Big Patient Data and the future of Real World Evidence: Learnings from BiotechX 2024

Big Patient Data and the future of Real World Evidence: Learnings from BiotechX 2024The Future of Oncology runs on Patient-scale Infrastructure

The Big Data in Oncology session started with an excellent presentation which made it clear how Top Pharma companies recognise the need for the industry to become more patient-centric.

How can this be achieved? One suggestion was a new perspective on "data ownership", with shared ownership between different stakeholders.

We found this reasonable, in that, while direct biological observations from a patient can hardly be claimed as a third party's property, downstream AI/ML applications do depend on private IP to generate beneficial insights.

Of course, "shared data ownership" is easier to discuss for inputs; what happens to the "outputs", those original algorithms, models, predictive tools whose development and improvement depends on patients feeding their data?

It was suggested the industry needs to enable patients and innovators to come to the same table with a more transparent, fairer agreement, a clear understanding on what happens with data and what benefits can be achieved from patients.

Phillip Pauli, Operations Lead, Takeda

Personally, I thought Phillip's perspective was refreshing and encouraging for the Healthcare and Life Sciences industry as a whole: after the collapse of 23-and-meme, it's great to see a Top Pharma learning and looking towards a patient-centric future.

From Axia's perspective, discussions about Big Data in Oncology gave the spotlight to two key requirements for that patient-centric future:

1) Next-generation data governance solutions, built to accommodate a complex, diverse set of stakeholders, especially new stakeholders which are closer to the patient;

2) Patient-centric infrastructure, which can enable patients to be central in R&D while escaping the opacity of third party, one-off database snapshots.

Digital Healthcare: When Country-scale means Citizen-scale

Our afternoon panel, chaired by Phillip, saw myself joined by Jacob Reimers ( Roche ) and Mieke Van Hemelrijck ( King's College London ). Our chat ended up focusing even more on infrastructure, especially on:

  1. the need to limit arbitrary, random expenditure in unnecessary research infrastructure, especially in public sector, possibly leveraging AI for benchmarking and optimisation of cloud and hybrid deployments;
  2. the challenges public sector organisations are going to face in the near future when it comes to unloading population-scale omics datasets - with the unaddressed challenge of managing exabytes of sensitive data from initiatives like the UK Newborn Sequencing concept;
  3. the future of data governance and the role patients should play, especially when researchers expect frictionless access to curated, high-quality, reliable and validated data from private citizens.

It was clear that public sector organisations still have a chance to learn from the past to ideate sustainable, effective, data infrastructure to manage citizens' private data. Whether the end solution should be built and delivered by a public sector entity, that's a question for another time.

So many skipped lunch to attend our RWE session. Kudos to all speakers.

 

Digital Health 2030: the journey from Real World Data to Real World Evidence

On Day 2, yours truly chaired BiotechX Europe's Real-World Evidence session.

For this session, we assembled a diverse, deeply experienced group of speakers, giving us an essential multi-facteted perspective on what RWE means for the industry.

A key aspect was the gap between the gathering of Real World Data, #RWD, and the creation of Real World Evidence, #RWE.

Abhishek Choudhary ( Bayer ) put things in perspective by introducing the audience to the complexity involved in building scalable data infrastructures to generate reliable evidence. The challenge for the future is to minimise costs and maximise efficiency in delivering infrastructure enabled for Machine Learning, with the aim of reducing the time to generate evidence from six months to just one week.

José-Tomás (JT) Prieto ( Apheris ) discussed the concept of federated data networks, where data stays local while computation moves to it. This approach solves challenges around data privacy and compliance, allowing pharmaceutical companies to generate insights without moving sensitive data. In an excellent case of "show, don't tell", JT was joined by Joel Federer-Gsponer ( Roche ) to discuss their recent collaboration.

Johann de Jong ( UCB ) highlighted the transformative role of big data and advanced analytics in RWE for areas like Crohn's disease and Alzheimer. One of the areas of interest here is to make Electronic Health Record data more widely integrated in RWE generation, to enable even more personalized approaches to drug discovery.

Digital Health indeed depends on data and infrastructure, and there's more to that: a third dimension, literally.

Gregoire Cosendai, Ph.D. ( Ypsomed AG ) brought us into a more material plane by discussing connected devices, providing real-time data on patient outcomes. Gregoire discussed Ypsomed’s innovative approach with autoinjectors that collect precise data during injections, offering valuable insights for RWE generation and clinical trial optimization.

For our closing talk, Feifei Wei ( Roche ) summarised almost a decade of pioneering work on #digitalhealth applications, generously sharing her learnings on what works and what doesn't when it comes to mobile app design and patient #adherence. Not all patients engage with an app in the same way, and Feifei's insights gave the audience a data-driven perspective on the impact of e.g. app reminders and #gamification.

 

Real World Evidence in 2030

Our speakers covered diverse, essential aspects of #RWE like

  • AI infrastructure
  • federation and privacy
  • multimodal data
  • connected devices
  • patient engagement

Regulatory bodies like the FDA are beginning to embrace RWE, creating more opportunities for quicker drug approvals. However, challenges like data privacy, adherence, and the cost of data generation remain significant hurdles to broader adoption.

Looking ahead, the future of #RWE generation lies in unlocking deeper integration of multimodal data, federated learning, and patient-centric, privacy-first data sources.

As these technologies evolve, the ability to harness real-world data will determine the future of clinical innovation.