COVARIATE SELECTION IN HIERARCHICAL MODELS OF HOSPITAL ADMISSION COUNTS: A BAYES FACTOR APPROACH

The Bayes factor is employed to select covariates for a hierarchical model applied to a collection of hospital admission counts. Integrals representing the Bayes factor numerator and denominator marginal probabilities are intractable for the model used. We examine three approaches to integral approximation: Laplace approximation, Monte Carlo integration, and a Markov chain Monte Carlo (MCMC) approach. Laplace-approximated Bayes factors are found to be nearly unbiased, and can be obtained very quickly. MCMC was used to implement an importance sampling scheme. The resulting approximate Bayes factors exhibited some bias and were obtained at a relatively high cost in computer time. Software to implement the Laplace approximation is available from StatLib.

[1]  M. Newton.  Approximate Bayesian-inference With the Weighted Likelihood Bootstrap , 1994 .

[2]  J. Eisenberg,et al.  Views from funding agencies. Agency for Health Care Policy and Research. , 1998, Medical care.

[3]  A. Raftery,et al.  How Many Iterations in the Gibbs Sampler , 1991 .

[4]  Jo Billings.  Consideration of the Use of Small Area Analysis as a Tool to Evaluate Barriers to Access , 1990 .

[5]  W. K. Hastings,et al.  Monte Carlo Sampling Methods Using Markov Chains and Their Applications , 1970 .

[6]  P. Diehr,et al.  Small area analysis of surgery for low-back pain. , 1992, Spine.

[7]  Eric R. Ziegel,et al.  Generalized Linear Models , 2002, Technometrics.

[8]  ジャーナルAMA日本版刊行会,et al.  The Journal of the American Medical Association. , 1983, JAMA.

[9]  J. Berger,et al.  Testing a Point Null Hypothesis: The Irreconcilability of P Values and Evidence , 1987 .

[10]  R. Deyo,et al.  The Back Pain Outcome Assessment Team. , 1990, Health services research.

[11]  Adrian E. Raftery,et al.  Bayes factors and model uncertainty , 1995 .

[12]  J. Freeman,et al.  ARE HOSPITAL SERVICES RATIONED IN NEW HAVEN OR OVER-UTILISED IN BOSTON? , 1987, The Lancet.

[13]  Margaret R. Spitz,et al.  Research methodology: Strengthening causal interpretations of nonexperimental data , 2004, Cancer Causes & Control.

[14]  Stuart Geman,et al.  Bayesian image restoration , 1986 .

[15]  L. Devroye.  Non-Uniform Random Variate Generation , 1986 .

[16]  D. Lindley,et al.  Bayes Estimates for the Linear Model , 1972 .

[17]  J. Besag,et al.  Bayesian image restoration, with two applications in spatial statistics , 1991 .

[18]  E. Hannan,et al.  Adult open heart surgery in New York State. An analysis of risk factors and hospital mortality rates. , 1990, JAMA.

[19]  Cloward Dowel,et al.  SPINE , 1995, Bone & Joint 360.

[20]  N. Breslow.  Extra‐Poisson Variation in Log‐Linear Models , 1984 .

[21]  R. Cheng,et al.  The Generation of Gamma Variables with Non‐Integral Shape Parameter , 1977 .

[22]  M. Young,et al.  Health services research , 2008, Journal of General Internal Medicine.

[23]  J. H. Schuenemeyer,et al.  Generalized Linear Models (2nd ed.) , 1992 .

[24]  P. McCullagh,et al.  Generalized Linear Models, 2nd Edn. , 1990 .

[25]  Brian D. Ripley,et al.  Stochastic Simulation , 2005 .

[26]  W. Wong,et al.  The calculation of posterior distributions by data augmentation , 1987 .

[27]  J. Q. Smith,et al.  1. Bayesian Statistics 4 , 1993 .

[28]  A. Stuart.  Gamma-distributed products of independent random variables , 1962 .

[29]  L. Tierney,et al.  Accurate Approximations for Posterior Moments and Marginal Densities , 1986 .

[30]  A. Raftery.  Approximate Bayes factors and accounting for model uncertainty in generalised linear models , 1996 .

[31]  L. Tierney.  Exploring Posterior Distributions Using Markov Chains , 1992 .

[32]  Adult open heart surgery in New York State. An analysis of risk factors and hospital mortality rates. , 1991, JAMA.

[33]  T R Holford,et al.  The estimation of age, period and cohort effects for vital rates. , 1983, Biometrics.

[34]  J. Lawless.  Negative binomial and mixed Poisson regression , 1987 .

[35]  L. M. M.-T..  Theory of Probability , 1929, Nature.

[36]  D. Rubin.  Using the SIR algorithm to simulate posterior distributions , 1988 .