Both: Some of our biggest challenges in the current healthcare setting include implementing data sharing capabilities, particularly in the hospital setting, without compromising on access to sensitive patient or sensitive hospital data. Current data, genomics or any other clinical data, is siloed and highly fragmented in different locations. We need to break these silos down and we are slowly beginning to do so. Precision medicine has tremendous potential and there is an enormous opportunity with aggregated data, but to fully understand how to ideally treat an individual patient, and therefore to implement and support precision medicine in the clinic, we need to look at large populations of people. This requires data aggregation and access to many small data silos while being able to deploy comprehensive compute capabilities that support rapidly evolving frameworks and applications across those data boundaries. With the current fragmented system in place, it is very hard to gain access to deep clinical, longitudinal patient records for treatment outcomes, such as genomics or test results. At this point in time this can only be achieved to a small degree within a health system, but even then most records currently collected are focusing on billing and they are not easily queried for precision medicine outcomes. Furthermore, the regulatory framework HIPAA (Health Insurance Portability and Accountability Act), that deals with patient privacy, creates a barrier for easily sharing results across hospital systems. While there are companies that have built a middle ground to support data aggregation for insight generation, trust and transparency is being challenged as they are monetizing their data providing limited value to the providers. Right now, we are in this odd situation, unless we figure out a chain of clear ownership, like others have in the areas of finance or the creative arts, including sharing the proceeds. There is a lack of trust and fundamental rules of data privacy, translating into inefficiencies and limitations of the current healthcare model, holding back precision medicine in terms of creating better patient outcomes. This also puts providers in a disadvantage in terms of not preserving their rights over the data, and not being able to appropriately value the data.

Both: Our common focus is to create a secure and encrypted cloud-based system with the necessary levels of data security and privacy that creates the trust and the foundation for a better future for medicine while never exposing individual level data.

Piers: I have been involved in building the National Cancer Institute’s Data Commons, a scenario where we used object storage from IBM. This is currently the only available and fully scalable completely secure object storage solution on the market. Fundamentally, the IBM cloud has been built with security from the ground up, including GDPR (General Data Protection Regulation) which comes into full effect May 25th of this year.

Jason: I am responsible for IBM’s US Systems healthcare and life sciences channels, strategic partnerships and solutions development and lead our efforts in the personalized medicine industry.

Regarding working with Health2047, we “share the same DNA” when it comes to building a data sharing model and supporting data privacy requirements. Sample size (millions), depth and diversity matter if we truly want to move from “One to Many” medicine to true personalized medicine. A key obstacle to achieving this goal is gaining patient’s and organizations trust that their data will be utilized in a secure and trusted fashion. Long before the current shifting personal data landscape (Facebook-Cambridge Analytica), we were working to build a data sharing environment with “trust” as a core function. Unfortunately, there has been a wild west mentality to data acquisition, use, and ownership in many industries, including the personalized medicine industry and if we are going to advance our industry we need to create a trusted and secure environment. We are in the middle of a massive macro shift that will be judged by history, organizations that embrace these values as foundational will thrive while others may cease to exist.

I am most excited how closely Health2047 and IBM goals and ethics are aligned, and how we want to address the data acquisition, aggregation, massive scale artificial intelligence and other emerging technologies to transform precision medicine. We are creating an architecture that will take advantage of the best of breed technologies available today and quickly adapt future technologies on existing data sets. If we do not start to fix this now, we will be nowhere sooner than we think. The upcoming conference will help us address this big elephant in the room. We have an opportunity to work with all the other stakeholders to truly make an impact, which in turn requires the alignment from the values and ethics perspective, and also specific technologies to be built from the ground-up.

Piers: The most useful patient data, such as deep longitudinal EHR data or genomic data can be localized at the individual patient level and is either identified or is easily re-identified from a HIPAA-compliant anonymized state. This presents clear issues of patient privacy. At the same time, having that data at scale has incredible potential and can result in a fundamental shift in medical outcome and care cost across a broad population. Right now I see the biggest challenges in relation to integrating that data at scale and building the underlying system that would allow people to share with confidence in a way it will be used appropriately and securely, and all this without exposing individual level data in an unencrypted fashion. It is a clinical ethics challenge that we need to appropriately address.

Piers: Yes, one of our goals is to achieve this basic FAIR data principle for private data where the data is made accessible and can be computed over in a secure manner. As of now, we do not really have that. As Jason will outline, IBM has already built pieces of the technology that supports this essential use case. Fundamentally, the way they are approaching security and creating storage and solutions for health data ultimately could allow hospitals to rethink about how they will want to deal with that data and how they build internal systems in support of it, including the design of centralized secure repositories. An appropriate system will allow us to move to secure data sharing with accompanying secure compute and data interrogation.

Jason: IBM will focus on pieces in support of the file system, storage and compute side. In my presentation at PMWC 2018 at the University of Michigan, I will touch upon where we need to go as a community to securely aggregate, compute and apply advanced artificial intelligence at scale and the technologies and capabilities required to achieve that.

Jason: Two short-term challenges facing the industry center around trust and the exponential deluge of data.

Even before Facebook-Cambridge Analytica, patients lacked sufficient trust to allow for the data acquisition at scale required for personalized medicine to fully realize the benefits of AI. We will face stronger headwinds as patients will rightly question what data is collected, how is it utilized, who owns the data and insights derived from that data. The industry will likely face increased oversight and regulation that will stifle the pace of innovation and societal benefits large-scale AI can provide. Now, more than ever our industry must embrace trust as our foundation and core strategy to ensure we can realize the full potential of precision medicine.

We are all familiar with the fact that it is becoming increasingly difficult for healthcare and life sciences organizations to manage, process, and store rapidly growing volumes of complex data that is coming from multiple data sources in a timely manner and control costs. Our challenges are tied to our customer’s challenges, and in the short and long term we are trying to come up with solutions that address their primary concerns; for example: how do they cope with performance and cost issues when dealing with petabyte/exabyte scale data? How do they manage the expected volume of genomic data that will be generated now that CMS has approved testing for metastatic stage 3 and 4 cancers? How do they turn massive data sets into medically actionable insights? And more…

One of our primary missions is to help healthcare and life sciences IT Leaders become trusted service providers to clinicians and researchers by mastering hybrid and multi-cloud service models for digital transformation in today’s data-driven era. This creates many challenges as the associated requirements for flexibility, scalability, and advanced data management disciplines are beyond the current capacity of many organizations. We believe that leveraging the cloud is critical to tackling the big data challenge.

The holistic approach we suggest is based on flexible deployment options which enable organizations to place data where it makes the most sense for them to run their applications and services — today and as business requirements evolve. The ability to manage unpredictable data growth while also balancing storage costs, location, performance and compliance demands across data sets and applications provides tremendous benefits.

Put differently; the precision medicine principal investigators imagination has been limited by many constraints that we have eliminated. Our collaboration with Health2047 is an example of how we are a tool that the Precision Medicine industry can utilize to imagine bigger and broader and most importantly make this vision a reality.

Jason: First of all, I want to thank Tal Behar and the rest of the PMWC team for creating the world’s leading forum bringing together the brightest minds focused on advancing precision medicine. The PMWC has created an environment that fosters creative collisions with academia, payer, provider, pharma, high tech, venture capital and dreamers like myself. It was only a year ago when I attended my first PMWC and quite honestly, I was very intimated as the data guy in the same room with the world’s brightest Ph.D. and MDs. They pulled me into the community and helped me realize we need people from different backgrounds and expertise to achieve the ultimate benefits of personalized medicine within our lifetime.

We are at the beginning of the AI medical revolution, and we must bring in cross disciplined experts that will accelerate the pace of innovation and ultimately ensure we see the fruits of our labors. Millions of lives depend upon us accelerating the pace of innovation and change and history will judge the ethical, technological and trusted foundation we lay, as well as how quickly we safely accelerated this medical revolution.