This article is part of the FT/ArcelorMittal Boldness in Business awards, recognising companies and individuals with novel answers to everyday needs.
Healthy.io styles itself as the “anti-Theranos”, the collapsed Silicon Valley blood-testing company headed up by the young, turtleneck-wearing Elizabeth Holmes.
The Israeli start-up aims to marry machine learning with medicine. But instead of raising hundreds of millions of dollars from high-profile investors, the company only wants to be paid if enough patients take its tests.
To become an expert for handling health technology tools that are totally based on Machine learning. Then Intellipat provides you best Machine Learning online course that serves you to get knowledge in this field.
Yonatan Adiri, Healthy.io’s chief executive, shies away from making grand claims about transformation or the company’s technology, which uses software and algorithms to allow anyone to analyse their own urine at home. “You’ve never heard me promise disruption or an era after Healthy.io where healthcare would look completely different and be solved,” the 38-year-old says in a vegan café in New York.
But the market Healthy.io is addressing is vast: it estimates urine tests are done 250m times a year in the US, for everything from urinary tract infections to monitoring during pregnancy and kidney disease. Millions who should take a test — for example, diabetics whose kidneys are at risk of failing — skip them because they are too much hassle, storing up health problems.
In the UK, the National Health Service has recognised the potentially huge savings from the product, and patients in 11 NHS areas are already using the kits.
The “smartest decision” the company ever made, says Adiri, was to not reinvent the dipstick, which has been around for 50 years and changes colour when exposed to urine.
Healthy.io’s innovation is to use a smartphone’s camera to detect problems such as bacteria in urine, or elevated protein levels, which could indicate kidney trouble. Users are guided by an app to dip the stick and take a photo of it on a panel that looks like a paint swatch. Image-focused machine learning — known as computer vision — does the rest. The kit has since gained approval from regulators, and convinced various health services to pay for the tests.
“It’s not so much magic as very, very strong, rigorous computer vision that knows how to turn this into a scanner,” says Adiri, pointing to the camera on his smartphone. By using a phone, which everyone has in their pocket, he says, people can do the test as part of their everyday routine, rather than having to go to a clinic for a test.
Adiri became interested in the potential of the smartphone when his parents were travelling in China. They had reached a small town when his mother became ill. She was about to be flown to Hong Kong when he asked his father to send him a photo of her CAT scan. Adiri forwarded this to a doctor, who advised his mother not to fly: she had punctured her lung and air travel could have been deadly.
Despite being neither a clinician nor a software engineer, Adiri decided to search for new ways to use smartphones in healthcare.
He had been Israel’s chief technology officer, meeting renowned scientists and global technology companies on behalf of then president Shimon Peres. As part of his role, he spent time in Silicon Valley, and co-founded Getaround, an app that helps users rent cars from people in their neighbourhood.
While searching for the right application for the “medical selfie”, Adiri landed on urine testing, which is based on colours, sometimes read visually by nurses or doctors, or else by devices using photometry. Healthy.io’s smartphone test matched the accuracy of these devices by using computer vision to address the problem shown most vividly by the “blue dress/gold dress” internet meme, where colours change depending on lighting conditions.
Josef Coresh, a professor at Johns Hopkins school of public health in Baltimore who specialises in kidney disease, is an adviser to the company. He says Adiri had also tapped into a major medical problem: more than 10 per cent of people in the US suffer from kidney disease. If caught early through regular testing, kidney damage can be reversed, he adds.
“The sad thing is, despite the universal recommendation to measure all diabetics’ urine, only about 50 per cent are [having it done] nationally,” Coresh says.
“The neat thing about the dipstick-type products is they allow for valid lab testing in the home,” he explains. “This changes the dynamic of something that has been completely stuck.”
Healthy.io is committed to being clinical grade — with regulatory approval for tests in the US and EU — and collecting scientific evidence to back up its claims. When an early trial failed because patients — mainly middle-aged men — failed to read the instructions and left the sticks in the pot for too long, the team created a chatbot that helps patients through the process step by step.
A study with Geisinger, one of the US’s largest private health groups, published in the journal of The Renal Association in the US, showed that 71 per cent of patients used the test they received, and 98 per cent rated it easy or very easy to use.
Healthy.io charges a fee only if at least 30 per cent of the people who are sent tests use them. So far, with projects in Israel, the Netherlands, the UK and the US, the lowest utilisation rate has been just over 60 per cent.
The strategy seems to be paying off. The number of tests done has soared from 10,000 in the first 13 months of the company’s life to 10,000 a month today. By the end of the year, as Healthy.io rolls out across the US, it expects to reach 25,000 a month. To fuel its expansion, the company raised $60m after a funding round last September.
Marissa Schlueter, a healthcare analyst at research firm CB Insights, says Healthy.io is “leading the pack” in a nascent category of smartphone diagnostics which includes Scanwell, a competitor in urine-testing, and companies that offer at-home fertility tests. The category has raised about $400m, she estimates, of which Healthy.io has garnered $86m.
Like many digital health companies, Healthy.io hopes to make money by taking a slice of cost savings. The NHS studied whether the at-home tests could cut costs and found that if kits were sent to diabetics, they could save the health service £660m over five years, by catching conditions before they required more expensive treatment.
In addition to NHS trials of the kits, Healthy.io is selling them through Boots, the UK pharmacy chain, in an unusual partnership that helps patients with urinary tract infections obtain antibiotics directly from a pharmacist, rather than waiting for a doctor’s appointment.
The next step is to persuade the NHS to adopt the kits more widely, when doctors will not necessarily feel the cost savings immediately. “On a macro level, it’s a huge opportunity for the NHS to make massive savings,” says Katherine Ward, Healthy.io’s chief commercial officer and head of the UK business. “But as an individual general practitioner [family doctor], would you pay [for the kits] in order to have extra time in your day?”
The NHS has already helped Healthy.io recognise the potential for its algorithms to move beyond urine. Simon Stevens, chief executive of the NHS, wrote to Adiri to explain a problem the service had in the care of chronic wounds, which have to be measured regularly to track progress. Nurses were relying on paper rulers, and measurements often varied so much they were inaccurate. Now, nurses in three clinics are experimenting with smartphone pictures of wounds Healthy.io’s algorithms measure and analyse.
For Adiri, Healthy.io’s technologies are about the decentralisation of healthcare, evolving from a top-down system designed to control communicable diseases to one where patients play a bigger role in managing chronic conditions.
“I think we’re going to create the best company in the world that transforms the smartphone camera into a clinical-grade device,” he says. “We’re not going to do anything beyond that.”