Determinants of Cesarean Section Deliveries in Public Hospitals of Addis Ababa, Ethiopia, 2018/19: A Case-Control Study

Objective The objective of this study was to assess the determinants of cesarean section deliveries in public hospitals of Addis Ababa, Ethiopia, 2019. Method A hospital-based unmatched case-control study was conducted to study 780 (260 cases and 520 controls) women who delivered in public hospitals of Addis Ababa from August 22 to September 20, 2019. The cases were all mothers who delivered through caesarean section, and controls were all mothers who delivered vaginally in the same time in the study area. Data were collected from the randomly selected women and looking into their cards. Data were entered on EpiData 3.1 and exported to SPSS version 20 for cleaning and analyzing. Binary logistic regression and AOR with 95% CI were used to assess the determinants of caesarean section. Results Majority of the study participants were in the age category 20–34 years. Nearly more than 1/3rd of the participants (32.7% cases and 34.6% controls) have attended primary school. Most of the cases 217 (83.5%) and few of the controls 21 (4%) possess previous caesarean section. One hundred three (52.3%) of the cases and 329 (63.6%) controls were multi-parous. Previous caesarean delivery (AOR = 6.93, 95% CI; (3.39, 14.16)), singleton pregnancy (AOR = 0.34, 95% CI; (0.12, 0.83)), birth weight less than 2500 gm (AOR = 0.29, 95% CI; (0.18, 0.92)), birth weight greater than 4000 gm (AOR = 16.15 (8.22, 31.74)), completely documented partograph (AOR = 0.13, 95% CI; (0.078, 0.23)), and pregnancy-induced hypertension (AOR = 2.44, 95% CI; (1.46, 4.08)) were significant determinants of caesarean delivery in this study. Conclusion Previous caesarean section, number of delivery, birth weight, partograph documentation, and pregnancy-induced hypertension had significant association with caesarean section delivery in this study.


Introduction
e proportion of caesarean section (CS) to the total births is considered as one of the important indicators of emergency obstetric care [1]. Currently, the WHO states that CS has paramount importance on reducing maternal and perinatal mortality and morbidity, provided that there is justified medical indication. However, in the absence of clear medical justification, CS has no medical benefit, rather it is associated with short-and long-term health risks as compared with vaginal delivery [2].
CS rates have been rising progressively worldwide with a wide variation between countries and regions. It is globally known that CS delivery can present several risks than a vaginal birth even if it has become nowadays one of obstetric intensive cares regarding its benefits for mother and baby [3].
ough the rate of CS is increasing in a normal pregnancy, CS has eight-fold higher maternal mortality and 8-12 times higher morbidity than vaginal delivery [11,12]. Nowadays, the CS rate has become more prevalent without a clear medical justification though it is associated with an increased morbidity and mortality of mothers and children [13][14][15][16][17].
In 2015, the WHO suggested CS can save the life of the mother and infant as well only when it is medically justified [2]. e Ethiopian Demographic Health Survey (EDHS) 2016 reported over utilization of CS rates in Addis Ababa (21.4%) [18]. However, there is uncertainty if the determinants leading to CS delivery are up to the WHO recommendation. us, this study is intended to determine what factors are really predisposing mothers to CS delivery. Moreover, this study would help to have an extensive and up-to-date picture on the problem.

Methods.
A quantitative facility-based unmatched casecontrol study design was applied to study a total of 780 participant mothers from August 22 to September 20, 2019. e mothers were interviewed after they gave birth either by spontaneous vaginal delivery or caesarean section. Mothers who delivered a baby after fetal viability (28 weeks) were included, while those who were admitted to the postnatal ward after home delivery were excluded from this study. e sample size was calculated using Epi Info version 3.5.1 statistical software using the double population formula for the unmatched case-control study. Hence, considering the previous exposure of macrosomia baby (p1 � 52% for cases and p2 � 1.6%) and 10% nonresponse rate, the calculated value was 260 for cases and 520 for controls. Finally, the total number of samples was 780 mothers. e study was conducted on six public hospitals administered by Addis Ababa Health Bureau. us, the sample size was allocated proportionally based on the average delivery load in the past three months (February, March, and April). Cases and controls were selected exclusively using the systematic random sampling technique. e data were collected using a questionnaire and checklist. e questionnaire was structured, pretested, translated, and adapted from previous articles. Medical records of parturient mothers were used to extract information such as partograph, gestational age at delivery, fetal presentation, indication of CS, and birth weight of the fetus, while the information of the other variables were primary data taken from interview of postnatal mothers. e data for this study were collected after ethical clearance was obtained from Gamby Medical and Business College and Addis Ababa public health research and emergency management care process. e participants were informed about the objective clearly and were also informed that the information they provided is only used for research purposes and kept confidential. Finally, the collected data were entered on EpiData3.1 and exported to SPSS version 20 for analysis. e variables with p < 0.25 on bivariate logistic regression were taken to multivariable logistic regression to control possible confounding factors. Finally, adjusted odds ratio with 95% confidence interval was used to measure strength of association between the predictors and occurrence of CS. Statistical significance was declared at p < 0.05.
Cases are mothers who delivered by caesarean section, whereas controls are those who delivered vaginally.
In this study, complete and partial partograph documentation is defined as if all the component of partograph was correctly filled and some components were missed, respectively, at the time of reviewing mother's card. Meanwhile, partograph is reported as undocumented if and only if the labor was not followed by Partograph at all.
Prolonged premature rupture of membrane is defined as duration of rupture of the membrane greater than 12 hours before the onset of labor [19].
Duration of labor: the duration of labor in this study is measured from the onset of true labor including the latent phase of first stage of labor to third stage of labor.

Socioeconomic and Demographic Characteristics.
e study recruited 780 mothers (260 cases and 520 controls) making a response rate of 100%. Majority of the participants (82.3% cases and 86% controls) were in the age category of 20-34 years. Nearly one third of enrolled mothers (32.7% cases and 34.6% controls) have attended primary school (Table 1).

Obstetric Characteristics of the Participants.
More than half of cases and nearly 2/3 rd of controls were multi-parous. Twenty-seven (10.4%) of cases and fifty (9.6%) of controls had previous stillbirth. Eighty-nine (34.2%) of cases and fifty-two (10%) of controls had faced fetal distress during labor. Most (83.5%) of the cases and 4% of controls had previous CS delivery ( Table 2).
Fully documented partograph was higher in controls than cases, whereas more proportion of partially and not documented partograph was higher in cases than controls ( Figure 1).
In Table 3, the Robson classification is used to compare the risk of delivering via CS among the ten groups of the delivering women. In this study, 2/3 rd (67.4%) of group five and more than half of group eight (54.5%) and group two (51.5%) of the Robson classification were delivered through CS delivery (Table 3).

Discussion
CS is a life-saving procedure. us, the procedure should not be done without clear medical indication. In this study, mothers who underwent previous CS were 6.93 times more likely to deliver by CS for the consecutive birth compared with those who gave birth via vaginal delivery. is finding is consistent with the studies conducted in Bahir Dar, Harar, Addis Ababa, Mekelle, and Dessie [20][21][22][23][24]. is might be due to the fear of uterine rupture associated with waiting longer time. Pregnancy-induced hypertension during labor was found risky for CS delivery. Hence, it is similar with the study conducted in Bukavu Provincial Hospital in Democratic Republic of Congo [25]. e possible reason could be long duration of vaginal delivery in case of pregnancy-induced hypertension.
In this study, neonatal birth weight ≥4000 gm was more likely to deliver through CS than birth weight of 2500-4000 gm. Perhaps, it is in line with the study done in eastern Ethiopia; Felege Hiwot Hospital, Oman, and DR Congo [20,25,26]. is is because macrosomia babies would face birth difficulty if vaginal delivery is allowed.
is study showed that completely filled partograph was 87% less likely to go to CS room than undocumented partograph. Studies conducted in Adigrat and Mekelle agreed with this study [23,27]. is might be due to close monitoring of partograph decreasing the likelihood of CS rate.
Furthermore, singleton pregnancy was found as a protective factor for caesarian delivery.

Limitations of the Study
(i) Some very important factors like maternal height and body mass index (BMI) were not included as they are not routinely registered on medical cards at the setups where the research was conducted (ii) Scare to add qualitative part would not have given a chance for women to discuss more information about the decision-making process (attitude and perception of physicians and family)

Conclusion
In this study, only the obstetric characteristics were significant. Previous CS, number of delivery, birth weight, partograph documentation, and pregnancy-induced hypertension had significant association with CS delivery. Future studies need to examine the attitude of service providers and their influence on the growing CS delivery rate.
Data Availability e datasets during and/or analyzed on the current study are available from the corresponding author on reasonable request.

Consent
Informed, written, and signed consent was obtained from each head of every facility involved in the study and participant mothers after the purpose and benefits of the study are discussed. Participants were informed about the minimal risk that it had in taking part in the study, their volunteerism, and the right to leave the interview at any time they want. Confidentiality of the study participant's information was also kept secure.

Conflicts of Interest
e authors declare that they have no competing interests.

Authors' Contributions
AG proposed and designed the study and developed the manuscript. TK supervised, advised, and helped in the analysis. GG assisted interpretation of the result and preparation of the draft of the manuscript. FT assisted on proposal development and the final write up of the data collection tool. NE performed the statistical analysis of the study. All authors read and approved the final manuscript.