Mapping Disease: Why The Details Matter for Spotting Health Trends
"Uncover how different methods of calculating expected cases impact disease maps and spatio-temporal analysis. See why accurate mapping is vital for public health."
Disease maps are powerful tools. They visually represent the spread and patterns of illnesses. These maps help us understand where diseases hit hardest and how they're changing. But how we create these maps matters—a lot. The choices we make in calculating the data can significantly change what the map shows us.
One key step is figuring out the 'expected cases' – how many cases we'd anticipate in a certain area based on general rates. This number becomes our yardstick for measuring if a region has more or fewer cases than expected. However, there's more than one way to calculate those expected cases. Do we use the average rates for the whole study period? Or do we use specific rates for each smaller chunk of time within the study?
Recent research digs into this very question. The researchers looked at stomach cancer mortality in Spain over two decades. They compared different methods of calculating expected cases and mapped the results. Their findings highlight how much these choices matter. Depending on the method, the maps could reveal different patterns and trends. This article breaks down their findings, showing you why choosing the right method is crucial for accurate disease mapping.
Cancer Statistics by Place and Population
CDC’s Cancer category in Chronic Disease Indicators reports incidence and mortality for selected cancer types, along with screening-test prevalence. The CDC USCS Explore Data module lets users examine incidence, mortality, trends, stage, survival, and prevalence across cancer sites, ages, racial and ethnic groups, years, and locations. SEER’s interactive tools also present measures such as incidence, mortality, prevalence, and cost by population and geography, while the NCI GIS Portal offers maps and downloads of cancer statistics, risk factors, and geographic patterns to support research and cancer control.
Mapping Rates Across Space and Time
Disease mapping gathers geo-referenced data to depict disease prevalence across time and space, using geographic information systems to store, analyze, and present multidimensional data. Mapping and analyzing incidence or mortality over space and time can reveal observed changes and suggest risk factors for further investigation. A key limitation is that results may change when analysts vary the geographic scale or the unit of aggregation, so a map’s patterns depend partly on how areas are defined.
From Recognizing Cancer to Mapping It
Cancer has been recognized in human history for millennia, and its history includes major discoveries, technological innovations, and shifts in how the disease is understood. In 2011, the article “Mapping Cancer Origins” described cancer as a diverse assortment of diseases and argued that this diversity hindered the goal of finding a cure for every patient. It pointed to the integration of genomic and stem-cell technologies as a promising route through that challenge. Today, the NCI Cancer Atlas lets users create maps of cancer statistics, demographics, and risk factors.
Different Roads to the Same Destination? How Calculation Methods Change the Map
Researchers in the study focused on municipal-level data of stomach cancer mortality in men in Spain from 1989 to 2008. They wanted to see how different calculation methods affected the resulting disease maps. They calculated "expected cases" in two main ways:
- Using average reference rates for the overall period: Here, they calculated one average mortality rate for the entire 1989-2008 timespan.
Spatial and Temporal Models in Cancer Research
A systematic review reports that studies including Adin et al. (2017), Carroll and Zhao (2018), and Goicoa et al. (2016, 2018) applied the Knorr-Held (2000) model to represent spatial-temporal interactions in cancer incidence and mortality. These interactions capture area-specific changes in disease risk over time, alongside spatial patterns modeled with conditional autoregressive (CAR) terms and temporal trends modeled with first- or second-order random walks. Separately, a study of mainland Portugal examined cancer mortality for 2009–2013 and concluded that the deaths formed clusters.
Mapping Choices Shape the Picture
Cancer mapping requires choices about scale, the base map, the number and length of map classes, and which data to display. These choices mean a mapped pattern is a starting point for further investigation rather than, on its own, a final explanation of mortality or morbidity. Data quality adds another challenge: the cancer-mortality review identifies collection and harmonization as major issues, noting that death certificates may not correctly record the actual cause of death.
The Takeaway: Why This Matters for Public Health
This research has important implications for anyone involved in public health and disease mapping. It demonstrates that creating disease maps isn't just about plugging in numbers. The choices we make in how we calculate those numbers—especially expected cases—can dramatically affect what the maps reveal.
Maps Make Cancer Patterns Visible
Cancer researchers have a long tradition of presenting population-based cancer statistics in maps. Historically, printed atlases left the map producer to select the content and format; geographic information systems later made comprehensive mapping and spatial analysis available. This history underscores that maps are not just displays of statistics: their selected content and format shape how people encounter geographic patterns.
Projecting Cancer Trends to 2050
The World Health Organization’s Cancer Tomorrow tool provides visualizations to predict future cancer incidence and mortality for a country or region, using current estimates for 2024 and projecting through 2050. The Global Cancer Observatory is an interactive platform presenting global cancer statistics for cancer control and research. The Cancer Atlas reports that GLOBOCAN’s 2024 estimates cover 34 cancers in 186 countries, with nearly 21 million people diagnosed and 9.8 million deaths worldwide that year.
Cancer as a Systemic Disease
A 2024 article describes cancer as systemic in several dimensions: its evolutionary course, its direct crosstalk with the microenvironment, and its communication with the macro-environment. It argues that this broader perspective can help address cancer’s complexity and connect what research can learn with the people it needs to help. The WHO’s 2026 Global status report on cancer assesses the global burden, prevention and control progress, and systemic challenges that continue to limit equitable outcomes.
So, what's the right approach? The answer depends on the goals. If you're trying to spot time trends in specific areas, using average reference rates might be the way to go. But, if you're more interested in seeing how the spatial pattern of a disease is changing, using local rates for each period is a better bet.
Ultimately, being aware of these nuances can lead to more accurate and insightful disease maps. And those maps can help us target resources, develop effective interventions, and improve public health outcomes.