Helen Signy writes:
With smartphone technology pervading even the lowest socio-economic groups in the US, the most vulnerable populations have become the target of ruthless campaigning by big tobacco and food companies
That’s the alarming conclusion drawn by Canadian health data science expert Associate Professor Nate Osgood, who visited Australia recently to discuss the potential of combining big data with simulation modelling to enhance our understanding of population health trends and public health policy design.
His research involves linking the vast amounts of data on human behaviour collected from technology like smartphones, social media, Fitbits and administrative records.
Osgood works across the Departments of Computer Science, Community Health, Epidemiology and Bioengineering at the University of Saskatchewan.
He has been applying dynamic modelling to improve decision making – with a particular focus on health – for more than 25 years, and has contributed to communicable, chronic and zoonotic disease areas, to social and environmental epidemiology, and to tobacco policy.
In the course of his work he’s unearthed two unexpected developments: first, that even in the lowest socio-economic communities, many people are now relying on smartphones to stay socially connected; and second, that corporates are actively targeting chat-rooms and online communities to undermine government health messages.
“We can monitor where consenting study participants are spending their time online, websites which promote smoking, apps developed by companies – and it appears these companies are far ahead of us in terms of maintaining these communities and apps online,” Professor Osgood told a seminar convened by The Australian Prevention Partnership Centre and the Sax Institute.
“Regulatory agencies are just waking up to this, they don’t have infrastructure or monitoring in place. They can monitor print ads but not this online manipulation of the population by tobacco and food companies.”
It appears that big business is actively targeting lower income groups, for example by populating forums expressing scepticism that smoking causes harm, or by encouraging people to buy sugary snacks, or by promoting these products to children.
Social networking matters
The Canadian researchers have worked with lower socioeconomic communities in Massachusetts, New York City and Saskatchewan. Researchers initially worried these populations might not have access to smartphones, but they were quickly proved wrong.
It turned out that for these populations, in particular, smartphones were being used in large numbers – they provided a lifeline to the social networks of family and friends on which many individuals relied to help out in difficult situations.
Researchers had initially been worried that they would not be able to include low SES communities in their big data collection, but their work with these communities found that this barrier simply did not exist, Professor Osgood said.
“We were worried smartphones were rare and wouldn’t be widely used, but the latest evidence in low socio-economic communities is that communications technology is up there with food and shelter in terms of priority – they are very heavily socially networked,” Professor Osgood says.
“In some of our work, we have found there are preferences as to which phones people want to use. Sometimes some phones don’t have enough storage space so we need to put in extra work to make them eligible, to strengthen their data plans. But sometimes we’ve offered a phone to individuals and they have turned it down because theirs is better.”
Whatever its downside, big data, collected every day from numerous sources ranging from social media to environmental sensors, lab test results, voice calls, browsing patterns and electronic sensors, is becoming fundamental to our understanding of how human behaviour plays into the complex web of interactions that determine health outcomes, says Professor Osgood.
What’s so exciting is that this kind of data provides insights that just aren’t available in the literature or when you ask people to self-report. People’s self-reported mobility levels, for example, can be vastly different to information gleaned from wifi or the GPS in their phone, and their self-reported physical activity can be very different from what is suggested via accelerometers and heart rate monitoring on smartphones and smartwatches.
Feeding all of this information into sophisticated computer models (‘dynamic simulation models’) provides researchers with a low-risk, low-cost way of testing the likely outcomes of different policy interventions before they launch them on the real world.
Tracking foodborne disease outbreaks
For example, one of Professor Osgood’s projects in Canada, the Ethica iEpi, used big data to effectively track foodborne disease outbreaks. Information collected from a smartphone app was used to understand when and where people ate from food vendors, those who developed stomach upsets that both did and did not require medical attention, and how long they stayed sick.
Feeding this information into a sophisticated dynamic simulation model helped researchers to understand which intervention features needed to be modified as the population evolved over time under different policy regimes.
The project found that if just four per cent of smartphone users were used as sentinels to report symptoms to public health authorities, the amount of food-borne illness could be reduced by 60 per cent, by enabling contaminated restaurants to be identified more quickly.
With corporate entities often collecting information on people’s behaviours without them realising, Professor Osgood says authorities and researchers can use big data to fight back – as long as they have informed consent from their participants.
“With more nimble adversarial corporate actors who are running circles around health authorities by tapping into social networking and the latest technologies, this is urgently needed,” he says.
“It is critical to understand how the system behaves to be able to develop reliable interventions and policy regimes that will yield long term gains.”
• Helen Signy, senior communications officer at The Australian Prevention Partnership Centre, prepared this report from a presentation by Professor Nate Osgood, co-hosted by the Partnership Centre and the Sax Institute. Watch his presentation below.
