The evolving digital diagnostic landscape

21 Mar 2023 13min read

The UK’s Department of Health and Social Care (DHSC) published their medical technology strategy in February 2023. The report focused on the UK and identified key healthcare areas to improve and strategic markets to focus on. Diagnostics (as a broad base) was identified as one of these key areas with a particular focus on distributed diagnostics, sharing of health data between different healthcare providers and digital technologies.

Another report from Deloitte published in October 2022 looked at the outlook for diagnostics technologies. This report mentioned that around 25% of companies interviewed reported integration of digital infrastructure as their biggest challenge to releasing new diagnostic products (50% of respondents identified the integration of digital infrastructure in their top three concerns).

It is clear that digital technologies are considered critical to the diagnostics industry.

With this in mind, the aim of this blog is to examine what we mean by ‘digital technologies’, what opportunities exist when using them and what challenges you can anticipate when adopting digital solutions. The blog will then look at how these technologies might be applied to address the strategic needs of the DHSC, focusing on:

  • Personalised diagnostics
  • Community diagnostics

Digital technologies and their challenges

What do we mean when we talk about ‘digital technologies’? At its most literal, ‘digital’ is a means of representing a signal, most commonly in a binary format. The general population largely uses the term to describe cloud computing, software systems (including smartphone apps), digital communications and wearable technologies. These are the core components which underpin any digital technology strategy.

Let’s look at a few examples of the core components underpinning digital technologies and some challenges diagnostics companies face in adopting them.

photo of person checking smartwatch with health data

Cloud computing for medical and diagnostics applications

Cloud computing underpins a significant number of web-services (like Netflix). Applications are built using a series of discrete services which are provided by the host (e.g. Amazon Web Services). These services can be configured and combined to form a coherent solution.

The primary advantage of cloud computing over traditional server-based applications is the ease with which applications can be scaled to meet changing levels of user traffic or application storage needs. In traditional ‘self-hosted’ network infrastructure, diagnostics companies manage the infrastructure themselves and can only increase the system capacity by purchasing more equipment at significant cost (servers, network adapters, databases etc.) By using a cloud platform, these companies can automate the provisioning of new resources to meet the demands of their users and quickly decommission resources when they are no longer needed. All of this can be achieved at a relatively low cost.

Many cloud platforms provide services designed specifically for medical and diagnostic applications. In some cases, they can even provide compliance documentation which can later be submitted to regulators and notified bodies. These services can be used to speed up the implementation of medical and diagnostic solutions in the cloud.

The key challenge of cloud computing for diagnostic organisations is around data interoperability between platforms. There is a wide range of communications standards when sharing data between devices and cloud services. More details in this blog further down.

Software systems and smartphone apps

The average person has more computing power in their pocket than the Apollo space missions of the 1960’s. A patient’s smartphone often contains a high-quality camera, provides an intuitive user interface, a reliable network interface and includes internal sensors (accelerometers, biometric sensors etc.) which can be easily utilised.

Using a patient’s smartphone to monitor their condition reduces the number of devices they need to manage. It also provides a single interface to engage with which collects data from external sensors in a simple way. Smartphone health apps can, if designed well, empower patients to manage their diagnosis and treatment.

Unfortunately, while smartphones have impressive capability, we’re still a long way from being able to use them to reliably produce diagnostic data without external devices, there’s just too much variability between devices.

Notwithstanding the variation between smartphones, we have the potential to use off-the-shelf computing and Software as a Medical Device (SaMD) to distribute diagnostic applications more broadly. SaMD gives us some opportunity to reduce the cost and complexity of traditional IVDs and update our data processing algorithms more easily and reliably when changes are identified.

Digital communications and the internet: a part of the digital diagnostic landscape

The internet itself is clearly not a diagnostic device, but it underpins almost all communications between devices and services. More patients, clinicians and operators expect that diagnostic devices will seamlessly connect to at least a local network and in many cases, wider health platforms via the internet.

We may not traditionally consider a point-of-care diagnostic or lab-based diagnostic instrument a ‘digital’ device. But when IVDs are integrated into the wider healthcare ecosystem (providing patient data to healthcare information systems and records platforms), they form a key constituent part of the overall digital diagnostics landscape.

Similar to cloud platforms, many diagnostic device manufacturers have identified that the integration of devices into networks with varying data format standards is a key challenge to address.

Diagnostic wearable devices

Wearables can collect data about a patient using small, reliable sensors to continuously monitor their condition in a user-friendly way. An example of a wearable device in a diagnostic context would be the continuous glucose monitors now available to diabetics. These are considered ‘digital’ solutions as they are integrated within a wider connected ecosystem.

Wearable devices rely on well-established communications technologies like bluetooth low energy and radio frequency ID, which can be integrated with smartphones or bespoke reader devices. In some cases, we could implement wearable diagnostics using a consumer device like the Apple Watch or Fitbit, implementing SaMD operating on the hardware. In this case, we would be launching SaMD applications on consumer devices.

The key challenge here is to ensure that wearable medical devices are user friendly enough that consumers want to use them and are able to do so reliably. The user interface and communication around the device needs to be well designed so that users can understand it. It also needs to be better at monitoring someone’s condition overall compared to their previous device or diagnostic.

Artificial intelligence and machine learning in a diagnostics context

Artificial Intelligence (AI) has been making headlines in recent months when OpenAI released ChatGPT: a natural language model used to open a dialogue with users in an (almost) conversational workflow.

In the diagnostics sector, during the peak of the COVID pandemic, the ZOE COVID tracking app (now the Zoe Health Study) used symptoms reported by patients together with the results of PCR and LF tests to identify and track infections. By asking patients to report symptoms on a regular basis, their AI platform was able to identify early warning signs and emerging symptoms of COVID below the limit of detection for lateral flow tests.

While this is an interesting area for development with highly variable applications, there is some need for caution when implementing AI for any medical purpose.

Biased or incomplete data sets can lead to your AI system inferring invalid or incorrect results. Team Consulting’s Thorbjorg Petursdottir explores this topic in detail in her blog on sexism in AI.

These risks aren’t insurmountable. Developers are aware of the requirements when releasing AI applications. Engineers must validate the training data used to train the model, in order to identify and eliminate bias. Later, they need to validate the algorithm using a different, high quality data set, to confirm that the AI is functioning correctly.

Regulators are expecting developers to take a pragmatic but risk-managed approach to managing AI.

photo of someone with aktiia’s wearable watch

The DHSC medical technology strategy

The DHSC’s medical technology strategy (published in Feb 2023) is broad reaching. The report aims to address the evolving needs of the UK healthcare sector and identifies targeted improvements across a wide range of health areas. It is split into four key priority points:

  • Resilience and continuity of the NHS supply chain
  • Enabling innovative and dynamic markets to increase choice and competition
  • Enabling infrastructure improvements within the healthcare system
  • Specific market focuses (such as diagnostics)

From all the identified healthcare needs, the DHSC strategy recognises that diagnostics is a key healthcare component, particularly the need for both community diagnostics centres and digital transformation.

The rest of this blog covers how digital technologies can help the transformation of community-based diagnostics and personal diagnostic testing.

Community-based diagnostics

Community-based diagnostics and what the DHSC report calls “Community Diagnostics Centres (CDCs)” allows the de-centralisation of testing to reduce the pressure on hospitals. Community diagnostic centres aim to be closer to patients (akin to local GPs surgeries) and offer specialist diagnostic capabilities. This has several advantages, including:

  • Minimising the spread of infectious diseases
    Consider the COVID testing sites established in the UK during the pandemic. With a potentially large number of people carrying a highly infectious disease, moving testing away from the hospital can slow down the spread of infection by minimising patient mixing and protecting the most vulnerable.
  • Reducing transport burden for patients
    Patients who require testing in hospitals may have to travel significant distances for tests, particularly if they use public transport. Moving diagnostics to de-centralised hubs ensures that patients have better access to vital diagnostics capabilities.

So how is ‘digital technology’ related to community-based diagnostics? With more distributed diagnostics capabilities as explained above, IVDs will need to be better integrated into hospital networks. This can be made possible by either continuously uploading results from test centres or uploading results on a regular basis. This could be done by using remotely hosted, cloud based SaMD to process test results efficiently and notify relevant clinicians that the results are available seamlessly.

These solutions rely upon the interoperability of devices and data which the industry has identified as a key concern moving forward.

Historically, IT providers carried out unique communications protocols, which IVD manufacturers implemented to receive data from remote devices. With the expansion of health IT platforms, there was a rapid proliferation of communications standards which had to be applied.

The industry has now largely standardised broad based communications standards such as FHIR and POCT1A for diagnostics. These protocols aim to ensure that diagnostics platforms and laboratory systems can share data and integrate seamlessly into clinical settings. We’re currently in a transitional period, with developers building an understanding of the protocols and IT providers adopting them – but with time, the interoperability issue should be improved.

Personal diagnostic testing

Personal testing is largely an extension of community diagnostics. Personal testing has the potential to empower patients to better manage their own symptoms while reducing their dependency on clinicians.

photo of someone adjusting wearable glucose monitor on arm

Going back to the example of wearable glucose monitors, diabetics can wear sensors attached to their skin, continuously reading their blood glucose concentration. The device pairs to their smartphone via Bluetooth and can provide warnings when their blood glucose is too high or too low. Because diabetic patients using these devices are continuously monitored, there is no need for repeated finger stick tests, which is much more comfortable and enables them to manage their condition more predictably.

We have recently seen the expansion of ‘virtual wards’ in the NHS, where patients are monitored by clinicians in their own homes. Patients on the ‘virtual ward’ will have very different care needs. They might need a wide variety of different parameters to be monitored e.g. blood pressure, pulse oximeter, heart rate, respiration rate. Gathering data to a central clinical dashboard that is able to be frequently updated, is critical to the patients’ care.

By facilitating this transfer of data, the diagnostic industry will ensure that clinicians on ‘virtual wards’ are able to focus their time and attention to patients’ care needs rather than completing test records.

To achieve this, there are some key challenges to tackle:

Patients are not experts in using complex diagnostics

While patients are more educated than ever in the digital era and highly knowledgeable about their condition, they are not trained experts in using complex diagnostics. If device developers are to empower patients to monitor their own conditions, diagnostic devices need to be intuitive and easy to fix when problems arise. You should also consider how familiar interfaces could be used to help patients use their diagnostics device safely.

Using smartphones both as an interface to home diagnostics and to report symptoms, provides a natural way for patients to interact with their devices, which your team can quickly adapt based on user feedback. Apps can provide the user with detailed, easy to access instructions which guide them through the process of using their device.

The need for cost-effective diagnostic systems

Expanding testing into the wider population may reduce the overall cost to health systems by spotting the early warning signs that a patient’s condition is changing. To ensure that the economics stack up, it is critical that these personal diagnostic systems are cost-effective not just at the point of purchase, but throughout its lifetime.

What can we do?

We’ve explored what we might mean by ‘digital technologies’ and the ways they might be applied to evolving healthcare strategies. By applying digital technologies in considered ways, medical device developers, like Team Consulting, hope to de-centralise healthcare and ensure that data can be quickly and easily shared with relevant clinicians for better access to care, quicker diagnostics and ultimately, better patient outcomes.

Join the conversation

Looking for industry insights? Click below to get our opinions and thoughts into the world of
medical devices and healthcare.