Frequent flyer points for daily steps, discounted premiums for meeting fitness goals, movie tickets for breaking a sweat – these are some of the incentives being offered by health insurers as they try to incentivise healthier behaviour and promote wearables among their customers.
The argument goes that we’ll get healthier, while insurers will pay out less.
But on the other hand, critics warn, we may be creating a dystopia where disadvantaged groups (who tend to have higher rates of health problems) cannot afford to pay for coverage, while corporations listen to our heart rates and track our sleep patterns.
Which should be something to be avoided.
“For centuries, the insurance model has primarily provided financial protection for families after death, without enhancing the very quality it hinges on: life,” Marrian Harrison, CEO of US insurance giant John Hancock said in a statement this week.
The company was announcing it would now only sell life insurance plans where customers have the option of paying less on premiums by exercising regularly.
If you didn’t want to wear a wearable, you’d effectively be paying more for life insurance.
It claims users of its “behaviour change platform”, called Vitality, live 13-21 years longer than the rest of the population and generate lower hospitalisation costs.
Aside from John Hancock, this week saw another development in the emerging alliance of big data, wearables and insurers: Fitbit has launched an app that processes data from wearables and other medical devices and makes it available to doctors and insurers.
The first users come from health insurance provider Humana, which already provides its customers with Fitbit devices.
A few other examples of the intersection of wearables and insurance:
And, yes, this is also happening in Australia:
In February, the health analytics company Vivametrica released a white paper on predicting a person’s mortality risk using steps per day.
It found it worked – steps per day is a “powerful predictor”.
“Insurers interested in adopting a wearables program should begin with a pilot to assess consumer adoption rates and to understand the physical activity characteristics of their customers,” the paper advised.
“The pilot serves as a baseline analysis to support a more comprehensive customer engagement or risk assessment program.”
And mortality is not the only thing steps per day may be able to predict.
In 2016, it was reported there were more than 20 trials underway using Fitbits to identifying signals of specific diseases in health data. An elevated resting heart rate, plus a sudden decrease in the number of steps the person takes per day, could signal the presence of the flu virus.
If an insurance company had access to this data, it could send a message to the patient and book a doctor’s appointment.
This sounds possibly quite handy, but consider the fact this same technology could be used to discriminate against people with pre-existing conditions.
Heart disease and diabetes, for example, are more common among lower socio-economic groups. Offering rewards to the healthiest would effectively discriminate against the poor and vulnerable – everyone other than the ‘white, worried and well’.
Experts say it would be analogous to setting rates of car insurance depending on the suburb where you live and the rate of crime.
Suneel Jethani, a lecturer at University of Melbourne’s School of Culture and Communication, who has recently submitted a PhD on wearables and big data, told Hack the idea of healthy active lifestyles was being used to sell a hidden commercial agenda.
You get cheaper premiums, but the company gets the data. And we don’t know how this data is being used, because the algorithms are protected by commercial IP.
“Insurance premiums are shaped by variables that can be used to predict mortality,” Suneel said.
“These include demographic things like age, gender, or smoking status along with things like BMI, blood pressure, cholesterol, drug, and alcohol use, family history, diabetes, and cardiovascular health.”
“These ‘indicators’ of health are highly contingent on socioeconomic factors and when predictive models are built on this type of data they can become biased. The biases in the predictive models can have implications for certain populations and groups.”
Insurance companies in Australia can’t discriminate on the basis of health. This protection is designed to ensure people with a history of health issues do not pay a much higher premium.
However, insurers can use health data to incentivise behaviour.
And this is where it gets tricky.
Paying rewards to customers who exercise more may – at some point – not be so different from selling them cheaper insurance.
In the US, under the Affordable Care Act (Obamacare), insurers are not able to deny insurance on the basis of pre-existing conditions. But there are moves to repeal Obamacare, and, if that happens, insurers might look to wearable devices for evidence they could use to refuse to pay for patient’s health care.
John Hancock is not denying life insurance to people who could die soon, but it would be in a good position to do this if the law was changed.
Health insurers have always tried to learn as much as possible about their customers, to minimise risk. They’ve also done their best to nudge customers towards healthier behaviour. The more customers that remain healthier for longer, the less the insurers pay out.
What’s unprecedented is the potential for mass surveillance and social engineering, behaviour monitoring and structural discrimination.
“People tend to develop certain patterns of behaviour,” Suneel said.
“They might avoid situations that produce ‘bad data’ or think about what a person might do if they really want to clock that 10,000 steps in a day.”
“It gets to 11:30pm and they’re only at 8,000 … jump up and down on the spot? take the dog for a walk?”
Responding to these developments, an art project called Unfit Bits provides products for tricking your wearable into believing that you have engaged in physical activity. The Desktop Stepper Smoker’s Edition is a pendulum from which you can hang your fitness tracker and let it register steps.
A guide details strategies for hacking your fitness trackers, such as using bicycle wheels and electric drills to simulate exercise.
“Free your fitness data from yourself,” the guide declares.