Diffusion and prediction of Leishmaniasis in a large metropolitan area in Brazil with a Bayesian space–time model

Abstract

We present results from an analysis of human visceral Leishmaniasis cases based on public health records of Belo Horizonte, Brazil, from 1994 to 1997. The main emphasis in this study is on the development of a spatial statistical model to map and project the rates of visceral Leishmaniasis in Belo Horizonte. The model allows for space-time interaction and it is based on a hierarchical Bayesian approach. We assume that the underlying rates evolve in time according to a polynomial trend specific to each small area in the region. The parameters of these polynomials receive a spatial distribution in the form of an autonormal distribution. While the raw rates are extremely noisy and inadequate to support decisions, the resulting smoothed rates estimates are considerably less affected by small area issues and provide very clear directions to implement public health actions.

An analysis of the spatiotemporal distribution of American cutaneous leishmaniasis in counties located along road and railway corridors in the State of Maranhao, Brazil

“…Several researchers have conducted spatial analyses of the dynamics of infectious diseases 4,5 . The analysis of relative risk (RR) over space and time has received a great deal of attention in epidemiological studies over the last few decades.…”

“…The analysis of relative risk (RR) over space and time has received a great deal of attention in epidemiological studies over the last few decades. Many studies assume that RR is composed of several random components, and these components explain different variations related to risk, such as temporal and spatial effects 4,6,7 .…”

Spatio-temporal modelling of coffee berry borer infestation patterns accounting for inflation of zeroes and missing values

“…In recent years, new methodologies have been developed for modelling disease incidence and mortality rates in space and time under a hierarchical Bayesian approach (Waller et al, 1997;Knorr-Held & Besag, 1998;Knorr-Held, 2000;Pickle, 2000;Sun et al, 2000;Assunção et al, 2001). The applications of those methods in human epidemiology have been numerous, particularly in disease mapping, to study variations in the risk of diseases in space and time and to visualize trends through time at a regional level (Kleinschmidt et al, 2002;Nobre et al, 2005;Chen et al, 2006;Mabaso et al, 2006).…”

“…We assume that the number of infested fruits, Y it , follows a binomial distribution with parameters n it and π it with a logistic link function, logit (π it ) = η it . To describe the temporal trend in the infestation rates we followed Assunção et al (2001), using as linear predictor polynomials of first (η it = δ i + γ i t), or second order (η it = δ i + γ i t + ν i t 2 ). For the linear trend model, the parameter γ i determines how the levels of infestation of the borer in each plant change over time.…”

Bayesian statistics in medicine: a 25 year review

“…These data featured, along with breast cancer data from Sardinia and infant mortality data from British Columbia in methodology using mixture models to identify high-risk areas [350]. An analysis of diffusion and prediction of Leishmaniasis in Brazil used hierarchical models to explore space-time interactions [351] and Bayesian testing for the presence of a cluster was proposed [352] It was no surprise that Bayesian methods featured strongly in a Statistics in Medicine tutorial paper on disease mapping [353].…”