Please use this identifier to cite or link to this item: http://archive.cmb.ac.lk:8080/xmlui/handle/70130/5462
Title: Model to Identify the Affect of Water Quality Parameters on Incidence of Hepatitis.
Authors: Madhubani, Peshala
Sooriyarachchi, M.R.
Keywords: Incidence, ordinal scale, Spearman’s correlation coefficient, Successive structure, Correlation, Forward selection
Issue Date: 2013
Publisher: CAFET-INNOVA TECHNICAL SOCIETY
Citation: Peshala Madhubani and Sooriyarachchi, M.R. (2013) Model to Identify the Affect of Water Quality Parameters on Incidence of Hepatitis. International Journal of Earth Science and Engineering (IJEE), India 6(4) : 751-756
Abstract: Viral Hepatitis is one of the main water born diseases and occurs primarily in third world countries. Hepatitis has been endemic in the country for many years. Several recent outbreaks were reported in last few years in Sri Lanka. The dangerous aspect of the disease is, more than 90% of the infected were children. The organisms are transferred through contaminated food, water and feces of infected people. Through the literature review, it was found that there were no broad research studies on viral hepatitis in Sri Lanka. Contaminated water is one of the major causes of water born disease. It is vital to identify the main water quality parameters which affect Hepatitis. The purpose of the research was to identify the factors affecting the incidence of Hepatitis. As the number of hepatitis cases are ordinal scale and the water quality parameters are continuous the Spearman’s correlation coefficients were calculated to explore the primary association of hepatitis cases and water quality parameters. Incidence of hepatitis was modeled as a function of explanatory variables in this research. Since data are collected on the same units (districts) across the successive structure and the responses are correlated within the cluster (district) and time (month), Generalizing Estimating Equation (GEE) methodology was used. The Negative binomial distribution was chosen for the response which is appropriate in cases such as when there are an excess of zeros. Significant variables for the model were selected by considering the p-value associated with the Wald test statistic, while following forward selection procedure. The parameter estimates of the model produces the contribution of each variable to the log of expected number of hepatitis cases recorded in each district of the country throughout the year.
URI: http://archive.cmb.ac.lk:8080/xmlui/handle/70130/5462
Appears in Collections:Department of Statistics

Files in This Item:
File Description SizeFormat 
02060431-Peshala_Sooriyarachchi_Full_Paper.pdf359.56 kBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.