Can data-driven medicine make us healthier?

18 Jun 2019 | News

At a Science|Business event in Norway, students, technology experts and teachers debated the merits of new breeds of healthcare delivery. There are benefits, but digital health must be thoughtfully controlled and users properly trained

The promises of data-driven healthcare abound. New technologies that run the gamut from artificial intelligence to health apps on your smartphone can lead to greater efficiencies in health care systems, allow those with chronic diseases to manage their illnesses, and educate people in healthier lifestyles.

But as a Science|Business meeting held during the Norwegian University of Science and Technology (NTNU) Big Challenge science festival in Trondheim heard this week, these new technologies can also have a dark side. While on the face of it, there are obvious benefits to data-driven medicine, data sources may be incomplete, data access inadequately controlled and objective evidence of positive effects non-existent.

As Signe Elisabeth Åsberg, a postdoctoral research coordinator at NTNU noted, it is not unrealistic to dream of the day when cheap technology in peoples’ homes will act as an early-warning system to flag imminent health problems.

A voice assistant, like Alexa, could signal when someone is displaying early symptoms of a disease, she said. “Ten years from now, the collection of health data will be more integrated, and less intrusive. A digital doctor – could be a smartphone, or Alexa – will be able to tell us when we’re deviating from our baseline; when we’re in the red, and our health at an increased danger.”

Data-driven techniques of this kind are being adopted, and with data analytics and machine learning, doctors are now getting a more detailed picture of the progression of diseases.

But more digital data in health can also increase the risk of unintended consequences, for example, enabling discrimination, the conference heard. “It seems like a great idea to volunteer our entire DNA to solve cancer,” said Marianne Synnes Emblemsvåg, molecular biologist and member of the Norwegian Parliament. “But if insurance companies one day discover your genetic profile, they might decide that, if you are at risk of being sick in 20 years, they won’t cover you.”

Companies are barred from hiring, firing or promoting workers based on genetic test results, and insurance companies from requiring or using such tests to decide coverage.

But private information about medical conditions held on health apps is not always confidential, and there have been cases where this information has been passed on to other organisations, such as credit rating agencies, or used to target advertising.

“Even if we have good regulations here, and we only give our genetic info to people we trust, we don’t know what the future will look like,” said Björn Gustafsson, dean of the Faculty of Medicine and Health Sciences at NTNU, drawing everyone’s attention to China. When Chinese authorities wanted more DNA information on its Uighur population, a predominantly Muslim ethnic group, they used equipment made by Thermo Fisher, a US company and genetic material provided by Kenneth Kidd, a Yale University geneticist. 

“China’s a totalitarian country, and it’s controlling its people,” Gustafsson said. “These people [the Uighur population] were promised a health examination for free. Instead, they got their faces, their blood and fingerprints recorded. This was biological material arriving into the wrong hands. This is the opposite side of what can happen if we don’t have good control over this,” he said.

Lost boys

As a result of concerns like these, getting hands on peoples’ data in Norway is massively time-consuming for researchers.

For all the talk about the growing mountains of health data, public health specialists say they are hardly any wiser about how they can apply it to improve overall population health.

“We have all this data, but none of it is linked to things like a person’s education or environment. Getting permission to combine sociological data, such as where you go to school, with health data, is really difficult. As a result, we have few studies looking at what interventions we can make,” said Camilla Stoltenberg, director general of the Norwegian Institute of Public Health.

In Norway and Finland, growing numbers of people report they are struggling with their mental health. The causes are unclear, Stoltenberg said. “We don’t know how, for example, kindergartens or high school affects children’s mental health,” she said. “We don’t know if the reforms we make to schools have changed things for the better, or for the worse. We need more access to data on individuals, but it’s very difficult to attain this data on a societal level.”

More boys are dropping out of school in Finland, said Anita Navaratnam, managing partner at Sygeny, a digital health analytics company in Helsinki. “They call them the lost boys. They’re young, they become marginalised. We need to understand what the sources of these problems are,” she said.

Legislators are not unaware of these data access problems, said Synnes Emblemsvåg MP, who is a member of a group of MPs trying to open up access for researchers to all kinds of data. “[Some] of my daughter’s peer group are being treated for depression,” Emblemsvåg said. “What’s the cause of this? It’s hard to say, because we’re still treating symptoms and not causes.”

She wants to build a “health analysis platform” to combine health registries across the Nordic countries, but says it could be a few years before this happens.

Education in Norway needs to reflect the data trend, with curricula more visibly focused on these new kinds of technologies and how to apply them, said Gustafsson. Health technology is moving fast and digital equipment in hospitals is now so advanced that medical staff are getting left behind, “We don’t know how to use it,” he said. “The average age of the people working in our hospitals is between 50 and 60. Of course, it’s not easy for them to adapt to new techniques.”

More time for patients

Artificial intelligence has been shown to be faster than doctors at correctly spotting lesions on hospital scans. “We have AI algorithms that are better than specialised radiologists,” said Håvard Ulsaker, a medical student at NTNU. “If we can make AI algorithms do groundwork, it will free up doctors, who can then have more time for patients.”

If doctors can use data to learn more about their patients, “They’ll be able to spend less time typing notes during a consultation, and more time engaging,” said Navaratnam.

The conference heard that it is too early to tell whether continuous monitoring tools, like apps and smartphones, will contribute to greater wellbeing. “There’s question marks over the reliability of this data,” said Gustafsson.

Worse, there is concern that apps and social media can trigger health problems. “Not all technology is good for us,” Gustafsson said. “With conditions like ADHD [attention deficit hyperactivity disorder], it’s not farfetched to think that tech which keeps people active – like smartphones – can also stimulate psychiatric disease. I think we have been slow to understand this.”

Apps that track eating could potentially contribute to eating disorders, said Åsberg. At the same time “Kids sleep less than they should; apps are keeping us awake. Social apps can also be a huge cause of stress,” she said.

While social media is “designed to be addictive,” apps that promote healthy living have not caught fire with young in the same way. “[The app developers] are trying to introduce more ‘addictive elements’, [by] making these apps more like a game. I’m not sure it’s going to work,” Åsberg said.

Never miss an update from Science|Business:   Newsletter sign-up