Software as a Medical Device
25 February 2025
Software as a Medical Device (SaMD) is a hot topic, with a predicted compound annual growth rate (CAGR) of 13.6% between 2024 and 2033 and predicted value of $5.0 billion in 2033 [1]. This expansion is driven by key trends such as the integration of wearable devices for remote patient monitoring, the convergence of consumer and healthcare technologies, and the application of artificial intelligence and machine learning (AI/ML) in diagnostics.
Wearable Devices for Remote Patient Monitoring
Fitness trackers that have flooded the consumer health market are capable of providing users with insights on heart rate (variability), body temperature and even blood oxygen saturation – used by fitness enthusiasts to optimise their training plans. Whilst these devices themselves cannot be used for medical purposes, they show just feasible it is to remotely monitor and report many metrics used for health monitoring.
As hospitals are ever more crowded, there is added impetus to treat patients at home – with the UK government launching the “NHS @home” initiative[2]. Patients are treated by healthcare professionals at home, but are required to monitor their own vital signs and report any issues. Given the burden of this on the patient, it’s no surprise that the medical devices telemonitoring market is expected to grow at a CAGR of 11.3%, with a market value of $8 billion in 2033[3].
By deploying a SaMD backend to allow healthcare professionals to remotely access and review the data collected by these devices, the appropriate medical support can be provided before the patient needs to be re-admitted to the hospital, reducing the stress on the patient and burden on the healthcare system.
A Smart Inhaler NFC app created by Springboard
The Blurring of Consumer and Healthcare Technologies
Whilst remote patient monitoring is well-established amongst the large medical device manufacturers, there are signs that the distinction between medical and consumer devices is becoming increasingly blurred – in part thanks to SaMD.
Apple’s 2024 showcase offered little in the way of revolutionary consumer products, but hinted at a new branch for the company, with new software updates for the Apple Watch and Airpods Pro 2 offering FDA approved medical features (ECG and sleep apnoea detection. and hearing aid functionality respectively) within their consumer products[4]. By incorporating SaMD, Apple is using their non-medically approved hardware for medical purposes – although the approval process itself has been carried out over a 2-year period in partnership with the FDA[5].
These capabilities are also only those developed by Apple, and do not account for the 40,000 health-related apps currently available for download within the Apple App store ecosystem[6]. These apps include tools for image post-processing and fusion, and for surgical planning and training – bringing medical software to consumer hardware.
Whilst the current healthcare features within Apple’s devices aren’t enough for a formal diagnosis, they can form part of the overall medical system, allowing for patients to be effectively triaged. It is estimated that Apple watches may be identifying a few thousand atrial fibrillation cases per year[7], which is both saving patient lives and helping the bottom-line for medtech companies. While the increase in patients identified through these devices is currently modest, it highlights the growing role of consumer technology in healthcare diagnostics.
AI/ML for Diagnostic Purposes
In 2023, 36% of all SaMD products were focused on screening and diagnostic purposes1, including software for the manual viewing of medical images, as well as software to aid in the detection of regions of interest.
As artificial intelligence (AI) and machine learning (ML) capabilities increase, the number of SaMD products in this area is going to increase. Recent models have shown a 94% accuracy[8] in the detection of cancers, and has therefore led to further governmental backing of AI[9] for radiology image analysis to improve patient outcomes – with the aim of reducing staffing shortages and shrinking diagnostic backlogs.
The integration of AI/ML extends beyond imaging, with further options to process the vast amounts of data that could be generated from wearables used in remote patient monitoring into actionable patient-centric insights. By proactively monitoring (at-risk) patients, clinicians may be able to intervene earlier, improving the overall patient outcomes and optimising resource usage within overstretched healthcare systems.
Challenges and Regulatory Considerations
Despite the promising opportunities afforded by SaMD, there are several regulatory challenges that need to be overcome by anyone attempting to develop these applications.
Despite being a software product, these applications are still necessarily medical devices; and therefore must go through the relevant regulatory approval process before entering the market, and then have sufficient subsequent monitoring once available to the public. If any vulnerabilities are found in the software, there is a 3-month window to get this fixed and still comply with section 524B of the FDA cybersecurity regulation [CITE] – so there must always be a software maintenance team on standby for the life of the product.
As these products are medical devices, all system changes – including performance updates to improve the AI algorithms – need to follow the pre-approved change control plan (PCCP). This is a key challenge for AI adaptivity, where new data/types of information may become available for training and re-optimisation of a model. To this end, the UK MHRA released the Software and AI as a Medical Device Change Programme[10] – which aims to clarify how AI interpretability and adaptivity can be implemented within medical systems. However, until a homogenised framework is established between the different regulatory agencies, updating SaMD products for different markets will be an arduous process.
Summary
The SaMD market is primed for rapid growth, driven by advancements in AI and the push to treat patients outside the hospital. How established SaMD can become though will depend on how regulators and manufacturers adapt to changes in technology and patient protections respectively. If done correctly, the sector shows great promise to revolutionise healthcare delivery.
By Mike Beck, Senior Electronic and Software Product Engineer, Springboard
References:
[1] Software as a Medical Device (SaMD) Market Size | CAGR 13.6%
[3] Telemonitoring Systems Market Growth Analysis | CAGR of 11.3%
[4] Think of the New Apple AirPods and Watch As Medical Devices
[5] The Apple Watch just made history with a feature that can save your life | Tom’s Guide
[6] Healthcare – Products and Platform – Apple (UK)
[7] Is Your Smartwatch Boosting Med Tech Stocks? – Barron’s
[8] New AI tool can diagnose cancer, guide treatment, predict patient survival— Harvard Gazette
[9] World-leading AI trial to tackle breast cancer launched – GOV.UK
[10] Software and AI as a Medical Device Change Programme – Roadmap – GOV.UK
