A Bayesian Zero-Inflated Mixed Lognormal Regression Model for Emergency Department (ED) Diversion Time

• Activity Number: 508
• Type: Contributed
• Date/Time: Wednesday, August 3, 2011 : 10:30 AM to 12:20 PM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Rongwei Fu*+ and Daniel A. Handel
• Companies: Oregon Health & Science University and Oregon Health & Science University
• Keywords: Bayesian model ; zero-inflated log-normal model ; mixed model ; ED diversion

Abstract

Ambulance diversion occurs when one Emergency Department (ED) is too crowded and the ambulance is re-routed to a different ED. Ambulance diversion was once thought to be a "novel" solution for the problem of ED crowding. However, it is now understood that diversion is not an effective means for alleviating over-crowding. It is important to investigate predictors associated the diversion. However, the distribution of diversion time is highly skewed with many days without diversion (diversion time = 0), and diversion time of days on divert still tending to right skewed. It is important to evaluate both the decision to go on diversion and the time on diversion across multiple EDs. Therefore, we propose a Bayesian zero-inflated mixed lognormal regression model to evaluate predictors for both the decision to go on diversion and the time on diversion. The model is illustrated using data from nine EDs across US.