Semiparametric efficient estimation in the generalized odds-rate class of regression models for right-censored time-to-event data (Q1277112)

This paper considers a generalized odds rate class of regression models \(g_p(S(t\mid Z))=\alpha (t)+\beta'Z\), where \(g_p(s) =\log(p^{-1} (s^{-p}-1))\) for \(p>0\), \(g_0(s)= \log(-\log s)\). \(S(t\mid Z)\) is the survival function of the time to event for an individual with \(q\times 1\) covariate vector \(Z\), \(\beta\) is a \(q\times 1\) vector of unknown regression parameters, and \(\alpha(t)\) is some arbitrary increasing function of \(t\). In the presence of right censoring, the estimators of \(\beta\) and \(\exp(\alpha(t))\) are constructed, and it is shown that they are consistent and asymptotically normal. Semiparametric efficient estimators are given.