In this paper, we propose a survival modeling approach to promoting ranking diversity for biomedical information retrieval. The proposed approach concerns with finding relevant documents that can deliver more different aspects of a query. First, two probabilistic models derived from the survival ana