Disease Mapping: From Foundations to Multidimensional Modeling

Description

Disease Mapping: From Foundations to Multidimensional Modeling guides the reader from the basics of disease mapping to the most advanced topics in this field. A multidimensional framework is offered that makes possible the joint modeling of several risks patterns corresponding to combinations of several factors, including age group, time period, disease, etc. Although theory will be covered, the applied component will be equally as important with lots of practical examples offered. Features: Discusses the very latest developments on multivariate and multidimensional mapping. Gives a single state-of-the-art framework that unifies most of the previously proposed disease mapping approaches. Balances epidemiological and statistical points-of-view. Requires no previous knowledge of disease mapping. Includes practical sessions at the end of each chapter with WinBUGs/INLA and real world datasets. Supplies R code for the examples in the book so that they can be reproduced by the reader. About the Authors: Miguel A. Martinez Beneito has spent his whole career working as a statistician for public health services, first at the epidemiology unit of the Valencia (Spain) regional health administration and later as a researcher at the public health division of FISABIO, a regional bio-sanitary research center. He has been also the Bayesian Hierarchical Models professor for several seasons at the University of Valencia Biostatics Master. Paloma Botella Rocamora has spent most of her professional career in academia although she now works as a statistician for the epidemiology unit of the Valencia regional health administration. Most of her research has been devoted to developing and applying disease mapping models to real data, although her work as a statistician in an epidemiology unit makes her develop and apply statistical methods to health data, in general. Cover Half Title Title Page Copyright Page Dedication Contents Preface Authors Part I: Disease mapping: The foundations 1. Introduction 1.1 Some considerations on this book 1.1.1 Notation 2. Some basic ideas of Bayesian inference 2.1 Bayesian inference 2.1.1 Some useful probability distributions 2.2 Bayesian hierarchical models 2.3 Markov Chain Monte Carlo computing 2.3.1 Convergence assessment of MCMC simulations 3. Some essential tools for the practice of Bayesian disease mapping 3.1 WinBUGS 3.1.1 The BUGS language 3.1.2 Running models in WinBUGS 3.1.3 Calling WinBUGS from R 3.2 INLA 3.2.1 INLA basics 3.3 Plotting maps in R 3.4 Some interesting resources in R for disease mapping practitioners 4. Disease mapping from foundations 4.1 Why disease mapping? 4.1.1 Risk measures in epidemiology 4.1.2 Risk measures as statistical estimators 4.1.3 Disease mapping: the statistical problem 4.2 Non-spatial smoothing 4.3 Spatial smoothing 4.3.1 Spatial distributions 4.3.1.1 The Intrinsic CAR distribution 4.3.1.2 Some proper CAR distributions 4.3.2 Spatial hierarchical models 4.3.2.1 Prior choices in disease mapping models 4.3.2.2 Some computational issues on the BYM model 4.3.2.3 Some illustrative results on real data Part II: Disease mapping: Towards multidimensional modeling 5. Ecological regression 5.1 Ecological regression: a motivation 5.2 Ecological regression in practice 5.3 Some issues to take care of in ecological regression studies 5.3.1 Confounding 5.3.2 Fallacies in ecological regression 5.3.2.1 The Texas sharpshooter fallacy 5.3.2.2 The ecological fallacy 5.4 Some particular applications of ecological regression 5.4.1 Spatially varying coefficients models 5.4.2 Point source modeling 6. Alternative spatial structures 6.1 CAR-based spatial structures 6.2 Geostatistical modeling 6.3 Moving-average based spatial dependence 6.4 Spline-based modeling 6.5 Modeling of specific features in disease mapping studies 6.5.1 Modeling partitions and discontinuities 6.5.2 Models for fitting zero excesses 7. Spatio-temporal disease mapping 7.1 Some general issues in spatio-temporal modeling 7.2 Parametric temporal modeling 7.3 Spline-based modeling 7.4 Non-parametric temporal modeling 8. Multivariate modeling 8.1 Conditionally specified models 8.1.1 Multivariate models as sets of conditional multivariate distributions 8.1.2 Multivariate models as sets of conditional univariate distributions 8.2 Coregionalization models 8.3 Factor models, smoothed ANOVA and other approaches 8.3.1 Factor models 8.3.2 Smoothed ANOVA 8.3.3 Other approaches 9. Multidimensional modeling 9.1 A brief introduction and review of multidimensional modeling 9.2 A formal framework for multidimensional modeling 9.2.1 Some tools and notation 9.2.2 Separable modeling 9.2.3 Inseparable modeling Appendix 1 Bibliography Index