Rapid advances in digital health, particularly in the last decade, are revolutionising the management of health conditions. From wearable gadgets to artificial intelligence (AI)-based software, end users span from patients to clinical lab personnel.

As with all medical products, digital health solutions with a medical purpose are subject to regulation. Existing regulatory frameworks initially developed to ensure the safety and effectiveness of traditional in-vitro diagnostics and medical devices, however, are not suited to the fast-paced innovation that characterises digital health.

A new white paper by the Asia Pacific Medical Technology Association (APACMed) provides best practice recommendations in digital health regulation to ensure greater access of software innovation. Varun Veigas, Regional Regulatory Affairs and Policy Lead at Roche Diagnostics Asia Pacific and Digital Health Regulatory Working Group Chair and China Centre of Excellence Chair at APACMed, was one of the authors.

How digital health regulation affects clinical labs

Lack of consistency on what qualifies as a medical device can have important implications on the accessibility of these products. Unnecessary regulatory oversight can delay or influence the use of innovative software in labs. In turn, this may impact their efficiency, ability to support physicians’ treatment decisions and patient outcomes.

“Not everything used in a laboratory setting is necessarily a medical device or should be regulated,” notes Veigas. Examples of software products that should not be regulated include medical device data systems (MDDS), laboratory information and management systems, and low-risk clinical decision support applications (CDS).

Furthermore, software is frequently modified after it has reached the market. If every modification to an existing laboratory software platform was required to undergo the same regulatory process, this would significantly increase the time needed for its implementation. Regulatory pathways should be streamlined to accommodate such software modifications so they can be deployed rapidly without compromising safety and effectiveness.

Key concepts in digital health regulation

In the paper, APACMed provides regulatory recommendations in five main areas:

1. Qualification

Not every software service used in a clinical environment is considered a medical device. Appropriate qualification of software allows regulators to focus their limited resources on high-risk products.

2. Risk classification

Software should be regulated according to its risk level, based on what extent the information generated is used in the healthcare decision and the severity of the health condition involved. For example, software that is used to treat or diagnose a critical health condition would be considered high risk, whereas software that is simply used to inform the clinical management of a non-serious condition would be low risk.

3. Software with multiple functions

Currently, software is regulated as a single device, regardless of its functional complexity. Software products with multiple functions may break down into a significant number of applications or modules that include medical device and non-medical device functions. For example, the software application of smart watches may be considered a medical device as it includes an ECG function although the watch itself is a consumer device and does not have a medical purpose. In such instances, it is important that regulators appropriately qualify and evaluate the intended use of each module or function independently, as the various modules may have medical or non-medical device functionality, even while residing on the same platform.

4. Alternative regulatory pathways 

Regulatory authorities currently use the same lengthy review process used for traditional medical devices and in-vitro diagnostics that can take months to years. Innovative protocols are needed to streamline the review of software products and their modifications so that they can be brought to market faster.

5. Frameworks for the regulation of AI/ML

Regulatory authorities create extra layers or insufficient regulation for software products that leverage AI or machine learning (ML). They should regulate AI/ML-based software as a medical device (SaMD) based on intended use and tailor regulatory frameworks to address their unique and iterative aspects.

Striving towards regulatory convergence in the Asia Pacific

There is growing interest in digital health regulation in the Asia Pacific region, evident with the rise of published guidance by local regulatory authorities in the last few years. The current extent to which best practices are encompassed in the regulatory frameworks of three major countries, Australia, Japan, and Singapore, is shown below.

Best practices and gaps in the Asia Pacific Region

While the Asia Pacific region is making progress on digital health regulation, more harmonisation and better guidelines are still needed. Greater consistency and predictability in regulatory review processes will enable the safe and widespread adoption of digital health solutions.

Key takeaways

• Digital health solutions should be regulated according to their intended use and within frameworks that accommodate fast-paced innovation
• Appropriate software qualification and alternative regulatory pathways will ensure faster access of software products in clinical labs
• Greater regulatory convergence and improved guidelines are needed for digital health regulation in Asia Pacific

To learn more, download Digital Health Regulation in Asia Pacific: Overview and Best Practices, a new white paper by APACMed.