Spline–based Hazards Regression Model for Current  Status Data: An Application to Simulated Data on  Renal Impairment

Gurprit Grover; Barnali Deka

Volume : III, Issue : XII, December - 2013

Spline–based Hazards Regression Model for Current Status Data: An Application to Simulated Data on Renal Impairment

Gurprit Grover, Barnali Deka

Abstract :

Regression modeling of current status data involves the unknown baseline cumulative hazard function. Estimation procedures yields non–smooth curves that complicates the process of understanding the behavior of survival function. Here we have proposed a sieve semiparametric maximum likelihood estimation method for the proportional hazards model for current status data. We have flexibly parameterized the unknown baseline cumulative hazards function using monotone splines. The developed estimation procedure has the advantage of being a computationally efficient one to produce smooth estimates of the survival (or hazard) function and regression parameters as well.