A log logistic model for survival time with covariates
M. A. Paiva Franco · Biometrika · 1984
A model for survival time where the effect of the explanatory variables is multiplicative on the median of a basic log logistic distribution is considered. Two different schemes of observation, namely the binary analysis of the proportion of survivors after a fixed time to and the analysis of all identified failure or censored times, are considered. Some results concerning the relative efficiency of estimators of the regression parameters under the two analyses are obtained.