Decoding the Data Dilemma

I recently had an amazing experience at the Aspen Ideas: Health Festival, where I spoke on a panel about turning patients into advocates. Not surprisingly, it didn’t take long for the subject of patient registries to come up, and I realized something: each of my fellow panelists was struggling to answer the same question that’s been nagging – and inspiring – me for years: how do we go from data to cures?

I arrived in Aspen with a mission to listen and learn. Specifically, I wanted answers to my burning data questions: Where is all of our health-related data going? Can we ensure its security? How can we share this data between organizations/researchers? If A.I. is our secret weapon in data analysis, what’s taking so long?

At the Multiple Myeloma Research Foundation, we go deep on data from a multiple myeloma perspective, wondering if everyone else in healthcare was just as excited and frustrated by data as we are. It’s a question we’ve also been exploring at the Harvard Business School Kraft Precision Medicine Accelerator. But at Aspen Ideas: Health, it was blatantly clear that our data-obsession isn’t unique. Every data panel was packed, and what’s more, it turns out we’re all facing similar challenges.

Believe it or not, this is actually encouraging. If we’re all focused on solving the same core data dilemmas, we can make data work for us to cure cancer and other diseases. Here are just a few of the points I found most thought-provoking from my colleagues at Aspen:

  1. Quality + Quantity = Better insights. This is a point Padmanabhan Anandan, CEO of Wadhwani Institute for Artificial Intelligence, really emphasized when talking about the relationship between data and machine learning. To help make machines more efficient and intelligent, we need to give them A LOT of high-quality, accurate and complete datasets to analyze. Then, they’ll be able to find patterns and breakthrough insights.
  2. Standardize and Share! As it stands, the data that researchers do have access to exist in disparate, non-standardized databases and electronic health records. This makes it hard to share datasets because they can’t be properly uploaded to (and analyzed by) different systems. We need to create common standards & practices across genomic and immune approaches.
  3. Prioritize The Right Use Cases. There are so many different ways to apply data & analytics. More often than you might think, patients, researchers and pharma companies aren’t aligned on what use cases are most critical. By connecting all of these stakeholders, we can consolidate resources and attentions to tackle high-priority use cases together.
  4. Finally…we need leaders. Right now, everyone is operating with their own organization’s priorities in mind. To figure out what the most critical use-cases are across the ecosystem, we need a trusted third party to take charge.

I’ll leave you with one more thought: Google Research’s Lizzie Dorfman was spot on when she said “Our results on data are incremental, but it’s building trust to allow us to become bolder and more innovative.” Many of us truly believe that data is the path to cure cancer and other diseases. Now, it just a question of…how soon can we do this?