PEP 2003-010

CIO Name: NCCDPHP

Title: Prostate Cancer ESDA and Spatial Statistics

Project Description:

The purpose of this project is to conduct research to systematically evaluate and compare the strengths and weaknesses of Exploratory Spatial Data Analysis (ESDA) and spatial statistical methods for prostate cancer for use by state-based Comprehensive Cancer Control (CCC) programs. Examples of ESDA and spatial statistical issues might include: identification of spatial outliers; discovery and evaluation of spatial patterns; analysis of changes in spatial patterns over time; methods for adjustment for confounders; smoothing methods; spatial regression methods; and simulation models. Multiple methods have been developed, but relatively little information exists about how these methods might be applied by CCC programs. The various methods do not always lead to the same conclusions. Research and field testing to compare the multiple methods will help improve understanding of how to select and apply these methods as part of CCC program efforts. Although field testing in this specific project will focus on prostate cancer, the results will be useful for many other types of cancer.

Project Activities:

Project activities should include: 1) convening a meeting of experts to identify key methodological issues in prostate cancer ESDA and spatial statistics; and 2) research and/or field testing to compare and evaluate different ESDA methods and spatial statistics for state-based CCC programs. The meeting of experts should include representatives from state-based CCC programs as well as academic researchers. Several examples of topics for the meeting agenda might include: 1) identification of outstanding questions regarding prostate cancer ESDA and spatial statistical methods; 2) discussion about limitations of existing methods; and 3) brainstorming about possible new methods.

Project research and/or field testing activities should be designed to systematically compare and evaluate key questions about prostate cancer ESDA and spatial statistics for CCC programs. Several examples of potential topics of interest might include:

What information should be included in an information frame describing the method?
What are the strengths and weaknesses of the various methods?
What methods should be used to adjust for confounders?
What methods should be used for spatial regression?
What methods should be used for analysis of spatial patterns of continuous variables (e.g., cancer survival)?
What methods should be used for analysis of variables with multiple categories (e.g., cancer histology and staging)?
What methods should be used for detection of irregular shaped clusters?
For CCC programs with multiple years of prostate cancer data, what methods should be used for early detection of changes in spatial patterns over time?
What decision trees should be followed by staff in cancer registry and CCC programs who need to evaluate prostate cancer spatial patterns (e.g., are multiple tests required? If so, which ones? and in what order?); and
From the perspective of practical problem solving for CCC programs, what are priority areas for additional research in ESDA and spatial statistics?

The first six (6) months of project activities should be reserved for activities that do not require data analysis, such as convening the meeting of experts, developing a protocol meeting CDC Institutional Review Board (IRB) requirements, and obtaining any necessary local or CDC IRB approvals. Project activities should include developing at least three articles suitable for publication in a scientific journal.

Preference will be given to proposals that include documentation of:

active participation in the project by a state-based CCC program; and
ability to successfully obtain and protect the confidentiality of the prostate cancer data to be included in the study.

Total Years of Project Length: Three (3) years

Total funding (direct and indirect) availability by project year: $250,000 per year

Is this PEP Non-research or Research: Research