Table of Contents Table of Contents
Previous Page  8 / 87 Next Page
Information
Show Menu
Previous Page 8 / 87 Next Page
Page Background

Bolet´ın de Estad´ıstica e Investigaci´on Operativa

Vol. 33, No. 1, Marzo 2017, pp. 4

-21

Estad´ıstica

Estimating the geographical distribution of diseases: a

statistical problem

Paloma Botella-Rocamora and Jordi P´erez-Panad´es

Subdirecci´on de Epidemiolog´ıa, Vigilancia de la Salud y

Sanidad Ambiental

DGSP - Conselleria de Sanitat - Generalitat Valenciana

B

botella

pal@gva.es

,

B

perez

jorpan@gva.es

Miguel A. Mart´ınez-Beneito

´Area de Desigualdades en Salud

Fundaci´on para el Fomento de la Investigaci´on Sanitaria y

Biom´edica de la Comunitat Valenciana (FISABIO)

B

martinez

mig@gva.es

Abstract

This paper introduces the estimation of the geographical distribution

of diseases as a statistical problem. The small size of the studied units

used in these applications usually leads to statistical problems of reliability

of the estimates. The objective of the models used in this field is to

improve the estimates by regularly sharing information between di↵erent

units of study. In this work we introduce both the motivation and the

basic modeling proposals used in this field. Subsequently, we also describe

spatio-temporal and multivariate modeling in this area, two of the main

lines of work developed during the last years. Finally, we introduce the

concept of multidimensional modeling as an incipient line of research in

which the previous models come together.

Keywords:

Disease mapping, mortality, spatial statistics, Bayesian hier-

archical models.

AMS Subject classifications:

62-07, 91-08.

c 2017 SEIO