A facility-based, retrospective, cross-sectional study design was used, including a review of 794 CS deliveries conducted over more than 12 months at Princess Marina Hospital (PMH) in Botswana. PMH is one of three tertiary and national referral hospitals in Gaborone, Southeast District, Botswana [23]. The hospital has 581 beds and serves a population of 231592 [24, 25]. The hospital provides comprehensive emergency obstetric care services, including CS interventions [23].

Princess Marina Hospital (PMH) was randomly selected from three National Referral Hospitals in Botswana. A total of 6737 deliveries were performed over 12 months at PMH. Among the total delivery registry, 1918 included all CS deliveries, including selective and emergency CS. A single population proportion was used for sample size calculation; n = Z² * P (1−P) / d², with proportion 50%, α= 0.05, 95% CI, and d=0.05, and a design effect of 2 was used to ensure national representation, since Princess Marina Hospital is a national referral hospital to which many pregnant women from different regions present. Additionally, 3.5% were included to address missing data and increase the sample size. The final sample of 794 randomly selected CS deliveries from the database was included to characterize indications and maternal and perinatal outcomes among the total CSs.

A structured, pretested data-collection tool was used to extract data from maternity registers [23]. The data collected included the mothers’ sociodemographic characteristics, such as maternal age and residence; antenatal care data, such as gravidity, parity, and gestational age; and maternal and fetal risk indications for CS deliveries, such as CPD, previous CS, fetal distress, and malpresentation. Immediate maternal and fetal outcomes such as maternal death, stillbirth, maternal and fetal complications, and morbidities were also included in the data collection [23]. Data completeness and consistency were assessed and cleaned prior to analysis. EpiData version 3.1 was used for double data entry verification, and analysis was conducted using SPSS version 25. A comprehensive data validation process was implemented to ensure dataset reliability, including checks for data integrity and logical consistency across variables. Discrepancies were resolved through data cleaning and cross-verification with source records. A pre-test was conducted to evaluate the effectiveness of the data collection instruments, leading to the identification and correction of errors, inconsistencies, and outliers.

As a retrospective study, data were extracted from maternity registers at selected health facilities. The extraction process involved reviewing records of cesarean section deliveries and entering relevant information into a pre-tested data collection tool. To minimize information bias, the tool was aligned with routinely documented variables, and cross-verification with source records was conducted during data cleaning.

The analysis focused on determining the rate of CS among total deliveries and characterizing the main indications and immediate outcomes of CS deliveries. Baseline characteristics, such as indications of CS deliveries, for example, labor-related complications, CPD, fetal distress, and malpresentation, were described via counts, frequency, proportions, means, and percentages.

Ethical approval to conduct the study was obtained from the University of the Witwatersrand Ethics Committee. The Health Research and Development Division, Ministry of Health and Wellness, and the PMH Research and Ethics Committee, Gaborone, Botswana, granted permission to conduct the study. Confidentiality was maintained by excluding patients’ identifiers from the data extraction tool.

For a total of 794 CS deliveries, the mean maternal age was 28.76 years, with a range of 15–45 years. More than half (50.6%) of the participants were aged 25-34 years, followed by more than a quarter (28.5%) aged 15-24 years, and one-fifth (20.9%) aged 35+ years. More than half (59.3%) of the participants were urban residents.

The CS delivery rate among women who attended deliveries for 12 months was 28.5% (1918/6737) among total deliveries attended during that period.

More than two-thirds (67.1%) (95% CI: 63.2 - 69.9) of the women were multigravidas, approximately two-thirds (65.1%) (95% CI: 61.3- 68.2) were multiparas, and one-fifth were less than 37 weeks of gestation (Table 1).

The most frequent indication for CS was fetal distress (23.6%, 187), followed by one previous CS (13.7%, 109), malpresentation (11.4%, 90), two previous CSs (10.8%, 86), and a CPD (6.7%, 53) (Fig. 1).

One previous CS (62%) was the most frequent CS indication for the 25–34-year-old age group, followed by labor-related CS (61%) and malpresentation (59%). More than two-thirds of cases of fetal distress were reported among those aged 15–24 years (Table 2).

With respect to women's obstetric history and CS, women with primigravida experienced more than two-thirds (37.6%) fetal distress as the most frequent indication for CS deliveries, followed by CPD (12.6%). Women with multigravidas with one previous CS accounted for 20.5%, followed by malpresentation (16.7%). With respect to parity, CPD (37.2%) was the most dominant indication of CS delivery for primiparas. In contrast, malpresentation accounts for 21.1% of CS indications for multiparas. With respect to gestation times of less than 37 weeks, the main indications for CS were hypertensive disorders (16%) and fetal distress (16%). A quarter (25.6%) of women with a gestation of 37 weeks and above reported fetal distress (Table 3).

Maternal outcomes were classified as those mothers who were alive, died/or developed a complication. All mothers who delivered were alive. Nineteen mothers (2.4%) (95% CI: 1.4 -3.5) had a material adverse outcome (i.e., developed complications but no deaths were reported). Among mothers who developed complications (18,95%), these complications were related to Peripartum Hemorrhage (PPH), and one mother experienced uterine rupture.

Perinatal outcomes were classified as live birth, stillbirth, and/or developed complications (i.e., low APGAR score, birth asphyxia, neonatal death, and fractures).

With respect to perinatal outcomes, 764 (96.2%) (95% CI: 95-97.5) live births, thirty (3.8%) (95%:CI: 2.5 – 5.0) were stillborn, and 126 (15.9%) (95%:CI: 14.4 - 19.9) of the newborns experienced perinatal complications (i.e., low APGAR score, birth asphyxia, neonatal death and fractures).

A total of 156 (19.6%) CS deliveries with adverse perinatal outcomes were reported. Thirty (3.8%) (95%: CI: 2.5 - 5.0) newborn babies died. One hundred twenty-six (15.9%) (95%:CI: 13.7 – 19.2) newborn babies presented with complications, which accounted for 93,74.0% of all complications due to a low APGAR score, followed by birth asphyxia (14, 11.0%) and neonatal death (NND) (8, 6.0%), (11, 9%) of which were fractures, and the other complications were major perinatal complications (Table 4).

There were no significant differences in maternal or fetal outcomes between primiparas and multiparas (p=0.08). CS at a gestational age of <37 years was associated with a greater risk of adverse fetal outcomes than CS at a gestational age ≥37 years (p=0.001).

This study revealed that PPH was the most frequent maternal complication, similar to national audit findings [39] and consistent with other studies [10, 22]. The Botswana Family Health Survey [39] revealed that most women in Botswana attend ANC and deliver in hospitals; thus, PPH is unlikely to be related to patient-related factors. More likely, they are related to key factors within the health system, such as inadequate quality of care or delays in providing the required interventions. Even though the majority of women attend ANC and deliver at health facilities, it is possible that PPH may be related to the fact that these women present late at health facilities. The current study was not designed to investigate the causes of maternal and fetal complications. However, to decrease the incidence of PPH, strengthening the capacity and training of health care workers in the provision of midwifery services, expanding community health education, and promoting early booking to enable early identification and management of this complication are critical.

This study also revealed that multiparous women had a higher rate of maternal morbidity, similar to findings in other studies [39]. It has been suggested that in this group of women, uterine muscle tone is decreased, and after delivery, uterine contractions may be ineffective and retraction may be inadequate to achieve hemostasis [39]. As most women requiring CS in this study were multiparous, the findings suggest the important role of family planning in reducing morbidity in this group [39]. Furthermore, it is essential to strengthen community health education on the importance of family planning and early ANC booking.

With respect to perinatal complications, this study revealed that a low APGAR score was the most common (74.0%) perinatal complication, which usually led to birth asphyxia and eventually NND. Although this observation is consistent with other studies, the rate is much higher than the 52% rate of low APGAR scores reported in these studies [10]. The high rate of low APGAR scores in this study could be attributed to several factors, such as a high referral number of complicated cases from lower-level health facilities, ineffective management of labor, delayed decision making, or delayed transfer of women from home or the referral of health facilities to referral hospitals [10, 40]. This study was not designed to analyze these factors, but the high rate of low APGAR scores with poor prognosis at PMH likely suggests that the neonatal intensive care unit (NICU) at this hospital lacks essential resources and is staffed by health care workers with limited expertise in neonatology. This has been documented in another study, which revealed that a lack of appropriate facilities in the NICU, an absence of neonatologists and neonatology nurses, and a shortage of pediatricians were contributing factors to high rates of low APGAR scores [10, 40]. Inadequate ANC utilization is associated with CS delivery and adverse perinatal outcomes [41], contributing to perinatal mortality [42], and ANC visits are significantly associated with lower neonatal death rates [43].

Therefore, to reduce these complications, it is critical for PMHs to institute quality antenatal care services, upscale comprehensive emergency obstetric and newborn care at lower-level health facilities, intensify neonatal resuscitative measures, and adequately train health care workers in intensive neonatal care. Furthermore, the community should be educated about the importance of early ANC booking and seeking medical help early in emergencies.

Although a large sample size was used in this study due to its retrospective design, it has limitations such as bias from data abstraction, incomplete data, missing information on confounders (e.g., BMI, HIV, gestational diabetes, history of medical condition), record quality, lack of detail in data, lack of active follow-up, and difficulty in establishing cause‒and‒effect relationships. A regression analysis to identify factors associated with CS delivery was not conducted.