Statistical Analysis of Florida Patients' Emergency Department Visits for Ambulatory Care Sensitive Conditions

• Activity Number: 675
• Type: Contributed
• Date/Time: Thursday, August 5, 2010 : 10:30 AM to 12:20 PM
• Sponsor: Health Policy Statistics Section
• Author(s): Anqi Tang*+ and Askar Chukmaitov and Xufeng Niu
• Companies: Florida State University and Florida State University and Florida State University
• Keywords: Ambulatory Care Sensitive Conditions ; Generalized Estimating Equations ; Bootstrap

Abstract

The growth in Emergency Department use for chronic Ambulatory Care Sensitive Conditions has been increasing and has drawn close attention from government and researchers. In order to identify patients' characteristics and other factors that may affect ED use, we study a random sample of 250,000 ED visits in the year of 2005 to explore associations between patients characteristics, factors such as insurance type, geographic locations and potentially preventable ED use. Multivariate logistic regression models, bootstrapping and generalized estimating equations are used and compared for assessing the association. We find statistically significant association between ACSC ED use and several factors such as patient age, gender and race. Results show that logistic regression is the most efficient in terms of computational cost and accuracy when we have large sample of administrative data.