SEMIPARAMETRIC METHODS FOR DATA FROM AN OUTCOME-DEPENDENT SAMPLING SCHEME - Quantitative Medical Data Analysis Using Mathematical Tools and Statistical Techniques - page 93

CHAPTER 5
SEMIPARAMETRIC METHODS FOR DATA FROM AN
OUTCOME-DEPENDENT SAMPLING SCHEME
Haibo Zhou a and Jinhong You
Department of Biostatistics,
University of North Carolina at Chapel Hill,
Chapel Hill, NC 27599-7400, USA
E-mail: a [e-mail]
Outcome-dependent sampling (ODS) is a cost eective way to enhance
study eciency. The case-control design for binary outcomes is a main-
stay of epidemiology research. As the eld of epidemiology expanding
and evolving, an increasing number of studies are conducted using the
ODS design with a \continuous" outcome. In an ODS design, obser-
vations made on a judiciously chosen subset of the base population can
provide nearly the same statistical eciency as observing the entire base
population. Dierent statistical inference procedures are needed in or-
der to reap the benets of such sampling. We review recently developed
methods that account for the ODS design. These methods are all semi-
parametric approaches.
1. Introduction
Observational epidemiologic studies are to evaluate the relationship be-
tween an exposure and a disease, taking into account the eects of addi-
tional covariates such as age and sex. Cohort and case-control designs
are the two most commonly used designs in such studies. In a cohort study,
subjects are randomly selected from the population. The selection may or
may not depend on covariates, but is independent of the response. In a
case-control study, sampling is conditional on the response. Both designs
allow one to evaluate the association between risk factors and disease. Some
large cohort studies cost hundreds of millions of dollars to conduct. Case-
control studies, on the other hand, are often preferred for rare diseases
because they can yield an equal number of diseased individuals in a much
smaller study. Since the work by Corneld 1 , the case-control method has