Εντοπίστηκε ένα σφάλμα στη λειτουργία της ΠΥΞΙΔΑΣ όταν χρησιμοποιείται μέσω του προγράμματος περιήγησης Safari. Μέχρι να αποκατασταθεί το πρόβλημα, προτείνουμε τη χρήση εναλλακτικού browser όπως ο Chrome ή ο Firefox. A bug has been identified in the operation of the PYXIDA platform when accessed via the Safari browser. Until the problem is resolved, we recommend using an alternative browser such as Chrome or Firefox.
Λογότυπο αποθετηρίου

Μεταπτυχιακές Εργασίες

Μόνιμο URI για αυτήν τη συλλογήhttps://pyxida.aueb.gr/handle/123456789/15

Περιήγηση

Πλοήγηση Μεταπτυχιακές Εργασίες ανά Θέμα "Adaptive design"

A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω
Φιλτράρετε τα αποτελέσματα πληκτρολογώντας τα πρώτα γράμματα
Τώρα δείχνει 1 - 1 από 1
  • Μικρογραφία εικόνας
    Τεκμήριο
    Adaptive clinical trial designs for survival outcomes testing the proportionality of hazards assumption
    (09/11/2018) Stavropoulos, Charalampos; Σταυρόπουλος, Χαράλαμπος; Athens University of Economics and Business, Department of Statistics; Karlis, Dimitris
    At this thesis we will examine a widely used assumption in the context of clinical trials, the assumption of proportional hazards. This assumption is about the relationship between the hazard functions of the groups that participate in the trial. More specifically, it is assumed that the ratio of hazard functions is constant through time. Based on that assumption, the researchers can use the Log Rank test and estimate the sample size that is needed. However, this is a very strict assumption and in this thesis we investigate its impact on the trial when it does not hold, in terms of power and sample size estimation. As an alternative, we investigate the Restricted Mean Survival Time (RMST), which is the mean survival time up to a certain point. We will use the difference between RMSTs for testing the difference between survival groups and we will compare this method to the Log Rank test under cases of proportional and non-proportional hazards. The comparison will be in terms of sample size estimation and power. Finally, we will provide a new clinical trial design that will start with the Log Rank test and at a certain point will test the proportionality assumption. If the assumption is rejected, the trial will adapt to testing the difference between RMSTs. The adaptive design’s performance will be compared to that of a simple design which does not test the proportionality assumption and uses the Log Rank test. The comparison will be in terms of sample size estimation and power.