Digital transformation by Software as a Medical Device (SaMD) in medicine and healthcare is not just a potential, it is a reality that is rapidly shaping the future of healthcare.

With the increasing use of devices and data, we are seeing a transformation in how health is identified, monitored, diagnosed, triaged, managed, and treated. However, the process of developing SaMD is not without its challenges and failure to plan from the outset will lead to difficulties down the road.

Although SaMD is only a small component of a larger regulated solution, mistakes made during the development phases can have significant impact on the whole solution and may slow down the ability to scale effectively, especially when the product has gained market momentum. Having successfully designed and delivered over 50 medical devices, we urge you to plan thoroughly as often there’s only one opportunity to succeed.

What is Software as a Medical Device?

SaMD (Software as a Medical Device) is software capable of running on general computing platforms devices – i.e. not on specialist medical hardware. It may be embedded in or connected to some kind of device, but it is the software that performs the medical function. Used in health and, sometimes, social care settings, it is subject to the same controls as any other medical device.

To identify whether your solution falls under the SaMD category, it’s helpful to use the MHRA flowchart. In essence though, you’ve got to work out whether the app has a medical purpose or not and this is very narrowly defined by the MHRA. If it doesn’t have a medical purpose – it’s not a medical device.

Software as a Medical Device examples

It can be difficult to understand what SaMD is without specific examples. So, here are a few solutions that would be considered Software as a Medical Device:

  • Symptoms checker in the NHS COVID-19 app
  • Apps that diagnose cognitive impairment
  • Solutions that automate clinical decisions (ex: can identify blood test anomalies)
  • Radiology diagnostics apps
  • Apps that calculate and adjust insulin dosage based on glucose levels (ex: CamDiab)

Key considerations for SaMD development

In our experience, there are seven key considerations to think about when designing and building SaMD.

1. Does your solution solve an entire problem?

It is tempting to take a technology-led approach to developing SaMD. This often results in narrow solutions that fail to address the user’s complete problem, thereby delivering limited value.

We recommend a user-centred approach, using technology as an enabler to solve the user’s problem completely, rather than relying on technology for the core value proposition.

It’s especially crucial to consider other, potential unintended consequences. Solving one facet of the problem may just create another, unintended problem. An example of this would be the creation of false positives in the early detection of cancers that increase unnecessary demand for biopsies or CT scans.

2. Can you measure the real impact?

Developing a functional solution in healthcare, particularly in SaMD, is not enough. In order to secure approval from both a CMO and CFO, it is crucial to also demonstrate its effectiveness. This can be a demanding process that may require extensive and costly studies or clinical trials. It is entirely possible to create an effective product but be unable to prove its benefits.

For example, a trial to establish whether a solution decreases maternal deaths by 25% would need to involve every pregnancy in England over a two-year period, which is clearly not feasible.

By thoroughly evaluating how you need to demonstrate the impact of your solution before beginning development, you can ensure that you are creating something genuinely game-changing that justifies the investment.

3. Do you understand how your SaMD fits within the users’ and data journeys?

Whilst your new solution may perform extremely well in testing, it’s important to consider whether it can provide the expected benefits in real-world scenarios? One of the key reasons why solutions fail in their transition from testing to implementation is a lack of understanding of the data and user journeys.

Take for example a wearable device that uses gait analysis to detect Parkinson’s disease at an early stage. In testing the device gets brilliant results as it detects the disease in its early stages.

However, it may not provide the expected benefit of early diagnosis. Why not? Because users need to recognise that they might have something wrong before they think of using the solution, which negates a lot of the benefit.

Similarly, data can play a crucial role in the success of your solution. If it’s not available in the right or at the right time, the expected benefits may not be realised. Before committing to your solution, it’s important to assess how it fits into the existing or future user and data journeys. 

4. Are you confident your SaMD solution can scale?

Having a valuable solution that is well-received by healthcare providers and their teams is a great start, but it’s not enough to guarantee success. To truly make an impact, your SaMD must be able to reach a substantial scale.

It’s common for healthcare solutions to be developed and tested in one setting, but this doesn’t mean they can be implemented in other locations or by other providers. By considering scalability from the very beginning, you’ll be able to ensure that your SaMD not only solves a specific problem but also has the potential for commercial viability.

You’ve got to plan for your SaMD being used not just in one location, but across the entire healthcare system from the start. It’s all about having a good understanding of what’s needed in the market and the underlying infrastructure of each of your potential customers (hospitals, GPs, etc.). If you can’t adapt your solution to suit the processes within all these organisations, you might build a really great product, but it’s only applicable to one of them.

So, how can you ensure business scalability? Essentially, don’t forget about the basics – market research. Engage with the market and be flexible. Don’t just go down a single road and focus on building for one organisation. Instead, take a step back and truly inspect what the market needs, wants, and how it’s behaving.

5. What is your path to reimbursement?

Reimbursement models via the NHS aren’t always obvious so it’s entirely possible to make a really good SaMD product which works, but monetising it is a struggle because there’s no mechanism to do so. Before you even start building your SaMD, get a clear understanding of how you’re going to get reimbursed for all your efforts. Essentially, you need to outline your business model.

To do this, you need to know who your target customer will be. In the UK, the direct-to-consumer route can be challenging. Our research on the future of the NHS illustrated that while people do want MedTech solutions, they aren’t willing to pay for them. So, you’ll likely target trusts in hospitals, general practitioners, or ICSs.

Once you’re clear about who it is you’re targeting, you need to understand what that group truly needs and how you can effectively approach them.

6. How will you ensure security?

Of course, your SaMD security considerations will depend on the nature of the solution you’re building. In the case of a symptom checker, like the one our team developed for the NHS COVID-19 app, the main thing you’ve got to do is protect the data you’re collecting because it’ll include special category personal data as outlined by GDPR.

Another element you’ve got to think about is protecting your system from cyberattacks and ensuring that it won’t be compromised or degraded in any way by a security incident. After all, the healthcare industry is a major target for cyber criminals due to the vast amount of information that hospitals, practitioners, and solution providers hold.

For example, if you’ve got a cloud infrastructure which helps alert clinicians when a patient’s condition is deteriorating and you fall victim to a DDoS attack that you can’t handle – consequences can be life-threatening. To avoid this, you need to not only design the solution effectively from the get-go but also implement strong cybersecurity measures.

To ensure security, you have to build your SaMD properly from the beginning and follow best practices, carrying out the risk assessment upfront, running thorough tests, and so on. Of course, if you’re a startup, it may seem like you don’t have the money or time to do this. But it’s well worth it, as cutting corners won’t work here.

7. Can you accelerate time to value?

Bringing SaMD solutions to market can be a lengthy process due to the need for evidence generation and meeting regulatory requirements. To maximise revenue, consider exploring your options where possible.

For instance, if you are developing a tool that assesses individual’s risk of complications, you could first deploy it at the population level to forecast demand and capacity.

By addressing an adjacent, non-regulated issue, you can start generating revenue while accumulating evidence of effectiveness, and effectively be paid to de-risk your product.

Building game-changing SaMD solutions

The development of SaMD presents a unique set of challenges. It’s therefore critical to adopt the right approach to develop SaMD, which delivers tangible benefits and high impact at scale, while keeping up with the fast-changing healthcare landscape. 

The key to success lies in understanding the importance of user adoption, data integration, and impact measurement. Anticipating unintended consequences and ensuring scalability is also crucial for long-term success.  

By following these principles and keeping the end-user in mind, you can create SaMD solutions that are genuinely game-changing and have a lasting impact on improving patient outcomes and streamlining healthcare processes.  

To find out how we can help you scope and assess your SaMD initiatives for success, book a discovery call with the Zühlke team today. 

To read this article in full, visit Zühlke’s website here.