The fortnightly newsletter that unpacks system leaders’ priorities for digital technology and the impact they are having on delivering health services. Contact Ben Heather in confidence here.

If you haven’t heard about the rise and fall of Theranos and its founder, Elizabeth Holmes, here is a brief precis.

The Silicon Valley start-up claimed to have developed a new cheaper blood testing technology and convinced enough investors to raise more than US$400m.

At its height in 2014, the company was valued at US$9bn and its founder’s personal worth at $4.5bn. Rupert Murdoch was an investor. In 2015, Ms Holmes was personally praised by both the then-US President Barack Obama and vice-president Joe Biden.

Except, as was eventually uncovered, the technology didn’t work.

Last month the company was formally dissolved and US prosecutors are pursuing criminal charges against Ms Holmes for defrauding investors of hundreds of millions of dollars.

The harm caused to patients in misdiagnosis to unnecessary operations arising from the company’s erroneous blood tests has yet to be quantified.

Why does it matter for the NHS?

The government is betting big (again) that technology, particularly digital technology, will help transform the NHS we have into the NHS we need.

Better digital tech is undoubtedly part of any future NHS but not all tech solutions are made equal and, as Theranos demonstrates, not all of them do what they say on the box.

Too risk adverse an approach can stifle innovation, a criticism frequently levelled at the NHS.

But investment in tech without a thorough evaluation risks public money and patient safety.

This is true of relatively simple technologies, such as vaginal mesh implants.

It remains no less true of the more complex, such as artificial intelligence-driven diagnostic software.

And the NHS, like Silicon Valley venture capitalists before it, is not immune to making big investments in over-hyped products.

So, what are the risks?


Patchy regulation

Regulation of digital technology in the NHS is fragmentary and accountability too opaque.

In theory, digital tech delivering medical care is the purview of the Medicines and Healthcare Products Regulatory Agency.

But for digital med devices deemed low-risk, an assessment the manufacturers make themselves, there is almost no active oversight.

The companies can register a software “device” with MHRA without telling the regulator what the device does, how it does it, or even what it’s called.

The only requirements are that the company declare it does indeed produce unspecified medical software and that it meets certain standards. Even then, The Download has been told MHRA doesn’t routinely check these claims.

There has been a sharp rise in these low-risk medical software devices registering with MHRA in the past few years.

After a piece of tech is out there delivering healthcare, MHRA will investigate complaints but their policy is to resolve concerns directly with manufacturer wherever possible without any public statements.

This means legitimate (and multiple) safety concerns can be raised about a company and resolved without the public ever knowing.

For apps specifically, NHS Digital has a separate quality assurance process for software included on its “NHS app library”. But most apps currently used in the NHS have not been through this process.

The Care Quality Commission regulates care delivered using digital technology, such as telehealth services, but not the technology itself.

Move fast and break things

A more fundamental disagreement about how to assess digital health technology could also increase the risk of a Theranos-like scandal in the NHS.

There is a school of thought that randomised control trials, peer-reviewed articles and methodologies of their ilk are too slow for the fast-moving world of technology.

Instead, we should deploy promising new technology in the NHS now, testing in quick cycles and improving the product, based on feedback, as we go.

Efficacy and safety are often conflated with user popularity in this argument.

When The Download asked one digital health company for independently validated evidence as to their product’s effectiveness and safety, it was referred to its overwhelmingly positive user reviews.

Some digital health companies operating in the NHS (but by no means all) and some people involved in developing NHS digital policy adhere to this view.

But it also has plenty of critics.

US clinical informatician Dr Eric Topol, who is leading an NHS digital workforce review, told HSJ in July that digital technology needed to meet the same standard of evidence as any other medical innovation.

Like an unproven drug, unproven algorithms can harm patients.

“There should be no exceptionalism.”

Which of these views holds most sway in the NHS will help heighten or lessen the risk of a Theranos-like scandal.

A change in the air?

In a broader DHSC “code of conduct” published last month, there was a quiet acknowledgement that a “more rigorous and sophisticated approach” was needed to assessing new health tech.

In the code, the government made a “commitment” to develop a “trusted approval scheme for digital technology in health and care”, based around Kitemark-like model.

In principle, this sounds sensible but when The Download approached DHSC it was told that the new scheme would not replace the existing “ecosystem” (MHRA, the app library).

Instead, the ”intention is to iterate the code and co-design the means of enforcing and supporting it with all of our stakeholders”.

What this actually means, and whether it will close existing regulatory gaps, is unclear. 

Changes at MHRA in response to EU regulations might offer more hope.

The changes will make it harder to register medical software as “low-risk”, thereby exposing more digital technology to the more rigorous independent assessment required in higher risk categories.

But they also won’t kick in till spring 2020.

Dr Topol’s review, expected by end of the year, could place more immediate extra pressure for better evidence before deploying new technology. But it also might not.

It the meantime, the best defence against a Theranos-like scandal in the NHS is a healthy scepticism for unproven technology at every level of the system.