Where is the Energy Behind the Health Data Graph?


There's big buzz around consumer behavior driving outcomes. But clinical data doesn't yet align with behavioral data.

Chris Schrader

8 min read

Editor’s Note: Take our anonymous reader poll at the end of this article and let us know where you stand.

Late last year, I had a disagreement with an executive working at a healthcare “incumbent”. We were discussing how Amazon’s entry into healthcare might impact the industry and the health of consumers. My friend thought Amazon was really far away from having an impact, lacking the claims and clinical data to manage someone to better health outcomes. I shot back that if I had all the data Amazon had today – what people buy and when, what they like and dislike, and their engagement preferences – I could manage a person’s health better than he could – even if he had immaculate, integrated claims and clinical data. 

Therein lies the debate – are our efforts in the right places when creating usable consumer data in healthcare? Where should we do more?

One Big Step Forward? Consumer Access.

Most action to date in mobilizing against the health data graph continues to side with my friend’s opinion about getting more accessible, better clinical and claims data. Earlier this year, The Centers for Medicare & Medicaid Services announced proposed rules to dramatically increase consumers’ access to their own data. The rule effectively proposed healthcare organizations provide health data to consumers in a machine-readable format – or face stiff penalties. The rule would also allow patients to electronically access all their electronic health record data – for free. Make no mistake, this is a huge, incredibly valuable step forward that’s not to be underestimated. Implementation of this standard will:

1)    Enable better care coordination and transitions in care for people who change care settings, providers, or health plans.

2)    Help drive increased efficiency and efficacy as providers treat illnesses.

3)    Allow consumers to push data to a third party that may use that data in a machine-readable format for a host of different reasons, including health management.

But Is This Data Really “Consumer Data”? 

There’s widespread recognition and understanding that behaviors drive most health outcomes. The idea behind the health data graph is that right place, right time intervention can drive behavior change and that other data sources beyond clinical and claims data are required to understand the consumer well enough to drive behavior change. Clinical and claims data alone don’t drive behavior change, but expose clinical solutions. Consumer data expression is a far larger data set, much more important in providing a longitudinal patient view in areas like engagement preferences. For example, if patients are more likely to respond to a text outreach than an email.

We’ve seen a lot of energy around the health data graph, but this energy has been spread unevenly…and NOT necessarily where it’s potentially the most impactful. The activity on the clinical and claims data side has not been matched by equal excitement, dialogue, and on the behavioral and motivational data side.

A Holistic, Intimate Patient View

Despite the fact that health’s data graph has yet to hit peak potential, there’s been one bright spot: social determinants of health. Here, health plans, providers, and even retail pharmacies are now incorporating data like a consumer’s income, education, zip code, living situation, and financial situation into the identification of high-risk individuals. Five percent of high cost, high-need individuals – people who aren’t just medically complicated, but also socially complicated – drive half of all healthcare costs, explained Dr. Karen DeSalvo on the Oliver Wyman Health Podcast. “There’s exciting work happening as systems dig into their high-cost, high-need population. Whether doing quantitative or qualitative data assessment, they’re learning it matters if somebody goes to bed hungry,” she said.

What’s Holding Healthcare Back

In other areas – like aggregating behavioral and motivational data from the “rest of someone’s life” – we’re still falling short of market potential. Why might that be, and what’s preventing us from turning the corner? Here are four reasons we haven’t made more progress (just yet):

1)      Focusing on what we have. The clinical and claims data challenges are front and center – and provide real value. We know what claims data and clinical data looks like. We know how to put it into practice. Much work is being done to optimize the use of that data. But…maybe we’re just prioritizing. And in doing so, we’re not yet focusing on other big impact areas.

2)      The collaboration stalemate. Most uses of behavioral data require collaboration with multiple parties. The thing is, most of the parties using and monetizing consumer behavioral data don’t have this data on hand.

3)      Privacy and ownership concerns. Although it’s unclear who will one day own a consumer data, it’s likely there will be some level of permissioning from the consumer to use their data. Although the complexity this creates will ultimately lead to the rise of a permissioning “mini-industry,” it also stifles near-term action.

4)      The chicken or the egg. Although we recognize value, we’re not quite sure what kind of data drives the most impact. But we need to start acting on data. We need use cases, and we need to learn from the data created through interactions.

The disagreement I had with my friend last year is one you’re probably in the midst of pondering yourself. As an industry, we must work together to improve consumer engagement – on consumers’ terms – in ways that resonate with them and align with their lifestyles. That’s something I can’t argue.

Oliver Wyman Health Asks Our Readers: Where Do You Stand?

  • Chris Schrader