The Covid-19 virus has changed our lives and is already transforming healthcare. After years of lamenting the slow spread of technology, we are witnessing an almost complete shift of primary care and outpatient services to remote consultations via video conference or WhatsApp. It is unlikely that we will ever go back to the status quo of just three months ago.
This massive experiment is an opportunity to monitor the strength and weaknesses of such technologies. We will go back to normal and at that point we will still be faced with the big question of digital health: what works, really?
Even before this crisis, digital health technologies – from telemedicine, to biosensors, to artificial intelligence algorithms – were slowly transforming the way we deliver healthcare. The digital health market was growing rapidly. In 2018, it was 12 times smaller than the pharmaceutical market. By 2023, it was expected to be a fifth of it. This rapid growth has attracted the attention of large technology and healthcare companies, as well as a whole galaxy of start-ups.
Yet, despite the billions flowing into the market, little was and is being done to understand the effectiveness of digital health technologies. A study by Harvard Medical School and Johns Hopkins Medicine showed that just 12 per cent of technologies from top-funded US digital health companies had been evaluated for clinical effectiveness. That is, just one in eight technologies had demonstrated that they could do what they were supposed to do. Meanwhile, Babylon Health – headquartered in London and one of the world’s most valuable digital health companies – has been accused by some critics of failing to supply enough convincing evidence for the performance of its chatbot.
In a well-known example of the dangers of hype, health tech company Theranos reached a valuation of $10bn before the bubble around its supposedly revolutionary blood test was punctured by scientists, journalists and regulators questioning its validity.
To be fair, regulators and healthcare providers have been slow to clarify what evidence they need from digital health innovators. Although recent guidance has come, for example, from the US Food & Drug Administration and NICE, there are still no clear rules, and there are several challenges in defining them. Traditional approaches that work for drugs don’t easily apply to digital technologies, which are constantly updated. How could you evaluate a drug that changes every few weeks? More pragmatic approaches to evidence generation and evaluation, which can adapt to each stage of development of a technology, are starting to emerge. They need to be properly codified and tested.
This is a complex and technical issue, yet it is crucially important to the future of digital health. If we do not solve it, we risk throwing away good solutions with the bad. With cases like Theranos, we have seen what happens when hype and marketing take the place of rationality and robust evaluation. We cannot afford any more examples like that, or the faith of the public and the medical community in new healthcare technologies will be severely compromised.
Dr Jack Halligan is a Policy Fellow and Gianluca Fontana is a Senior Policy Fellow at Imperial College London’s Institute of Global Health Innovation.