Back to Annual Meeting

Back to Annual Meeting

APHA Scientific Session and Event Listing

Jorge Quiroz, Preclinical Biostatistics, Wyeth Research, Preclinical Biostatistics, Pearl River, NY 10965 and Jeffrey Wilson, School of Health Management and Policy, Arizona State University, BA318, W. P. Carey School of Business, Tempe, AZ 85287, 4809655628, jeffrey.wilson@asu.edu.

It is customary to find censored count data in analyses pertaining to health related data. Modeling these types of data while ignoring the censoring often results in biased parameter estimates. In addition to censoring, correlated observations may also be present. In this paper, we propose a censored Poisson regression model with normal random effects that is suitable to adjust for censoring as well as overdispersion in count data. Data from dilution assay with count response were analyzed using a censored Poisson regression model with normal random effects. In addition, a simulation study was conducted to compare the performance of a censored Poisson regression model with normal random effects as opposed to other approaches.

Learning Objectives:

• Simulation
• Censored observations
• Adjusting for correlation

Keywords: Simulation, Biostatistics

Presenting author's disclosure statement:

Any relevant financial relationships? No