Mapping cancer risks – what can (and can’t) it tell us?

For centuries humans have used maps to chart the world around us. As well as capturing the planet’s geography, they can also help us make sense of the lives of the people that inhabit it.

Maps tracking diseases such as the Black Death go back to the Middle Ages, and John Snow’s depiction of the 1854 Broad Street cholera outbreak has a special place in the hearts of medical mappers.

Information about who gets what illnesses and where they live is vital when it comes to understanding and researching health risks, and planning public services. Today we have more data about health and disease than ever before, yet no-one but the most hardened statistician would want to trawl through endless spreadsheets to pull out meaningful patterns.

So it’s not surprising that researchers are using computer software to help process and display this kind of data, making it easier to navigate and understand. One example is our own Local Cancer Statistics portal, which we launched last year to enable politicians, healthcare providers and the public to access understandable data about cancer diagnosis and treatment in their area.

Today, the Small Area  Health Statistics Unit (SAHSU) at Imperial College London launches their Environment and Health Atlas – a new set of high-resolution maps displaying data about diseases such as cancer alongside environmental risk factors including pollution and pesticides.

Users can zoom into neighbourhoods and flip between different health outcomes and local environmental factors. But while it’s tempting to pop your postcode into the site and start scaring yourself silly, it’s important to step back and look at what these maps can – and can’t – tell us.

What is it?

The atlas details the distribution of 14 health conditions, including lung cancer, breast cancer, heart disease and leukaemia, across England and Wales. Alongside sit maps revealing local variations of certain environmental agents such as air pollution, sunshine and pesticides.

This is the first time that this kind of map has been created at such high granularity, drilling down to local neighbourhoods made up of 3,000 to 6,000 people. The online portal is accompanied by a hefty print version packed with extra detail – a perfect gift for the statistician in your life, perhaps?

It’s also the first UK atlas of its kind to adjust for age and social deprivation, revealing underlying health risks that aren’t just the result of ageing populations or poverty.

What can – and can’t – it tell us?

The most important thing to know about these maps is that they reflect overall health risks and trends in different areas, rather than an individual person’s risk of getting any particular disease. They also show relative rather than absolute risks, which can be difficult to understand in the absence of wider context.

Relative risks compare outcomes between different groups – for example, saying that people living in one town are 20 per cent more likely to get a particular type of cancer than those in another when the overall risk might still be relatively low. Whereas absolute risks make more definitive “You have a 15 per cent chance of getting disease X if you live here”-type statements.

To focus particularly on cancer, there are many different factors that influence a person’s risk, including their age and sex, genetic makeup and lifestyle (such as smoking, diet and physical activity). As an example, if you have a strong family history of breast cancer, moving to an area with a lower risk on the map is unlikely to make a difference.

But, like our own Local Cancer Statistics portal, the maps reveal patterns and trends that will be of vital importance to regional healthcare planners and policy-makers. For example, there are higher health risks across the board in parts of the North West, Yorkshire and South Wales, highlighting the need for further research into why this is happening and what can be done about it.

Correlation or causation?

It’s also important to note that the maps can’t tell us if there is a direct link between a particular environmental factor and a disease. This is because, as any good statistician will tell you, correlation does not equal causation.

Correlations can easily be drawn between all kinds of trends, but they don’t necessarily prove that one thing causes the other. A country’s chocolate consumption reflects its number of Nobel prizewinners, the falling German birthrate appears to be due to a decline in the stork population, and the performance of the Welsh rugby team can be linked to dead Popes.  None of these are true causal relationships.

In contrast, the maps show that respiratory problems, such as lung cancer and chronic pulmonary disease are worse in urban areas, which fits with what we know about the evidence from large-scale studies of a link between the two.

Related to the story we released earlier this week about increasing melanoma rates and UV exposure from the sun and sunbeds, the atlas also shows that skin cancer risk is highest in the South West. However, sunshine duration is highest in southeast England.

Images from the atlas showing sunshine duration from shortest (green) to longest (orange) and malignant melanoma skin cancer rates in men, from lowest (purple) to highest (brown)

Are people in Devon and Cornwall worse at protecting themselves from sun overexposure than the citizens of the Kent coast, or is something else going on? And is that pocket of increased skin cancer risk in the Liverpool area linked to sunbeds? More research will help to answer questions like this.

Not only do the maps highlight possible behavioural differences across the country, they could be useful in cases where the environmental causes of cancer are less clear – such as the impact of chlorine by-products in water supplies – illuminating potential areas for future research.

As you might expect with this kind of project, there are a lot more questions that people might have about the data and what it means, so the Imperial researchers have teamed up with Sense About Science to produce a handy set of FAQs, which are worth a browse.

Why is this kind of data important?

There’s no denying that this kind of data visualisation has become increasingly prevalent – even trendy – over recent years. There’s even a fascinating exhibition about it at the British Library right now. But there’s more to these types of online portals than pretty pictures.

We believe it’s important that decisions and discussions about cancer care are based on genuine scientific evidence, not opinion and speculation. Providing easy access to understandable statistics and data about risk is essential for the people who decide healthcare and research policies, and plan health services across the country.

As the well-worn phrase goes, knowledge is power. And this kind of knowledge can be literally life-saving.

Kat

• Image from Wikimedia Commons