American Academy of Microbiology Board of Governors

Rita R. Colwell, Ph.D. University of Maryland (Chair)

Joseph M. Campos, Ph.D. Children's National Medical Center, Washington, D.C.

R. John Collier, Ph.D. Harvard Medical School

James E. Dahlberg, Ph.D. University of Wisconsin, Madison

Julian E. Davies, Ph.D. University of British Columbia, Vancouver, Canada

Harold S. Ginsberg, M.D. National Institute of Allergy and Infectious Diseases, National Institutes of Health

Martha M. Howe, Ph.D. University of Tennessee, Memphis

Mary E. Lidstrom, Ph.D. University of Washington

Eugene W. Nester, Ph.D. University of Washington

Mary Jane Osborn, Ph.D. University of Connecticut Health Center

Moselio Schaechter, Ph.D. San Diego State University

Colloquium Steering Committee

Rita R. Colwell, Ph.D. University of Maryland (Co-Chair)

Jonathan A. Patz, M.D. Program on Health Effects of Global Environmental Change, Johns Hopkins University (Co-Chair)

Robert Correll, Ph.D. Global Change Research Program, National Science Foundation

Mary Gant National Institute of Environmental Health Sciences, National Institutes of Health

Duane J. Gubler, Sc.D. Division of Vector-Borne Infectious Diseases, Centers for Disease Control and Prevention

J. Michael Hall, Ph.D. Office of Global Programs, National Oceanic and Atmospheric Administration

Stephanie James National Institute of Allergy and Infectious Diseases, National Institutes of Health

Dennis Lang, Ph.D. Enteric Diseases Branch, National Institute of Allergy and Infectious Diseases, National Institutes of Health

Nancy Maynard, Ph.D. Mission to Planet Earth, National Aeronautics and Space Administration

Michael A. McGeehin, Ph.D. Health Studies Branch, Centers for Disease Control and Prevention

William A. Sprigg, Ph.D. National Academy of Sciences

Colloquium Participants

Rita R. Colwell, Ph.D. (Co-Chair, Steering Committee), University of Maryland

Jonathan A. Patz, M.D. (Co-Chair, Steering Committee), Johns Hopkins School of Public Health

John Balbus, M.D., George Washington University

Menno Jan Bouma, M.D., Ph.D. London School of Hygiene and Tropical Medicine

David J. Bradley, D.M. London School of Hygiene and Tropical Medicine

JoAnne Burkholder, Ph.D. North Carolina State University

Kristie L. Ebi, Ph.D. Electric Power Research Institute

Paul Falkowski, Ph.D. Brookhaven National Laboratory

J. Michael Hall, Ph.D. National Oceanic and Atmospheric Administration

Scott B. Halstead, M.D. Naval Research and Development Command

Anwarul Huq, Ph.D. University of Maryland

Md. Sirajul Islam, Ph.D. International Center for Diarrheal Research, Bangladesh

Bradford A. Kay, Dr.P.H. Centers for Disease Control and Prevention

David le Sueur, Ph.D. University of Natal, South Africa

Simon A. Levin, Ph.D. Princeton University

Steven E. Lindow, Ph.D. University of California, Berkeley

M. Leonardo Lizarraga-Partida, Ph.D. CICESE, Mexico

Nancy Maynard, Ph.D. National Aeronautic and Space Administration

Michael A. McGeehin, Ph.D. Centers for Disease Control and Prevention

Mercedes Pascual, Ph.D. University of Maryland Center of Marine Biotechnology

Andrew Pearson, M.D. St. Thomas Hospital, London, England

William C. Reeves, Ph.D. University of California, Berkeley

Paul Reiter, Ph.D. Centers for Disease Control and Prevention, San Juan, Puerto Rico

Joan Rose, Ph.D., University of South Florida

Estelle Russek-Cohen, Ph.D. University of Maryland

R. Brad Sack, M.D., Sc.D. Johns Hopkins University

David M. Sander, Ph.D. ASM Congressional Fellow, Washington, D.C.

Ramon J. Seidler, Ph.D. U.S. Environmental Protection Agency, Corvallis, Oregon

Robert E. Shope, M.D. University of Texas Medical Branch, Galveston

Kassem Siddique, M.P.H. International Center for Diarrheal Research, Bangladesh

William A. Sprigg, Ph.D. National Academy of Sciences, Washington, D.C.

Juli Trtanj National Oceanic and Atmospheric Administration

Thomas F. Wilson, Ph.D. Electric Power Research Institute

Byron L. Wood NASA Ames Research Center, Moffett Field, California

Catherine E. Woteki, Ph.D. U.S. Department of Agriculture

Staff

Carol A. Colgan Director, American Academy of Microbiology

Peggy McNult Manager, American Academy of Microbiology

Report Editor

Robert Loper Washington, D.C.

Temperature:

Microorganisms carried by vectors, such as mosquitoes, ticks, and other blood-sucking arthropods, are strongly influenced by temperature of the microenvironment within their cold-blooded vector hosts. The survival rates of vectors and the rates of multiplication and transmission of the microorganisms that infect them are temperature dependent (Fig. 1, Reeves, et al., 1994). The physiological relationship between temperature and the malaria parasite and its mosquito vector exemplifies this dependency. The geographical distribution of the Anopheles mosquito and the four species of malaria parasites which infect humans is restricted by low—and probably also high—temperature. Over the low temperature threshold, the rates of development of the parasite and the vector population increase with temperature, thereby increasing transmission capacity. Temperatures in the 20 to 30˚C range and humidity above 60% provide optimal conditions for mosquitoes to incubate and transmit the malaria-causing parasites.

Figure 1.

Effect of temperature on transmission rates (the percentage of bites that result in transmission) of Western equine encephalitis (WEE) and St. Louis encephalitis (SLE) viruses by the mosquito, Culex tarsalis.

Precipitation:

Precipitation, especially in the form of rainfall, can affect disease transmission via the effects of normal, as well as severe (i.e., flooding and drought), events on vector populations. Again, malaria serves as an example. In the drier regions of the tropics, low rainfall restricts the breeding of vector mosquitoes, and low humidity reduces the number of mosquitoes that survive to the age required to transmit malaria. A preceding drought may play an important role in the development of malaria epidemics in areas such as the Indian subcontinent, South America, and southern Africa. However, the effects of the amount and temporal distribution of precipitation on the transmission intensity of malaria are not well understood. Existing models for predicting malaria incidence as a function of rainfall have a high degree of inaccuracy. Flooding can influence disease transmission in a number of ways, most notably by increasing run-off and disturbing breeding grounds and habitats. The drowning of rodents that are reservoirs for diseases can lead to increased transmission by increasing human contact with contaminated water.

Wind and Ocean Currents:

The potential for variations in wind and wind patterns to promote infectious diseases is a generally neglected aspect of climate-human health interactions. A notable example of this potential occurs in the “meningitis belt” of Sub-Saharan Africa, where an epidemic form of meningococcal meningitis is associated with the climate conditions of the Sahel. Stretching along the southern rim of the Sahara Desert from Mauritania to Somalia, the Sahel is an area of low rainfall and seasonally harsh temperature variations. Meningitis epidemics, accompanied by high mortality, occur during the hot, dry season with 8- to 14-year cycles (Moore and Broome, 1994). Generally, the disease begins with bacterial colonization of the nasopharynx. It has been postulated that the dry, sand-carrying harmattan winds cause mucosal damage that facilitates the entry of the meningitis-causing bacteria into the body.

Finally, recent research has demonstrated that ocean currents and tides are connected with plankton blooms and cholera epidemiological patterns (Colwell, 1996). Sea-surface temperature, height, and concentration of nutrients in seawater are associated with cholera incidence through their effects on plankton blooms, phytoplankton being a food source of the zooplankton that carry Vibrio cholerae as part of its natural microbial flora.

El Niño-Southern Oscillation:

The unseasonably warm temperatures, increased amounts of rainfall, and/or flooding associated with El Niño have been linked to outbreaks of leptospirosis, Rift Valley fever, hantavirus pulmonary syndrome, malaria, and Ross River virus, among other diseases.

The El Niño-Southern Oscillation (ENSO) is a dramatic and well-documented recurrent climatic variation that involves a warming of surface waters in the equatorial Pacific, a decrease in barometric pressure in the eastern Pacific, and the weakening of easterly surface winds. These events alter the distribution of rainfall in the tropics and lead to changes in weather patterns over much of the globe. The unseasonably warm temperatures, increased amounts of rainfall, and/or flooding associated with El Niño have been linked to outbreaks of leptospirosis, Rift Valley fever, hantavirus pulmonary syndrome, malaria, and Ross River virus, among other diseases. Recent studies have found an apparent link during the unusually long 1991–95 El Niño between higher seawater temperatures off the coast of Peru and in the Bay of Bengal and an increase in cholera cases in parts of South America and the Indian subcontinent, respectively.

Vectorborne Diseases:

Climate factors and weather conditions affect the prevalence and distribution of vectorborne diseases by influencing the reproductive success of arthropods and small rodents that transmit pathogenic organisms. By extending the normal range of the disease vector and/or by providing a more hospitable environment for its reproduction, variations in rainfall and temperature can be instrumental in the occurrence of epidemics in areas in which the particular disease is not endemic. Temperature can also be a significant factor in the transmission of viral diseases spread by mosquitoes because warmer temperatures shorten the extrinsic incubation period of the virus, i.e., the time it takes for the virus to transport through the gut and reach the salivary glands of the mosquito.

Malaria.

Malaria is endemic throughout most of the tropics, with a high prevalence in Africa, the Indian subcontinent, Southeast Asia, many South Pacific islands (excluding Polynesia, and parts of South and Central America and Mexico. (It should be recalled that both malaria and yellow fever were endemic in some regions of the United States until the turn of this century.) Some 2.4 billion people live in areas of risk (WHO, 1996), and approximately 350 million new infections occur annually, resulting in over 2 million fatalities. Four species of the Plasmodium protozoa, transmitted by Anopheles mosquitoes, cause this disease in humans. Untreated, malaria tends to be a lifelong affliction, with recurring episodes that last one to four weeks. Its symptoms include fever, headache, and malaise; death (predominantly in young children) may result from neurological damage and progressive coma, pulmonary edema, kidney failure, or shock caused by the collapse of the vascular system.

According to the WHO, malaria is one of the most climate sensitive among vectorborne diseases (WHO, 1996). Malaria provides a well-documented example of the interactions among parasites, vectors, and hosts that affect the transmission of vectorborne diseases. Within the bioclimate envelope, outside of which malaria is not transmitted, transmission intensity depends on interrelations among factors, including vector density, biting rate, and survival of the mosquito. The longevity of the vector is particularly significant because the parasite must undergo a process of development within the mosquito before the disease can be transmitted. This process, termed sporogony, can take 8 to 35 days and occurs more quickly at an ambient temperature between 20 and 27˚C. Outbreaks of epidemics in highlands or desert fringes, where malaria is not endemic, are strongly influenced by temperature or rainfall, respectively. Figure 2 shows the cyclical pattern of epidemic malaria associated with El Niño in Venezuela.

Figure 2.

Malaria in Venezuela During El Niño

Waterborne Diseases:

Waterborne microorganisms have been estimated to account for as much as 80% of the annual mortality due to infectious disease (Clark, 1993). Evidence suggests that warmer temperatures are correlated with an incidence of some of these pathogens, including the bacterium that causes cholera (Colwell, 1996). Flooding and increased run-off are also causes of increased disease transmission to humans.

Cholera.

Cholera is a debilitating, potentially fatal, disease caused by the bacterium Vibrio cholerae. It is spread by bathing in or drinking contaminated water and by ingestion of contaminated food. There have been seven cholera pandemics (worldwide epidemics) since 1816. The seventh and current pandemic began in 1961 and is the longest one to date. Cholera symptoms include explosive, watery diarrhea; vomiting; and abdominal pain. Mortality rates are highest among untreated patients under 2 years of age. Cholera is now affecting more nations worldwide than at any other time during the 20th century (Watson, Zinyowera, and Moss, 1996), and it remains endemic in India, Bangladesh, and Africa. In the United States, the bacterium is present in the Gulf Coast of Texas, Louisiana, and Florida, the Chesapeake Bay, the California coast, and coastal waters to the north.

Cholera outbreaks occur seasonally and are associated with monsoon seasons, warm temperatures, heavy rainfall, and increased plankton populations. V. cholerae occurs in riverine, brackish water, and estuarine ecosystems, being part of the natural flora of plankton. V. cholerae is found in the gut of, and attached to the surface of, both freshwater and marine copepods. The bacterium reproduces on the copepods and appears to be able to enter a dormant, nonculturable state during seasons of zooplankton decline. The factors that trigger entrance into and exit from this dormant state are as yet undetermined, although evidence for genetic regulation of the process has been obtained. Associations among sea-surface temperature, sea-surface height, and disease incidence have been found in Bangladesh (Colwell, 1996) (Fig. 3, Colwell, 1996), and links between El Niño and the 1991 and 1997–98 epidemics in South America are being investigated. A current hypothesis suggests that outbreaks are related to plankton blooms associated with warmer sea-surface temperatures (Colwell, 1996). The phytoplankton blooms are a food source for the copepods upon which the cholera bacterium thrives. Movement of tidal waters carry the algal blooms and copepods towards land and into rivers, bringing the bacterium into contact with humans who use this water for bathing or as a source of drinking water.

Figure 3.

Relationship between sea-surface temperature and cholera case data in Bangladesh from January to December 1994.

Foodborne Diseases:

Transmission of infections via food causes significant illness in both developed and developing countries. The microorganisms responsible include viruses, bacteria, and protozoa. Morbidity depends on the pathogen, the susceptibility of the human population, and the medical care available. Diarrheal diseases show such a strong cyclical occurrence that climate is widely assumed to play some role in all epidemic outbreaks. During periods of higher ambient temperatures, the prevalence of foodborne illnesses caused by pathogens that replicate on foods at ambient temperatures is likely to rise. Production of staphylococcal toxins is also associated with temperature.

Other Foodborne Diseases.

Cyclospora cayetanenis is a recently identified diarrheal disease pathogen. Data collected in Peru indicate an association between temperature and prevalence of infection (Fig. 4, Madico, 1997). Disease incidence during the winter of 1997 was higher than usual, due to warmer conditions resulting from the strong El Niño. Salmonella enteritidis is a leading cause of food poisoning worldwide. Infection results from eating improperly prepared, contaminated food, usually meat or dairy products, but also uncooked or undercooked eggs. Seasonal peak of its incidence coincides with peak climatological temperatures.

Figure 4.

The prevalence of Cyclospora cayetanesis infection in children younger than 18 years of age and the air temperature in Lima, Peru (September 1992–July 1994). Lima is in a desert region where the rainfall is <2 cm/year.

Airborne Diseases:

Most airborne diseases are spread through respiratory tract secretions and can be carried by the wind or be transmitted within buildings by air-handling systems. Inhaled wind-blown spores are also a source of infectious disease. The seasonal variability of many respiratory diseases suggests a weather-related factor in their transmission.

Plant and Fish Diseases:

Weather-related conditions can indirectly influence human health by stimulating plants to produce a range of toxic chemicals, called mycotoxins. Consumption of such plants can cause negative health effects. Depending upon the plant, inadequate water or excessive salt or moisture at harvest can increase the likelihood of fungus infection and the production of mycotoxins. Perhaps the best known mycotoxins are the aflatoxins, a group of potential liver cancer-causing chemicals produced by Aspergillus flavus, a mold that grows on grains and peanuts. Mycotoxins may indirectly influence the susceptibility of humans to other diseases because chronic exposure to mycotoxins can depress the immune system. Moisture can also facilitate infection of plants by a variety of pathogens that induce the plants to produce defensive compounds (phytoalexins) to protect themselves. These compounds are known to be toxic to humans and animals. Fish diseases are cyclical, but the causes of periodicity are relatively unknown. Diseases in fish that are caused by rhabdoviruses have been cured by raising, or in some cases lowering, the temperature of the water in which the fish are grown. This technique is well described for four fish diseases and is used in aquaculture and mariculture (Shope and Tesh, 1987).

Field Studies and Long-Term Data Sets:

Field studies are generally designed to be intensive, short-duration analyses of selected phenomena. Often they are undertaken in response to an outbreak of a specific disease and may focus on the immediate factors involved at the epidemic site. Long-term data sets are developed over time and provide more extensive information on the pathways connecting climate, ecosystems, and human health. There are important intrinsic relationships between field studies and the development of long-term data sets. Not only do field studies provide the basis for creating long-term data sets, but the results obtained in each type of study are used to refine further studies of the same or similar phenomena. Thus, the patterns and associations evident in long-term studies can be used to orient the quick-response, intensive investigation of specific disease outbreaks.

The issue of data quality is of paramount importance for field studies. The following discussion is limited to general concerns about the circumstances under which data on climate variables and disease incidence are collected. Determination of the well-foundedness of particular theoretical explanations of disease-climate interactions was not within the scope of the colloquium. A subsequent section of this report addresses the importance of databases for research on links between infectious disease and climate.

Meteorological data tend to be of comparable quality, despite being gathered at different recording stations around the globe. These data include information on day and night temperatures, winds, relative humidity, ice and cloud cover, sea-surface and cloud-top temperatures, and atmospheric CO₂ and nitrous oxide concentrations. Morbidity and mortality data are much less standardized, in terms of collection method and recording format, and are subject to variations in reliability. Worldwide, only a fraction of infectious disease cases are treated, a consideration that affects interpretation of the data gathered in medical or public health settings. The reliance upon passive reporting can also lead to problems in the under-reporting of the disease burden, for reasons of national chauvinism or economic considerations, such as tourism. There are, however, good-quality data sets on human disease. The World Health Organization (WHO) is the primary agency for the worldwide reporting of infectious disease, and its program on polio eradication, for example, is a model for the collection of reliable data.

Important questions remain to be answered in terms of whether the right environmental, ecological, and weather data are being collected for meaningful analysis. There are also uncertainties to be resolved regarding the integration of data from diverse studies, the design of interdisciplinary analyses, and the establishment of more complete and systematic monitoring and surveillance programs. One promising method of investigating the pathways connecting climate, pathogens, and human health is the measurement of genetic markers of particular micro-organisms as a means of tracing the pathogen through the ecosystem. Large-scale weather disturbances, such as El Niño, offer opportunities for cross disciplinary collaboration on research projects using different methodologies to analyze the effects of such phenomena. One example of such research is an ongoing study being coordinated by the National Oceanic and Atmospheric Administration (NOAA) and the International Research Institute to examine the effects of the 1997–98 El Niño-related weather events on infectious disease and human health (The ENSO Experiment).

Well-designed field studies are also needed to assess the impact of different microenvironments on the incidence of infectious disease. Microenvironmental variables can affect the viability of pathogens and can create or enhance conditions that are conducive to disease transmission. Although laboratory studies have demonstrated the adverse effects of high temperatures on survival of mosquitoes that carry dengue fever, field studies have not yet documented this relationship. This is most probably a result of the vector's survival in microhabitats, such as underground storm drains, houses, or even leaf cover.

Modeling Synthesis:

Models and field studies are used in complementary ways to analyze the connections among climate and weather factors, disease incidence, and human health.

Modeling studies are essential to research on the links between climate and infectious disease. Models are used to represent relationships among variables and to organize data and formulate hypotheses. Models and field studies are used in complementary ways to analyze the connections among climate and weather factors, disease incidence, and human health. To paraphrase the philosopher Immanuel Kant, without field studies models are empty, but without models field studies are blind. Improving the availability, quality, and applicability of data is key for enhancing accurate modeling. One of the primary goals of model building for research on climate-disease links is the development and refinement of predictive models relating climatic variables to disease incidence. Such models have been developed most extensively for vectorborne diseases and are useful for demarking areas of potential risk.

Models are tools designed for specific purposes and range from simple conceptual representations to complex mathematic systems. In the type of research being discussed here, models are used to relate data on weather, ecology, and disease to support explanations and predictions of events such as disease outbreaks. The data used as the informational foundation for these models come primarily from field studies and contemporary data sets, but are also drawn from historical meteorological data series and health statistics, as well as geographical comparisons. The gathering of data through diagnostic studies of disease and disease parameters in real time is an essential activity in both constructing models and testing their validity. Over time, a “virtuous circle” of interactions among model development, data collection, and model testing can generate more sophisticated models of the dynamic relations within the ecosystem. Existing data on weather variables and disease incidence go back more than 100 years for many of the developed industrialized countries and for some of the developing countries as well. This historical record is a valuable source of information on baselines and trends in incidence patterns and climate variables. Geographical comparisons of the disease burden in different climates are another important resource for gaining insight into the effects of variables, such as temperature, rainfall, and humidity, on disease prevalence.

Most of the modeling that is done of linkages between environmental variables and human diseases has a single-disease focus. In this approach, the potential impact of various climate- or weather-related factors on the incidence of a particular disease, such as cholera, is investigated. Quite often, the variables of interest are disease specific. Another approach, which is more commonly used in analyzing plant rather than human diseases, is known as deterministic modeling. Instead of focusing directly on one disease, a set of system variables—such as temperature, humidity, ultraviolet radiation, cloud cover, salinity, and vegetation—is modeled to gauge how these variables determine such things as, for example, the size of a vector population. These models are quite complex and have proved to be difficult to validate.

Issues of geographical scale and duration of disease transmission and questions of relevant environmental parameters and disease indicators are central to model design. Depending upon the disease and the interactions of importance, the geographical scale for models of weather-disease links can range from less than 1 to more than 200 km₂. In the case of malaria, an event such as El Niño can affect the normal geographic distribution of both the disease vector and parasite. Consequently, a large-scale model is needed to assess the range of such an event's effects. Certain diseases exhibit peak incidence periods that correspond to seasonal cycles of temperature and rainfall. Models must be designed to take into account the occur-rence of the peaks, their duration, and the interval between them. Biotic factors, such as vector species and vectorial capacity, can also alter the dynamics of what constitutes a minimum window of transmission. A variety of models are needed, not only to organize and assess the new data that are being collected, but also to extract relevant information from the data that are already available.

The impact of human behavior is a potentially significant confounding variable for assessments of the effects of weather-related conditions on disease prevalence and occurrence. Individual and group behavior can cause changes in transmission patterns or exposure to vectors and can affect the survival and distribution of the pathogens themselves. Inadequate sanitation and drinking water, overcrowding, and improper use of antibiotics are some of the factors that can contribute to an increase in disease incidence. Alternatively, expanded access to clean water and improvements in housing and medical care can reduce the infectious disease burden. Fully integrated climate/disease models must incorporate the impacts of human activities, as well as variations among vector species and their breeding ecologies, in predicting disease risk.

Key Databases:

A catalogue of existing databases relevant to research into climate-disease links would be a useful addition to the literature. Compiling such a list, however, would be a serious undertaking in itself. There are numerous such databases, ranging from those maintained by international agencies to small data sets compiled by discrete research communities. The following overview is merely illustrative and serves to highlight the importance of developing data information systems that will allow researchers to make optimal use of various kinds of available information.

The World Weather Watch of the World Meteorological Organization (WMO) encompasses ca. 800 stations operated by the weather services of nations around the globe. The network comprises a number of weather, climate, and hydrology efforts intended to support government, commercial, and academic interests and activities. These interests include the optimal scheduling of agricultural planting and irrigation, planning for energy distribution systems, and basic research on environmental problems. Information on current weather, changes of seasons, and long-term climatologic trends and characteristics is documented.

The WHO supports a major international database focused on the incidence of infectious diseases worldwide. The CDC maintains data on infectious diseases in the United States, including, for example, information on dengue fever in Puerto Rico. Various databases of the National Oceanographic and Atmospheric Administration (NOAA) hold important information on environmental parameters, including cloud cover and atmospheric and sea-surface temperatures.

Many more data sets are maintained by smaller communities of scientists and may not be familiar to other potential users. For example, there are Japanese, Indian, and Brazilian databases composed of satellite meteorological observations. A global change master directory is being produced which is intended to become a resource for international weather and climate information. Global climate information and related data being collected in the United States are now required to be maintained in a format compatible with this global change master directory. Satellites that collect information on weather conditions over land and sea provide an increasingly important source of information. The Tropical Rainfall Measuring Mission, a joint project of the United States and Japan designed to measure rainfall in the tropics, is an example of a new database that is being developed. The data obtained from this project will be correlated with disease incidence to establish a matrix for analyzing the correlations between rainfall and infectious disease.

Database Accessibility:

Few of the existing databases are either coordinated or designed to communicate with one another, a situation that hinders the exchange and use of information among research teams within and across scientific disciplines. The CDC is experimenting with a data information system capable of “translating” electronic data from various medical data systems. This system requires that the electronic data be in a data format known as Health Level 7 (HL7). HL7 is not an information program per se, but is a method of formatting electronically transmitted information. This data format defines where in the data stream certain information bits will be located. It does not define the information itself. Thus, scientists can use their own local codes or standardized dictionaries to download the information. CDC, commercial software companies, and health care providers have agreed to use HL7 as the standard means for formatting electronic data. CDC's demonstration project, employing the HL7 format, in collaboration with several state public health departments, allows public health practitioners to access medical information from hospitals, physicians, and private clinical laboratories. This system of “translating” information from diverse data programs, thereby making data from diverse sources accessible to difference users, is promising and should be expanded.

Progress in research depends on the availability and accessibility of quality data, yet financial considerations concerning access to data are becoming increasingly an impediment. The flow of information to researchers is threatened by attempts to commercialize data that historically have been freely available for scientific inquiry. The WMO has proposed mechanisms for charging for the cost of reproducing and mailing data compiled by global monitoring systems. The World Intellectual Property Organization has advocated the copyrighting of digitized data sets. Such proposals are harbingers of restrictions that may ultimately determine who can afford to do this kind of research. At the very least, this situation indicates the economic value of data sets and the need to view data management as a legitimate cost of research, much like re-agents and laboratory equipment. Funding agencies should acknowledge these costs and be prepared to accept them as part of the cost of doing research.