Can regulatory systems built for the industry over 70 years ago keep pace with today’s world of biologics, AI-driven devices, personalised diagnostics and combination products? As data science, technological convergence and lifecycle evidence reshape how risks and benefits are understood, regulators and industry alike must question how traditional silos, reactive post-market surveillance and compliance-driven regulatory roles can keep up.
John Wilkinson OBE, Principal at MedTech Vision and former Director of Devices at the UK Medicines and Healthcare products Regulatory Agency, examines the key drivers of change and discusses how regulatory models must evolve to unlock the future of the regulatory affairs profession.
Lessons from the past
The regulation of both medicines and medical devices is a post-World War II phenomenon and, like many regulatory frameworks, was largely driven by events. In Europe, medicines regulation was formulated as a result of the thalidomide tradegy. Device regulation was later to the party and evolved over several years, culminating in the first comprehensive European framework, the Medical Devices Directives, which came into force in the early 1990s. Similarly, issues with breast implants, metal-on-metal hip prostheses and surgical meshes have been catalysts in the evolution of device regulation.
As a former medical device regulator, I have long been conscious of the often uncomfortable fit of device regulation within agencies that were primarily designed and resourced for medicine regulation. The fundamental structure of most national agencies in Europe and beyond has changed little over the 70 years that medicine regulation has been in serious development. Process and structure were designed primarily for compounds manufactured in chemical factories. The structure was composed of pre-market licensing, post-market vigilance and inspection. In the context of its original challenge, that structure and those activities have broadly served their purpose well.
Fast forward from those origins and we have seen tensions begin to emerge, in part as a result of devices being merged with medicines activities and force-fitted into agencies. This has caused the effectiveness of regulatory operations to be sub-optimised, and the form and function of the 70-year old medicines model to become increasingly challenged.
Drivers of change
The world we now live in is witnessing unprecedented change and the arena of medicine is at the forefront. Chemical medicines are being complemented or replaced by biologicals. Vaccines, immune and gene therapies are creating new risk management challenges along with unimagined therapeutic benefits. Devices are becoming ever more sophisticated with embedded technology that enhances outcomes. Diagnostics - both in vitro and in vivo - are paving the way for increasingly personalised medicines which change the cost-benefit profile of many traditional care pathways and are highly disruptive to both industry and care delivery. Layer on top of that the impact of software and Artificial Intelligence at all stages of product development and deployment, and you end up with regulatory and risk management challenges that cannot easily be pigeon-holed and posted to one group of experts to resolve.
All of this is evidenced by the rapidly increasing number of combination products emerging on the market and the increasingly complex combinations of products deployed in clinical practice. From a regulator’s perspective, this translates into an enormous increase in expert inputs that are needed in the regulatory process, whether that is pre- or post-market. While complexity is challenging, there is a very positive element of this environment that should contribute to both better outcomes for patients and better regulatory decision-making. That element is the emergence of increasing amounts of well-structured data and the evolution of the relatively new discipline of data science. Better, quicker and more actionable data is the Holy Grail for regulators and this should create new opportunities for risk management across the whole product lifecycle. In addition, it brings greater focus on the continuous improvement of patient outcomes.
Reliance on adverse event reporting has been nearly the only tool at regulators’ disposal for post-market surveillance and, from a devices perspective, this has repeatedly generated slow and incomplete data which is difficult to act upon. Data science working off well-structured datasets offers the possibility to identify both good and bad outliers and provides the opportunity to remove the bad while highlighting and propagating the good. The transparency of this data to multiple stakeholders offers the potential to combine competence and perspective and, ultimately, achieve better outcomes at lower cost. This may require both behavioural and legislative change which is difficult, but this is achievable through wide-ranging stakeholder agreement on the ultimate goals.
What does this mean for regulators and regulatory professionals?
For regulators, the world of traditional silos is over, and it is essential to ensure that the right skills are picked and mixed to address different challenges. They will need increasingly agile organisations and access to skills and capability that are at the leading edge of technological change. They’ll also need to acknowledge that it’s impossible to outrun the pace of change and install all the necessary expertise in-house.
Collaboration is the word of the future and this extends from academia, through industry, to health systems. Regulation is not just the domain of the professional regulator but instead requires collective input from all involved in product development. Shared goals, transparency and partnership is the mantra that will be needed if regulation is to be viewed as an essential partner in risk management and the continuous improvement of healthcare delivery.
For regulatory professionals in industry, the same leadership and collaborative skills will be needed in order to engage people, both within their organisation and outside it. The days of regulation being a tick-box compliance exercise at the end of a development process are well and truly over, as lifecycle risk management, accompanied by sourcing the data to perform that role, come to the fore.
The challenge for all is huge but worthy. We cannot continuously layer new wedges of regulation on top of old while hoping to address the overall affordability challenges of healthcare; innovation in the way that we regulate and deploy resources is essential.
The affordability of healthcare systems shows no signs of improvement as expectations and capability forge ahead of the capacity to pay. To that end, regulators and health economists will need to work together to finesse the data gathering that takes place across the product lifecycle. Safety and effectiveness come first but the boundary between that and developing strong cost-effectiveness information will further blur; post-market data is critical to both. Those datasets also contain the seeds of future improvement for medical devices as they continue to evolve.
A vision for 2036
What will regulatory affairs look like in ten years? I see the following points as critical to future success:
- Greater technological convergence and overlap, which requires the integrated deployment of specialists
- Cross-specialist collaboration rather than silo working
- Transparency and shared goals with stakeholders
- Advanced data science capability and insight-driven decision-making
- Collaboration and alignment between regulatory and health technology assessment data
- Regulatory leadership equipped to drive partnerships, agility and trust
