Adaptive Conditional Hazard Regression Modeling of Multiple Event Times

Recurrent event time data and more general multiple are commonly analyzed using extensions of Cox regression, or proportional hazards as used with single data. These methods treat covariates, either time-invariant time-varying, having multiplicative effects while dependence on is left un-estimated. An adaptive approach formulated for analyzing Conditional hazard rates modeled in terms both covariates fractional polynomials restricted so that the conditional positive-valued excess probability functions (generalizing survival times) decreasing. Maximum likelihood to estimate parameters adjusting right censored times. Likelihood cross-validation (LCV) scores compare models. Adaptive searches through alternate rate models controlled by LCV combined tolerance parameters. identify effective underlying regression demonstrated times between tumor recurrence bladder cancer patients. Analyses theory-based these provide conflicting results treatment group initial number tumors. On other hand, polynomial analyses consistent identifying significant tumors model-based robust empirical tests. further distinct moderation effect order an additive after controlling nonlinear order. Results example indicate modeling can generate useful insights into