Geography Helps Epidemiologists To Investigate Spread Of Disease

'CLEAR AWARENESS': Keith Clark says epidemiologists have recongized the importance of geography in studying infectious diseases. Adventurers of the 18th and 19th centuries in search of gold and new trade routes were not the only ones to value a good map: Early epidemiologists inspected the lay of the land in attempts to discern the causes and spread of diseases. But as unexplored frontiers slowly disappeared, geography came to be taken for granted. In fact, the number of classic epidemiology papers that take spatial patterns into account can be counted "on the fingers of two hands," according to Keith Clark, a geographer at the National Center for Geographical Information and Analysis in Santa Barbara, Calif., one of three centers in a National Science Foundation-funded research consortium. The emergence of AIDS and antibiotic-resistant bacteria has fueled a rush to reestablish monitoring programs that were practically abandoned. Previous attempts to revive an interest in so-called medical geography had foundered, notes Baruch Blumberg, a senior member at the Fox Chase Cancer Center in Philadelphia. He recalls the brief life of the Geographic Medicine and Genetics section of the National Institutes of Health that he formed around 1960. Set up to study the geographic distribution of polymorphisms (or variations) of specific human genes, the section soon found itself an outsider at NIH. The problem, he believes, stemmed from a conflict between biochemistry, which considers only a small number of variables in a given experiment, and epidemiology, which is necessarily more complex. "Few others in the institute were interested in bridging that gap," he says. However, a new report from the National Research Council entitled Rediscovering Geography (Washington, D.C., National Academy Press, 1997), released this year, is calling for geographers to work closely with other scientific disciplines. Clark is just one of many geographers participating in the reemerging partnership between geography and the fields of epidemiology and public health. "[Epidemiologists] are interested in being able to model the spread of disease by correlating populations or the movement of potential natural hosts, and that is dependent on an understanding of geography," says Stephen Morse, director of the program on emerging diseases at the Columbia University School of Public Health. "There seems to be a clear awareness among the [epidemiological] community that . . . we need to incorporate [geography] into the suite of tools" for studying emerging infectious diseases, adds Clark, who has served on several NIH funding panels. Not surprisingly, much of that funding is focused on projects that sit prominently in the public eye, such as surveillance of hantavirus and Ebola. But the routine stuff-improvement of facilities and labs, and building the overall infrastructure-tends to be overlooked, says Clark. Meanwhile, new technological advances-such as the hand-held global positioning system (GPS), a military by-product that can display the user's position to within a few meters-allow public health officials and epidemiologists to track diseases and outbreaks with far greater ease than the compasses and star-charts employed by their predecessors. The GPS instrumentation relies on a constellation of orbiting satellites to pinpoint locales. Inexpensive models, which cost about $200, are able to name the user's position to within 100 meters. The fancier versions, which come with a higher price tag of around $10,000 to $15,000, are accurate to within a few meters. Today, researchers are using the GPS system in remote areas throughout the world to "map on the run." Surveyors can plug the GPS data into a global information system (GIS) workstation, which processes it into customizable maps. Scientists from both sides of the fence agree that the technology is bringing fruitful collaborations. Geography allows researchers to examine the effects of climate patterns, weather patterns, and spatial clusters of disease cases "to see where there might be vulnerability to outbreaks," explains Paul Epstein, associate director of the center for health and the global environment at Harvard Medical School. MAPMAKER, MAPMAKER: Allen Hightower of the National Center for Infectious Diseases uses differential GPS equipment in front of a field station in Western Kenya. "In any large-scale field study, one of the first things you do is make a map of the area. We used to hire local people to make sketch maps, but updating them was awkward, and they were difficult to archive. [With GPS] you can [take map coordinates] while you're taking a census of a village," points out Allen Hightower, chief of data management for the division of parasitic diseases at the National Center for Infectious Diseases in Atlanta. Custom mapping using GIS workstations is a snap, reports Hightower: "If you want just where the schools are [for example], you can do that." The instrument and software package together give surveyors a powerful tool to map the progress of epidemics and the organisms that cause them, says Hightower, who has used the system for projects in Kenya aimed at halting the spread of malaria and its carrier mosquitoes. The team began by trapping mosquitoes in areas hit hard by malaria, mapping the abundance of the mosquitoes in relation to the nearest rivers, roads, and facilities. They also collected altitude, rainfall, and temperature data to complete the picture of the area. "We wanted to include spatial factors [that] a researcher could include in any [malaria project]," he explains. He is also working on a related study to determine whether insecticide-impregnated bednets reduce childhood malaria mortality. Since the mosquitoes that carry the malaria parasite bite only at night, many propose these barriers as a deterrent. Using a test area of nearly 150 villages in a 75-square-mile area, the team hopes to verify whether or not such efforts will be worth their considerable expense. In the first year of the study, half of the houses will be equipped with bednets, and the others left as controls. In the second year, the team will provide bednets to the remaining households, and then note any reductions in childhood mortality. The GPS mapping will help demonstrate whether bednets protecting one village divert the mosquito populations to unprotected neighbors. It will also target problem areas, helping officials better allocate their resources, Hightower points out. TECHNOLOGICAL AID: Field staff operate GPS units in Western Kenya. GPS data from field sites and the computer center are processed by a computer program that computes the calibrated locations. In Egypt, the GPS/GIS combination has already contributed to the control of schistosomiasis-the result of a waterborne parasite, carried by snails, that invades through the skin and causes lethargy and kidney damage. According to Clark, who consulted on the project, there were a "huge number" of paper records of schistosomiasis testing that had been done at schools and health clinics throughout Egypt. The records included good demographic and patient health data, Clark recalls, but they had one rather telling weakness: The only geographical information on each record was the name of a local school where the information had been gathered. Egyptian officials wanted to find the hot spots in the canal system so that they could control the responsible parasite in problem areas. The team went to all of the schools with a GPS receiver, generated a map, and then entered the data. They next overlaid the schistosomiasis data onto a map of Egypt's canal system, bringing into clear focus the trouble spots in the canals. "It was a proof of concept" that GPS can be used to improve public health treatments, contends Clark. The GPS techniques will likely need refinement, as researchers collect data and set up models to track disease transmission, only to find later, perhaps, that one or two crucial details were left out of the survey, notes Hightower. "A lot of what we're doing [with the GPS system] is not going to work very well-we'll have to step back and reconsider what we do." GPS isn't the only tool that geography has to offer public health officials and epidemiologists. The complexities of a fast-moving world and easy international travel may require more complex statistical models to account for disease spread in today's geography. FLIGHT PATTERN: Penn State's Peter Gould has modeled the spread of HIV using an air travel matrix. For example, Pennsylvania State University geographer Peter Gould has modeled the spread of HIV using an air travel matrix. "Except for cases of contaminated blood transfusions and hemophiliac injections, HIV needs human beings and their [cells] to exist, and it goes where they go," Gould notes. "If geographers can predict to some degree how people move, they ought to be able to predict how HIV moves." This would help researchers anticipate future health care needs. In one model, Gould inserted HIV "infections" into one or more population centers, then assigned a transmission probability, and allowed the virus to spread through the airline system. The epidemic ends up with similar distributions of cases no matter where the virus is first placed: A starting point of Peoria, Ill., eventually achieves the same distribution as when the epicenter is New York City or San Francisco. The models closely match the known historical distributions of the disease itself, suggesting that HIV depends on the movement of people-not environmental barriers or time factors. Gould has also developed three-dimensional maps of the early years of the AIDS epidemic, depicting case concentrations in large cities as peaks rising above a flat plain. The peaks soon rise to greater heights as more cases develop, and ridges arise along roadways to connect the cities. Interstates 87 and 90 in New York State form one corridor, soon joined by the Interstate-5 artery connecting San Diego and Seattle. "Such effects should come as no surprise: Major highways are built to facilitate the movement of people, and therefore the HIV," says Gould. FINDING DIRECTION: Harvard's Paul Epstein notes that geogrpahy allows researchers to spot areas vulnerable to outbreaks. Indeed, an emerging disease that would have remained a local problem 50 years ago now has the means-through international air travel-to rapidly become a global threat. "Particularly if you have a disease whose incubation period is longer than its travel time . . . if you have someone on a plane, and they're not symptomatic, [sick] people will soon be scattered all over," points out Jonathan Mayer, an adjunct professor of geography and public health at the University of Washington. MODEL COLLABORATION: Columbia's Stephen Morse says geography is essential for epidemiologists in modeling the spread of disease. The potential for rapid spread of an easily transmitted disease has led to a new concern about two arcane killers: influenza and tuberculosis. With the possible exception of Stephen King, who made the "superflu" the carrier of destruction in his novel The Stand, few in the popular media want to hear about the flu virus. But the time and effort that the Centers for Disease Control and Prevention puts into selecting the right flu vaccine every year is inspired by a simple fact, say epidemiologists: The flu virus has killed far more people in history than AIDS, hantavirus, Ebola, and any number of other headliner microbes. Every 20 years or so a new strain of the flu arises, bringing with it an increased death toll and the threat of a new epidemic to match the 1917-18 outbreak, which claimed 18 million to 20 million people worldwide. Tuberculosis and the flu are especially worrisome because they reproduce on the outer endothelial cells, which gives their progeny easy access to the respiratory tract, and hence to the air we breathe, says Mayer. "That makes it much easier to transmit than something like Ebola, where you practically have to put your hand in a surgical incision to get it," he states. As barriers of perception and technology continue to fall, geography will likely play a more prominent role in epidemiology, predicts Morse. But there is much work yet to be done: "There's always been interest in [employing geography in epidemiology], but there hasn't been much impact because people are only gradually becoming aware of how to analyze these phenomena. . . . Modeling geographic processes and the spread of diseases is still a difficult [task]." James Kling, a science writer based in Bellingham, Wash., can be reached online at jkling@nasw.org.

Geography Helps Epidemiologists To Investigate Spread Of Disease

Jul 20, 1997

Interested in reading more?

Become a Member of