| Autor: |
Munda, Marco, Legrand, Catherine |
| Přispěvatelé: |
UCL - SSH/IMMAQ/ISBA - Institut de Statistique, Biostatistique et Sciences Actuarielles |
| Jazyk: |
angličtina |
| Rok vydání: |
2013 |
| Předmět: |
|
| Popis: |
Multicentre clinical trials allow the required number of participants to be recruited within a fairly short time frame. Conducting a clinical trial at more than one centre also deepens the generalisability of its findings. But the other side of the coin is that this raises the issue of whether and how to adjust for centre effects in the statistical analysis. In this paper, we discuss a number of strategies for time-to-event outcomes. We balance the conventional fixed effects and stratified Cox models against the frailty modelling approach with the objective to provide a clear picture of the weaknesses of the current practice, and to encourage the use of frailty models. A special attention is paid to the problem of misspecification of the frailty distribution. Based on simulations, we illustrate the performances of the frailty model over its competitors, and we argue that it can generally be used advantageously to draw inferences about the treatment effect, even if the frailty distribution is misspecified. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
