Evaluation of Bayesian Spatial-Temporal Latent Models in Small-Area Health Data

• Activity Number: 547
• Type: Invited
• Date/Time: Wednesday, August 3, 2011 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics and the Environment
• Author(s): Jungsoon Choi*+ and Andrew B. Lawson
• Companies: Medical University of South Carolina and Medical University of South Carolina
• Keywords: Spatial cluster ; diagnostic ; Spatial temporal mixture model ; latent model ; small area health

Abstract

Health outcomes are linked to air pollution, demographic, or socioeconomic factors which vary across space and time. Thus, it is often found that relative risks in spatial health data have locally different patterns. In such cases, latent modeling is useful in the disaggregation of risk profiles. In particular, spatial-temporal mixture models can help to isolate spatial clusters each of which has a homogeneous temporal pattern in relative risks. Mixture models are assumed as they have various weight structures and considered in two situations: the number of underlying components is known or unknown. In this paper, we compare spatial-temporal mixture models with different weight structures in both situations. For comparison, we propose a set of spatial cluster detection diagnostics which are based on the posterior distribution of weights. We also develop new accuracy measures to assess the recovery of true relative risk. Based on the simulation study, we examine the performance of various spatial-temporal mixture models in terms of proposed methods and goodness-of-fit measures. We examine two real data sets: low birth weight data and chronic obstructive pulmonary disease data.