Beta 1


Title Describing hypoglycemic events using extensions of the Cox model
Author Nielsen, Rebecca Zachariae
Davidsen, Charlotte Kampp
Supervisor Brockhoff, Per B. (Mathematical Statistics, Department of Informatics and Mathematical Modeling, Technical University of Denmark, DTU, DK-2800 Kgs. Lyngby, Denmark)
Institution Technical University of Denmark, DTU, DK-2800 Kgs. Lyngby, Denmark
Thesis level Master's thesis
Year 2011
Abstract This thesis deals with statistical modeling of hypoglycemic events experienced by diabetics. Novo Nordisk has provided the data which contains information about diabetes patients assigned one of the two insulin products BIAsp30 and Insulin Glargine. The main purpose of the study is to investigate if there is any difference between the two insulin products in regard to the hypoglycemic event rate. Furthermore an association between other possible explanatory variables and the hypoglycemic event rate are tested. To explore the hypoglycemic event data this thesis deals with a variety of statistical survival methods both classical and extended. The applied models are the ordinary Cox model, Andersen-Gill model, Conditional Andersen-Gill model, Frailty model and Conditional Frailty model. The models are thoroughly introduced and then applied. The main purpose of implementing the recurrent event models is to investigate whether event dependence and/or heterogeneity between patients explain the variety in the hypoglycemic event rate. The ordinary Cox model revealed that a patient’s risk of experiencing a hypoglycemic event was 15% lower assigned to Insulin Glargine compared to BIAsp30. However, the difference was not statistical significant. When applying the Conditional Frailty model, including event dependence and heterogeneity between patients, the treatment effect turned out to be significant having a 22% lower risk of hypoglycemia for patients utilising Insulin Glargine instead of BIAsp30.
Imprint Technical University of Denmark (DTU) : Kgs. Lyngby, Denmark
Series IMM-M.Sc.-2011-05
Fulltext
Original PDF ep11_05.pdf (1.66 MB)
Admin Creation date: 2011-02-07    Update date: 2011-02-07    Source: dtu    ID: 274780    Original MXD