OR WAIT 15 SECS
Sean Milmo is a freelance writer based in Essex, UK.
EU regulators have accelerated their efforts to use the mass of data emerging from the lifecycles of drugs as an effective basis for both the development and control of medicines.
Editor’s Note: This article was published in Pharmaceutical Technology Europe’s March 2020 print issue.
The European Union medicines licensing network, headed by the European Medicines Agency (EMA) and covering all of its 27 member states, has been falling behind the US Food and Drug Administration (FDA) and Japan’s Pharmaceutical and Medical Devices Agency (PMDA) in the application of Big Data in the authorization of medicines.
Now a joint Big Data taskforce (BDTF) of EMA and the Heads of Medicines Agencies (HMA), representing national licensing authorities, has completed a twoâphase report on a proposed framework for realizing the potential of Big Data for enhancing public health and innovation (1). It aims to aid decision making not just by regulators but also by stakeholders, such as industry, academia, and healthcare professionals.
Although the proposals are mainly focused on raising the efficiencies of authorization and postâmarketing surveillance procedures, they are also aimed at strengthening the development of new medicines and processes, especially with advanced therapy medicinal products (ATMPs). The introduction and implementation of the framework will be a huge task in pulling together the data management activities of EU countries’ healthcare systems, while also establishing links with the Europe-wide data infrastructures of different research communities, especially in life sciences.
However, the EMA–HMA initiative will be greatly helped by the impetus coming from a programme announced by the European Commission (EC), in February 2020, with the objective of making the EU’s data economy a leading force in the world (2). Through substantial advances in connectivity, processing and storage of data, computing power, and cybersecurity, Europe will be able to make large amounts of quality data available for use and re-use to improve health, well-being, the environment, and public services, according to the EC’s plans (2).
“[The digital transformation] will be a truly European project-a digital society based on European values and European rules-that can truly inspire the rest of the world,” said the EC in a communication on Europe’s digital future in February 2020 (2). “But the benefits arising from digital technologies do not come without risks and costs. Citizens no longer feel in control over what happens with their personal data.”
Hence the realization of the EU’s digital strategy will require new legislation, changes to existing regulations, guidelines, and other regulatory actions. This will particularly be necessary in healthcare and regulation of products like medicines, where issues such as standards of data quality, accountability, data management, transparency, and interoperability will be important.
A key area will be the protection of personal data used in the development and improvement of new and existing medicines and treatments such as gene therapies, which may require access to real-world data (RWD) in patient records. The BDTF’s report is helping to shape the digital future in healthcare by putting forward 10 main recommendations, the implementation of which will become the responsibility of an HMA–EMA Big Data steering group (1).
The taskforce suggested that the immediate focus should be on preparing the EU’s existing regulatory model for the new data environment. But it warned that the European regulatory network has currently both limited capacity and capability to access and analyse Big Data’s large, mixed, and unstructured data sets. The objective of using Big Data to raise the efficiency of regulatory decision-making “still sit closer to aspiration than reality,” the taskforce said (1). Nonetheless regulators had no alternative but to exploit the growing diversification of data generation.
The BDTF’s recommendation aimed to tackle what the taskforce saw as the regulatory system’s main weaknesses in dealing with Big Data. These weaknesses ranged from the fragmentation of the European healthcare sector with countries pursuing their own policies in areas such as data generation and distribution. There was a lack of standardization in data management and controls and uniform metrics for gauging data quality (1).
The General Data Protection Regulation (GDPR), approved by the EU in 2016 to replace the 1995 Data Protection Directive but which started to be implemented only last year (3), is already showing inadequacies in the protection of personal data, especially in respect of the application of artificial intelligence (AI) and machines learning with pharmaceuticals, as well as the use of algorithms in healthcare.
To sort out many of the issues undermining effective use of data, the EU has had to find a means for dealing with the disparities between the healthcare data provided by its member states. Europe has a richness of healthcare data because in most countries there is universal healthcare coverage but the region lacks the means to take advantage of it, according to BDTF (1).
Differences in health services, national guidelines, clinical practices, and even terminologies have resulted in the generation of heterogeneous data. Also, there is a shortage of funding at the national and EU levels to promote the creation of data, such as real-world databases, which could be used for regulatory purposes.
BDTF claimed that in comparison to European countries, the United States, Canada, and Japan had committed ‘significant resources’ to establishing RWD systems to which regulators would have routine access (1). BDTF recommended that to encourage a centralized RWD operation, a European network of databases of ‘known quality and content’ with the highest level of data security should be set up. It would be called the Data Analysis and Real-World Interrogation Network (DARWIN) (1).
In addition to worries about data inconsistencies due to differences between national healthcare services, the taskforce was also concerned about poor data quality. BDTF cited a study which found ‘disappointingly low’ numbers of European databases able to meet minimum requirements on content and accessibility (4). BDTF suggested the establishment of a data quality system under which Big Data sources would have to undertake a renewable certification process.
Operators of Europe’s data infrastructures, particularly those in Open Science sections, have been applying for around five years a standardization system (5). It is called FAIR, for findability, accessibility, interoperability, and reusability, which when bound together should ensure acceptable levels of data quality.
Although FAIR has now been endorsed by governments not only in Europe but globally, the BDTF considered that it was only a long-term option for creating Big Data standards for the EU regulatory network. An interim report on four pilot assessments of the FAIR-compliance of datasets from research work by Brussels-based Innovative Medicines Initiative (IMI), which is jointly funded by the EU and the European Federation of Pharmaceutical Industries and Associations (EFPIA), found that compliance levels were ‘variable’ (6). Furthermore, the assessment processes themselves were underdeveloped and immature, leaving a lot of scope for improvement.
Instead the taskforce proposed that for the short-term, EU regulators should use the European Network of Centres for Pharmacoepidemiology and Pharmacovigilance (ENCePP) and other repositories, such as Genomic Alliance for Genomic Health (GA4GH), for genomics data and Elixir, a European inter-governmental data-coordination platform for life sciences (1).
A major barrier to effective regulatory use of Big Data is the inability of current legislation to give full privacy protection to patient data. Despite its tough controls on personal data, the EU’s GDPR is already considered to be in need of tightening up because of weaknesses particularly in protection of patient data (7). In a review of the legislation issued in November 2019, Germany’s federal and state data protection supervisory authorities concluded that, despite the success of its regulatory concepts, “some uncertainty remains and there are shortcomings when it comes to implementation” (7).
Because of a lack of clarity in parts of the GDPR, ‘countless’ data protection consultants have been offering customers interpretations of the text, the German authorities complained. These views were echoed in a briefing in on GDPR by GA4GH (8), which claimed that with key issues, such as consent and public interest, the legislation does not “give (enough) flexibility but reinforces a conclusion about the need for clarity.”
As part of its data strategy, the EC is planning a series of measures, many of them legislative. With so many legislative initiatives in the pipeline, it could be several years before the EU introduces a full regulatory framework that determines how successful medicines regulators will be in using Big Data in their decision making.
1. HMA–EMA Joint Big Data Taskforce, “Evolving Data-driven Regulation,” ema.europa.eu (Amsterdam, January 2020).
2. EC, “Shaping Europe’s Digital Future,” ec.europa.eu Communication-COM (2020) 67 final (Brussels, 19 Feb. 2020).
3. EU, The General Data Protection Regulation. Regulation (EU) 2016/679 (Brussels, 23 May 2018).
4. A. Pacurariu, et al., BMJ Open, 8 (9) e023090 (2018).
5. EC’s Expert Group on FAIR Data, “Turning FAIR into reality,” ec.europa.eu (Brussels, 2018).
6. IMI and FAIRplus, “First phase exemplar IMI projects FAIRified” (Brussels, 2019).
7. Independent German Federal and State Data Protection Supervisory Authorities, “Report on Experience Gained in the Implementation of the GDPR,” datenschutzkonferenz-online.de (Berlin, November 2019).
8. GA4GH, “GDPR Brief: ’At Least One’ Legal Basis for Processing under the GDPR: Clarifying Article 6 (1),” ga4gh.org, 6 Jan. 2020.
Pharmaceutical Technology Europe
Vol. 32, No. 3
When referring to this article, please cite it as S. Milmo, “Playing Catch Up with Big Data in Europe,” Pharmaceutical Technology Europe 32 (3) 2020.