A quantitative cross-sectional analytical study design conducted to determine the acceptance of CS and its associated factors among pregnant women in Lusaka, Zambia seeking antenatal care at WNH-UTHs.

The study took place at the WNH-UTHs in Lusaka Zambia. It’s a national referral centre with a 562-bed capacity and also serves as the main medical training institution for healthcare professionals. The hospital provides secondary and tertiary care services.

The study population consisted of pregnant women.

The study recruited only consented and assented pregnant women irrespective of trimester seeking ANC services at the WNH-UTHs.

A systematic random sampling technique was used to select all eligible pregnant women seeking antenatal care at WNH-UTHs. The sampling frame of the study population was all pregnant women seeking antenatal services at WNH-UTHs.

The study included

· Pregnant women aged 15 years to 49 years seeking ANC services at WNH-UTHs with or without a history of previous CS.

· Pregnant women brought in as an emergency.

The sample size was determined using Cochrane’s formula (single proportion) for sampling size calculation of the unknown population as shown below:

n = z 2 p ( 1 − p ) e 2

The primary data source was used in this study. An interview schedule was adapted from other studies [16] [17] , that have researched factors associated with caesarean section acceptability and were used to collect information by the researcher. Before data collection, the questionnaire was pre-tested to ensure its validity and reliability.

The respondents were given full information about the purpose and significance of the study. Informed consent was obtained before the interviews were conducted. The respondents were made anonymous as no identification detail was requested or recorded. The researcher ensured all interview questions were answered and questionnaires were stored in a secure place.

According to Creswell [18] reliability and validity focuses on examining the stability or consistency of responses or generalizability (the external validity of applying results to new settings, people, or samples). The face and content validity of the questionnaire was established through examination and scrutiny by independent researchers’ experts in obstetrics and gynaecology. Supervisor and co-supervisor for inspection and Likert scale to ensure validity was used. The instrument underwent pilot testing for its reliability and was subsequently revised for clarity. The data collection tool was pre-tested. A comprehensive literature review was conducted, and privacy during data collection was maintained. Last but not least consistency was maintained in the manner questions were asked to ensure reliability.

Before the actual study, the researcher conducted a pilot study at Levy Mwanawasa Teaching Hospitals, WNH in which the validity and reliability of the data collection tools were ascertained. The pilot study comprised 10% (35 respondents) of the total study sample size. The pilot study enabled the researcher to check for respondent’s ability to understand and answer questions, it also helped to clarify grey areas in the tool.

The data collected were examined for consistency and completeness, cleaned, coded, and analysed using R software version 4.3.2 [19] . Frequency distributions and summary statistics were used to describe categorical and numerical variables. The median and interquartile range were used to describe numeric variables, as they were skewed. Normality tests were done with the Shapiro-Wilk test and graphically using a histogram with a superimposed normal curve. Associations between each categorical variable and the outcome variable were tested using a chi-square test. However, for variables with expected frequencies less than 5, Fisher’s exact test was used. For associations between the dependent variable and numeric independent variables, a Wilcoxon rank sum test was used. To identify the adjusted effects of independent variables, an investigator-led backward stepwise multivariable binary logistic regression analysis was done. The modelling process was guided by the Akaike Information Criterion (AIC) and the likelihood ratio test. All statistical analyses were conducted at a 5 per cent level of significance, thus, p-values not greater than 0.05 were considered significant.

Table 1 shows that respondents had a median age of 31 years (IQR, 26 - 36), the majority 351 (84%) were married, and Christianity accounted for the largest religion 408 (97%). Around half, 225 (54%) attained high-level education (secondary or tertiary), and 240 (57%) lived in medium-density areas. Being in formal 184 (44%) and informal 148 (35%) employment was common among respondents,

with most respondents 237 (56%) coming from a low economic status. Around two-thirds, 281 (67%) had 1 - 4 children, and most 287 (68%) respondents were in their third trimester of pregnancy at the time of the study. About a quarter, 113 (27%) had a previous CS delivery, with the majority of these deliveries being emergencies 81 (72%). The median time to recover was 7 weeks (IQR, 4 - 8) for respondents who had a previous CS delivery.

Acceptability was assessed using a five-item Likert scale, which focused on respondents’ willingness to undergo CS delivery if indicated. Acceptability was then dichotomized into agreeable and not agreeable based on the overall scores from the Likert scale.

As shown in Table 2, over a third of the respondents, (36%, n = 150) would not accept CS if it were indicated and two-thirds (66%, n = 277) would not request to delivery by CS. Over a third of the respondents, (39%, n = 163) would not willingly consent to an emergency or elective CS delivery, (30%, n = 126) indicated that CS was an abnormal means of birth, and (30%, n = 127) suggested that CS was dangerous to both the mother and baby. Overall, CS delivery was not agreeable among most of the respondents, (71%, n = 299), whereas (29%, n = 121) expressed acceptability of CS delivery.

Results in this section focused on respondents’ knowledge of CS delivery by assessing awareness of aspects such as indications, possible effects, and benefits of CS. The overall knowledge score was then used to categorize knowledge levels into low, medium, and high.

As shown in Table 3, (57%, n = 238) of the respondents indicated that prolonged labour was an indication for CS, and (75%, n = 313) indicated abnormal lie, abnormal placenta position, and carrying a big baby as the indications. About two-thirds, (65%, n = 275) suggested that vaginal delivery was still possible after CS, (34%, n = 142) agreed that CS delivery may require blood transfusion and that it may prolong hospital stay (77%, n = 323). A third of the respondents, (33%, n = 139) indicated that drugs given during CS were harmful to a

woman’s health, and most, (74%, n = 311) were aware that CS required a woman’s consent. The majority, (87%, n = 364) denied ever receiving IEC on CS at ANC, and (96%, n = 405) indicated that recovery was faster after vaginal than CS delivery. Many respondents indicated that CS could save the mother’s (69%, n = 288) and baby’s (83%, n = 350) lives.

Overall, over half of the respondents, 232 (55%) expressed high knowledge about CS, 111 (26%) had medium knowledge, whereas 77 (18%) expressed low knowledge.

This section presents findings on the attitude of respondents towards CS delivery. Attitude was assessed using a seven-item Likert scale, which focused on women’s feelings and thoughts on CS. Based on the overall scores, it was then dichotomized into positive or negative attitudes.

As shown in Table 4, (53%, n = 224) agreed that it was right to undergo CS for medical reasons, (47%, n = 199) agreed and (30%,n = 124) strongly agreed that a woman could give birth vaginally after CS. Around a third of the respondents, (34%, n = 141) disagreed that it is weak women who undergo CS, (29%, n = 121) disagreed that it is lazy women who undergo CS, whereas (25%, n = 104) agreed that CS was an abnormal means of delivery. More than a third, (38%, n = 161) strongly agreed that CS was more painful than vaginal delivery, and 174 (41%) disagreed that healthcare workers deliberately opted for CS. About half, (47%, n = 199) agreed to the possibility of a vaginal delivery after CS.

Overall, about half of the respondents, (48%, n = 203) expressed a negative attitude towards CS, whereas (52%, n = 217) expressed a positive attitude.

Results in this section focus on respondents’ perception towards CS delivery, assessed using a Likert scale, which focused on respondents’ opinions on the safety and effects of CS, and how they perceived women who deliver through CS. The overall perception was dichotomized into positive and negative based on the overall scores from the Lickert scale.

Table 5 shows that (41%, n = 173) of the respondents strongly disagreed that CS was a safe method of delivery, around half, 214 (51%) disagreed that women with previous CS cannot deliver again, (47%, n = 96) and (44%, n = 183) indicated that babies born by CS were normal. Less than half [(42%, n = 176), and (42%, n = 175)] were worried about the competence of health professionals in CS. Similarly, most respondents agreed that CS delivery was not a taboo (46%, n = 193) and that women who delivered by CS were not sub-humans (49%, n = 206). Less than a quarter, (23%, n = 96) perceived CS to be expensive, most disagreed that CS was done only on cursed women [(36%, n = 152) and (42%, n = 175)], and under half of the respondents, (44%, n = 183) agreed that CS is sometimes done for medical reasons.

Overall, about a quarter of the respondents, (23%, n = 98) had a negative perception towards CS, while most, (77%, n = 322) expressed a positive perception.

This section presents results on socio-cultural beliefs on CS delivery among study respondents. This focused on respondents’ religious and traditional convictions and fears about CS delivery.

Table 6 shows that only (16%, n = 69) of the respondents had some sort of beliefs that affected their preferred choice mode of delivery, (22%, n = 93) believed

that only wealthy women undergo CS, and (21%, n = 89) thought CS was a punishment and not God’s will. Most respondents, (64%, n = 269) needed their partners to decide on undergoing CS, and the majority, (80%, n = 336) would undergo CS if a partner or significant other supported it. About (39%, n = 163) believed that women did not want to undergo CS because of fear of being rejected by their husbands. Many respondents, (92%, n = 388) had fears about CS, with common fears being bleeding, death, wound gapping, infections, and scarring 205 (53%), and that practitioners could forget instruments and materials in the uterus/abdomen (25%, n = 96).

This section presents cross-tabulations and tests for associations between the acceptability of CS and various independent variables using the chi-squared and Fisher’s exact tests.

Table 7 shows that acceptability of CS was common among respondents who were married (94%, n = 114), those who attained a high education level (69%, n = 83), and those who lived in medium-density areas, (63%, n = 76). The results showed that marital status (P < 0.001), education attainment (P < 0.001), and area of residence (P = 0.018) were significantly associated with the acceptability of CS. Furthermore, most respondents who were in formal employment (45%,n = 55), those with 1 - 4 children (69%, n = 83), and most respondents with no history of CS (66%, n = 80) expressed acceptability of CS. Similarly, CS was commonly accepted among those with high knowledge, (76%, n = 92), positive attitude (77%, n = 93), positive perception (93%, n = 112), and those with no negative socio-cultural beliefs about CS (89%, n = 108). Employment status (P < 0.001), parity (P = 0.003), history of previous CS delivery (P = 0.040), knowledge of CS (P < 0.001), attitude towards CS (P < 0.001), perception of CS (P < 0.001), and socio-cultural beliefs on CS (P = 0.045) were all significantly associated with acceptability of CS in the study.

Results in this section focus on regression analysis estimates using both univariable and multivariable regression analysis. The findings presented below were arrived at using an investigator-led backward step-wise model selection approach.

Table 8 shows that unmarried respondents had significantly lower odds of CS acceptability compared to those in marriage at both univariable (cOR = 0.23, CI = 0.10, 0.50, P < 0.001) and multivariable (aOR = 0.22, CI = 0.08, 0.53, P = 0.002) analysis. High relative to low education attainment was significantly associated with increased odds of accepting CS (cOR = 2.41, CI = 1.56, 3.80, P < 0.001) at univariable but not at multivariable analysis. Similarly, at multivariable analysis, respondents with a high socio-economic status had significantly lower odds of accepting CS compared to those with a low status (aOR = 0.58, CI = 0.33, 1.00, P

^wWilcoxon rank sum test; ^cPearson’s Chi-squared test; ^fFisher’s exact test.

= 0.050). Compared to respondents with no children, those with 1 - 3 children (aOR = 0.44, CI = 0.23, 0.82, P = 0.011) and four or more children (aOR = 0.19, CI = 0.06, 0.55, P = 0.004) had 66% and 81% lower odds of accepting CS respectively, and this was significant taking account of other variables.

The odds of accepting CS were significantly higher for respondents in their third trimester of pregnancy compared to those in the first trimester (aOR = 3.94, CI = 1.49, 11.8, P = 0.008), controlling for all other variables. History of a previous CS was significantly associated with increased odds of accepting CS independently (cOR = 1.62, CI = 1.02, 2.55, P = 0.041) and taking into account other characteristics (aOR = 2.50, CI = 1.38, 4.59, P = 0.003). Having high knowledge about CS significantly increased the odds of CS acceptability 5.67 times (aOR = 5.67, CI = 2.14, 18.1, P = 0.001) compared to low knowledge. Similarly, respondents with a positive attitude towards CS had higher odds of CS acceptability at both univariable (cOR = 4.69, CI = 2.93, 7.69, P < 0.001) and multivariable (aOR = 3.87, CI = 2.24, 6.85, P < 0.001) analysis compared to those with a negative attitude.

In this study, the respondents had a median age of 31 years. This aligns with the expected demographic, as the subjects primarily consisted of women of reproductive age. These findings are consistent with previous research, where antenatal clinics commonly saw women in the age range of 26 to 30 years [20] . More than three-quarters of the respondents, 351 (84%), were married this could be attributed to cultural norms and expectations as marriage is considered a significant life event in one’s life, the other reason could be the result of Christianity being the dominant religion in Zambia which often supports marriage as sacred. The high levels of marriage in the study are consistent with the ZDHS [21] report which shows that marriage among women in Zambia starts early, as early as 19 years and over half of women in the country are married. Additionally, societal norms demand that pregnancy occur in marriage and because we are targeting pregnant women this was a likely finding.

The majority of respondents (97%) were identified as Christians this is so because of the strong reflection of Christian historical, cultural and social factors identified in Zambia.

About 57% of the respondents reside in medium-density areas while 56% were from a low economic status. In this study, approximately two-thirds (67%) of respondents were a combination of primipara and multipara. Additionally, more than half (68%) of the respondents were in the third trimester of pregnancy. Notably, the current study found a significant association between gestational age and the acceptability of caesarean section (CS). Specifically, those in the third trimester were more likely to agree to undergo CS (P ≤ 0.008). The finding of this study is in line with those found by Himalowa et al. [22] , who conducted a study at WNH-UTHs Lusaka Zambia although the study population was different from the current found that gestational age and maternal age were significantly associated factors associated with Cesarean Section. The current study found that both gestational age and parity play a crucial role. Parity and the frequency of antenatal visits significantly influence pregnant women’s decisions and reshape existing views regarding delivery.

Furthermore, the history of previous CS was found to be significant in the current study which is contrary to the study by Omotayo et al., [23] , where there was no significant relationship between previous obstetric experiences and acceptability of caesarean section. The discrepancy in the findings could be attributed to the study setting, study design, and population recruited only multiparous women, as well as the sample size which was smaller than the current study. Economic status has been found to influence the acceptability of CS in this study with a P ≤ 0.050. This implies that women with high economic status and women from medium-density areas are more likely to accept CS and this finding is similar to a study conducted by [24] , in Vietnam where they found that women with high socio-economic status are likely to accept CS among pregnant women. In the present study educational status has also been found to significantly influence CS acceptance among respondents (P ≤ 0.001) similar to a study by Oshodi et al., [25] , which found that educational status positively influenced acceptance of CS (P ≤ 0.001). These findings suggest that there is a need to address the socio-economic and demographic factors that affect the acceptance and preference of pregnant women towards CS and to ensure that women have access to respectful and evidence-based care that meets their needs and preferences.

The study findings revealed lower acceptability of CS and about half of the respondents would not be willing to consent to an emergency or elective CS delivery. On the other hand, more than half of respondents indicated CS was an abnormal means of birth while (44%, n = 186) respondents stated that CS was dangerous to both mother and baby. Despite the study indicating high knowledge and positive perception of CS, these did not influence the acceptability of CS among women. This could be attributed to the lack of IEC during ANC (Table 3).

The findings of this study are consistent with a study conducted in Ghana in Cape Coast Metropolis by Ansah, [16] , who also indicated a lower acceptance level of 34%. The current study is also in agreement with the study conducted by Begum et al., [15] in a rural community in Bangladesh where women had a low preference for CS which was attributed to a lack of IEC on the medical indications for caesarean section. However, these findings are contrary to the findings of a study conducted in Nigeria that indicated about 68% of pregnant women received knowledge on CS during ANC care from health personnel. The current study is in line with the study conducted by Schantz, [24] , in France, where 2.5% of respondents had a preference for CS which was perceived as more painful, riskier, and less natural compared to vaginal delivery.

The findings of the current study are contrary to those of Omobolanle et al. [26] , who found a high acceptance of CS (80.3%) in Nigeria. This could be explained by the differences in context and region as well as the study being a multi-facility study. The Nigerian study utilized a descriptive cross-sectional survey and a purposive sample based on specific criteria related to the research. It also involved high-risk pregnant women, who had medical or obstetric conditions that increased the likelihood of complications during pregnancy or delivery.