National Health Data Resources from NCHS

Abstract:  This presentation will include an overview of the major data systems from NCHS, and examples of their uses in current research. It will also explain the differences between restricted and unrestricted data, and how the Research Data Center facilitates access to these data resources.

Location:	AE 1002 (Biostatistics Seminar Room - Pavilion I)
Date:	Tuesday, September 08, 2009
Time:	3:00 – 4:00 PM
Contact:	Lifang Zhang (706) 721-4453 or Biostat@MCG.edu

Refreshments and socializing: 2:30 - 3:00 PM
 

Some Semiparametric Bayesian Techniques for Cancer Surveillance

Abstract:  Ascertaining the effectiveness of current cancer-control strategies and projecting upcoming trends for optimal allocation of future resources are some of the important challenges faced by cancer surveillance community. In this talk, we will discuss several semi-parametric Bayesian techniques that have been found useful for such purposes. In particular, we will use the Dirichlet process and its variants to (i) develop improved methods of short-term prediction of cancer mortality counts, (ii) cluster geographic locations in terms of “similarity" of cancer trends, and (iii) summarize individual cancer series. We will illustrate some of these methods using data on U.S. cancer mortality obtained from the Surveillance, Epidemiology and End Results (SEER) program of the National Cancer Institute.

Refreshments and socializing: 10:30 - 11:00 AM