Background

Caesarean section (CS) is a life-saving obstetric surgery, which may be necessitated (sometimes the only feasible option) in high-risk pregnancies such as those with multiple/large fetuses, breech presentations, obstructed labour, as well as in women with transmissible infections such as HIV/AIDS.1 The adequate population-based prevalence for this essential obstetric intervention remains a subject of strong contentions, worldwide, revealing a lack of consensus.1–3 However, evidence suggests that a population-based CS prevalence <5% indicates unmet needs (lack of access to women in need of it), while prevalence >15% may show no additional benefit for mothers and babies.4 5

In 1985, the WHO recommended CS rates―as a percentage of live births―between 10% and 15% as the optimal range, with a declaration that ‘there is no justification for caesarean section rates in any region to be higher than 10%–15%’.6 This position has been contested given the data on which the recommendation was based were limited and drawn primarily from northern European countries.3 In a more recent position statement, the WHO maintains that population-based CS rates >10% are not associated with a reduction in maternal and neonatal mortality rates.1 7 Nonetheless, the world health body emphasises the need of CS service provision to every woman in need of it regardless of the prevailing population-based rates.1 7 When medically indicated, CS has the potential for reducing maternal/neonatal mortalities and morbidities including delivery complications such as obstetric fistula.1 7 8 However, a non-medically indicated CS has no associated additional benefits for mothers and newborns, rather like any surgery, it carries both short-term and/or long-term health risks.1 7 8

Caesarean delivery is over-utilised in many middle-income to high-income countries.9 For instance, the rate is as high as 25.9% in China, 32.3% in Australia/New Zealand and 45.9% in Brazil.2 4 9 It has been argued that many of the caesarean deliveries in these countries were in excess, medically unjustifiable and thus unnecessary.4 However, in several low-income countries, where over 60% of the world’s births occur, the population-based prevalence of CS is low—for example, 3.0% in West Africa.4 9 This low prevalence may reflect poor availability of-/accessibility to comprehensive essential obstetric care services (EOC) in the countries/region.2 Comprehensive EOC refers to a package of clinical services for managing pregnancy/childbirth-related complications of which CS is a critical component.5

Available evidence pertaining to the population-based prevalence of CS in Nigeria reveals a threshold, that is, far below the 10% recommended by the WHO. Moreover, there has been no significant increase in the population-based CS rates for several years in the country.10 For instance, in 2008, merely 2% of births were delivered through CS in Nigeria,11 and the rate remained unchanged in 2013.10 This prevalence is substantially lower than for many African countries including Ghana (12.80% in 2014), Lesotho (9.70% in 2014) and Uganda (5.22% in 2011).12–14 The considerably low population-based prevalence of CS in Nigeria suggests unmet needs which may contribute to poor maternal and neonatal outcomes in the country.4 10 Consistent with this premise, Nigeria currently accounts for the highest absolute number of maternal mortality and the second highest number of neonatal mortality in the world.15–17 Hence, the importance of investigating factors associated with the utilisation of this life-saving obstetric surgery in the country.

Some studies have been conducted on CS utilisation in Nigeria18–20 including a survey which examined the views of pregnant women and found that a high proportion of the study participants were averse to caesarean delivery.19 A significant association between CS and parity, maternal weight, child’s birth weight and previous CS were reported in another study.20 However, studies to date are institutional-based and limited by small sample sizes. Nationally representative studies on this crucial subject are necessitated in the country. The present study, thus, assesses the prevalence and factors associated with CS utilisation in Nigeria. Findings will provide evidence-informed knowledge for decision-making on the provision and utilisation of caesarean delivery in Nigeria.

Methods

Results

Prevalence of caesarean delivery

Out of the total number of deliveries, 659 were through CS, representing a prevalence of 2.1% (95% CI 1.8 to 2.3) (table 1). The highest prevalence of caesarean delivery was observed among women who had access to health insurance (10%), followed by those who delivered in private health facilities (7.2%), women who were obese (6.9%) and those who had multiple births (6.4%). CS prevalence was comparatively higher in women who had acquired at least a secondary level of education (4.8%), and in rich households (4.5%). Women in Christian religion (4.1%) or residing in the South-West region (4.7%) or who had attended at least four antenatal care sessions (3.9%) or living in urban areas (4%) had a comparatively higher prevalence of caesarean delivery. Conversely, the lowest prevalence of caesarean delivery was observed among women professing Islam (0.6%), or in poor households (0.6%), or whose husband had no education (0.4%) or did not attend antenatal care at all (0.4%) or were not educated (0.5%). Women residing in rural areas (figure 1), as well as the North-West and North-East regions (figure 2), had a substantially lower prevalence of CS at 1%, 0.6% and 0.9%, respectively. Significantly lower prevalence of CS was recorded among women who had genital cutting (1.6%) compared with their counterparts who did not (2.3%, p=0.011).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Prevalence of caesarean section by rural–urban residence in Nigeria.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Prevalence of caesarean section by region of residence in Nigeria.

Discussion

We determined the national prevalence of CS to be 2.1% in Nigeria which indicates under-utilisation of the service in the country. Factors associated with low prevalence and decreased odds of CS include residence in rural areas, lack of antenatal attendance, affiliation with Islamic religion, lack of health insurance coverage, lack of husband/partner’s formal education and birth order ≥4. Maternal age ≥35 years, large birth size, multiple births and maternal overweight/obesity were similarly associated with higher prevalence and increased odds of CS. Previous studies have reported a much higher prevalence than the 2.1%, ranging from 11.3% in the North-West to 18.8% in the South-East and 40.1% in the South-West regions in Nigeria.20 30 31 However, all these studies were institutional-based; and, do not give a true reflection of the prevalence of CS at the population level in Nigeria. Health facilities, particularly, tertiary and regional healthcare centres in Nigeria, where some of the studies were conducted, receive a greater proportion of high-risk patients and would more likely perform a greater number of caesarean deliveries.

A range of factors may explain the low prevalence of CS found in the present study. First, is limited access to- and availability of obstetric care services in Nigeria. The WHO’s guideline recommends at least five EOC facilities per 500 000 people, one of which should be capable of providing comprehensive EOC services, and, these need to be evenly spread in the population.5 This level of facility and service coverage has yet to be realised in Nigeria.32 33 Available evidence indicates that facilities and expertise for EOC are inadequate and/or sparsely distributed in the country.32–34 Access to facilities could be poor, coverage low and the needed manpower for anaesthesia and caesarean delivery may be lacking/insufficient in many facilities.35 Second, is the challenge of low acceptance of CS among women in Nigeria, blameable on fear of death, concern about complications, the negative perception of CS as an abnormal mode of delivery and the high cost of the surgery in the country.19

Following the multivariable analysis, the odds of caesarean delivery were 50% higher in urban compared with rural residence, and this may be due to the urban advantage in access to obstetric care services in Nigeria.34 Caesarean delivery is one of the nine life-saving signals that constitute comprehensive EOC,5 36 and studies agree on the poorer coverage/availability of services in rural Nigeria.32 33 For example, in Abia state, South-East Nigeria, only ~19% of the health facilities surveyed met the requirements for EOC services and 77% of those were sited in urban centres.34 A similar finding has been reported in other parts of the country.32 33 Findings in a nationwide study further indicate that EOC services are inadequate in rural Nigeria.37 Promoting equitable access to quality and accessible obstetric services including CS should indeed be the focus of future interventions and women in rural Nigeria need to be especially prioritised.

Previous studies have shown disparities in the use of CS between the poor and the rich13 38 and, factors related to financial capability and access to health insurance were strongly associated with increased use of CS in this study. For instance, women with health insurance coverage had the highest prevalence of CS―10%. Also, compared with their counterparts with no health insurance coverage, women who enjoyed the facility were twice as likely to utilise caesarean delivery following adjustment for other factors/confounders. Similarly, the odds of a CS were twofold higher among women whose husband had at least a secondary school education―a possible indication of a higher socioeconomic status. These results compare well with previous findings.18 39 Considering that the cost of CS is rather high,4 and may not be within the reach of an average Nigerian family, it is likely that financial constraints contributed to the low utilisation of CS in this country. The results of our χ² and simple logistic regression analysis lend credence to this argument indicating that women in rich households had a much greater prevalence and increased odds of utilising CS compared with their counterparts in poor households.

However, wealth index did not attain statistical significance in our multivariable analysis. A follow-up analysis showed that the effect of the variable waned and disappeared following adjustment for antenatal visits and health insurance. This finding suggests that antenatal attendance and health insurance coverage may modulate the effects of socioeconomic status in respect of CS utilisation in Nigeria. Similar to the present finding, previous studies have shown that access to health insurance coverage increased the odds of healthcare facility delivery and antenatal care services utilisation in Nigeria.25 40 Hence, interventions targeted at enhanced coverage of the insurance may prove an important entry point for improved utilisation of CS and other maternal healthcare services, particularly, among the poor and underprivileged women in Nigeria.

The strong association found between antenatal attendance of at least four times and increased prevalence/odds of CS may be explained by the unique opportunity that antenatal care services offer in identifying clients with high-risk pregnancy for appropriate obstetric intervention.18 25 Antenatal services provide the best avenue for counselling and awareness creation thereby empowering pregnant women to make informed decisions in matters of their health, including, when necessary, the utilisation of CS.18 25 While the present finding underscores the relevance of antenatal care attendance to the uptake of CS, antenatal care is equally under-utilised at 46.5% in Nigeria, 61.1% in rural Nigeria and 22.4% in urban Nigeria.25 Intervention efforts aimed at improving CS utilisation, therefore, need to further prioritise antenatal care attendance among pregnant women in the country.25

Other factors, including maternal age ≥35 years, multiple births and maternal overweight/obesity, were associated with increased odds of CS and the findings are consistent with previous studies.20 41 The named factors are known risks for adverse pregnancy outcomes,20 41 and pregnant women in any of the categories are more likely to undergo a life-saving CS. The findings of a significant increase in the odds of CS among women with low parity and those whose babies were perceived as being large have been reported in studies.20 42 Cephalopelvic disproportion commonly associated with fetal macrosomia may explain the finding in respect of large birth size.43

Corroborating the reports of previous studies in respect of maternal healthcare services utilisation,25 40 44 45 our study reveals over four-fold higher prevalence and more than twofold increased odds of CS among Christian women compared with their Muslim counterparts. Several factors may contribute to this finding. First, preference for female healthcare providers is common among Muslim women, and, where it cannot be guaranteed, may result in low utilisation of healthcare services.25 44 45 Second, religious belief/obligation which discourages women from undue exposure of their bodies has been suggested in explaining low use of maternal healthcare services among Muslim women,44 and this may be relevant to the present finding. Another important factor, perhaps, borders on maternal autonomy, women empowerment and gender inequality as several Muslim women often need to take permission from their husbands and/or religious leaders before making health-related decisions.25 44 45

In the Nigerian context, the present result may also relate to differences in geographic location and education level between Christian and Muslim women. For example, our descriptive statistics and simple logistic regression analysis show a significantly lower prevalence and decreased odds of a CS in northern Nigeria, where Islam is predominant and many states are educationally less-developed, compared with the southern regions.10 Notably, our follow-up analysis—cross-tabulation of maternal education level and religion (data not shown in table)—reveals that Muslim women accounted for 90.4% of the respondents with no education compared with only 7.7% among Christian women (p<0.001). In contrast, 71.9% of women who had acquired secondary/higher education were Christians compared with 27.4% among Muslim women (p<0.001).

These results suggest a possible contribution of disparities in educational attainment in the observed CS utilisation difference between Christian and Muslim women in Nigeria. In support of this position, lack of maternal and husband/partner’s education was significantly and overwhelmingly associated with low prevalence and decreased unadjusted odds of a CS. Granted that maternal education did not attain significant status in the multivariable analysis, husband/partner’s education retained its significance, underpinning its importance in the present context. Education does not only contribute to an improved socioeconomic status, it enhances skills, knowledge and confidence for appropriate healthcare services utilisation.46 Hence, when a need arises, better-educated husbands would more readily appreciate and support their wives for a life-saving caesarean delivery use.

There is consistent evidence that vaginal delivery is associated with many complications in women with FGC, which may result in an increased risk of a CS.47 Our study, however, shows that FGC was not associated with an increased prevalence or unadjusted odds of CS. A similar finding has been reported.48 Given the low CS prevalence in the present study, limited access to the obstetric surgery in Nigeria may have contributed to our findings for FGC highlighting issues related to the supply side of services.

The national representativeness of the 2013 NDHS means our findings are generalisable to all women of reproductive age in Nigeria. Low missing data, use of complex sample and high response rates are additional strengths of this study. To the best of our knowledge, the is the first population-based study to examine factors associated with CS utilisation using nationally representative data in Nigeria. Nonetheless, our findings need to be interpreted taking into consideration a few limitations. First, the data utilised were self-reported, collected retrospectively, and so liable to recall bias. Restricting our analysis to the most recent live deliveries, however, reduces the chances of this limitation. Second, given the cross-sectional design of the data analysed, causal relationships between our outcome and explanatory variables could not be ascertained. Lastly, the 2013 NDHS data are at least 5 years old and may not reflect the current state of things in Nigeria. Nevertheless, the data remain the most recent edition in the series of NDHS available at the time of this study and our findings provides a suitable comparison for future studies on this subject.

Conclusions

We found a considerably low prevalence of caesarean delivery in Nigeria. Rural residence, Islamic religion, lack of antenatal visit, lack of health insurance coverage, lack of husband/partner’s education and birth order ≥4 were significantly associated with lower prevalence and decreased odds of caesarean delivery. While there is justification for keeping CS rates as low as possible, this study highlights the critical need for increased provision and better utilisation of life-saving CS in Nigeria. The present prevalence suggests unmet needs which are a known risk for higher maternal and newborn mortalities. Our study reveals the need to address geographic, and socioeconomic factors associated with the low prevalence of CS in Nigeria.

A faith-based approach, as well as interventions, focused on women empowerment/maternal autonomy may prove beneficial in improving the uptake of CS, particularly, among women with Islamic affiliation in Nigeria. Improved availability and access to obstetric care services need to be further pursued by meeting the WHO’s recommendations on EOC in all the regions in Nigeria. This will entail increasing the number of comprehensive EOC facilities and promoting even distribution of same, improving staff strength and enhancing their skills as well as equipping and upgrading the existing facilities in Nigeria.

Considering that CS is costly in Nigeria, delivery services need to be made freely available or at the very least, substantially subsidised to address the challenge of inequitable access between the rich and the poor in the country. Based on our findings, the provision of universal health insurance coverage is an important, and practical intervention in this respect. On the other hand, caesarean deliveries associated with maternal overweight and obesity are rather avoidable/preventable. A short-term to long-term intervention efforts would be to implement health promotion programmes targeted at preventing/reducing maternal overweight/obesity—known risk for CS and several chronic diseases. Future disaggregated studies are recommended for a better insight into the within-country variations in access to- and utilisation of CS in Nigeria. Also, future population-based studies need to explore the contribution of fear and cultural practices to the utilisation/non-utilisation of CS in Nigeria.