Environmental Factor, March 2010, National Institute of Environmental Health Sciences
Linking Disease Incidence with Environmental Exposure Data over Time
By Erin D. Hopper
Epidemiological studies are often hindered by the challenges associated with visualizing environmental exposures and disease incidence as they evolve over time. To address this challenge, Elena N. Naumova, Ph.D., and her research team describe a dynamic mapping strategy for visualizing complex spatio-temporal data, in a new study (http://www.ncbi.nlm.nih.gov/pubmed/20042115?ordinalpos=1&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_MultiItemSupl.PMC_FreeArticle_ad&linkpos=1&log$=pmcad6_article) funded by NIEHS and the National Institute of Allergy and Infectious Disease (NIAID).
Using dynamic maps, researchers can examine the timing and geographic locations of disease incidence, the duration of time that the disease persists, and the dynamics of disease transmission (see sample, ) - all important elements for understanding an outbreak of infectious disease and for planning an effective public health intervention. In a description of her research, Naumova commented that "dynamic mapping creates a visual representation of data over time, allowing us to detect relationships between disease and environmental factors that cannot be observed in static maps."
Applying dynamic mapping to a food-born pathogen
In the study, Naumova and her team sought to determine if dynamic mapping could be used to investigate complex data sets and develop new hypotheses. To this end, the researchers applied the technique to a data set from 2002 relating environmental exposures to Salmonella infections. They compared infection rates from season to season, across geographic locations in the U.S. and among locations with varying rates of broiler chicken sales.
This analysis led them to conclude that Salmonella infection rates are generally higher during the summer months, are often particularly concentrated in the South, and are frequently correlated with levels of broiler chicken sales. Levels of these sales are an indicator for the amount of livestock production in a particular geographic location.
Principles of user-friendly dynamic mapping
Creating a successful dynamic map requires careful consideration of several parameters, and as part of their study, Naumova and her team defined a set of principles to aid researchers in constructing these maps. For example, researchers must choose an effective color scheme that is well-labeled and that allows for easy differentiation between colors. Another critical factor is the selection of an effective aggregation scheme for the organization of data. Researchers must carefully choose a frame speed that allows viewers to absorb the data in a particular image before moving on to the next image, and the map should incorporate a control interface to allow viewers to move through the images at their own pace.
A tool with potential for epidemiological studies
Dana Hancock, Ph.D., a postdoctoral IRTA fellow at NIEHS, agrees, saying "There is a complex array of environmental factors, time, and geography to consider when investigating patterns of disease occurrence. Dynamic maps provide a simplified way to explore these data and to generate new hypotheses, and perhaps future development of dynamic maps will provide a mechanism to test such hypotheses," said Hancock.
A professor of public health and community medicine at Tufts University School of Medicine, Naumova also serves as the director of the Tufts University Initiative for the Forecasting and Modeling of Infectious Diseases (Tufts InForMID). She plans to present the results of this study in November at the American Society of Tropical Medicine and Hygiene (ASTMH) 59th Annual Meeting.
Citation: Castronovo DA, Chui KK, Naumova EN. (http://www.ncbi.nlm.nih.gov/pubmed/20042115?ordinalpos=1&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_MultiItemSupl.PMC_FreeArticle_ad&linkpos=1&log$=pmcad6_article) 2009. Dynamic maps: a visual-analytic methodology for exploring spatio-temporal disease patterns. Environ Health 8:61.
(Erin D. Hopper, Ph.D., is a postdoctoral fellow in the NIEHS Laboratory of Structural Biology Mass Spectrometry Group.)