NEONATAL DEATHS IN EMPOWERED ACTION GROUP STATES: MULTILEVEL MODELING TO STUDY THE WITHIN DISTRICT EDUCATION EFFECT ON NEONATAL SURVIVAL

Abstract

Neonatal Deaths are deaths that occur during rst 28 days of life. It is still very high in demographically poor performing States in India termed as Empowered Action Group (EAG) States. Study using the National Family Health Survey-4 data suggests high neonatal death burden in EAG States including Odisha. Uttar Pradesh recorded the highest neonatal death at 51.6 deaths per 1000 live births and the lowest was recorded in Uttarakhand at 32.3 neonatal deaths per 1000 live births. Neonatal deaths in Chhattisgarh were recorded as 48.2 deaths per 1000 live births, followed by Odisha at 46.9, Madhya Pradesh at 44.9, Bihar at 41.9, Jharkhand at 39.2, Assam at 36.1 and Rajasthan at 34.4 per 1000 live births. The parametric Cox Hazard model was used to study the unobserved heterogeneity. The Weibull distribution shape parameter value greater than one suggests considerable unobserved heterogeneity in the data. Considering district as unit of analysis (group level variable), attempt was made to capture the between district individual variations in neonatal deaths with regard to women educated 12+grades within districts. The low levels of Intra Class Correlation are suggestive of individual level variations than group level differences in studying the variations in neonatal deaths.