Towards Optimizing Caesarean Section: Robson Ten Group Analysis of Caesarean Section and It’s Determinants in a Tertiary Hospital in South-South, Nigeria

Background: To optimize caesarean section (CS), Robson classification is useful for audit of CS rates within and across hospitals and regions. Valid conclusions are also possible by analyzing other characteristics of parturients that determine CS rates based on Robson groups. This study aimed to analyzed CS rate and determine the impact of parturients’ age and booking status on delivery by CS based on Robson classification. Subjects and Method: A cross-sectional descriptive study that classified parturients into Robson 10-groups using data from hospital records at the Federal Medical Centre Yenagoa (FMCY) in Bayelsa State, south-south, Nigeria. The dependent variable was mode of delivery. The independent variables were parturients’ age and booking status by Robson group. The study instrument was a predesigned spreadsheet used to collect real-time relevant data on all deliveries from patients’ hospital records. Descriptive statistics were presented using frequencies, percentages, mean and standard deviation. Chi-square, Exact test and logistic regression were used to determine association of parturients’ age and booking status with mode of delivery. Level of significance was p <0.05. Results: There were 556 deliveries during the study period and 269 CSs, giving a CS rate of 48.4%. Robson group 3 made the highest (27.9%) contribution to CS rate, followed by group 10 (22.3%), 5 (13.8%) and 1 (11.2%). The commonest indication for CS was cephalopelvic disproportion, followed by severe preeclampsia. Booked parturients in Robson groups 1 and 3 had 61.0% reduced odd (OR= 0.39; 95% CI = 0.15 to 0.99; p = 0.050) and 74.8% reduced odd (OR= 0.25; 95% CI= 0.14 to 0.45; p < 0.001) of delivery by CS, respectively. Conclusion: The CS rate at the FMCY was contributed largely by group 3, 10, 5 and 1 parturients. Using Robson classification, CS rate can be focused to targeted intervention to optimize CS.


BACKGROUND
While it is important to ensure adequate access to caesarean section (CS) for medically justified indications, to save the lives or protect the health of the mother and or the baby, it is also important to avoid unnecessary CSs.One CS increases the risk of further CSs which is associated with increasing risk of surgical maternal morbidity.Increasing CS rate has become a global trend (Betran et al., 2016a), especially in developed countries and also among the population above low social class and in the urban settings of developing countries (Makinde et al., 2020).The World Health Organization (WHO) recommends a CS rate of 10-15% at the population level and stated that "there is no justification for any region to have CS rates higher than 10-15%", as above this threshold there is no further reduction in maternal and perinatal morbidity and mortality (Betran et al., 2016b;WHO, 2019).
There is a need for continuous audit of CS rates at the hospital level.The WHO proposed the Robson classification system as a tool for hospital-level audit of CS rates, based on standardized data, enabling reproducible comparisons within and across populations (WHO, 2019).The Robson classification system groups all deliveries into one of ten totally inclusive and mutually exclusive groups on the basis of six parameters: parity, previous CS, onset of labour, gestational age, number of fetuses and fetal lie and presentation, which are usually routinely collected by obstetric care providers worldwide (Voget et al., 2015;WHO, 2017).Also known as the 10-group classification system, the Robson classification system is for grouping all women who are admitted for delivery at a specific setting and not only for the women who deliver by CS (WHO, 2017).
Using the Robson classification system, groups of parturients who contribute most and least to CS rates can be identified, and specific group-focused interventions relevant for each health care facility can be adopted.Analysis of hospital-level CS rate on the basis of the Robson classification system is already being carried out in hospitals in Nigeria (Ochejele et al., 2021;Bello and Agboola, 2022;Okonta et al., 2022;Akadiri et al., 2023) and other countries.Country (Kelly et al., 2013;Nakamura-Pereira et al., 2016;Kacerauskiene et al., 2017) and regional (Zeitlin et al., 2021) data is also being published.Besides CS rates, other pregnancy outcomes have also been successfully analysed based on Robson classification (Savchenko et al., 2022), suggesting that the utility of the Robson grouping system is yet to be fully explored.Valid conclusions are possible if other characteristics of parturients that may determine CS rates are analysed on the basis of the Robson groups.This study aimed to analyse CS rate and determine the impact of parturients' age and booking status on delivery by CS based on Robson classification.

SUBJECTS AND METHOD 1. Study Design
This was a cross-sectional descriptive study conducted at the Federal Medical Centre Yenagoa (FMCY), Bayelsa State, southsouth, Nigeria, from July 2020 to April 2021.FMCY is a tertiary level health facility in Yenagoa the capital city of Bayelsa State in the south-south geopolitical zone of Nigeria.Bayelsa State has a population of about 2,700,000 (Bayelsa State government, 2023).
The indigenous people of Bayelsa are collectively referred to as Ijaws and the state is also home to a sizable community of nonindigenous tribes including the Igbos, Ibibios, Efiks, Urhobos, Itsekiris, Isokos, Edos, Yorubas, Hausas etc.Most people in Bayelsa State engage in trading, subsistence farming and small-scale commercial farming.Others work in the State and Federal civil service.The hospital receives patients directly and also serves as a referral hospital to the entire primary and secondary level health facilities in the state; to private hospitals in Yenagoa and its environs, and to traditional birth attendants in Yenagoa and its environs.The department of obstetrics and gynaecology conducts an average of 1800 deliveries annually.However, the COVID-19 global pandemic caused a reduction in the patient load received by the department during the study period, just as it was in other settings globally.

Population and Sample
Study population were parturients managed at the obstetrics and gynaecology department of the FMCY.All women who did not deliver at the FMCY but were managed following referral from another health care facility post-delivery were excluded from the study.The calculated required minimum sample size was 375, using the formula for sample size estimation for a population prevalence; N= (Z1-α/2) 2 (P) (q)/ d 2 (Sharma et al., 2020), at a confidence level of 95%, 42.4% prevalence of the dependent variable (Allagoa et al., 2021) and margin of error of 0.05.All eligible parturients during the period of the study were enrolled.

Study Variables
For assessment of determinants, the dependent variable was mode of delivery and independent variables were parturients' age and booking status by Robson group.

Research Ethics
The study was performed in accordance with the 2013 Helsinki declaration; study protocol was approved by the research ethics committee, FMCY, and all participants gave verbal consent for use of their data.

RESULTS
Five hundred and fifty-six (556) women delivered in the study centre during the study period.Table 1 shows the sociodemographic and obstetric characteristics of the parturients.The mean age of the women was Mean = 31.16;SD = 5.4 years.Most (54.7%) had at least two previous deliveries.Majority of them were booked (62.8%).Only 55 (9.9%) of the parturients had a previous CS.Majority (59.9%) of the women who delivered by CS in index pregnancy had an urgent CS, while another 21.9% had emergency CS.27.5%) relative contribution to overall CS rate.The relative contribution of women in group 10 was the second highest, followed by group 5 at 14.1% and 13.8% respectively.Women in group 1 also made a notable contribution to overall CS rate at 11.2%.Group CS rate was 84.1% in group 5 and 76.9%, 79.2%, 79.3%, 85.2% and 100% in groups 2, 4, 7, 8 and 9 respectively.
Table 3 shows the common indications for caesarean section by Robson groups.The commonest indication was cephalopelvic disproportion (20.1%), followed by severe preeclampsia with unfavourable cervix (12.3%).The uncommon indications for caesarean section were failed induction of labour (IOL), abnormal biophysical profile, cord prolapse with live fetus, three previous caesarean sections, triplet gestation, uterine scar rupture, retained second twin, severe oligohydramnios, poor progress in labour, delayed second stage of labour, IVF conception, fetal hydrocephalus, severe chronic hypertension and sickle cell anaemia.

Figure 1. Robson classification chart as derived from the WHO implementation manual
Table 4 shows the mode of delivery of parturients with previous CS.Of the 55 parturients that had a previous CS, 48 delivered by caesarean section, giving a repeat CS rate of 87.3%.Of the 16 parturients with one previous CS that had a trial of labour after a caesarean section (TOLAC), nine had urgent and emergency CSs and six had vaginal birth, giving a 43.8%rate of successful vaginal birth after caesarean section (VBAC).
While parturients' age was not associated with mode of delivery, booking status of parturients was significantly associated with mode of delivery (p <0.001).By Robson groups, the association of booking status with mode of delivery was seen only in Robson groups 1 (p = 0.047) and 3 (p < 0.001).See Table 5.On logistic regression analysis, booked parturients in Robson group 1 and 3 had 61.0%reduced odd (OR = 0.39; 95% CI = 0.15 to 0.99; p = 0.050) and 74.8% reduced odd (OR = 0.25; 95% CI = 0.14 to 0.45; p <0.001) of delivery by CS respectively.See Table 6.

DISCUSSION
This study found that the hospital-level CS rate at the FMCY was 48.4% in the period of the study.Parturients in Robson groups 3, 10, 5 and 1 consecutively made the highest contributions to CS rate.The commonest indication for CS was cephalopelvic disproportion, followed by severe preeclampsia with unfavourable cervix.Only 9.9% of the parturients had a previous CS.Among them, 87.3% had a repeat CS.The rate of successful VBAC following a trial of labour after one previous CS was 43.8%.Booking status was associated with mode of delivery in Robson groups 1 and 3.
The highest rates from this comparison were more from private hospital settings in Bangladesh, Brazil and Nigeria.This is largely because threshold for CSs for medically unjustifiable indications is more likely to be low in private hospitals, largely for social reasons, profit and fear of litigation that may attend adverse perinatal outcome (Akadiri et al., 2023).The variation in hospital-level CS rates within and across hospital settings and regions also reflects differences in obstetric case mix.

Robson groups with the highest contributions to overall caesarean section rate
The groups making the largest contributions to CS rate in this study; 3, 10, 5 and 1 consecutively, is similar in composition to results from previous Nigerian studies done in Ibadan (Bello and Agboola, 2022), Ilisan-Remo (Akadiri et al., 2023) and Makurdi (Ochejele et al., 2021), and from studies done in some other African countries (Mbaye et al., 2015;Tura et al., 2018;Makhanya et al., 2015), Bangladesh (Begum, 2019), andBrazil (Nakamura-Pereira et al., 2016).It is closest in both composition and order to the study done in Ethiopia (Tura et al., 2018) that reported 3, 5 and 1.
The much larger contribution from group 3 is largely a result of the predominantly multiparous obstetric population in this study, which is reflected in the combined group size of Robson groups 3 and 4, which more than doubled the size of groups 1 and 2 (45.3% vs 17.6%).Besides the contribution of the large multiparous population to CS rate, group CS rate of Robson group 3 in this study is also higher than an average expected among multipara as indicated in the Robson classification implementation manual (WHO, 2017).An attributable reason is that the study centre being a referral hospital receives parturients with labour dystocia and maternal and/ or fetal compromise which significantly increases the need to deliver by CS.
Robson group 10 made the second largest contribution to overall CS rate in this study.The contribution from group 10 is also explainable by the study centre being a referral hospital and having to deal with pregnancy-related complications leading to spontaneous or iatrogenic preterm deliveries.The size of group 10 in this study larger than an expected average of 5% (WHO, 2017), shows that a relatively large population of parturients had preterm births and the large CS rate within the group shows that most of them required CS to manage the risk of maternal and/ or perinatal mortality.
Only 9.9% of parturients had a previous CS in this study.This suggests that the CS rate within the obstetric population served by the study centre is low.This inference can also be buttressed by the size of group 5 in this study in the range of values expected for populations with low CS rates, that is, less than half of the overall CS rate (WHO, 2017).This picture is also possibly the effect of the fraction of women with previous CS who avoid hospital delivery because of the fear of a repeat CS.The practice in the study centre is that women with two or more previous CS are not allowed a TOLAC, and women are counselled on the increased risk of CS in the subsequent pregnancies after one CS.It is observed that some women who want to avoid CS at all costs visit traditional birth attendants or churches in labour, where CS can be jettisoned (Utuk et al., 2017).Some of these parturients present with uterine scar rupture as seen in the results of this study.Despite the low population of women with previous CS, the risk of a repeat CS after a previous CS is reflected in the 84.1% group CS rate of Robson group 5 and the 87.3%rate of repeat CS when those in other Robson groups with previous CS are factored in.This is attributable to a large fraction of women with one previous CS being classified as not eligible for VBAC.This is commonly because of short interpregnancy interval, suspected fetal macrosomia or other coexisting complications like prelabour rupture of membranes (PROM) and prolonged pregnancy.The 43.8% rate of successful VBAC among women with one previous CS who had a trial of labour in this study is consistent with a study in Southeast, Nigeria (Ugwu et al., 2014).
The ratio of the size of Robson group 1 to group 2 in this study (>2:1) shows that the nulliparous population were low risk, until labour and delivery.The lower contribution to CS rate from group 1 compared to 3 reflects the small size of group 1.However, group CS rate was larger in Robson group 1 compared to group 3, showing that despite the smaller group size, nulliparous women tend to be more prone to labour complications requiring CS in this study.The group CS rate in Robson group 1 was larger than an expected rate of <10%, if CS rates are to be kept under control (WHO, 2017).This higher-thanexpected CS rate within group 1 is attributable to the referral hospital status of the study centre, handling parturients with complicated labour and increased likelihood of CS.Robson groups with high group caesarean section rates Despite making lower contributions to overall CS rate, group CS rate was high within the other Robson groups in this study.This study did not subclassify Robson groups, thus limiting interpretation of high CS rate within groups 2 and 4. Nonetheless, larger sizes of 2b and 4b relative to 2a and 4a is expected to result in a high CS rate within groups 2 and 4 (WHO, 2017), as a result of more parturients with pregnancy-related complications in which pre-labour CS is a safer mode of delivery.Maternal request also drives up CS rate in this category, and in this study, it accounted for two CSs out of ten (20%) in group 2 and four out of 19 (21.0%) in group 4. Another possible driver of high CS rate within groups 2 and 4 is the possibility of not enough IOL because of preference of obstetricians for CS even when IOL is a justifiable option; perhaps due to a need to practice defensive medicine.
The combined size of groups 6 and 7 is larger than an expected average of 4% (WHO, 2017) in this study.This can be attributed to a large rate of preterm deliveries especially as the size of group 10 is also large.Preterm breech delivery has implication of a high CS rate within groups 6 and 7 as there is a concern for entrapment of the aftercoming head during vaginal delivery.Although the factors underlying the line of management of the breech presenting fetus at term was not part of the scope of this study, there is a chance that counselling and selection for external cephalic version (ECV) or assisted vaginal breech delivery (AVBD) may be affected by the obstetrician's preference between CS with a relatively lower perinatal risk and AVBD with a relatively lower maternal risk.While the latter helps reduce CS rates, the former increases CS rates and may be a contributing factor in this study.
In Robson group 8, apart from noncephalic leading twin, coexisting pregnancy related complications and previous CS also contributed to CS rate within the group.The 100% CS rate within group 9 in this study is expected by convention and is an indication of good quality data.However, the size of group 9 is larger than an expected average of 1% (WHO, 2017).There were risk factors for abnormal lie in the population studied that can possibly account for the larger than expected size of group 9, these include multiparity (including grand multiparity) and a relatively large number of preterm deliveries.

Indications for caesarean section
The five leading indications for CS in this study; cephalopelvic disproportion, severe preeclampsia with unfavourable cervix, persistent fetal heart rate abnormality, one previous CS not eligible for vaginal delivery and obstructed labour, reflects the Robson groups with the largest contribution to CS rates.Beyond indications for CS, analysis of CS based on Robson groups helped to expand the information obtainable on CS and bring to the fore the areas requiring intervention for reducing CS rates.For example, rather than focus on primigravid parturients, this study shows that cephalopelvic disproportion and obstructed labour can drive CS rates in previously parous women, especially with poorly managed labour.

Impact of parturients' age and booking status on mode of delivery
The age of parturients in this study did not show any significant association with mode of delivery.Booking status was significantly associated with mode of delivery in the population of parturients studied.Analysis based on Robson groups identified that booking status was associated with mode of delivery in Robson groups 1 and 3 but not in other Robson groups.Smaller group sizes may have prevented any association in other groups from being statistically demonstrable in this study.
Pregnant women with an unbooked status are at higher risk of pregnancyrelated complications and usually they do not set out for hospital delivery (Utuk et al., 2017).Besides, a proportion of women who received antenatal care, thus classified as booked choose to deliver in unorthodox facilities for reasons of cost or fear of caesarean section.Most of them go to traditional birth attendants' facilities or churches in labour (Utuk et al., 2017), and some of them are poorly managed.The study centre, being a referral hospital receives this category of complicated pregnancies and labour cases and most of the parturients undergo emergency and urgent CSs.This phenomenon explains the high group CS rates in groups 1 and 3, and perhaps underlie the small size of group 5, the number of preterm pregnancy complications and Apgar score <7.This is buttressed by the result of this study showing reduced odd of delivery by CS by as much as 61.0% and 74.8% in booked parturients in Robson group 1 and 3 respectively.

Implications for optimizing caesarean section
The CS rate at the Federal Medical Centre Yenagoa has risen steadily from an average of 28.6% as at 2017 (Osegi and Makinde, 2020) to an average of 42.4% as at 2020 (Allagoa et al., 2021), and to 48.4% from this study.This study directs the focus of interventions to optimize CS on parturients in Robson group 3, 10, 5 and 1. Interventions to reduce CS rates in Robson groups 1 and 3 in this population has to first address antenatal care access and utilization, then address utilization of in-hospital delivery services or access to delivery supervised by a skilled birth attendant.Particular attention must be given to multiparous women in the study setting, to seek to understand the determinants underlying their attitude to antenatal care and inhospital delivery.In addition, policies and protocols at the facility level for labour management that optimizes the chances of vaginal delivery and those that promote alternatives to CS when possible and safe, can improve CS rates in groups 1 and 3 (Makinde et al., 2020;Allagoa et al., 2021).
Besides the benefits from health information, screening, risk-based prevention and early detection and management of pregnancy complications offered by antenatal care, CS rates in group 10 can be improved by studies that explore the determinants of good maternal and perinatal outcome from vaginal delivery in women with severe preeclampsia with unfavourable cervix, building on the work of Nassar et al. (1998).Improved contraceptive uptake to help achieve adequate interpregnancy interval, improved utilization of antenatal services and improved utilization of in-hospital delivery services will help reduce repeat CS rates among parturients with one previous CS in group 5. Interventions aimed at avoiding or limiting CSs done for medically unjustifiable indications, while promoting safe alternatives to CS in Robson groups 2, 4 and 7 will further reduce CS rates.
The CS rate at the FMCY was 48.4% in the period of the study.Women in Robson groups 3 (parity ≥1, without a previous CS, singleton, cephalic, term pregnancy, in spontaneous labour), 10 (all singleton, cephalic, preterm pregnancy, irrespective of previous CS or onset of labour), 5 (one or more previous CS, singleton, cephalic term pregnancy, irrespective of onset of labour) and 1 (nulliparous, singleton, cephalic term pregnancy, in spontaneous labour), made the highest contribution to CS rate.Using the Robson classification system, audit of CS rate can be taken beyond just a list of indications and percentages, to a focus on the groups of women contributing to CS, their characteristics that determine CS rates, which can be a focus of targeted intervention to optimize CS.

Table 2
is the Robson classification report table.The overall hospital-level CS rate was 48.4%.Robson group 3 made the highest (

Table 6 : Probability of a caesarean section over vaginal delivery based on booking status of parturients in Robson groups 1 and 3 at FMCY, Nigeria Variable Regression Coefficient
*Significant