A Bayesian model for repeated measures zero-inflated count data with application to outpatient psychiatric service use

In applications involving count data, it is common to encounter an excess number of zeros. For example, in the study of outpatient service utilization, the number of utilization days will take on integer values, with many subjects having no utilization (zero values). Mixed distribution models, such as the zero-inflated Poisson and zero-inflated negative binomial, are often used to fit such data. A more general class of mixture models, called hurdle models, can be used to model zero deflation as well as zero inflation. Several authors have proposed frequentist approaches to fitting zero-inflated models for repeated measures. We describe a practical Bayesian approach which incorporates prior information, has optimal small-sample properties and allows for tractable inference. The approach can be easily implemented using standard Bayesian software. A study of psychiatric outpatient service use illustrates the methods.

[1]  David C. Heilbron,et al.  Zero-Altered and other Regression Models for Count Data with Added Zeros , 1994 .

[2]  Alan E. Gelfand,et al.  Bayesian statistics without tears: A sampling-resampling perspective , 1992 .

[3]  Joseph L Schafer,et al.  A Two-Part Random-Effects Model for Semicontinuous Longitudinal Data , 2001 .

[4]  P. Green.  Reversible jump Markov chain Monte Carlo computation and Bayesian model determination , 1995 .

[5]  LambertDiane.  Zero-inflated Poisson regression, with an application to defects in manufacturing , 1992 .

[6]  Gary K Grunwald,et al.  Analysis of repeated measures data with clumping at zero , 2002, Statistical methods in medical research.

[7]  A. Gelfand,et al.  Bayesian Model Choice: Asymptotics and Exact Calculations , 1994 .

[8]  Andrew Gelman,et al.  General methods for monitoring convergence of iterative simulations , 1998 .

[9]  J. Berger,et al.  The Intrinsic Bayes Factor for Model Selection and Prediction , 1996 .

[10]  John Hinde,et al.  Models for count data with many zeros , 1998 .

[11]  Irene Lena Hudson,et al.  Finite Mixture, Zero-inflated Poisson and Hurdle models with application to SIDS , 2003, Comput. Stat. Data Anal..

[12]  A. Gelman.  Prior distributions for variance parameters in hierarchical models (comment on article by Browne and Draper) , 2004 .

[13]  David J. Spiegelhalter,et al.  Bayesian graphical modelling: a case‐study in monitoring health outcomes , 2002 .

[14]  P. Gustafson,et al.  Conservative prior distributions for variance parameters in hierarchical models , 2006 .

[15]  Alan Agresti,et al.  Random effect models for repeated measures of zero-inflated count data , 2005 .

[16]  W. Greene,et al.  Accounting for Excess Zeros and Sample Selection in Poisson and Negative Binomial Regression Models , 1994 .

[17]  C. Czado,et al.  Modelling count data with overdispersion and spatial effects , 2008 .

[18]  Peter Congdon,et al.  Bayesian Models for Categorical Data: Peter Congdon/Bayesian Models for Categorical Data , 2006 .

[19]  Brian J. Smith,et al.  boa: An R Package for MCMC Output Convergence Assessment and Posterior Inference , 2007 .

[20]  Adrian E. Raftery,et al.  Bayesian model averaging: a tutorial (with comments by M. Clyde, David Draper and E. I. George, and a rejoinder by the authors , 1999 .

[21]  Nicola J Cooper,et al.  Predicting costs over time using Bayesian Markov chain Monte Carlo methods: an application to early inflammatory polyarthritis. , 2007, Health economics.

[22]  By W. R. GILKSt,et al.  Adaptive Rejection Sampling for Gibbs Sampling , 2010 .

[23]  Alan E. Gelfand,et al.  Model Determination using sampling-based methods , 1996 .

[24]  Andy H. Lee,et al.  Zero‐inflated Poisson regression with random effects to evaluate an occupational injury prevention programme , 2001, Statistics in medicine.

[25]  William J. Browne,et al.  Implementation and performance issues in the Bayesian and likelihood fitting of multilevel models , 2000, Comput. Stat..

[26]  K. Wannemuehler,et al.  On the Use of Zero-Inflated and Hurdle Models for Modeling Vaccine Adverse Event Count Data , 2006, Journal of biopharmaceutical statistics.

[27]  Bradley P. Carlin,et al.  Bayesian measures of model complexity and fit , 2002 .

[28]  N. Duan.  Smearing Estimate: A Nonparametric Retransformation Method , 1983 .

[29]  D. Hall.  Zero‐Inflated Poisson and Binomial Regression with Random Effects: A Case Study , 2000, Biometrics.

[30]  Kathryn Roeder,et al.  Modeling Uncertainty in Latent Class Membership: A Case Study in Criminology , 1999 .

[31]  W. Gilks,et al.  Random-effects models, for longitudinal data using Gibbs sampling. , 1993, Biometrics.

[32]  R. Rosenheck,et al.  Service systems integration and outcomes for mentally ill homeless persons in the ACCESS program. Access to Community Care and Effective Services and Supports. , 2002, Psychiatric services.

[33]  I. Katz,et al.  Using a Bayesian latent growth curve model to identify trajectories of positive affect and negative events following myocardial infarction. , 2005, Biostatistics.

[34]  J. Mullahy.  Specification and testing of some modified count data models , 1986 .

[35]  J. Rodrigues.  Bayesian Analysis of Zero-Inflated Distributions , 2003 .

[36]  Xiao-Li Meng,et al.  POSTERIOR PREDICTIVE ASSESSMENT OF MODEL FITNESS VIA REALIZED DISCREPANCIES , 1996 .

[37]  Hui Liu,et al.  Growth Curve Models for Zero-Inflated Count Data: An Application to Smoking Behavior , 2007 .

[38]  S. Geisser,et al.  A Predictive Approach to Model Selection , 1979 .

[39]  Hakjin Jang,et al.  Bayesian Analysis for the Zero-inflated Regression Models , 2008 .

[40]  C. Robert,et al.  Deviance information criteria for missing data models , 2006 .

[41]  Hong Chang,et al.  Bayesian Approach for Nonlinear Random Effects Models , 1997 .

[42]  W. Gilks,et al.  Adaptive Rejection Metropolis Sampling Within Gibbs Sampling , 1995 .

[43]  Dani Gamerman,et al.  Sampling from the posterior distribution in generalized linear mixed models , 1997, Stat. Comput..

[44]  Nairanjana Dasgupta,et al.  Bayesian Models for Categorical Data , 2007, Technometrics.

[45]  Nicola J Cooper,et al.  Use of Bayesian Markov Chain Monte Carlo Methods to Model Cost-of-Illness Data , 2003, Medical decision making : an international journal of the Society for Medical Decision Making.

[46]  Katherine E. Masyn,et al.  General growth mixture modeling for randomized preventive interventions. , 2001, Biostatistics.

[47]  Sylvia Richardson,et al.  Markov Chain Monte Carlo in Practice , 1997 .

[48]  L. Fahrmeir,et al.  Structured additive regression for overdispersed and zero‐inflated count data , 2006 .

[49]  David R. Jones,et al.  How vague is vague? A simulation study of the impact of the use of vague prior distributions in MCMC using WinBUGS , 2005, Statistics in medicine.