Joint Modeling of Longitudinal and Time-to-Event Data

Longitudinal studies often incur several problems that challenge standard statistical methods for data analysis. These problems include non-ignorable missing data in longitudinal measurements of one or more response variables, informative observation times of longitudinal data, and survival analysis with intermittently measured time-dependent covariates that are subject to measurement error and/or substantial biological variation. Joint modeling of longitudinal and time-to-event data has emerged as a novel approach to handle these issues.Joint Modeling of Longitudinal and Time-to-Event Data provides a systematic introduction and review of state-of-the-art statistical methodology in this active research field. The methods are illustrated by real data examples from a wide range of clinical research topics. A collection of data sets and software for practical implementation of the joint modeling methodologies are available through the book website.This book serves as a reference book for scientific investigators who need to analyze longitudinal and/or survival data, as well as researchers developing methodology in this field. It may also be used as a textbook for a graduate level course in biostatistics or statistics.

Introduction and ExamplesIntroductionMethods for Ignorable Missing DataIntroductionMissing Data MechanismsLinear and Generalized Linear Mixed ModelsGeneralized Estimating EquationsFruther topicsTime-to-event data analysisRight censoringSurvival function and hazard functionEstimation of a survival functionCox's semiparametric multiplicative hazards modelsAccelerated failure time models with time-independent covariatesAccelerated failure time model with time-dependent covariatesMethods for competing risks dataFurther topicsOverview of Joint Models for Longitudinal and Time-to-Event DataJoint Models of Longitudinal Data and an Event timeJoint Models with Discrete Event Times and Monotone MissingnessLongitudinal Data with Both Monotone and Intermittent Missing ValuesEvent Time Models with Intermittently Measured Time Dependent CovariatesLongitudinal Data with Informative Observation TimesDynamic Prediction in Joint ModelsJoint Models for Longitudinal Data and Continuous Event Times from Competing RisksJoint Alaysis of Longitudinal Data and Competing RisksA Robust Model with t-Distributed Random ErrorsOrdinal Longitudinal Outcomes with Missing Data Due to Multiple Failure TypesBayesian Joint Models with Heterogeneous Random EffectsAccelerated Failure Time Models for Competing RisksJoint Models for Multivariate Longitudinal and Survival DataJoint Models for Multivariate Longitudinal Outcomes and an Event TimeJoint Models for Recurrent Events and Longitudinal DataJoint Models for Multivariate Survival and Longitudinal DataFurther TopicsJoint Models and Missing Data: Assumptions, Sensitivity Analysis, and DiagnosticsVariable Selection in Joint ModelsJoint Multistate ModelsJoint Models for Cure Rate Survival DataSample Size and Power Estimation for Joint ModelsAppendicesA Software to Implement Joint ModelsBibliographyIndex

Robert Elashoff, Gang Li, Ning Li