ICBayes: a Package for Bayesian Semiparametric Regression Analysis of Interval-Censored Data

• Activity Number: 574 - Recent Advances in Software
• Type: Contributed
• Date/Time: Wednesday, July 31, 2019 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistical Computing
• Author(s): Chun Pan* and Bo Cai and Lianming Wang and Xiaoyan Lin
• Companies: Hunter College and University of South Carolina and University of South Carolina and University of South Carolina
• Keywords: interval-censored data; Bayesian semiparametric; survival analysis; R package

Abstract

This presentation introduces the R package ICBayes, which aims to be a comprehensive statistical package for analyzing interval-censored data under different survival regression models. Interval-censored data occur when the time-to-event of interest is not directly observable, but falls within a time interval. The current version of ICBayes incorporates four Bayesian functions for analyzing current status data and general interval-censored data under the proportional hazards model, the proportional odds model, and the probit model. These Bayesian functions all adopt a monotone spline for the unknown non-decreasing baseline function in each semiparametric model and are developed based on model-specific data augmentations. In addition, the package provides the log pseudo marginal likelihood for each method and thus allows users to perform model comparison and selection for a given data set. The use of the package is illustrated through real data examples.