Renato Assunção

A Space–Time Permutation Scan Statistic for Disease Outbreak Detection

Background The ability to detect disease outbreaks early is important in order to minimize morbidity and mortality through timely implementation of disease prevention and control measures. Many national, state, and local health departments are launching disease surveillance systems with daily analyses of hospital emergency department visits, ambulance dispatch calls, or pharmacy sales for which population-at-risk information is unavailable or irrelevant. Methods and Findings We propose a prospective space–time permutation scan statistic for the early detection of disease outbreaks that uses only case numbers, with no need for population-at-risk data. It makes minimal assumptions about the time, geographical location, or size of the outbreak, and it adjusts for natural purely spatial and purely temporal variation. The new method was evaluated using daily analyses of hospital emergency department visits in New York City. Four of the five strongest signals were likely local precursors to citywide outbreaks due to rotavirus, norovirus, and influenza. The number of false signals was at most modest. Conclusion If such results hold up over longer study times and in other locations, the space–time permutation scan statistic will be an important tool for local and national health departments that are setting up early disease detection surveillance systems.

A simulated annealing strategy for the detection of arbitrarily shaped spatial clusters

We propose a new graph-based strategy for the detection of spatial clusters of arbitrary geometric form in a map of geo-referenced populations and cases. Our test statistic is based on the likelihood ratio test previously formulated by Kulldorff and Nagarwalla for circular clusters. A new technique of adaptive simulated annealing is developed, focused on the problem of finding the local maxima of a certain likelihood function over the space of the connected subgraphs of the graph associated to the regions of interest. Given a map with n regions, on average this algorithm finds a quasi-optimal solution after analyzing snlog(n) subgraphs, where s depends on the cases density uniformity in the map. The algorithm is applied to a study of homicide clusters detection in a Brazilian large metropolitan area.

A new proposal to adjust Moran's I for population density

We analyse the effect of using prevalence rates based on populations with different sizes in the power of spatial independence tests. We compare the well known spatial correlation Morans index to three indexes obtained after adjusting for population density, one proposed by Oden, another proposed by Waldhör, and a third proposed by us in this paper. We find an effect of spatially correlated populations in the type I error probability on the test based on Morans and Waldhörs indexes. We conclude also that the test proposed by Oden is powerful to test risk heterogeneity, but it has disadvantages when the interest is solely on the spatial correlation of morbidity risks. In this latter case, we recommend using our proposed test which is more powerful than the usual Morans index applied directly to the rates.

Efficient regionalization techniques for socio‐economic geographical units using minimum spanning trees

Regionalization is a classification procedure applied to spatial objects with an areal representation, which groups them into homogeneous contiguous regions. This paper presents an efficient method for regionalization. The first step creates a connectivity graph that captures the neighbourhood relationship between the spatial objects. The cost of each edge in the graph is inversely proportional to the similarity between the regions it joins. We summarize the neighbourhood structure by a minimum spanning tree (MST), which is a connected tree with no circuits. We partition the MST by successive removal of edges that link dissimilar regions. The result is the division of the spatial objects into connected regions that have maximum internal homogeneity. Since the MST partitioning problem is NP‐hard, we propose a heuristic to speed up the tree partitioning significantly. Our results show that our proposed method combines performance and quality, and it is a good alternative to other regionalization methods found in the literature.

Simulating fire regimes in the Amazon in response to climate change and deforestation

Fires in tropical forests release globally significant amounts of carbon to the atmosphere and may increase in importance as a result of climate change. Despite the striking impacts of fire on tropical ecosystems, the paucity of robust spatial models of forest fire still hampers our ability to simulate tropical forest fire regimes today and in the future. Here we present a probabilistic model of human-induced fire occurrence for the Amazon that integrates the effects of a series of anthropogenic factors with climatic conditions described by vapor pressure deficit. The model was calibrated using NOAA-12 night satellite hot pixels for 2003 and validated for the years 2002, 2004, and 2005. Assessment of the fire risk map yielded fitness values > 85% for all months from 2002 to 2005. Simulated fires exhibited high overlap with NOAA-12 hot pixels regarding both spatial and temporal distributions, showing a spatial fit of 50% within a radius of 11 km and a maximum yearly frequency deviation of 15%. We applied this model to simulate fire regimes in the Amazon until 2050 using IPCCs A2 scenario climate data from the Hadley Centre model and a business-as-usual (BAU) scenario of deforestation and road expansion from SimAmazonia. Results show that the combination of these scenarios may double forest fire occurrence outside protected areas (PAs) in years of extreme drought, expanding the risk of fire even to the northwestern Amazon by midcentury. In particular, forest fires may increase substantially across southern and southwestern Amazon, especially along the highways slated for paving and in agricultural zones. Committed emissions from Amazon forest fires and deforestation under a scenario of global warming and uncurbed deforestation may amount to 21 +/- 4 Pg of carbon by 2050. BAU deforestation may increase fires occurrence outside PAs by 19% over the next four decades, while climate change alone may account for a 12% increase. In turn, the combination of climate change and deforestation would boost fire occurrence outside PAs by half during this period. Our modeling results, therefore, confirm the synergy between the two Ds of REDD (Reducing Emissions from Deforestation and Forest Degradation in Developing Countries).

[Maps of epidemiological rates: a Bayesian approach].

This article presents statistical methods recently developed for the analysis of maps of disease rates when the geographic units have small populations at risk. They adopt the Bayesian approach and use intensive computational methods for estimating risk in each area. The objective of the methods is to separate the variability of rates due to differences between regions from the background risk due to pure random fluctuation. Risk estimates have a total mean quadratic error smaller than usual estimates. We apply these new methods to estimate infant mortality risk in the municipalities of the State of Minas Gerais in 1994.

Homicide clusters and drug traffic in Belo Horizonte, Minas Gerais State, Brazil from 1995 to 1999

The article presents a spatial analysis of homicides in Belo Horizonte according to the Minas Gerais Military Police records from 1995 to 1999. The authors identify clusters of high mortality risk and relate them to areas with drug traffic and associated violence. SaTScan software is used to locate the clusters.

Spatial Scan Statistics Adjusted for Multiple Clusters

The spatial scan statistic is one of the main epidemiological tools to test for the presence of disease clusters in a geographical region. While the statistical significance of the most likely cluster is correctly assessed using the model assumptions, secondary clusters tend to have conservatively high P-values. In this paper, we propose a sequential version of the spatial scan statistic to adjust for the presence of other clusters in the study region. The procedure removes the effect due to the more likely clusters on less significant clusters by sequential deletion of the previously detected clusters. Using the Northeastern United States geography and population in a simulation study, we calculated the type I error probability and the power of this sequential test under different alternative models concerning the locations and sizes of the true clusters. The results show that the type I error probability of our method is close to the nominal level and that for secondary clusters its power is higher than the standard unadjusted scan statistic.

Neighborhood Dependence in Bayesian Spatial Models

The conditional autoregressive model and the intrinsic autoregressive model are widely used as prior distribution for random spatial effects in Bayesian models. Several authors have pointed out impractical or counterintuitive consequences on the prior covariance matrix or the posterior covariance matrix of the spatial random effects. This article clarifies many of these puzzling results. We show that the neighborhood graph structure, synthesized in eigenvalues and eigenvectors structure of a matrix associated with the adjacency matrix, determines most of the apparently anomalous behavior. We illustrate our conclusions with regular and irregular lattices including lines, grids, and lattices based on real maps.

EMPIRICAL BAYES ESTIMATION OF DEMOGRAPHIC SCHEDULES FOR SMALL AREAS

In this article, we analyze empirical Bayes (EB) methods for estimating small-area rate schedules. We develop EB methods that treat schedules as vectors and use adaptive neighborhoods to keep estimates appropriately local. This method estimates demographic rates for local subpopulations by borrowing strength not only from similar individuals elsewhere but also from other groups in the same area and from regularities in schedules across locations. EB is substantially better than standard methods when rates have strong spatial and age patterns. We illustrate this method with estimates of age-specific fertility schedules for over 3,800 Brazilian municipalities.