Statistical Analysis of Familial Aggregation of Adverse Outcomes

Luise Cederkvist Kristiansen

AbstractIn survival analysis, the survival times of the subjects in the study population are generally assumed to be statistically independent, conditional on the covariate information. However, situations where the survival times are correlated due to a natural clustering of the study subjects may arise. In this study, di fferent statistical methods for analysis of clustered survival data are evaluated and compared using data from a Danish register-based family study of the psychological e ffects of exposure to childhood cancer. In addition to assessing the eff ect of exposure to childhood cancer whilst coping with familial clustering, two of the presented models are applied in order to estimate familial correlation of ages at onset. The models show that individuals diagnosed with cancer and individuals with a family history of admission have an increased hazard rate. Furthermore, a signi ficant correlation of age at onset within families is identifi ed. Of the models presented, the shared gamma frailty Cox proportional hazards model and the Clayton-Oakes copula model with the marginal Cox proportional hazards model as margin are the most applicable in this study.
TypeMaster's thesis [Academic thesis]
Year2012
PublisherTechnical University of Denmark, DTU Informatics, E-mail: reception@imm.dtu.dk
AddressAsmussens Alle, Building 305, DK-2800 Kgs. Lyngby, Denmark
SeriesIMM-M.Sc.-2012-13
NoteSupervised by Professor Per Bruun Brockho ff, pbb@imm.dtu.dk, DTU Informatics, and co-supervised by Klaus Kaae Andersen and Kirsten Frederiksen, SBR
Electronic version(s)[pdf]
Publication linkhttp://www.imm.dtu.dk/English.aspx
BibTeX data [bibtex]
IMM Group(s)Mathematical Statistics