Joint Semiparametric Mixed Effect Model of Hospital Admission and Readmission

• Activity Number: 611
• Type: Contributed
• Date/Time: Wednesday, August 12, 2015 : 2:00 PM to 3:50 PM
• Sponsor: Section on Medical Devices and Diagnostics
• Author(s): Yuqi Chen* and Yuedong Wang
• Companies: UC Santa Barbara and UC Santa Barbara
• Keywords: mixed outcome ; mixed effect model ; latent random variable ; joint model ; hospitalization ; hospital readmission

Abstract

Hospitalization and readmission are an increasing focuses for medical researchers and policy makers. They will help identify the patients who would benefit most from intervention as well as evaluate utility of treatment. However there is a lack of statistical methods that model hospitalization and readmission jointly. We propose a semi-parametric mixed effect model for the two mixed discrete outcomes, accounting heterogeneity between and correlation within a patient using a shared latent random variable. Number of hospitalization will serve as the largest possible value for the number of readmission in our model and the effect will be modeled semi-parametrically using B-splines. Our methods will be applied to analyze hospital admission and readmission of dialysis patients.