Back to browse results
Multilevel Logistic Regression Analysis of the Determinants of Stillbirth in Ethiopia Using EDHS 2011 Data
Authors: Kidanemariam Alem Berhie, and Habtamu Gebremariam Gebresilassie
Source: Turkiye Klinikleri Journal of Biostatistics, 9(2):121-42; DOI: 10.5336/biostatic.2016-54437
Topic(s): Antenatal care
Country: Africa
Published: JUL 2017
Abstract: Objective: Stillbirth is often defined as fetal death after 24 weeks of gestation, but a fetus greater than any combination of 16, 20, 22, 24, or 28 weeks gestational age and 350g, 400g, 500g, or 1000g birth weight may be considered stillborn depending on local law. Once the fetus has died, the mother may or may not have contractions and undergo childbirth or in some cases, a Caesarean section. Most stillbirths occur in full-term pregnancies. If all causes of stillbirth are taken together, the new estimates would place stillbirths fifth on the list of causes of deaths (COD) worldwide. Each year, about 3 million families worldwide will experience a stillbirth. This study has intended to model determinants of experiencing stillbirth among women in child bearing age group of Ethiopia using the Ethiopian demographic and health Survey data. Material and Methods: First, the bivariate chi-square test of association was fitted to the data and significant variables were considered for further investigation in binary logistic regression model. Lastly, the multilevel models were fitted. Results: This study revealed that the rate of experiencing stillbirth among women of child bearing age was about 25.5 per 1000 deliveries in Ethiopia. From binary logistic regression, region of residence, maternal age, place of residence, education level, parity, antenatal care utilization, place of delivery, body mass index (BMI) and anemia level were found to be significantly associated with experiencing stillbirth. From multilevel logistic regression, it was found that the random intercept model provided the best fit for the data under consideration. The variance of the random component related to the intercept term was found to be statistically significant implying differences in prevalence of experiencing stillbirth among the regions. And in this random intercept model, age group, type of place of residence, antenatal care visit and delivery place were found to be statistically significant factors for experiencing stillbirth among regions. Conclusion: Mothers should prefer and people who are around them should advise them to give birth at health centers than delivering at home. Especially older age women, above 35 years, should be more careful for difficulties that come with age, like hypertension and should visit antenatal care during pregnancy. Further studies should be conducted to identify other correlates of stillbirth that are not included and confirm the variables which are insignificant in this study because of many reasons and since regional variation are found significant spatial models can be applied to investigate spatial variations of experiencing stillbirth. Keywords: Stillbirth; antenatal care visit; multilevel logistic regression; Ethiopia