
Accession Number : ADA180750
Title : Survival Analysis Using Additive Risk Models.
Descriptive Note : Technical rept.,
Corporate Author : FLORIDA STATE UNIV TALLAHASSEE DEPT OF STATISTICS
Personal Author(s) : Huffer,Fred W. ; McKeague,Ian W.
Report Date : APR 1987
Pagination or Media Count : 41
Abstract : Cox's (1972) proportional hazards model has so far been the most popular model for the regression analysis of censored survival data. However, the additive risk model of Aalen (1980) provides a useful and biologically more plausible alternative when large sample size makes it application feasible. Let lambda(t,Y) be the hazard function for a subject whose covariates are given by Y = (Y sub 1,...,Y sub p). Aalen's model stipulates that lambda(t) =Y alpha(t), where alpha = (alpha sub 1,..., alpha sub p) is an unknown vector of hazard functions. This paper discusses inference for alpha sub 1,..., alpha sub p) based on continuous and grouped data. Asymptotic distribution results are developed using the theory of counting and used to provide confidence bands for the cumulative hazard functions. The method is applied to data on the incidence of cancer mortality among Japanese atomic bomb survivors. Keywords: Monte Carlo method; Multivariate analysis; Martingale methods.
Descriptors : *REGRESSION ANALYSIS, *SURVIVAL(PERSONNEL), *MATHEMATICAL MODELS, *NUCLEAR RADIATION, COVARIANCE, RADIATION EFFECTS, ADDITIVES, ASYMPTOTIC SERIES, CANCER, COUNTING METHODS, HAZARDS, JAPAN, MONTE CARLO METHOD, MORTALITY RATES, MULTIVARIATE ANALYSIS, NUCLEAR BOMBS, RISK, SURVIVAL(GENERAL), THEORY, VECTOR ANALYSIS
Subject Categories : Numerical Mathematics
Radiobiology
Distribution Statement : APPROVED FOR PUBLIC RELEASE