother to child transmission of HIV [56]. The WHO also recommends that infants receive antiretrovirals during the post-natal period if their mother is HIV positive [56].

Acknowledgments

We are grateful to the Catholic Bishop of Agwu Diocese, Anglican Bishop of Enugu; Catholic Bishop of Enugu; and Anglican Bishop of Oji-River. Their support was instrumental to the successful implementation of HBI. HBI implementation would not have been possible without the support and tireless effort of the priests in the participating churches. Church-based Volunteer Health Advisors took ownership of the program and made the process of recruitment and implementation smooth for our study team and participants. This study would have been impossible to conduct without the support of PeTR-GS (our PEPFAR-supported partner) staff and volunteers.