Show simple item record

dc.contributor.advisorLesaffre, Emmanuel
dc.contributor.advisorMarshall, Guillermo
dc.contributor.advisorCarbonez, A
dc.coverage.spatialHeverlee-Santiago
dc.creatorGarcía-Zattera, María José
dc.date.accessioned2017-03-27T22:21:43Z
dc.date.available2017-03-27T22:21:43Z
dc.date.issued2011
dc.identifierhttp://creativecommons.org/licenses/by-nc-nd/3.0/cl/
dc.identifier.urihttp://hdl.handle.net/10533/179855
dc.description.abstractOral diseases, such as dental caries, are a major health problem worldwide. Even though the prevalence of dental caries in children of Western countries has declined considerably in the last three decades, the disease has now become concentrated in a small group of children. Only a small proportion of the population experiences 50% of all caries lesions. The most likely explanation for the difference in oral health seems to be socio-economic environmental factors and it occurs during childhood. Therefore, to improve dental health, early identification of groups at a particular risk of developing caries becomes essential. The identifieation of risk groups for dental caries is challenging because often oral health data show a complex structure. Caries experience ( defined as past or present caries on teeth) data have a hierarchical structure (mouth, jaw, tooth, surface on tooth) with the lowest levels ofhighest interest to oral health researchers. This leads to the analysis of high dilnensional correlated data, because events on tooth surfaces of the same child are dependent and, therefore, the conclusions arising from statistical methods ignoring such an association may be misleading. On top of that, the cletection of dental caries might be difficult for a variety of reasons. HE-mee, miselassification of dental caries is likely to happen in practice. The fact that scoring caries is done with considerable error further complicates inference. The previous complexities of dental caries data sets, make necessary the development of adequate statistical methods that take into account all these aspects of the data at the same time in order to obtain valid inferences. The understanding of the association structure of the caries process is important for the understanding of the etiology of the disease and can help the dentists in optimizing their clinical examination of the patient and clirect preventive and restorative approaches. Motivated by dental data gathered from a longitudinal oral health study conductecl in Flanders (Belgium), the Signal-Tandmobiel® study, we have studicd the interpretation and the cffcct of thc misclassification on the association parameters associated with two models for the analysis of multivariate binary data. We have also proposed uni- and multi-variate Markov models for the analysis of longitudinal monotone binary data subject to misclassification. These models account for the effects of the covariates on the prevalences and incidences, and allow for the existence of different classifiers. Empirical and t heoretical evidence are provided to show that the model parameters can be estimatecl from the main data \Vithout the need of external information on the misclassification pa.rameters. In the multivariate Markov model, the joint distributions are ddined through the specification of compatible full conditional distributions. The proposed multiva.ria.te hiclden Ma.rkov moclel permits the stucly of the within- and acrosstime a.ssocia.tion para.meters a.mong the responses.
dc.language.isoeng
dc.relationinstname: Conicyt
dc.relationreponame: Repositorio Digital RI2.0
dc.relationinstname: Conicyt
dc.relationreponame: Repositorio Digital RI2.0
dc.rightsAtribución-NoComercial-SinDerivadas 3.0 Chile
dc.titleMultivariate models for the analysis of caries experience data subject to misclassificatión
dc.typeTesis Doctorado
dc.description.degreeDoctor en Ciencias
dc.contributor.institutionPontificia Universidad Católica de Chile
dc.description.statusTERMINADA
dc.country.isobel-chi
dc.description.conicytprogramPFCHA-Becas
dc.description.pages193p.
dc.relation.projectidinfo:eu-repo/grantAgreement/PFCHA-Becas/RI20
dc.relation.setinfo:eu-repo/semantics/dataset/hdl.handle.net/10533/93488
dc.rights.driverinfo:eu-repo/semantics/openAccess
dc.type.driverinfo:eu-repo/semantics/doctoralThesis
dc.relation.programhandle/10533/108040
dc.description.shortconicytprogramPFCHA-Becas
dc.type.tesisTesis
dc.type.openaireinfo:eu-repo/semantics/publishedVersion


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record