A Spatial Poisson Hurdle Model for Exploring Geographic Variation in Emergency Department Visits

• DOI: 10.1111/j.1467-985x.2012.01039.x
• Publication Date: 2012-06-28T11:07:12Z
• Authors: Brian Neelon; Pulak Ghosh; Patrick F. Loebs
• Keywords: Geographic variation; Variation (astronomy)

We develop a spatial Poisson hurdle model to explore geographic variation in emergency department (ED) visits while accounting for zero inflation. The consists of two components: Bernoulli component that models the probability any ED use (i.e., at least one visit per year), and truncated number given use. Together, these components address both abundance zeros right-skewed nature nonzero counts. has hierarchical structure incorporates patient- area-level covariates, as well spatially correlated random effects each areal unit. Because regions with high rates are likely have expected counts among users, we via bivariate conditionally autoregressive (CAR) prior, which introduces dependence between provides smoothing sharing information across neighboring regions. Using simulation study, show modeling between-component correlation reduces bias parameter estimates. adopt Bayesian estimation approach, can be fit using standard software. apply study patient neighborhood factors influencing Durham County, North Carolina.