present state of knowledge, some address methods. My presentation today will, I

Thank you for including me in the Metalund series of seminars for this year. I have attended and enjoyed a number of the presentations covering indoor air pollution in children’s bedrooms and schools, cooking stoves and indoor air pollution, light and health, mercury and children’s health, Jonas Björk’s excellent presentation on acyclic digraphs, and last week’s talk by Kim de Jong on genetics, behavior, the

environment, and COPD.

From this series, I have seen that Metalund seminars are presented in a variety of formats, some are straight descriptions of research, some provide overviews of the present state of knowledge, some address methods. My presentation today will, I hope, be in line with the idea of presenting information to stimulate discussion and the exchange of ideas to help us do the kinds of research that will advance human health.

1

This seminar started with a question posed by Kristina Jakobsson. When we were discussing my plan of work for the Visiting Professorship and scheduling a Metalund lecture, I asked her what she thought I should talk about.

She told me she wanted me to talk about what we can do with GIS in occupational and environmental health.

After my meeting with Kristina, I took some time to reflect on this assignment. GIS have been used in health research for several decades, yet they are still not widely and deeply embedded in clinical care or public health and epidemiology.

So, the purpose of my presentation today, in order to answer Kristina’s question, is to help us understand what GIS and related technologies are and the key difference they have made in our ability to describe, analyze, and report information on our living world, to reflect on the different perspectives of clinical medicine, public health and epidemiology, and environmental science, and to show how the spatial perspective enabled by GIS can integrate these perspectives to advance human health.

3

At the outset, it is important to distinguish the terms GIScience, geographic data, and GIS.

GIScience is the research field seeking answers to fundamental questions about how we represent and analyze geographic data [in Europe and in Canada, this is

sometimes referred to as “geomatics.”]

Geographic data are data resulting from observation and measurement of earth phenomena, referenced to their locations on the earth.

A GIS is a digital system for integrating and analyzing geographic data.

Geographic Information Systems are the tool for modeling our living environment.

They are to people who study the earth’s surface what the telescope was to

astronomers, what the microscope was to biologists. GIS help us see our environment in new ways.

Today, anyone who is concerned with the environment comes into contact with GIS and related geospatial technologies for capturing, integrating, and analyzing data.

Most researchers who are not actively engaged with GIS think of it primarily as a means for producing maps or other spatial representations of data. Today, I hope that I broaden our understanding of the role that GIS play in implementing spatial

analyses of geographic data and the role that these systems can play in disseminating data. Methods of spatial statistical analysis have been the focus of my visiting

professorship here in Lund.

All data on human health are implicitly geographic data. The people who breathe our air, commute on our buses, work in our factories and fields, and visit our clinics can all be located in space. GIS requires us to make these data explicitly geographic; the locations of observations become part of the data we collect, manage, and analyze.

One barrier to the greater use of GIS in occupational and environmental health is that the clinicians, epidemiologists, and environmental scientists concerned with health do not all view the environment in the same way.

5

In clinical medicine, the focus is on the individual person, perhaps the family or household. When the patient is seen in a clinic, the environment of the person may not be well understood by the clinician or of concern to the clinician at all.

Occupational medicine recognizes that where a person works may be an important factor exposing the person to health risks. Here, for example, we may have a clinician with expertise in ergonomics assist the person in setting his chair at a proper height to ease strain from working at a desk with a computer.

The information of interest is recorded in a medical record which may include location data such as residence or clinic location or workplace or not.

In public health and epidemiology, the focus shifts from the individual to the population. Information on the health of populations is usually recorded in the form of tables with one record for each individual. Sometimes, the table contains

geographic data, most often describing where the person is living at the time the data were collected.

In public health and epidemiology, these data are usually aggregated for analysis and reporting. Sometimes, the aggregation is spatial and most often based on political and administrative units, for example, aggregating individuals to one of the 33

municipalities or communes of the county. The analysis itself, however, is rarely spatial, meaning that distance relationships between places are not explicitly accounted for in the analyses.

7

In environmental health, we place individuals in their real environmental contexts.

We seek to understand both where people live (Malmö) and work (Lund), and their movement to other places in the environment, for example, their commuting routes including the walking route from the Lund C station to Alwallhuset and where they play football.

We also model the environmental risks to which people are exposed and the health and social resources to which they have access within their living environments.

Our view of the environment recognizes its complexity as the product of physical processes and human activity. We retain as much as possible the individual-level perspective of the clinician on human health, mapping individual locations of residence, work, and so on, as well as the distributions of individual places like the set of places where workers butcher meat using different methods, or cut hair, or perform surgery.

This map, downloaded from the VASYD website this morning and showing the area where the recommendation to boil water for drinking and cooking is in effect due to the recently detected levels of E. coli in the public drinking water supply, reminds us that the world has an interesting way of calling our attention to the importance of taking the spatial view of data.

http://www.vasyd.se/SiteCollectionDocuments/Vatten%20och%20avlopp/

Dricksvatten/Lund/Karta_Påverkade_Områden_Lund.pdf

9

The differences between clinical medicine, public health, and environmental health are not limited to views of the environment. This table is adapted from a presentation by Ted Schettler of the Science and Environmental Health Network given in 2005.

The differences in focus, scale in time and space, education, ethics, evidence, and goals across these disciplines help us, perhaps, to understand why GIS is not more widely used in health research or not used as fully as it could be.

Despite these differences, there are important similarities in clinical medicine, public health, and environmental health.

These disciplines are empirical, that is, we make observations and take measurements.

These disciplines are analytical, meaning that we analyze the data we collect.

We are also practical, meaning that we use the results of our analyses to influence individuals and societies. Advancing human health means finding some combination of changing the individual, changing behavior, changing the environment, or

changing behavior in the environment.

11

Here are some examples of how researchers in different disciplines observe and measure the phenomena of interest. In the clinical setting, there is increasing use of instruments to collect data and this has lead to an enormous increase in the amount of data clinicians must analyze. Epidemiologists have traditionally used surveys of samples or populations to collect data and then develop measurement scales. In environmental health, we use direct and remote sensing methods to capture data on environmental quality. Very often, we use data collected by other agencies, especially government agencies, on environmental conditions.

We also use a wide range of methods to analyze the data we collect.

Camilla Dahlqvist was kind enough to speak with me about some of the studies she has worked on with her colleagues in AMM. Methods like calculating prevalence odds ratios and confidence intervals are used in some of these studies.

Camilla and I discussed the importance of understanding the total context of the person. For example, although teachers might be considered to experience mostly psychological stress in their jobs, rather than physical stress, Camilla learned that one of the teachers had horses and spent a lot of time caring for them. This physical work, though not job related, might have an impact on neck and shoulder pain.

Epidemiologists often use a wide range of multivariate inferential statistics to analyze the data they collect. These statistics are usually non-spatial, although there are growing numbers of epidemiologists embracing GIS and spatial analytic methods. An interesting aspect of this study of PFC exposure in groups of men is that the data include individual, behavioral (food consumption), and general environmental or place differences. Men from three different countries were included. We can think of health problems in terms of how individual, behavioral, and environmental factors come together in particular places to affect health in the local population.

Environmental health analysts integrate data from a wide range of sources and the preparation of each database often involves multiple analyses. Special software like GIS is usually required to manage, integrate, and analyze the spatial databases on individuals and environmental conditions, as in this study of exposure to noise.

13

The results of clinical studies are often preserved in individual medical records and summary reports. In addition, as in epidemiology, scholarly journals are also an important outlet for reporting the results of research. In environmental health

research, spatial representations including maps are important for presenting results.

Maps have been used for observation and measurement, analysis, and reporting of health problems for a long time This map shows us part of the rich history of geospatial methods implemented before the advent of the new technologies. The map, from 1819, plots yellow fever cases near Old Slip in Lower Manhattan. A version of this map, which is in the public domain, is in the U.S. National Library of Medicine along with earlier yellow fever maps for cities dating to 1796. They were made to investigate the cause of yellow fever at a time when its etiology was unknown.

There are strong connections between medical mapping and the development of thematic cartography. Thematic maps are maps that show a spatial pattern. They are a form of statistical graphic and they are not like topographic maps which are used for wayfinding.

These maps show highly disaggregate, individual-level data mapped by residence.

The cases on this map are also labeled to show the temporal order of “sickening” so we know that time-space patterns were of interest in 1819. Unfortunately, we will not be able discuss time and time-space methods today. I hope you will be able to

consider these in another lecture.

A map like this makes the data spatially extensive; we can observe the individual- level phenomenon in many places.

But, at the time this map was made, we could only see this level of detail for small geographic areas.

15

A key technical advance associated with the development of GIS is breaking the link between spatial extent or the size of the area we are looking at and the level of detail we can observe.

GIS have enabled us to represent and analyze large volumes of individual-level data for members of populations across large geographic areas.

As such, they enable us to create a link between the individual, as viewed by the clinician, the population, as viewed by the epidemiologist, and the environments where people are exposed to risks and resources affecting their health, as viewed by the environmental health analyst.

GIS can help with data collection.

In clinical studies, individuals may be enrolled because they come to the hospital seeking help. This may lead back to a workplace. In addition to this approach, GIS can be used to map locations of, for example, all of the schools in the county or all of the places in the county where meat cutters work, or all hairdressing salons. Then, a sampling scheme could be designed. Of critical importance here is to recognize that we cannot assume that populations are uniformly distributed by characteristics such as occupation and commuting patterns in the environment.

In epidemiology, GIS can help collect and manage data at an individual level throughout the various phases of the study so that we do not have to aggregate data beforehand, perhaps obscuring the very patterns of health and disease we are trying to observe.

On the environmental side, GIS can help us collect and analyze data on the built environment as well as the natural environment, and on how people themselves evaluate environmental quality.

17

For data to be useful in GIS and spatial analysis, they must be explicitly spatial. We need to understand that location is part of the data we must collect. For individuals, this includes home location, locations of other activity sites, and pathways of travel connecting them.

One individual may not travel to every hairdresser and one teacher may not work in every school. So, we can understand how our study subjects use the entire

opportunity set by mapping the locations of all relevant activity sites and pathways in relation to the ones used by our study subjects.

In addition, we collect spatial data on risks and resources in the environment. These may be due to individual characteristics, individual behavior in the environment, environmental characteristics, or, most likely, a combination of all three sets of factors.

Geospatial technologies such as the Global Positioning System are enabling new forms of digital data collection. Cell phones, for example, along with digital cameras and many other devices are GPS enabled. Applications on cell phones can be used to capture real time data on human movement in the environment. This information can be automatically streamed to servers and then analyzed.

These technologies offer AMM new opportunities for data collection in the workplace and in the larger community.

19

GIS can help with data analysis.

When I was speaking with Camilla about her work, she pointed out that, in the study of people working in hospital operating rooms, processes and outcomes were not the same across all hospitals. This is the idea of spatial heterogeneity, something

important to understand. Individual, behavioral, and environmental factors may come together in different ways in different places to explain patterns of health and disease.

In addition, we can use GIS in clinical studies to understand whether or not the individuals enrolled in the studies are similar to or different from other individuals in the local population. If we see a worker with a particular kind of neck and shoulder pain, is he or she unusual or typical of other workers in the same setting?

In epidemiological studies, we can use GIS to explore and visualize spatial and temporal variability in data so that we can design better studies. In addition, we can use global and local spatial-temporal methods to analyze our data, if we have collected spatial-temporal data.

On the environmental side, we can work to develop systems dynamic approaches for studying health problems. These models incorporate feedback loops governing whether the numbers of people with particular health problems are increasing or

Explicitly spatial data analysis allows us to uncover and understand spatially varying relationships.

For example, the amount of time spent outdoors varies spatially across people and communities. Air quality also varies spatially.

But, the relationship between amount of time spent outdoors, air quality, and health outcomes may itself vary from place to place.

This approach helps us to understand the spatial domains to which we can generalize our conclusions.

21

Finally, once we have data, analyses, and results, we want to share the information with others. GIS have a role to play in dissemination and access.

Two examples, the Malaria Atlas Project in the UK and the NIH RePORT Atlas in the U.S., serve as illustrations.

Although malaria is not at present a significant public health threat in Sweden, this project offers an interesting model of the kinds of work we could be undertaking.

Mapping malaria is challenging because transmission intensity is geographically heterogeneous. Researchers interested in malaria realized that we did not have a good map of malaria prevalence (reports of the proportion of a sampled population that is confirmed positive for malaria parasites), and this was making it difficult to evaluate the impact of different malaria control projects being implemented in different parts of the world. They gathered prevalence reports, evaluated them based on strict inclusion criteria for the map project, and georeferenced them. The MAP researchers worked with a global database of almost 8,000 survey reports to model a continuous endemicity surface on a 5 x 5 km grid based on initial work to define the global spatial limits of malaria transmission. In assembling the prevalence reports for the mapping project, the researchers noted an increasing tendency for national surveys to be conducted so that they would be representative of all areas within a country, not just areas of high prevalence, and there were many zero prevalence values recorded in the reports analyzed. Researchers used a combination of data provided by the source of reports, an online database of geographic names, online gazetteers, and paper maps to create lon/lat point references for the reports. Surveys that could not be georeferenced or that could be georeferenced only to larger areas (greater than 25 km sq) were excluded. The temporal structure of the data was also taken into account.

The reports covered different times throughout a period from 1985 to 2009. A report was referenced temporally by the midpoint in decimal years between the start and end months of the report.

Using geostatistical methods and gridded population data, they constructed a continuous, age-standardized, urban-corrected malaria prevalence map.

23

This map shows estimated levels of Plasmodium vivax malaria endemicity within the limits of stable transmission. [Malaria Atlas Project, The spatial distribution of Plasmodium vivax malaria endemicity map in 2010 in South and Central Americas, available

www.map.ox.ac.uk/browse-resources/endemicity/Pv_mean/south-and-central- americas/].

The mapped variable is the age-standardised P. vivax Parasite Rate (PfPR_0-99)which describes the estimated proportion of the general population infected with P. vivax at any one time, averaged over the 12 months of 2010. Values range from 0 to more than 7%.

Estimates are made based on data from parasite rate surveys which feed into a geostatistical model that produces a range of predicted endemicities for each location (a probability distribution). The model also uses data from environmental covariates which help predict more accurately, especially in areas far from any actual

For some maps, uncertainty maps are also provided. This is the uncertainty map associated with the modeled distribution of Plasmodium vivax endemicity presented in the previous slide.

Areas in bright yellow are areas of higher uncertainty in the endemicity estimates; the dark blue areas are areas of lower uncertainty.

25

The Research Portfolio Online Reporting Tools (RePORT) website provides a central point of access to reports, data, and analyses of NIH research.

The downloadable RePORT brochure provides a brief description of the various major modules of RePORT.

There are online tutorials for teaching people to use RePORT.

It would be interesting to think about how AMM could use GIS in a dissemination project to develop an AMM Reporting Atlas.

A simple online application for searching AMM research might be useful to the public, public officials, and other agencies to find out where and when AMM projects took place, what the projects were about, what was learned, and what the next steps might be.

27

We began with Kristina’s question: what can we do with GIS in occupational and environmental health?

We can conclude that AMM is already doing a lot of great work, some of it involving GIS and spatial analysis.

And, if people want to, we can do more.

I acknowledge the support of the Swedish Council for Working Life and Social Research (FAS) and colleagues here at AMM.

29