Snakes on a (graphic) plane
1 Dec 2017
Nina Cromeyer Dieke

“Challenges in life can either enrich you or poison you.” So said a motivational speaker.

Presenting data in a creative, accurate and accessible way is no easy feat. Those wanting to produce a good data visualisation, for whatever purpose, should be aware of several challenges lurking underfoot.

Data collection

For some diseases, collecting data in the first place is a challenge. Snakebite data availability is very low compared to even other NTDs in large part because most victims live in poor areas in the tropics that are difficult to access. Those bitten may not go to a hospital, where researchers normally collect data, so the need for community surveys is crucial.

“Numbers of bites, numbers of patients attending hospital, number of patients receiving antivenom, numbers of deaths, days spent in hospital, cost of treatment, long term morbidity data are all important pieces of data to collect so that a full picture relating to disease burden and the socioeconomic costs of snakebite can be estimated,” says Dr Nicholas Casewell, Senior Lecturer at the Liverpool School of Tropical Medicine (LSTM) and part of the Alistair Reid Venom Research Unit (ARVRU) there.

Only then could a truly representative data visualisation even begin to uncurl.

Data access

When the data you want does exist, access can still be a problem.

A widely used statistic says up to half of all clinical trials never report their data. Reasons abound: drug companies don’t want to publish unflattering results or give away ideas to competitors; universities and other research institutions don’t uphold funders’ data-sharing requirements, or funders don’t enforce them; and prestigious journals sometimes aren’t as keen to publish disappointing results.

With snakebite, Casewell says data is being collected to some extent in many regions of the world, but not at a comprehensive national level. It is often collected from hospitals mostly in countries where snakebite is notifiable (a few Latin American countries with Costa Rica as a leader, Cameroon, Kenya, Burkina Faso and Sierra Leone, although some include snakebite along with other animals).

We at Manta Ray Media know this problem all too well. We’re working with the ARVRU team in Liverpool to create a snakebite atlas, but the data just isn’t there! We have had to reduce the scope of our initial idea and start by mapping medically important snakes and their distribution. Our longer term aim is to grow the atlas to include more data on bites and deaths as that information becomes available.

Data skills

If you’ve managed to obtain data and the permission to use it, the next challenge is knowing what to do with it. Producing a data visualisation requires technical skills: programming, statistics, information architecture, plus the ability to know how to tell a good story to the right audience.

Many researchers use data analysis software like R or STATA to crunch numbers and graph conclusions. But who is their audience? They’ll likely publish in research journals, so only other experts.

Recruiters and hiring companies are increasingly demanding “data storytellers” who are versed in data, visuals and narrative to tell data stories. Demand for people with this cross-disciplinary skillset will continue to grow, according to Forbes, and global health funders, research institutions and advocacy organisations should seriously consider their team structure if they want to be competitive in data visualisation production (they should).

Et voilà!

If you overcome these challenges, you may just well have yourself a dazzling data viz.

A great example is this widescreen image of causes of death in the 20th century, in which snakebite features (look closely!). It was commissioned by the Wellcome Collection and made by well-known data storyteller David McCandless. The graphic works on a few levels:

-       The message is clear at first glance.

-       You don’t need to know anything about global health to understand it.

-       The design invites you to focus on specific areas at a time.

Some other good ones:

The Institute for Health Metrics and Evaluation produces many visualisations. The design is still closer to academic than artistic, but they work.

John Snow’s life-saving map of Soho in 1854, which helped end the cholera outbreak. Perhaps the first ever data visualisation about health.

The Guardian contrasted accessibility of metro systems in different cities, and it’s pretty shocking.

Information is Beautiful illustrated scientific evidence for popular health supplements (including snake oil)

And this isn’t health-related but it’s awesome. It lets you compare the tree canopy in different cities across the world.