Using Covariates to Model Repeated Clustered Data with Informative Cluster Sizes

• Activity Number: 604
• Type: Contributed
• Date/Time: Thursday, August 7, 2014 : 8:30 AM to 10:20 AM
• Sponsor: Biometrics Section
• Author(s): Ana Maria Iosif*+
• Companies: University of California, Davis
• Keywords: clustered data ; informative cluster size ; covariate ; recurring episodes

Abstract

Many chronic diseases or health conditions manifest with recurring episodes, each of which can be characterized by a measure of intensity or severity. Both the number of episodes and the severity of each episode can depend on the latent severity of an individual's underlying condition. Furthermore, the underlying condition severity might be impacted by observed covariates, or risk factors. Data like this are commonly collected for a broad range of chronic conditions for which repeated data are available on both the number of episodes and their severity. From a clinical standpoint it is important to assess both these outcomes, since they both might inform about the condition severity. We propose models to analyze data collected repeatedly on both the frequency of an event and its severity when both of these are informative about the underlying latent severity. We propose a general ML framework that allows the use of covariates and derive estimators with good asymptotic properties. Our simulations under scenarios with sample sizes in a small to moderate range show that the estimators also have good finite sample properties. We implement this method to a real life data set.