A random-effects Markov transition model for Poisson-distributed repeated measures with non-ignorable missing values

UCLA-Department of Statistics, PO Box 951554, Los Angeles, CA 90095-1554, USA.
Statistics in Medicine (Impact Factor: 1.83). 05/2007; 26(12):2519-32. DOI: 10.1002/sim.2717
Source: PubMed


In biomedical research with longitudinal designs, missing values due to intermittent non-response or premature withdrawal are usually 'non-ignorable' in the sense that unobserved values are related to the patterns of missingness. By drawing the framework of a shared-parameter mechanism, the process yielding the repeated count measures and the process yielding missing values can be modelled separately, conditionally on a group of shared parameters. For chronic diseases, Markov transition models can be used to study the transitional features of the pathologic processes. In this paper, Markov Chain Monte Carlo algorithms are developed to fit a random-effects Markov transition model for incomplete count repeated measures, within which random effects are shared by the counting process and the missing-data mechanism. Assuming a Poisson distribution for the count measures, the transition probabilities are estimated using a Poisson regression model. The missingness mechanism is modelled with a multinomial-logit regression to calculate the transition probabilities of the missingness indicators. The method is demonstrated using both simulated data sets and a practical data set from a smoking cessation clinical trial.

Download full-text


Available from: Steve Shoptaw, Aug 04, 2014
  • Source
    • "Here, the inverse Wishart prior for Σ unnecessarily incorporates too many parameters. Some customized structures, say the intra-class correlation matrix (ρ 1 [i = j] ) and the Toeplitz matrix (ρ |i− j| ), are difficult for sampling-based inference since they inherently require a restricted domain where ρ is always between 0 and 1. Li et al. [16] applied a first-order Markov transition to repeated measures Poisson model, where the measurement at time point t for subject i, y i,t is only dependent on y i,t−1 , covariate x i,t and a subject-specific random effect ξ i , e.g., y i,t ∼ Poisson(λ i,t ), log(λ i,t ) = x i,t β + (log(max(1, y i,t−1 )) − x i,t−1 β i )c + ξ i . "
    [Show abstract] [Hide abstract]
    ABSTRACT: We review Bayesian analysis of hierarchical non-standard Poisson regression models with an emphasis on microlevel heterogeneity and macrolevel autocorrelation. For the former case, we confirm that negative binomial regression usually accounts for microlevel heterogeneity (overdispersion) satisfactorily; for the latter case, we apply the simple first-order Markov transition model to conveniently capture the macrolevel autocorrelation which often arises from temporal and/or spatial count data, rather than attaching complex random effects directly to the regression parameters. Specifically, we extend the hierarchical (multilevel) Poisson model into negative binomial models with macrolevel autocorrelation using restricted gamma mixture with unit mean and Markov transition covariate created from preceding residuals. We prove a mild sufficient condition for posterior propriety under flat prior for the interesting fixed effects. Our methodology is implemented by analyzing the Baltic sea peracarids diurnal activity data published in the marine biology and ecology literature.
    Statistical Methodology 07/2007; 4(3):354-370. DOI:10.1016/j.stamet.2006.11.006 · 0.64 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Biomedical research is plagued with problems of missing data, especially in clinical trials of medical and behavioral therapies adopting longitudinal design. After a literature review on modeling incomplete longitudinal data based on full-likelihood functions, this paper proposes a set of imputation-based strategies for implementing selection, pattern-mixture, and shared-parameter models for handling intermittent missing values and dropouts that are potentially nonignorable according to various criteria. Within the framework of multiple partial imputation, intermittent missing values are first imputed several times; then, each partially imputed data set is analyzed to deal with dropouts with or without further imputation. Depending on the choice of imputation model or measurement model, there exist various strategies that can be jointly applied to the same set of data to study the effect of treatment or intervention from multi-faceted perspectives. For illustration, the strategies were applied to a data set with continuous repeated measures from a smoking cessation clinical trial.
    Statistics in Medicine 07/2008; 27(15):2826-49. DOI:10.1002/sim.3111 · 1.83 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: A continuous time three-state model with time-dependent transition intensities is formulated to describe transitions between healthy and unhealthy states before death. By using time continuously, known death times can be taken into account. To deal with possible non-ignorable missing states, a selection model is proposed for the joint distribution of both the state and whether or not the state is observed. To estimate total life expectancy and its subdivision into life expectancy in health and ill health, the three-state model is extrapolated beyond the follow-up of the study. Estimation of life expectancies is illustrated by analysing data from a longitudinal study of aging where individuals are in a state of ill health if they have ever experienced a stroke. Results for the selection model are compared with results for a model where states are assumed to be missing at random and with results for a model that ignores missing states.
    Journal of the Royal Statistical Society Series A (Statistics in Society) 04/2010; 173(2):331-349. DOI:10.1111/j.1467-985X.2009.00610.x · 1.64 Impact Factor
Show more