Estimation for paired binomial data with application to radiation therapy

Citation
Jmg. Taylor et al., Estimation for paired binomial data with application to radiation therapy, STAT MED, 20(22), 2001, pp. 3375-3390
Citations number
20
Categorie Soggetti
Research/Laboratory Medicine & Medical Tecnology","Medical Research General Topics
Journal title
STATISTICS IN MEDICINE
ISSN journal
02776715 → ACNP
Volume
20
Issue
22
Year of publication
2001
Pages
3375 - 3390
Database
ISI
SICI code
0277-6715(20011130)20:22<3375:EFPBDW>2.0.ZU;2-2
Abstract
We compare and contrast several different methods for estimating the effect of treatment when responses are paired binomial observations. The ratio of binomial probabilities is the parameter of interest, while the binomial pr obabilities are nuisance parameters which may vary between pairs. The appli cation is a meta-analysis of the treatment of rectal cancer, with observati ons in each study indicating the number of recurrences of the cancer in eac h of two groups, one with radiation therapy and one without. The ratio of t he probabilities of recurrence in the radiation to non-radiation groups is of substantive interest, and is modelled as a logistic or complementary log -log function of an unknown linear combination of the covariates. The three methods we consider are maximum likelihood, a Bayesian approach and an app roach based on estimating equations. For the MLE and Bayesian approach the potentially large number of nuisance parameters are estimated together with the parameters of interest, whereas for the estimating equation approach o nly the parameters of interest are estimated. A simulation study is perform ed to compare the methods and evaluate the impact of overdispersion. Copyri ght (C) 2001 John Wiley & Sons, Ltd.