Login

Publications  •  Project Statistics

Glossary  •  Schools  •  Disciplines
People Search: 
   
Title/Abstract Search: 

Dissertation Information for Joel B. Greenhouse

NAME:
- Joel B. Greenhouse
- (Alias) Joel Bruce Greenhouse

DEGREE:
- Ph.D.

DISCIPLINE:
- Biostatistics

SCHOOL:
- University of Michigan, Ann Arbor (USA) (1983)

ADVISORS:
- None

COMMITTEE MEMBERS:
- None

MPACT Status: Incomplete - Inspected

Title: ANALYSIS OF SURVIVAL DATA WHEN A PROPORTION OF PATIENTS ARE CURED: A MIXTURE MODEL

Abstract: "Probability models for survival times of patients treated for a disease are often interpreted as though all patients who die within a certain period of time after treatment die from the disease of interest and are not cured. Furthermore, those patients who survive beyond a given time after treatment are assumed to be cured. For example, an index of cure often quoted for patients treated for cancer is the proportion of patients alive five years after treatment, i.e., the five year survival rate. Another basic assumption in the analysis of mortality data is that the cause of death is known and due to a single cause.

The objective of this research is to develop and to study an alternative model for the analysis of survival data which is based on a more realistic concept of cure and deals with the setting where the exact cause of death may not be known. Following Boag (1949), Berkson and Gage (1952), and Chen et al. (1981), we model the overall survival function as a mixture of survival functions where the mixing parameter is the probability of cure.

We relate the work of Chen et al. to the work of the previous authors and, in addition, present an alternative derivation of the mixture model based on the theory of competing risks. We obtain a general likelihood function for the parameters and consider several special cases of this likelihood function. The Weibull survival model for patients who are not cured is considered in detail. In order to better understand the behavior of the likelihood function, we obtain and interpret several forms for the maximum likelihood equations and for the observed information matrix. We also derive bounds and approximations for the solutions to the likelihood equations. The concavity of the likelihood function is investigated. Finally, we apply the theory developed for the mixture model to data from the Stanford Heart Transplant Program."

MPACT Scores for Joel B. Greenhouse

A = 0
C = 1
A+C = 1
T = 0
G = 0
W = 0
TD = 0
TA = 0
calculated 2008-01-31 06:24:55

Advisors and Advisees Graph

generating graph, please reload

Students under Joel B. Greenhouse

ADVISEES:
- None

COMMITTEESHIPS:
- Linda Hogan - University of Pittsburgh (2001)