What is the rationale of using three different types of log-rank equivalent tests (Mantel-Cox, Breslow, and Tarone-Ware). How can I examine the confounding effects of other factors when running the survival analysis?