Regression analysis for current status data using the EM algorithm

• Published in: Statistics in medicine Vol. 32; no. 25; pp. 4452 - 4466
• Main Authors: McMahan, Christopher S., Wang, Lianming, Tebbs, Joshua M.
• Format: Journal Article
• Language: English
• Published: England Blackwell Publishing Ltd 10.11.2013; Wiley Subscription Services, Inc
• Subjects: African Americans - statistics & numerical data; Algorithms; Bayes Theorem; Computer Simulation; data augmentation; European Continental Ancestry Group - statistics & numerical data; Female; Humans; Incidence; Leiomyoma - diagnostic imaging; Leiomyoma - epidemiology; Likelihood Functions; Mathematical functions; Mathematical models; maximum likelihood; monotone splines; North Carolina - epidemiology; Poisson Distribution; Pregnancy; Pregnancy Complications, Neoplastic - diagnostic imaging; Pregnancy Complications, Neoplastic - epidemiology; Pregnancy Trimester, First; proportional hazards model; Proportional Hazards Models; proportional odds model; Prospective Studies; Regression Analysis; semiparametric regression; Simulation; survival analysis; Tennessee - epidemiology; Texas - epidemiology; Ultrasonography; monotone splines; data augmentation; survival analysis; proportional hazards model; maximum likelihood; semiparametric regression; proportional odds model
• ISSN: 0277-6715; 1097-0258; 1097-0258
• DOI: 10.1002/sim.5863

We propose new expectation–maximization algorithms to analyze current status data under two popular semiparametric regression models: the proportional hazards (PH) model and the proportional odds (PO) model. Monotone splines are used to model the baseline cumulative hazard function in the PH model and the baseline odds function in the PO model. The proposed algorithms are derived by exploiting a data augmentation based on Poisson latent variables. Unlike previous regression work with current status data, our PH and PO model fitting methods are fast, flexible, easy to implement, and provide variance estimates in closed form. These techniques are evaluated using simulation and are illustrated using uterine fibroid data from a prospective cohort study on early pregnancy. Copyright © 2013 John Wiley & Sons, Ltd.