Our UK life sciences vision needs an end-to-end innovation pathway
This would enable research findings to be translated into improvements in patient care more quickly and safely. The UK can leverage key advantages: world-class research institutions; national agencies that can create a coherent research landscape (such as Genomics England, HDR-UK, NIHR, MHRA and NICE); excellent funding ecosystem (Wellcome Trust, MRC, etc); a thriving life sciences sector; and the position of the NHS as a single payer health system with the potential to capture an integrated picture of care across the life cycle for 67 million people.
If we get this wrong and focus solely on enabling better research, we will miss the economic opportunity promised by the UK Government’s Life Sciences Vision (July 2021). The UK is too small a market to be globally significant to pharmaceuticals, medical device and med-tech companies, so beyond world-class research, we need a broader value proposition. For this, I believe our opportunity is to support the end-to-end innovation pathway: beginning with leading research, faster technology appraisals and regulation, and building a robust safety net for post market surveillance; through to strong health economic models to underpin the expansion of an innovation; then digitising and monitoring guidelines and recommendations to drive best practice and remove unwarranted variation in care. We also owe it to the public to ensure that as we increase the volume and speed of bringing innovations to market, so we will also need a robust safety harness. This end-to-end innovation pathway will provide a leading global incubator to enable international expansion of industry innovations carrying the kudos of our globally-revered regulatory organisations.
I believe we should conceptualise and focus on delivering this end-to-end innovation pathway as the desired outcome, where we can assess gaps today and address them. There has already been much work with speeding up regulatory approval, including the Cancer Drug Fund, Innovative Medicines Fund, Innovative Licensing and Access Pathway. But we have other gaps to address beyond this, such as:
- nationally-accessible Real World Data environments to assess new technologies and reassess following contingent approvals pending more data;
- digitally-enabled surveillance of new technologies used in practice with digitally-enabled pharmacovigilance programmes that go beyond yellow card reporting;
- integrating future national panel testing for multiple pharmacogenes associated with adverse drug reactions back into clinical workflows at point of prescribing;
- executing NICE’s strategy for live guidance including the MCBK movement to computable guidance;
- audit of NICE guidance adoption in clinical practice, enabling Trusts and ICSs to focus on unwarranted variation in care; and
- an ability to monitor and measure uptake of technologies against outcomes such as will be needed with new national population health contracts, like Inclisiran.
In parallel to this pathway innovation, attention will need to be paid to reimbursement models, not least for medtech.
Will Secure Data Environments/Trusted Research Environments get us there?
Closing these gaps in our innovation pathway is going to require data platforms across the NHS and social care. To this end, it is pleasing to see the shift in NHS thinking away from the use case of Trusted Research Environments (TREs) to a broader concept of Secure Data Environments (SDEs), as this will be needed to support a breadth of use cases across this innovation pathway, as well as already envisioned use cases around planning, population health management, research and operations. This autumn, we are expecting the NHS Centre for Improving Data Collaboration (CIDC) to publish guidelines and specifications for SDEs to be followed later in the year with guidance on how these environments will be accredited.
Despite concerted efforts, significant barriers remain to realising the potential of data in the UK. Data is fragmented, there are major bureaucratic constraints and the public lacks confidence in data sharing initiatives. The repercussions from care.data and GPDPR have generated a reluctance for central storage of data initiatives and encouraged a move to federation, where there are hoped to be a small number of accredited SDEs in geographical locations, with national specialist TREs. Indeed, a large number of TREs have already been created at national and at regional levels, and rather than consolidation there appears to be proliferation of such environments.
Unfortunately, this means we now have a fragmented data landscape where data cannot be optimised for research and innovation across the NHS or life sciences sector. Leading to a number of challenges, such as:
- central agencies including CQC, MHRA, UKHSA and NICE require defined anonymous real-world data from right across the country for surveillance and regulatory needs;
- TREs are focused on leading academic centers so we do not have an equitable approach to research across the country that is representative of all cohorts;
- researchers cannot easily understand what data might be available that is relevant to their purposes, although the HDRUK Innovation Gateway and metadata models being harmonised will help with this;
- industry attracted to the UK for access to Real World Data environments will need to repeat their research many times over in each sub-national SDE; and
- data cannot easily be shared between TREs, which limits the scope for analysis of related datasets or requires a lot of repeat work in each TRE.
We can also ask ourselves: how much research is even conducted in a single locality? The trend is the opposite - to pool data across collaborative research networks.
The current landscape for TREs has resulted in multiple copies of data being pooled and as a result there are now calls for interoperable TREs and reproducible analysis pipelines. The linked Trusted and Connected Data and Analytics Research Environments (DARE UK) programme envisages that TREs will be harmonised and joined-up to support and scale research. Whilst these initiatives are important, they will not fully address the needs and close the gaps in the innovation pathway.
What strategies could we deploy?
We have seen a similar approach attempted with the establishment of Local Health and Care Record Exemplars (LHCREs), which developed shared health and care records for people in a particular region for better coordination of care. The vision was to integrate these across the country with common standards recognising issues around borders and people moving between regions or being seen for care our home LHCR region. However, this vision was never realised: LHCRs are not operating to common data standards or routinely sharing data, which demonstrates the difficulty of achieving integration in practice.
The innovative approach to a Federated Data Platform proposed by NHS England has the potential to show the direction for data management for healthcare, population health management and research. In that model, patient data continues to be owned by the data controller in a hospital or primary care. It is approved for collection and linking in other data platforms for specific uses, of which SDEs could be one use, but it always maintains its link to the original data. Transparency as to which data has been used for what purpose is available for patients and information governance. Finally, as the data becomes richer and cleaner for one purpose it benefits the whole of the health and life sciences. If the life sciences industry really wants to leverage the unique opportunity of NHS data, it needs to be in lock-step with the NHS.
In contrast, the initial definition of a TRE suffered from being “anything you wanted it to be”, so hasn’t yet driven standardisation. Palantir argued in our National Technical Architecture to Support TREs White Paper last December that the NHS needed to separate the concept of secure, curated, linked platforms for research data, which could be federated, but would exist within the NHS, from solutions used by researchers to connect to multiple sources of data and enable complex research programmes that would connect to these data sources. This two-layer model could enable national data platform scale, whilst researchers could be empowered and public trust maintained. We hope that the launch of SDEs will capture this vision and refine the definition of TREs as the solutions that sit on top of an ecosystem of SDEs, so that we have a coherent whole to the national strategy.
If the UK could create a federated, integrated data layer across the country, it could deliver on the vision for an end-to-end innovation pathway for life sciences, whilst still enabling a proliferation and then natural selection process for TRE application environments. It could bring greater coherence, clarity and control to the way health data is managed and shared across the UK and build trust and accountability with the public. Technology is now available to support such a framework in a way that wasn’t possible a few years ago. Building on the experience Palantir has had supporting large collaborative networks in Life Sciences to speed research and commercialisation, the above white paper explores this topic in more detail.
Disclosures: Justin is a member of the techUK Life Sciences Forum, on the vendor advisory panel supporting CIDC in shaping requirements for Secure Data Environments, a Visiting Professor in health informatics at UCL, and is a non-executive director of NICE.
If you would like to find out more about techUK's work in life sciences, email [email protected].