A: Digital twinning is a very interesting area and particularly applicable for biomarkers.

Digital twinning has quite a broad application in a healthcare sense. As you might have guessed from the name it’s the idea of creating a realistic virtual model of a patient, process, or even an entire hospital. From a process or hospital sense it's used for things like tracking equipment, looking for efficiencies and staffing management.

But the real value is modelling a patient.

By creating a digital twin, doctors can do things like try multiple treatments on the model to identify which would work best for a particular patient, or compare the model to the real patient to identify when someone is falling ill, or even monitoring how the patient is complying with a particular treatment plan.

Digital biomarkers come into this because a model is only as good as the data that's fed into it therefore the more continuously the patient is monitored the closer the comparison between the twin and the patient allowing more accurate assessments to be made in real time.

A: The main difference is regulation and the implications on validation and safety.

A medical device would typically make a claim of a health benefit and this claim needs to be demonstrated through robust clinical trials. A consumer device on the other hand will generally deliberately steer away from making a specific claim so they won’t need to go through the same process. Paradoxically, consumer devices can sometimes seem more advanced as they quickly exploit emerging technologies without the lengthy process required to prove benefit.

From a safety point of view, while consumer devices also need to be safe, the threshold for acceptable safety tends to be a lot stricter on medical devices and this also puts a much higher burden on the design of these types of devices.

A: Increasingly, all of them.  Digital services are powerful tools to help health insurers move towards partnership rather than transactional models as insurers are increasingly recognising that ‘we help you be well’ is a preferable approach over ‘you get ill, we pay you’.

Considering different stages of the care cycle, it's in the interest of insurance companies to keep people healthy and stop them developing disease conditions in the first place: it’s a great opportunity for general health and predictive devices. Nevertheless, if people do fall ill it’s clearly better for insurance companies to have the disease condition identified early so the most appropriate treatment chosen as soon as possible to ensure the patient spends the least time possible in a healthcare setting.

Insurance companies will also want to ensure adherence to the treatment plan and also monitor the efficacy of a particular treatment, so that in the future, they can support efficacy-based reimbursements which ultimately, is best for patients too.

A: Security is a big topic in healthcare, particularly after the WannaCry attacks a few years ago. Now, the key thing is to consider cybersecurity as a patient safety concern, so it needs to be handled in the same way as any other patient safety concern through things like risk analysis and risk management.

There can be lots of strategies, but examples could be things like filtering data at source so that only limited, strictly necessary (and consented) data is transferred off a device, data encryption, and data anonymisation at source.

The FDA is quite active in this area and publishes security vulnerabilities of all registered devices so it’s a good starting point to have a look at these to find out what not to do.  They also have strategies like requiring the submission of software bill of materials that can help developers identify security vulnerabilities.

A: Unfortunately, there isn't a simple one size fits all template for this. The important thing is to consider all the challenges early on during the requirements and at the architecture stage of any development, so that informed choices can be made. It will almost certainly take longer and cost more when a restriction caused by one of these challenges needs to be weaved into a design that is already far down the development road. While it is sometimes true that a lot of things can be fixed in software, it's always better if you can avoid the problem arising at all through an intelligent design choice at an earlier stage.

For example, with the challenge of data quality. If there is a way of ensuring that a wearable device is always worn properly by design, then that can save on having to do a lot of complicated post-processing in software to filter things out to avoid skewed results.

A: Of course, the patient is at the centre of everything we do in healthcare and it's fundamental in digital therapeutics as it increasingly offers the opportunity to treat patients as consumers of healthcare services.  These products and services demand the highest quality of patient engagement, usability, and product stickiness. At Sagentia Innovation, to fulfil these sorts of projects, we must ensure our team expertise and skill base covers human factors, usability, patient studies, behavioural science, and user centric design. It’s important to have all those skill sets in a full integrated, multi-disciplinary team to successfully deliver developments in the digital therapeutics space.

These disciplines will ensure the right decisions are made up front in the definition of the solution independent requirements, and then run right through the development at every stage.

A: Reimbursement for devices is moving increasingly towards a value-based system. Therefore, these devices will need to go beyond what is required for regulatory approval and demonstrate the meaningful impact of using the device over alternative options. This in general may make adoption slower due to the time required to gather evidence for some of the longer-term conditions that these devices are targeting. However, with the increased shift towards preventive and personalised healthcare this transition is inevitable.

A: In general, lower risk devices will tend to go down the predicate route of showing substantial equivalence to a previously approved product. This does reduce the regulatory burden on device manufacturers. However, on the flip side the FDA is increasingly asking for more evidence to demonstrate equivalence, including evidence across the range of operating conditions the device may be specified for.

A: The main benefit for the drug development process will be in streamlining and potentially speeding up the clinical trial process. Monitoring clinical trial participants through digital biomarkers takes out participant data entry errors and biases, allows for constant monitoring and identification of side effects or adverse reactions and potentially enables faster insights into effectiveness.

A: Yes, we have a very well-defined software process according to IEC 62304, within our Quality Management System.

Sagentia Innovation has a medical device software development code of contact for us to facilitate compliance with European and FDA external standards and guidelines, and this includes the 62304 medical device software, software life cycle processes, principles of software validation, guidance for industry and FDA, content guidance for pre-market submission, as well as continued medical devices. Across our client base we’ve developed medical software in all medical safety classifications, so that’s 62304, class A and B and the highest level of concern, class C.