Regulating medical devices in the ‘Internet of things’

Any device that can collect its own data and communicate via a network is now part of the billions of gadgets collectively known as the Internet of Things (IoT).

Common uses include things like FitBits, automated security systems and stock inventory, but health technology arising from the IoT is also advancing rapidly.

And this raises issues around the adequacy of regulation and highlighting concerns of healthcare professionals.

Increasingly, people with type 1 diabetes use interconnectivity of sensing and actuating devices – those that sense and respond- to manage their blood glucose.

People with type 1 diabetes need to monitor their blood glucose levels multiple times a day, about 40 per cent of children use a continuous glucose monitor (CGM), with insulin being delivered through a pump or by injection to regulate these levels.

Emerging technologies that provide a ‘Do It Yourself’ approach are changing the way some people manage this disease.

People can build their own ‘hybrid closed loop’ system using open source software – software for which the original source code is made freely available – providing access to real-time continuous glucose monitoring data via a personal website or smart phone.

These ‘loopers’ still have CGMs with alarms and they still monitor them - there is no completely closed artificial pancreas - but the DIY system allows users to better manage glucose levels so they can focus on everyday life.

Although a commercial hybrid closed loop system, Medtronic 670G was regulated by Australia’s Therapeutic Goods Administration (TGA) in 2019, it has hard-wired safety features, resulting in a less personalised approach.

The #WeAreNotWaiting movement is also growing, a few steps ahead of the commercial automated insulin delivery systems. Our research has considered this user-led technology in children with type 1 diabetes.

Currently, ‘looping’ software hasn’t been registered on the Australian Register of Therapeutic Goods.

In Australia, software can be listed as a medical device if it undergoes a review of its safety and effectiveness by experts. The TGA fee for a Class III device is around $A100,000, but developers of the software aren’t in a position to pay this.

While it’s unknown exactly how many people are using DIY open-source technology in Australia, we know from the University of Melbourne study that it is being used by parents for their children.

DIY looping is a great case study of the challenges healthcare professionals face in fulfilling their legal and ethical duties where parents modify a standard treatment (CGM integrated with an insulin pump) by customising it using open-source software.

Our research team, with collaborators from Royal Children’s Hospital and Diabetes Australia, are finalising a two-year research project exploring this issue.

We have investigated the perspectives and attitudes of key stakeholders including the healthcare professionals who provide care to children with type 1 diabetes, parents who are using the new technology, lawyers, medical indemnity insurers and software developers.

We had a wide range of responses from the stakeholders we interviewed, reflecting the complexity of the issue and the uncertainty everyone feels about what they can or should do.

Our study found that parents turn to DIY looping because the system provides better control of their child’s diabetes, and commercial hybrid closed-loop systems cost a lot of money.

Many parents we interviewed reported a benefit for the child and for the family overall. They also commented on the technical ability required to set up and run the system.

“…you really do need a good sense of technology. It’s quite tricky to set it up. It’s not anyone who can do DIY.”

It also shows how far parents go to do the best for their children - even if that goes beyond what healthcare professionals recommend.

Healthcare professionals are concerned about the potential risks of the open-source software and the lack of regulatory approval.

“…if we don’t know it’s entirely safe, you’re potentially putting them [the patient] at risk.”

The software developers we interviewed said that there were enough safety checks in the open-source algorithm. They weren’t concerned about hacking or a failure of the software or system.

“….assembling software is not hacking. It’s assembling software. It’s software engineering which is something that doctors know very little about.”

The lawyers discussed the potential for legal liability in negligence of healthcare professionals who supported looping for their paediatric patients.

They felt it was important for diabetes educators and paediatric endocrinologists to engage in a discussion with parents about the risks, so that the parents can decide whether to take those on board.

“...how is it any different from the risks that are associated with parents or self-administering?”

Although there may be concerns about legal liability and medical indemnity, it’s important that a therapeutic relationship is maintained with parents to enable the ongoing care of the child.

The gap between regulation and practice is at the heart of the discussion.

If the software was regulated as a medical device, then healthcare professionals may feel much more at ease as it is likely that they would be covered by medical indemnity insurance.

One of the parents we interviewed said “this technology is literally life changing.” If DIY looping continues to be used, then some sort of resolution of the gap between regulation and practice needs to occur.

Software used with medical devices is susceptible to regulation and we need to think about how innovative technologies, like looping, can be regulated, and the most appropriate forms of doing that.

Does the looping community even want it to be regulated?

In the US, the Food and Drugs Administration collaborated with the diabetes nonprofit organisation, Tidepool, to develop a safe and effective form of mobile looping systems.

Traditionally, we think about safety or efficacy of new technologies as a medical or scientific endeavour. But, is the TGA missing an important source of information, such as data from patients about their experiences, which can be used as evidence of the benefit of innovative technologies?

As new technologies are constantly developed, we require open dialogue to incorporate different perspectives to consider the regulation of IoT devices.

People with an interest in this topic are welcome to attend our Virtual Panel Presentation on 4th November when we will discuss the way forward for regulation with experts in software development, regulation and medical indemnity. The research project team can be found here.

Banner: Getty Images