Artificial intelligence is incredibly buzzy in health care right now, and for a good reason. Other industries are already experiencing AI-enabled radical transformations, like real-time fraud monitoring and detection in banking and finance and instantaneous image recognition across the web and social media. Health care now stands similarly positioned to capitalize on the transformative power of machine learning, and the massive potential is driving levels of investment expected to reach $6.6 billion by 2021.
Exactly what problem is such a massive capital infusion betting AI can solve in health care? The answer is simple.
Today’s approach to health care in America is, by design, reactive. We react to cancer, to heart attacks and strokes, to sudden acute harm and to the slow, downward spiral of chronic physical and mental health conditions to the tune of over $3 trillion a year. The Commonwealth Fund reported that just 4 percen of Medicare’s total population accounted for over $6,500 per capita in excess annual preventable healthcare spending — again, that’s preventable health care spending. If America is ever going to move beyond reacting to health issues after they’ve become problems, we need to enable more proactive, holistic approaches to solving for health.
Moving upstream
It’s this challenge — namely, solving for health by moving upstream of an emergency room visit, upstream of chronic disease, upstream of an unnecessary procedure, upstream of a “Code Blue” — that AI stands well-positioned to take on. AI offers the ability to find patients at a rising risk of acute and chronic health events with striking accuracy. Moreover, AI’s predictive capacity can provide weeks, months, even years of lead time to connect each rising risk patient with the individualized, targeted interventions most likely to redirect his or her health trajectory.
This isn’t science fiction. AI is already driving versions of effectively this same optimization function nearly everywhere we look. A recent New York Times article highlighted how Amazon’s massive vault of customer purchase data is being used for hyper-targeted local advertising. Instead of helping a baby formula company target consumers who recently purchased diapers or bottles, imagine using this power to deflect the health trajectory of an entire community positively. By pairing remarkably accurate risk projections with tailored, personalized interventions, AI shrinks health care’s ever-challenging “last mile” to perhaps the “last hundred yards,” and in doing so allows health care’s human providers to focus their human time on closing this final gap.
What’s still in the way?
Despite the revolutionary buzz — which some may argue is rapidly approaching Gartner’s classic “peak of inflated expectations” — AI in health care has one huge hurdle to clear before realizing its revolutionary potential. Because AI offers the ability to capitalize on previously inconceivable proactive, predictive and prescriptive insights, using AI-enabled solutions also by definition requires changing how end-users think, train, act, connect and engage. To borrow from General Colin Powell’s book “It Worked for Me,” realizing the power of AI in health care doesn’t just require changing hardware and software. It requires changing people’s “brainware” — and changing brainware is hard.
This is because, as humans, our habits are hardwired. Truly changing a habit takes both dedicated time and dedicated energy. Until now, the motivation to spend the sizable amounts of time, energy, and resources needed to shift America’s reactive health care paradigm to a proactive, health-centric approach has been leaden at best.
Today, that motivation has finally been kindled by the alignment of alarming population health needs (best evidenced by America’s falling life expectancy), the maturation of risk-based reimbursement models, and the advent of AI-enabled health solutions. With the social, fiscal, and technological means now available to drive real change, shifting our brainware represents the single-hardest remaining challenge to flipping America’s health care paradigm.
How do we change our brainware?
The only way to change brainware is to start with purpose. Outside of a “command and control” environment, interventions that fundamentally change someone’s personal or professional life are rarely successful unless they first connect to a simple, powerful purpose. For AI in health care, that purpose squares solidly with the health care system’s loftiest goals — to improve long-term health, decrease long-term cost, and enable clinicians to spend more human time engaging 1:1 with their human patients.
After solving for purpose, changing brainware requires bringing the ultimate end-users of any AI-enabled solution into the development process from its earliest stages. The input of the people who live at the pointy end of health care’s stick — be they doctors, nurses, pharmacists, care managers or patients themselves — is crucial for two reasons. First, it clearly improves design. Second and more importantly, deep end-user engagement is needed to both fully understand and then re-engineer all the human processes that go along with using an AI-enabled solution. Remember that (at least in health care), people drive most of the value in any value-stream map. Engaging them with purpose early and often is the only real way to align their brainware with novel AI-enabled approaches to solving for health.
Conclusions
With the social, fiscal and technological stars all finally aligned, it’s time for AI-enabled solutions in health care to live up to their buzz. If we make investments in not just technology but also the in people who can drive this paradigm-shifting change, AI’s long-term impact in health care will reverberate throughout the lives of clinicians and patients alike. We can solve for health, move interventions upstream, and redirect millions of people out of harm’s way and toward lower-cost, long-term well-being.
Chris DeRienzo is chief medical officer, Cardinal Analytx and author of Tiny Medicine: One Doctor’s Biggest Lessons from His Smallest Patients.
Image credit: Shutterstock.com
