These days the opportunities are promising to make beneficial use of the vast amounts of health (or health-related) data collected across healthcare and other settings and change research and practice. Realistically appreciating the balance between expectations and limitations is important, especially as the recent financial challenges in research and development have led all stakeholders involved to reconsider fundamental ways of operation. In this context, ‘collaboration’ is the new mantra, not just within the pharmaceutical industry, but also across the not-for-profit research sector.
The recent history
2010 marked a clear shift of mind-sets in the UK as universities’ budgets were cut down by almost £1bn while record low rates of drug approvals were reported. Big data in health research has been identified as a window of opportunity and one of the major transformative forces in these challenging times, and various settings are still exploring its role in a much-needed collaborative approach between pharmaceutical industry and academia. Collaboration between industry and academia is nothing new of course; the Bayh-Dole Act in 1980 set guidelines for such partnerships. However, in 2003 it was suggested that further development was still needed. While in several countries such collaborations have just started to come to life, in the UK progress in this area over the past decade has been impressive.
There have been incentives for both sides as, on the one hand, the pharmaceutical industry is seeking stronger ties with academia looking to speed-up drug development, especially the early phases. On the other side, after Research Excellence Framework 2014 defined that demonstrating the impact of research is worth 20% of an institution’s final score, more discussions about universities working with industry have been triggered. Noteworthy international examples to this direction include Pfizer’s collaboration with seven New York medical institutes and eight Boston research institutions, Gilead’s with Yale University, AstraZeneca’s with the National Institute of Health and Medical Research in France, and Takeda’s with Kyoto University in Japan. Most notably in the US, the Regenstrief Institute has launched more than 50 projects relying on data from the records of the Indiana Network for Patient Care, although it may still be too early to assess the impact of this partnership.
Evidence of opportunity
In this context, however, one could argue that in the UK collaborations using big data have existed for many years. Research has been a primary use of data in this country for more than two decades, based on the integrated structure of the NHS, especially in the context of CPRD (Clinical Practice Research Datalink) and NHS Digital (previously Health and Social Care Information Centre), the governmental centre designed to maximise the potential of the routinely collected data across all NHS endpoints.
Lessons from the UK can be extrapolated to other settings and countries. For example, data in the NHS primary care databases is pseudonymised at source and researchers don’t have access to identifiable information. Multiple Research Ethics Committee approvals have to be granted when running observational studies. Further, all projects, regardless of origin and funding, can only be enabled after approval of a standard research protocol by the MHRA Independent Scientific Advisory Committee (ISAC) or other relevant streamlined committees. As an example, from research that my team has done, of all 526 research projects enabled by ISAC over a past closed period of three years (January 2011 to December 2013), 67 reflected collaborations between academic and industry researchers (the number is higher if we include industry-sponsored research). In addition, of more than 1,000 published projects enabled by database research, over 35% were internationally co-authored.
Addressing the challenges
Naturally there are challenges when promoting such partnerships, namely potential polarisation within academic centres and fear of questionable scientific credibility. Industry is very focused on financial goals while academia has been criticised for being less focused on this area. This is where the problem of ethics arises. While it is certainly an important issue, it is not insurmountable and academia’s objectives can be respected. Maintaining academic values, including the right to publish and assuring research objectivity, is extremely important in this context. This is why it is imperative for researchers involved in medical informatics research to be transparent when publishing their results.
Another emerging challenge has been the lack of transfer of academic innovations into commercial projects, which has been termed as the ‘translational gap’ between the university and industry. This is where new UK structures like the Farr Institute for Health Informatics Research can play a major role. There is a clear incentive for the gap to be bridged, both for the academics who want to see their innovations become effective tools in future drug development (among others), and for industry scientists seeking to gain new technologies and bolster ever-shrinking pipelines. The faults of the conventional model have pushed the pharmaceutical industry to also consider government-funded structures, like the Medical Research Council. The most recently created partnerships feature far greater levels of interaction, better defined goals and more appropriate incentives for all parties involved. Notably, certain UK structures – like the Technology Strategy Board (TSB) or CASE Studentships (Collaborative Awards in Science and Engineering) – have provided incentives or directly fund such collaborative projects. The three-way academia-industry-government partnership, known as the ‘Triple Helix Model’, is seen by many as the latest evolution of strategic alliances which is being employed to bridge the translational gap.
Brexit means…
Of course, some of the challenges above can be addressed by refreshing the existing structures and establishing some strict and transparent means of working. However, one UK-specific overarching challenge that is not straightforward to overcome is Brexit. Almost half of all the UK academic publications stem from international collaborations and, despite vague reassurances, many university leaders see some clearly restrictive new barriers to such international partnerships.
The exact extent to which this will disturb cross-sector partnerships remains to be seen, but it will most certainly have a detrimental effect on progress. All the evidence shows that collaborations like the ones outlined above are crucial to demonstrate real-world impact. Hence, clarity is needed on funding and policy stability if the UK is to continue to benefit (and at the same time offer public health benefits) from making the best use of its databases. For now, we know that the UK will roll out the new General Protection Data Regulation in 2018, as the rest of Europe will, so at least there will be a level of consistency from this perspective. A Universities UK report recommends as a starting point that a mechanism to facilitate effective research collaboration could be to create more agency-level bilateral agreements, built flexibly and with funding attached. The role and position of UK or international companies based in the UK post-Brexit will play a key part in understanding the implications for pharmacoepidemiology and real-world data research.
Maintaining momentum
Cross-sector collaboration is fundamental in order to further capitalise on the momentum of big data and address well-documented challenges of expertise, public trust, complexity and expectations versus capacity. Programmes and incentives to enable partnerships across academia and industry, as well as healthcare systems, regulatory authorities, research funders and patient groups, have to be explored. Big data may indeed be the catalyst to enable academic-industry collaborations and the UK is endeavouring to promote such synergies at a global level. But the country needs to make sure that it can maintain this momentum post-Brexit.
Acknowledgement: I thank my colleague Dr Stephen Mulgrew for his contribution to an earlier version of this article.