Whether you call it—"population health," "digital health" or "value-based care"—the way to fix U.S. healthcare is to build a system that focuses on keeping people healthy rather than treating them only after they get sick.
This is harder than it sounds. First, we have to understand the health status of a population and determine who in that population is on a trajectory towards becoming sick. Then we must engage with those people to help them avoid getting sick or—if that’s not possible—get them to the right level of care at the right time.
Each step requires data: data to understand the population, data to see what interventions work and when they are best applied, and data to measure and continuously optimize the system. Seems so simple. But unfortunately, limited access to this data makes it anything but.
I have lost track of how many times I have been in board meetings and heard some version of the following conversation:
CEO: “Our product is amazing. Our client [often a health insurance company] loves it. It is helping their members; it is improving outcomes, even saving lives and it’s saving them lots of money.”
Board: “Great! Implementations must be cranking along. Can we keep up with demand? Should we start interviewing bankers for an IPO?”
CEO: “Hold on, there’s one issue.”
Board: “What’s that?”
CEO: “We haven’t been able to get the necessary data from our client to begin a large-scale rollout.”
Board: “What? I thought you said this saves money and lives. They have the data right? Why can’t they just get it over to us and we can get started?”
CEO: "Yes, they definitely have all the data. It’s just that they aren’t able to get it pulled together in a way that we, or anyone else for that matter can use.”
Board: “That’s insane! Why can’t they just throw some resources at this? You said it saves money and lives, right?”
CEO: “It’s complicated!”
What usually ensues is a painful reminder that most payers have antiquated systems, find it challenging to recruit tech talent, and are dealing with the complexity caused by years of duct-taping together systems and data sources. These are legitimate obstacles, but having heard about them so often, from so many different sources, and at such great expense to the greater good, I have come to believe that the healthcare data problem may be the biggest single roadblock to healthcare innovation today.
In the short term, necessity has driven hard-to-believe "work-arounds," including one case, in which one our companies (which, at that time had approximately $3 million in revenue) went out to Best Buy and bought its client (a multi-billion-dollar health plan) a new computer with a 1 terrabyte removable hard drive that could be FedExed back and forth. This worked, but it was slow, the data was stale and obviously it wasn’t a long-term scalable solution.
Health plans are sitting on some of the most valuable data in the world. Unlocking that data could save hundreds of billions of dollars and improve the lives of tens of millions. COVID has brought these deficiencies and the opportunities into even starker contrast.
With the near shutdown of the broader healthcare system, many health plans realized just how data-blind they were. Many scrambled to identify which of their members were most vulnerable. For those who had it available, simple data, like age, pre-existing conditions and geography, directed outreach programs, which had remarkable results.
As members avoided hospitals, plans who had clarity on their members’ individual situations were able to ensure that critical activities were not skipped. As elective procedures, many of them unnecessary, were delayed, some plans have wisely used data-driven insights to engage with members about better alternatives.
Finally, as the lasting behavioral effects of the pandemic emerge, plans that have invested in data-driven surveillance have been proactively helping members, especially in areas like behavioral health.
As investors, we focus on backing companies that capture, organize and unlock invaluable healthcare data. Some are pure plays like Abacus Insights, which focuses on the expansive data held by payers or Redox, which works in electronic medical record interoperability. Others, like ClosedLoop, use artificial intelligence and machine learning to sort and analyze healthcare data, allowing plans and providers an ability to do the analysis themselves.
It is my hope—dare I say expectation?—that at some board meeting in the not too distant future, I will hear the new launch enthusiasm and ask “how’d we figure out the data issues so painlessly?” and the response will be “Oh, that was easy, the client was using Abacus and Redox so we were pre-integrated. Up and running in a week.”
That’s when I’ll know we are focused on the really important things: improving outcomes, reducing costs and fixing U.S. healthcare.
Liam Donohue is co-founder and managing partner of .406 Ventures.