Induced Smoothing for the Semiparametric Accelerated Failure Time Model with Clustered Data: Marginal Methods and Frailty Models

• Activity Number: 64
• Type: Topic Contributed
• Date/Time: Sunday, August 4, 2013 : 4:00 PM to 5:50 PM
• Sponsor: Biometrics Section
• Author(s): Lynn Johnson*+
• Companies: Cornell University
• Keywords: induced smoothing ; censored data ; clustered data ; marginal methods ; frailty model

Abstract

The semiparametric accelerated failure time (AFT) model provides a useful, easy to interpret method for characterizing the relationship between failure time and covariates. Clustered failure time data can be handled in the AFT model by using marginal estimation methods or by incorporating a random cluster-level frailty term. However, regression parameter estimation for both approaches requires optimization of a non-smooth objective function. Induced smoothing can be used in the marginal approach to construct a procedure that permits fast, accurate computation of regression parameter estimates and standard errors. The regression parameter estimates are strongly consistent, asymptotically normal and, additionally, the asymptotic distribution of the smoothed estimator coincides with that obtained without smoothing. For the AFT frailty model, induced smoothing can be used in conjunction with an EM-type algorithm to construct a procedure which permits simultaneous estimation of the regression parameters, the baseline cumulative hazard, and the parameter for a general frailty distribution. We will review these results with an emphasis on the role of the induced smoothing procedure.