Detailseite
Projekt Druckansicht

Assessment of time-varying long-term effects of therapies and prognostic factors (T: Time-varying)

Fachliche Zuordnung Epidemiologie und Medizinische Biometrie/Statistik
Förderung Förderung von 2004 bis 2010
Projektkennung Deutsche Forschungsgemeinschaft (DFG) - Projektnummer 5470786
 
The availability of large datasets with long-term follow-up from different breast cancer studies has opened a controversial discussion on possible time-varying effects and their implication on clinical practice. Several approaches for the modeling of such time-varying effects have been proposed, but up to now, no standard procedure has been established. This project will compare some of these methods and assess their characteristics to establish these methods to practical application. We will consider proposals that build on the semi parametric Cox model, but relax the assumption of proportional hazards (PH). A special focus will lie on fractional polynomials. Other possibilities evaluated will include parametric survival and frailty models. Within breast cancer studies with long-term follow-up we will assess the different approaches. The analyses and interpretation of results will help to better understand the complex mechanisms affecting treatment response and prognosis in the framework of breast cancer.
DFG-Verfahren Forschungsgruppen
 
 

Zusatzinformationen

Textvergrößerung und Kontrastanpassung