The study was conducted among health facilities in Bahir Dar city. Bahir Dar is the capital city of the Amhara National Regional State in the Federal Democratic Republic of Ethiopia. According to the Amhara Bureau of Finance and Economic Development (BOFED), the population of Bahir Dar city was estimated to be 339,683. Among these, 156,376 of them are females. The city has one specialized, one referral and one primary government hospitals (Tibebe Giwon, Felege Hiwot,and Adiss Alem respectively), 11 health centers (including one private health center), 10 health posts and one family guidance association clinic, 4 private general hospital, and 35 medium private clinics, among all 4 public and 5 private health facilities provide Emergency Obstetric and Newborn Care service. According to the Bahir Dar city zone health department 2010 E.C report, there were 15,208 annual deliveries and among this 4,160 had CS deliveries (19).
Study Design and period
An institutional-based comparative cross-sectional study was conducted from March 1 to April 15, 2019.
All Women who gave birth in public and private health facilities of Bahir Dar city were the source of population. Women who gave birth in selected public and private health facilities in Bahir Dar city during the study period was the study population
Sample Size Determination
Sample size using double proportion formula
The estimated sample size was calculated considering CS proportion( public facility 34% and private facility 47%) from previous study finding in Addis Ababa ((17) by using a double proportion formula(N (in each group) = (p1q1 + p2q2) (f (a,b)) / ((p1 - p2)². Taking assumptions of power 80, design effect 1.5 then after the calculated sample size was 657. The sample size for factors associated with CS delivery was also calculated using Epi info version 7.2.2 and found to be less than the sample size for the proportion of CS.
Therefore, by adding a possible 10% none response rate the estimated sample size for this study was 724 (362 and 362 subjects study for the public and private facility).
A multistage systematic random sampling procedure use used. From a total of 9 health facilities (4 public and 5 private) which provide a comprehensive obstetric and newborn care in Bihar Dar city, 3 public and 4 private health facilities were selected using random sampling and included in the study. The required sample size from each group was proportionally allocated using stratified sampling for selected health facilities in each group, based on the previous year's week's average number of client flow. The six-week average client flow of the selected health facilities, Flege Hiwot Referral Hospital, Addis Alem hospital, Bahir Dar health center, GAMBY hospital, Mari stop, Addinas General hospital, and Dr. Amiro MCH specialty clinic were 521,254,363,72,393,173 and 124 respectively. The proportional allocation was done for each facility in each stratum. Systematic random sampling was used to select each study subject. The first case was randomly selected after calculating the interval for both public and private facility and then every 3rd case for public and every 2nd case for private health facilities were selected from the delivery record till the required sample size was achieved for each facility.
Medical factors:- Obstructed Labor, Cephalo-pelvic disproportion, Antepartum hemorrhage, Previous cesarean section scar, Mal-presentation, Preeclampsia/ eclampsia syndrome, Failure to progress, Failed induction, Suspected uterine rupture, Cord prolapsed, Non-reassuring fetal heart rate pattern and post-term (2).
Cesarean delivery on maternal request: - defined as a primary cesarean delivery done on request from the mother in the absence of any medical or obstetric indication(6).
Data Collection Tool and Procedure
The questionnaire and checklist were adapted through reviewing of different works of literature and previous similar studies (17, 20-24). The questionnaire was initially prepared in English, then translated to Amharic, and then translated back into English to check for consistency. A structured questionnaire was used to collect the data through a face to face interview and checklist for reviewing client charts. The Amharic version of the questionnaire was used for data collection. Main variables included in the questionnaire and checklist for assessment, including socio-demographic characteristics of the respondents, previous and current pregnancy history, indication and fetal conduction.
The data was collected by trained seven diploma midwives working in obstetric wards of other health facilities. The data collection process was supervised by two BSc holder senior staff working in the obstetric department. The participants were interviewed after they gave birth and stable to communicate. The checklist was filled after the delivery summary was written. The completed questionnaires were collected every day after checking the completeness, consistency of the data and by providing timely feedback by supervisors.
Data Quality Assurance
To maintain quality of the data, data collectors and supervisors were trained for two days on the objective of the study, the content of the questionnaire, how to fill the questionnaire, respondent rights, informed consent, and technique of interview and how to keep confidentiality and privacy of the study subjects. Before one week of the actual data collection period, the data collection tools were pretested 36 individuals in Tibebe Giwon and Dream Care Hospitals, thereby possible adjustment or modification was made. The principal investigators & supervisors gave feedback and correction daily for the data collectors. The data was cleaned, coded and entered to Epi info window version 7.2.2.
Data Processing and Analysis
Each completed questionnaire was coded on a pre-arranged coding sheet by the principal investigator to minimize errors. Data were entered into a computer using Epi info window version 7.2.2 statistical programs. The data were cleaned accordingly and then exported to statistical package for social science (SPSS) version 23.0 the data were cleaned and analysis was done. Frequencies and percentages were generated. The findings presented by text, tables and graphs and Chi-square tests.
Initially, bivariable logistic regression analysis was performed between the dependent variable and each of the independent variables. Then all variables from bivariable logistic regression analysis found to be statistically significant at (at p-value<0.05) were fitted into the multivariable logistic regression model to control possible confounder. Adjusted odds ratio (AOR) with 95% confidence interval (CI) was used to measure the strength and significance of the association. If P-value <0.05, indicating the presence of a statistically significant association between cesarean section delivery and independent variables.