Catchment Area Analysis Using Bayesian Regression Modeling

Authors

Aobo Wang, Virginia Commonwealth UniversityFollow
David C. Wheeler, Virginia Commonwealth UniversityFollow

Document Type

Article

Original Publication Date

2015

Journal/Book/Conference Title

Cancer Informatics

Volume

2015

Issue

14

DOI of Original Publication

10.4137/CIN.S17297

Comments

Originally published at http://dx.doi.org/10.4137/CIN.S17297

Date of Submission

December 2015

Abstract

A catchment area (CA) is the geographic area and population from which a cancer center draws patients. Defining a CA allows a cancer center to describe its primary patient population and assess how well it meets the needs of cancer patients within the CA. A CA definition is required for cancer centers applying for National Cancer Institute (NCI)-designated cancer center status. In this research, we constructed both diagnosis and diagnosis/treatment CAs for the Massey Cancer Center (MCC) at Virginia Commonwealth University. We constructed diagnosis CAs for all cancers based on Virginia state cancer registry data and Bayesian hierarchical logistic regression models. We constructed a diagnosis/treatment CA using billing data from MCC and a Bayesian hierarchical Poisson regression model. To define CAs, we used exceedance probabilities for county random effects to assess unusual spatial clustering of patients diagnosed or treated at MCC after adjusting for important demographic covariates. We used the MCC CAs to compare patient characteristics inside and outside the CAs. Among cancer patients living within the MCC CA, patients diagnosed at MCC were more likely to be minority, female, uninsured, or on Medicaid.

Rights

Copyright © 2015 the authors, publisher and licensee Libertas Academica Limited. This is an open-access article distributed under the terms of the Creative Commons CC-BY-NC 3.0 License.

Is Part Of

VCU Biostatistics Publications