Random Effects Cox Models: A Poisson Modelling Approach

We propose a Poisson modelling approach to random effects Cox proportional hazards models. Specifically we describe methods of statistical inference for a class of random effects Cox models which accommodate a wide range of nested random effects distributions. The orthodox BLUP approach to random effects Poisson modeling techniques enables us to study this new class of models as a single class, rather than as a collection of unrelated models. The explicit expressions for the random effects given by our approach facilitate incorporation of relatively large number of random effects. An important feature of this approach is that the principal results depend only on the first and second moments of the unobserved random effects. The application of proposed methods is illustrated through the reanalysis of data on the time to failure (tumour onset) in an animal carcinogen esis experiment previously reported by Mantel and Ciminera (1979).