الفهرس 
INTRODUCTIONOnly 14 pages are availabe for public view
 |  | from | 145 |  |  |
| | | from | 145 | | |
Abstract
Caesarean section delivery has a significant role in reducing maternal and child death. However, unnecessary utilization has adverse health effects. The World Health Organization (WHO) reported that the optimal proportion of CS should range between 10-15%. Globally, Egypt has the third highest rate of CS delivery.
The present study was conducted to examine the spatial patterns of the caesarean section delivery in Egypt by identifying the clustering of high values (hot-spots) or low values (cold-spots), and to apply mixed effect modelling and Generalized Estimating Equation to determine the best approach to develop a multilevel model for the factors associated with caesarean section delivery in Egypt.
The study was based on secondary data analysis of EHIS 2015 which was cross-sectional study. A total of 5550 women aged 15-49 years who had children ever born, and had heard about caesarean delivery from 25-governorate was included in the study. Women were level-1 observations and governorates as a group of analysis (level-2).
The study revealed the following main finding:
Concerning overall Egypt, 43.8% of deliveries were CS delivery. Demietta had the highest weighted percentage of caesarean delivery (64.8%) while Suhag had the lowest (28.4%). Concerning urban areas, all the governorates had a proportion above 30%. Regarding rural areas, only one governorate had a proportion mostly near 15% (Matrouh).
The hot-spot analysis revealed that, the 99% confidence significant hot-spots are found at one governorate (Demietta). On the other hand, the 99% confidence significant cold-spots are found at 2 governorates (Qena and Luxor). The rural areas didn’t observe any hot-spots or cold-spots.
The weighted mean age of women was 33.47 ± 8.11 years, more than half (66.5%) of women had higher education, 22.3% of women had middle wealth and women who had more than one CS constituted 43.8% of the sample.
A multivariable analysis showed that the most significant factors related to CS delivery in mixed-effect approach were age, education, wealth index, BMI, parity, and governorate-level poverty rate. In the GEE approach; the associated factors were the same as the mixed-effect model. In the mixed-effect model, only women’s education had a significant indirect effect through the attitude, while in the GEE model the women’s education and parity had a significant indirect effect. The percentage correctly classified of CS delivery in the current study was 65.8% for the mixed-effect model and 64.27% for GEE model.
For the whole Egypt, 62.3% of women had all of their births via CS in 2015. The Generalized Estimating Equation (GEE) with Poisson distribution was considered to identify the factors associated with a greater rate or lower rate of CS. The factors that were associated to greater rate of CS are age (Young), education (High), wealth index (High), parity (primipara or multipara), timing of CS delivery (Early in pregnancy), previous CS, and multiple birth.
6.2. Conclusion
from this study we can conclude the following:
o The prevalence of CS delivery in Egypt was higher than that has been recommended by WHO.
o Demietta is the highly significant hot spot while Qena and Luxor are the highly significant cold-spots.
o On the women-level, women’s age, education, wealth index, BMI, parity, and attitude toward CS delivery were significant predictors of mode of delivery.
o On the governorate-level, governorate poverty rate was significant predictor of mode of delivery.
o Attitude toward CS delivery mediate the relation between women’s education and CS delivery.
o On the women-level, women’s age, education, wealth index, parity, timing of CS delivery, previous CS, and multiple births were significant associated with a greater rate of CS.
o Attitude toward CS delivery doesn’t mediate the relation between women’s personal characteristics and CS rate.
o Choosing between mixed-effect or GEE model depends on the goals of the analyses. Knowing the assumptions of each method and how these assumptions affect the inferences from the analysis will enable researchers to determine the best approach to analyzing their data.
6.3. Recommendations
o Simulation study is needed to compare the mixed-effect and GEE approaches using different values of Intraclass correlation (ICC) and working correlation matrix.
o Simulation study is needed also to compare the mixed-effect and GEE approaches when adding mediation.
o Researchers are advised to use the mixed-effect model if they want to assess the variation between clusters (level-2).
o The GEE is recommended if the aim is estimation of the effects of individual-level risk factors while adjusting for between cluster heterogeneity.
o It is recommended to use GEE approach if they know the correlation matrix within the cluster.
o Clinics should be audited regularly, and there must have some strict instructions and proper guidelines when conducting CS is justifiable.
o Intervention programs are needed to raise public awareness regarding the harmful consequences of unnecessary CS delivery by considering the founded associated factors.