The report “Evolving Data-Driven Regulation” by a European strategic advisory group on Big Data has outlined the need for medicines regulation to adapt as healthcare data and technology evolve. The objective is to provide the European regulatory network and stakeholders with recommendations on how to leverage the potential of Big Data for public health and innovation. This will be important to support robust, data-driven and evidence-based decision-making regarding the development, authorisation and monitoring of medicines.
Advances in science and technology, driven by innovations such as diagnostics, wearables, sensors and connectivity, generate vast amounts of data. Computational power, bioinformatics tools, algorithms, machine learning and artificial intelligence are gaining increased access to healthcare systems. This shift means that evidence generation must keep pace with scientific innovation. Although randomised, double-blind, controlled clinical trials remain the benchmark standard for most regulatory use cases, the regulatory system must evolve to incorporate complementary evidence generated by Big Data.
The European Medicines Agency (EMA) — the EU body responsible for the scientific evaluation, supervision and safety monitoring of medicines — and the Heads of Medicines Agencies (HMA) — a network comprising the heads of national competent authorities across the EU and EEA — established a Joint Big Data Taskforce consisting of two phases.
Phase I involved reviewing the Big Data landscape, identifying opportunities, and conducting a gap analysis of expertise across the European regulatory network. This phase also yielded a summary report containing 47 core recommendations and 138 supporting actions.
Phase II, which is the core of this article, prioritised the recommendations from Phase I and developed practical suggestions for their implementation by the European regulatory network in collaboration with stakeholders.
The problem statement that guided the Joint Big Data Taskforce highlighted that, although Big Data offers novel insights, the acceptability of this evidence for regulatory decision-making requires clarification. Furthermore, the EU regulatory network lacks the capacity and capability to access and analyse large, heterogeneous and unstructured datasets.
The resulting vision for this taskforce is to strengthen the regulatory system, so that it can efficiently integrate data analysis into its assessment processes and improve decision-making. This will require building expertise, fostering external collaboration, ensuring an ethical culture of data sharing, and understanding the quality and relevance of data sources for the European population.
The 10 Priority Recommendations
Phase II has consolidated the extensive recommendations into 10 key priorities, organised for action by the European regulatory network, specific committees/working parties, or through collaboration with stakeholders. These recommendations are fully compatible with the current EU legal framework.
- Deliver a sustainable platform to access and analyse healthcare data from across the EU. Develop a robust business case in collaboration with relevant stakeholders and secure funding to establish and maintain a secure EU data platform that facilitates better decision-making on medicines by providing robust evidence from healthcare professionals.
- Establish an EU framework for data quality and representativeness. This includes developing guidelines, strengthening the process for data qualification and promoting the uptake of electronic health records, registries, and genomics data across member states.
- Enable data discoverability. Key metadata should be identified for regulatory decision-making on the choice of data source, and the use of the FAIR principles (Findable, Accessible, Interoperable and Reusable) should be promoted.
- Develop EU network skills in Big Data. This requires a skills analysis, developing a Big Data training curriculum, collaborating with external experts (e.g. academia), and targeted recruitment of specialists such as data scientists, epidemiologists, and AI experts.
- Strengthen EU network processes for Big Data submissions. A ‘Big Data learning initiative’ should be launched to track submissions involving Big Data, review outcomes, and feed these learnings into guidelines and reflection papers. In this regard, the existing HMA-EMA Catalogue of real-world data sources should be enhanced to increase transparency on study methods.
- Build EU network capability to analyse Big Data. This involves building computing capacity to receive, store, manage, and analyze large datasets, including Patient Level Data; establishing a network of analytics centres; and strengthening the network’s ability to validate AI algorithms.
- Modernise the delivery of expert advice. Build on the current working party structure by creating a Methodologies Working Party, covering biostatistics, modelling and simulation, extrapolation, pharmacokinetics, real-world data, epidemiology, and advanced analytics. In parallel, establish an Omics Working Party to expand and strengthen the existing pharmacogenomics group.
- Ensure data is managed and analysed within a secure and ethical governance framework. Actions include engaging with EU data protection regulations to deliver data protection by design, consulting patients and healthcare professionals, and establishing an Ethics Advisory Committee.
- Collaborate with international initiatives on Big Data. Support the development of guidelines at international multilateral fora, a data standardisation strategy delivered through standards bodies, and bilateral collaboration and sharing of best practice with international partners.
- Create an EU Big Data stakeholder implementation forum. Engage in constructive dialogue with key EU stakeholders, including patients, healthcare professionals, industry representatives, payers, device regulators and technology companies. Key communication points should be established in each agency, and a resource of key messages and communication materials on regulation and Big Data should be built.
From Recommendations to Action: Building a Data-Driven Regulatory Future
Implementing the recommendations will require a range of actions. These include providing training and development opportunities for staff, alongside targeted recruitment of new personnel, and conducting demonstration pilots to test and refine approaches. It will also involve establishing centres of excellence in analytics and regulatory science, developing new guidance, and strengthening existing regulatory tools such as qualification advice.
In addition, investment in fit-for-purpose information technology will be essential, as will the delivery of a major Big Data initiative, such as DARWIN, which aims to create a framework for accessing and analysing EU healthcare data, with an initial focus on real-world evidence.
Moving forward, the following factors will be key to our success: building on the strengths of the current system, working collaboratively within the EU regulatory network and with EU and international stakeholders, providing clear requirements for the regulated industry, and targeting the network’s efforts to where the maximum health and innovation benefits can be delivered.
Big Data is not necessarily the solution to all the challenges faced by regulators in reaching appropriate decisions. However, new Big Data sources generate complementary evidence to clinical trials that may facilitate, inform and improve our decisions.
It is evident that the data landscape is evolving and, as such, the regulatory system must evolve in tandem. This way, we can identify opportunities for public health and innovation by providing better evidence to inform decisions on the development, authorisation and on-market safety and effectiveness monitoring of medicines.
By working together in a smart and collaborative way, and by embracing change, we can evolve to deliver better regulation for patients. This will also help establish the EU medicines regulatory network as a reference for data-driven decision-making.
- Laura Avogaro FRESCI
- No Comments