The New Epidemic - Big Data and America's AddictionTom Keefe, Director, Oxeon Partners

Big Data and America's AddictionTom, a Director at Oxeon, focuses on building leadership teams for high-growth Healthcare IT and Services clients, including athenaHealth, ABILITY Network, Crossover Health, Intermountain Healthcare, Pamplona Capital, and many others. Most recently, Tom has driven Oxeon Holdings internal growth, working to shape the company's go-to market efforts in the organizational design and development space, as well its portfolio expansion into the data science and machine learning community. He is also involved in internal operations and Oxeon's ongoing efforts to support portfolio investment's commercial activity.

The numbers are staggering. From 1999 to 2014, over half a million Americans died from overdoses of drugs, both legal and illegal, and more than half of the overdose deaths in 2014 involved some type of opioid.1 The heroin crisis that preceded this current epidemic leveled communities across the country, and even today, 125 Americans will die from heroin-related deaths.2 As we go about our lives in the era of 24-hour news cycles, we all too often find ourselves attempting to comprehend crises like this one, but unable to do so, we shake our heads and move on. Not in this case.

These numbers aren't just statistics — these numbers are personal. While writing this article, I've had two teammates experience tremendous loss at the hands of this country's prescription drug problem, and it hurts. The article that follows benefited from tremendous input, support and collaboration with Oxeon teammates, and outstanding executives and mentors across our network. What's remarkable here is that many of these same people are in a position to help SOLVE the problem, and that potential for change is where I want to focus.

Late last year, the Obama administration enlisted the help of over 40 national provider groups to tackle the nation's growing prescription drug abuse problem, and the epidemic of heroin-related deaths that it has precipitated. The NBA, Google, and the New York Times lined up to donate ad space for PSAs, and retail pharmacies accelerated their community outreach and training efforts across the country. Dr. Oz has even taken time off from debating late night television personalities to support the cause.

Here's what we've learned from early efforts at addressing the epidemic: the startling data makes a clear case that the proliferation of prescription opiates has created a largely dependent class, a majority of whom will at some point — should they survive addiction and the specter of overdose — be thrust into the illicit drug trade in search of their next dose after being denied their prescription fix. We as a nation now know how this generation of opioid abuse has formed. The old adage "admitting you have a problem is the first step..." could not ring more true. With nearly 32% of all opioid prescriptions subsidized by employers (resulting in $8 billion in additional spend per year)3, we have the data and resources required to identify pre-existing patterns and symptoms of those already trending or actively engaged in a cycle of addiction.

Established organizations like Anthem and Cigna began their own formal outreach to patients this summer who fall into prescription patterns that indicate the potential for abuse. Other groups such as the Pharmacy Quality Alliance and Prime Therapeutics have done tremendous work examining the clinical implications of multi-prescription overdose victims and their interactions with their pharmacy benefits, as well as the performance of CMS' Overutilization Monitoring System in patient populations. Their work has already enabled managed care organizations to identify patients for case management and intervention. We've taken the first steps, but we're leaving too much on the table, with a gap in connectivity - combating this epidemic will require us to work smarter, and be coordinated across the healthcare ecosystem.

Employing the same methods Facebook uses to serve us ads, and Amazon uses to accurately predict our optimal laundry detergent ordering cycle, we can "capture" patient signals and patterns that exist before the first prescription is filled and sold off to the highest bidder, or worse, over-utilized in a fatal outcome. Instead of falling back on traditional means of examining usage, we need to double-down on future-state, data-driven approaches to saving lives.

Oxeon client Cogitativo puts these practices to work, combining a team of world-class data scientists and best-in-class machine learning production. Using descriptive, predictive, and prescriptive analytics, CEO Gary Velasquez and his team have identified thousands of patients seeking poor performing, out-of-network medical and mental health services that impact quality of care.4 Using data science and machine learning, we will better understand how changes in care coordination impact our ability to intervene and drive recovery, moving to proactively engage those at risk and get them out of harm's way.

We need to build on this approach, and encourage national payers and providers that are focused on utilization to invest more deeply in acting on available data that clearly indicates patterns of abuse. The next frontier requires us to proactively identify our neighbors who are at risk and get them out of harm's way, all while expanding our knowledge of what's working in the fight against the cycle of abuse. A critical first step will require a thorough examination of how to advance our efforts by leveraging data science and machine learning.

Working in this data-driven way to combat the opioid epidemic also presents us with opportunities to apply our learnings to other disease states. We could begin to identify patients with depression or other sources of fraud and abuse tied to prescription drugs. We could use this data to create engaged, disease-specific care management and community-based, tech-enabled solutions. Oxeon client and portfolio investment Quartet has worked on the disease state front and done just that, partnering with leading Blues plans like Highmark to identify patients with undiagnosed mental illness and addiction challenges, coordinating specialty treatment before acute symptoms emerge.

In March of this year, a STAT-Harvard poll found that 41% of Americans knew someone that had abused prescription painkillers in the last 5 years.5 Efforts are underway to mobilize public resources to combat this public health crisis. But the epidemic will not slow if legacy organizations and new ventures do not leverage existing data assets in new ways to drive treatment and identification to combat the crisis. We can begin to recover from one of America's bleakest public health moments and spare countless neighbors and friends from the pain of addiction and its grip on their future.

It is my hope that we embrace the opportunity to begin or continue to reach into their communities, both local and national, to solve this crisis. All of the data and science cannot replace the human element of support for those who most require our empathy and compassion.

Get help for and seek resources to support those battling with substance abuse problems: 1-800-662-HELP.

1 CDC. Increases in Drug and Opioid Overdose Deaths — United States, 2000–2014. January 1, 2016 / 64(50);1378–82