RISK FACTORS FOR ANEMIA LEVELS AMONG PREGNANT WOMEN IN ETHIOPIA: ORDINAL LOGISTIC REGRESSION AND MULTILEVEL MODELS

KASSAHUN, ANIMUT

dc.contributor.author	KASSAHUN, ANIMUT
dc.date.accessioned	2019-09-25T05:33:04Z
dc.date.available	2019-09-25T05:33:04Z
dc.date.issued	2019-09-25
dc.identifier.uri	http://hdl.handle.net/123456789/9758
dc.description.abstract	Abstract Background: Lower concentration of the blood hemoglobin than the normal level is defined as anemia. It is a public health problem for all countries having different income levels with significant consequences of health and economic growth. About 95.7% of the global load of prenatal anemia is originated from developing countries. The main purpose of this study was to identify factors of anemia levels among pregnant women under ordinal logistic regression models by considering regional discrepancy of anemia levels in Ethiopia. Methods: The EDHS 2016 conducted by the CSA of Ethiopia was used as main data source. The survey considered 1122 pregnant women at the reproductive age; of which 1053 pregnant women with complete anemia levels were included in this study. Ordinal logistic regression and multilevel models were employed to explore factors of anemia levels among pregnant women. Results: From the total pregnant women included in this study 3.04% were severe, 20.32% were severe or moderate and 37.51% were severe, moderate or mild anemic. Under single level ordinal analysis partial proportional odds model was best fitted the pregnant women data. Ordinal logistic regression model the best model for the data as PPOM was selected as it had smallest AIC value. The anemia levels of pregnant women vary among regions of Ethiopia and random intercept multilevel PPOM was best fitted the data. The intercept variance decreased from 0.531 in null to 0.165 in random intercept PPOM indicating the predictive power of the predictors. Rural pregnant women were more anemic, taking iron was 3.71 times decreased the risk of anemia. Education level had invers relation with risk of anemia while parity had direct relation. Conclusions: The prevalence of anemia levels of pregnant women among regions indicated that the proportions of severe anemia were 10.98%, 4.08% and 3.51% in pregnant women from Somali, Dire Dawa and Afar respectively. Education, iron take, wealth index, residence, births in 5 years, parity, visit health facility and antenatal visit were significant factors of anemia level. Polices and strategies established by the government will better to enhance women education by considering regional variability of anemia and women should participate in different decisions, which affects their health status due to anemia. Keywords: Anemia levels, Ordinal logistic regression, PPOM, Ethiopia, Multilevel, Region	en_US
dc.language.iso	en	en_US
dc.subject	Statistics	en_US
dc.title	RISK FACTORS FOR ANEMIA LEVELS AMONG PREGNANT WOMEN IN ETHIOPIA: ORDINAL LOGISTIC REGRESSION AND MULTILEVEL MODELS	en_US
dc.type	Thesis	en_US