Show simple item record

dc.contributor.authorRonoh, Alban
dc.date.accessioned2013-05-22T06:40:24Z
dc.date.available2013-05-22T06:40:24Z
dc.date.issued2009
dc.identifier.citationM.Sc (Biometry)en
dc.identifier.urihttp://erepository.uonbi.ac.ke:8080/xmlui/handle/123456789/24302
dc.descriptionMaster of Science thesisen
dc.description.abstractHIV/AIDS have remain the leading global health challenge.Its dynamics and spread is the concern of all the sectors of the society.In research, many studies continue to be carried out to really try and understand the key determinants of its dist ribution, which areas and groups are most vulnerable. This is aimed at designing effective intervention measures and seeking cure and development of a vaccine. The goal of modeling is to extract much information from the available data in order to provide an accurate representation of knowledge and uncertainty of the epidemic.Many models have been put forward to understand the level of prevalence, which include a mathematical model called t he back calculation, the WHO and UNAIDS have developed a computer program called the EPP and spectrum to provide projections and mortality due AIDS. The sentinel surveillance data from ANCs still remain the crucial source of prevalence data though they are reports suggesting that it normally overestimate the level of prevalence.Other modeling techniques can be developed to give short term projections of the prevalence level in various settings of the pandemic.Mixed effect models if well fitted can give a useful insight into the prevalence. This is where certain covariates are held as fixed while others are random for example, the rural and urban settings could be random while thinks like clinics are fixed and so on. Other time covariates should also be considered, for instance the incorporation of things like condom use, circumcision, coverage of ARVs and awareness campaigns. One of the key to modeling time varying data is the consideration of the correlation in the model.en
dc.description.abstract
dc.description.sponsorshipUniversity of Nairobien
dc.language.isoenen
dc.titleLongitudinal modeling of Hiv prevalence in Kenyaen
dc.typeThesisen
local.publisherSchool of Mathematics, University of Nairobien


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record