SmoothHazard: Fitting illness-death model for interval-censored data

SmoothHazard is a package for fitting illness-death (and survival) model with possibly interval-censored data for transition toward transient state. Left-truncated and right-censored data are also allowed. Proportional intensities regression models allow for covariates on each of the 3 transitions. We use either a parametric approach with Weibull baseline intensities or a semi-parametric approach with M-splines approximation of baseline intensities in order to obtain smooth estimates. Estimation of parameters is conducted by likelihood maximization in the parametric approach or penalized likelihood maximization in the semi-parametric approach.

Downloads: