Un Modelo jerárquico Bayesiano espacio-temporal con variable de conteos: aplicación de VIH/SIDA en Costa Rica

• Autores: Shu Wei Chou, Ricardo Alvarado
• Localización: Comunicaciones en Estadística, ISSN 2027-3355, ISSN-e 2339-3076, Vol. 11, Nº. 1, 2018, págs. 9-35
• Idioma: español
• DOI: 10.15332/s2027-3355.2018.0001.01
• Títulos paralelos:
  • A bayesian hierarchical spatio-temporal model with count data: application to the HIV/AIDS in Costa Rica
• Enlaces
  • Texto completo (pdf)
• Resumen
  • español

    Los modelos espaciales que suavizan las tasas de mortalidad estandarizada o los riesgos relativos son utilizados ampliamente en el mapeo de enfermedades, con el objetivo de explorar y describir patrones espaciales de un evento de interés. Generalmente, la estimación de estos riesgos relativos es imprecisa cuando los eventos son raros. Cuando se quiere incluir la tendencia temporal, el problema es aún másgrave pues el conteo de las defunciones en el perı́odo dado se divide en varios años, lo que resulta en que los conteos sean aún más bajos. En este trabajo, se analizan los modelos Bayesianos espacio-temporales que toman en cuenta la información geográfica y temporal, además de algunas covariables como el porcentaje de viviendas urbanas, porcentaje de personas entre 24 y 49 años y la tasa de mortalidadinfantil de cada cantón en el año 2011. Se concluyó que estos modelos producen mejores estimaciones de riesgos relativos por cantón y año, además de que el modelo que asume una interacción espacio-temporal más simple ajusta mejor. Finalmente, se comparan los riesgos relativos estimados con el modelo seleccionado, contra la estimación obtenida vı́a máxima verosimlitud, y resulta que el método propuestoes más eficiente y preciso.

  • English

    Spatial models that smooth standardized mortality ratios are widely used in disease mapping. Usually, estimation is imprecise when events are rare. In situations where each areal count splits into different time periods, this problem is more evident because of the presence of even lower counts for the areal units for each time period. In this work, we analyze models that include geographic and temporal information and some covariates such as percentage of urban household, percentage of people between 24 and 49 years old, and infant mortality ratio of each county in 2011. As a result, these models produce better estimations, especially for the model with the simplest space-time interaction. Finally, HIV/AIDS mortality data in Costa Rica (1998-2012) are used as an illustration to compare classic standardized mortality ratios and posterior means of relative risk. The proposed method is more efficient and more precise than the maximum likelihood.

• Referencias bibliográficas
  • Abellan, J. J., Richardson, S. & Best, N. (2008), ‘Use of space-time models to
  • investigate the stability of patterns of disease’, Environmental Health Pers-
  • pectives 116(8), pp. 1111–1119.
  • *http://www.jstor.org/stable/25071151
  • Agencia de los Estados Unidos para el Desarrollo Internacional (2011), ‘An over-
  • view to spatial data protocols for HIV/AIDS activities: why and how to in-
  • clude the “ where” in your data’.
  • Altman, B. (2011), Continuing promise 2011: Host nation health brief, Technical
  • report, National Center for Disaster Medicine & Public Health. Recuperado
  • de http://ncdmph.usuhs.edu/Documents/2011-CRC.pdf el 17 de marzo del
  • American Cancer Society (2014), ‘Infección con VIH, SIDA y cáncer’, Recupera-
  • do de http://www.cancer.org/acs/groups/cid/documents/webcontent/
  • -pdf.pdf el 13 de marzo del 2015.
  • Bailey, T. & Gatrell, A. (1995), Interactive Spatial Analysis, Prentice Hall, Harlow.
  • Banerjee, S., Carlinl, B. & Gelfand, A. (2014), Hierarchical Modeling and Analysis
  • for Spatial Data, Chapman and Hall/CRC, Boca Raton.
  • Besag, J., York, J. & Mollié, A. (1991), ‘Bayesian image restoration, with two
  • applications in spatial statistics’, Annals of the Institute of Statistical Mathe-
  • matics 43(1), 1–20.
  • *http://dx.doi.org/10.1007/BF00116466
  • Bivand, R. & Lewin-Koh, N. (2014), maptools: Tools for reading and handling
  • spatial objects. R package version 0.8-30.
  • *http://CRAN.R-project.org/package=maptools
  • Brooks, S. P. & Gelman, A. (1998), ‘General methods for monitoring convergen-
  • ce of iterative simulations’, Journal of Computational and Graphical Statis-
  • tics 7(4), 434–455. http://www.stat.columbia.edu/~gelman/research/
  • published/brooksgelman2.pdf.
  • Cai, B., Lawson, A. B., Hossain, M. M. & Choi, J. (2012), ‘Bayesian latent
  • structure models with space-time dependent covariates’, Statistical Modelling
  • (2), 145–164.
  • *http://smj.sagepub.com/content/12/2/145.abstract
  • Centro Centroamericano de Población (2014), Consulta del 10 de octubre del 2014,
  • de la base de datos del Censo del 2011 y de defunciones de Centro Centroa-
  • mericano de Población: http://ccp.ucr.ac.cr.
  • Fortunato, L., Abellan, J., Beale, L., LeFevre, S. & Richardson, S. (2011), ‘Spatio-
  • temporal patterns of bladder cancer incidence in utah (1973-2004) and their
  • association with the presence of toxic release inventory sites’, International
  • Journal of Health Geographics 10(1), 16.
  • *http://www.ij-healthgeographics.com/content/10/1/16
  • Gelfand, A., Dey, D. & Chang, H. (1992), ‘Model determination using predictive
  • distributions with implementation via sampling based methods (with discus-
  • sion)’, Bayesian Statistics 4, 147–67.
  • Gelman, A. & Rubin, D. B. (1992), ‘Inference from iterative simulation using
  • multiple sequences’, Statistical Science 7(4), pp. 457–472.
  • *http://www.jstor.org/stable/2246093
  • González-Ramı́rez, V. (2009), ‘Intervención psicológica en VIH/SIDA’, UARICHA
  • Revista de Psicologı́a (13), pp. 49–63.
  • Hunter, L. M., Souza, R.-M. D. & Twine, W. (2008), ‘The environmental dimen-
  • sions of the HIV/AIDS pandemic: A call for scholarship and evidence-based
  • intervention’, Population and Environment 29(3/5), pp. 103–107.
  • *http://www.jstor.org/stable/40212350
  • Hyndman, R. & Khandakar, Y. (2008), ‘Automatic time series forecasting: the
  • forecast package for R’, Journal of Statistical Software 26(3), 1–22.
  • *http://ideas.repec.org/a/jss/jstsof/27i03.html
  • Lagazio, C., Biggeri, A. & Dreassi, E. (2001), ‘A hierarchical bayesian model for
  • space-time variation of disease risk’, Statistical Modelling 1, 17–29.
  • Lagazio, C., Biggeri, A. & Dreassi, E. (2003), ‘Age-period-cohort models and di-
  • sease mapping’, Environmetrics 14(5), 475–490.
  • *http://dx.doi.org/10.1002/env.600
  • Langford, I. H. (1994), ‘Using empirical bayes estimates in the geographical analy-
  • sis of disease risk’, Area 26(2), pp. 142–149.
  • *http://www.jstor.org/stable/20003399
  • Lawson, A. & Williams, F. (2001), An Introductory Guide to Disease Mapping,
  • John Willy & Sons, New York.
  • Ministerio de Salud (2004), ‘La situación del VIH/SIDA en Costa Rica’.
  • Organización Mundial de la Salud (2014), ‘Health for the world’s adolescents: A
  • second chance in the second decade’.
  • Organización Panamericana de la Salud (2004), ‘La situación del VIH/SIDA en
  • Costa Rica’.
  • Paradis, E., Claude, J. & Strimmer, K. (2004), ‘APE: analyses of phylogenetics
  • and evolution in R language’, Bioinformatics 20, 289–290.
  • Pebesma, E.J., R. B. (2005), ‘Classes and methods for spatial data in r’, R News
  • (2).
  • *http://cran.r-project.org/doc/Rnews/
  • Plummer, M. (2008), ‘Penalized loss functions for bayesian model comparison’,
  • Biostatistics 9(3), 523–539.
  • *http://biostatistics.oxfordjournals.org/content/9/3/523.abstract
  • Poundstone, K. E., Strathdee, S. A. & Celentano, D. D. (2004), ‘The social epide-
  • miology of human immunodeficiency virus/acquired immunodeficiency syn-
  • drome’, Epidemiologic Rev 26(1), pp. 22–35.
  • Press, J. (2003), Subjective and Objective Bayesian Statistics, John Wiley & Son,
  • New Jersey.
  • Richardson, S., Abellan, J. J. & Best, N. (2006), ‘Bayesian spatio-temporal analysis
  • of joint patterns of male and female lung cancer risks in yorkshire (uk)’,
  • Statistical Methods in Medical Research 15(4), pp. 385–407.
  • Richardson, S., Thomson, A., Best, N. & Elliott, P. (2004), ‘Interpreting poste-
  • rior relative risk estimates in disease-mapping studies’, Environmental Health
  • Perspectives 112(9), pp. 1016–1025.
  • *http://www.jstor.org/stable/3838103
  • Ripley, B. (2004), Spatial Statistics, John Wiley & Sons, Ltd., New Jersey.
  • Schmid, V. & Held, L. (2004), ‘Bayesian extrapolation of space-time trends in
  • cancer registry data’, Biometrics 60(4), pp. 1034–1042.
  • *http://www.jstor.org/stable/3695483
  • Spiegelhalter, D. J., Best, N. G., Carlin, B. P. & Van Der Linde, A. (2002), ‘Ba-
  • yesian measures of model complexity and fit’, Journal of the Royal Statistical
  • Society: Series B (Statistical Methodology) 64(4), 583–639.
  • *http://dx.doi.org/10.1111/1467-9868.00353
  • Team, R. C., Wuertz, D., Setz, T. & Chalabi, Y. (2014), fBasics: Rmetrics -
  • Markets and Basic Statistics. R package version 3011.87.
  • *http://CRAN.R-project.org/package=fBasics
  • Thomas, A., O’Hara, B., Ligges, U. & Sturtz, S. (2006), ‘Making bugs open’, R
  • News 6(1), 12–17.
  • *http://cran.r-project.org/doc/Rnews/
  • Waller, L. A., Carlin, B. P., Xia, H. & Gerfand, A. E. (1997), ‘Hierarchical spatio-
  • temporal mapping of disease rates’, Journal of the American Statistical As-
  • sociation 92, 607–17.
  • Wickham, H. (2009), ggplot2: elegant graphics for data analysis, Springer New
  • York.
  • *http://had.co.nz/ggplot2/book
  • Zanakis, S. H., Alvarez, C. & Li, V. (2007), ‘Socio-economic determinants of
  • HIV/AIDS pandemic and nations efficiencies’, European Journal of Opera-
  • tional Research 176, pp. 1811–1838.
  • Zeileis, A. & Hothorn, T. (2002), ‘Diagnostic checking in regression relationships’,
  • R News 2(3), 7–10.
  • *http://CRAN.R-project.org/doc/Rnews/