The Washington fight over the future of Obamacare will have enormous repercussions for our health care system, which now accounts for nearly 18 percent of the U.S. economy. But that impact pales before the transformation already underway due to the rise of artificial intelligence (AI) and machine learning.
AI and machine learning are forcing dramatic business model change for all the stakeholders in the health care system.
What does AI (and machine learning) mean in the health care context? Simply put, it’s the ability to aggregate newly available and vast amounts of disparate data from electronic health records, consumer media and purchasing trends, smart devices, the social sphere, and other sources to make predictions about patient outcomes. What is the best way to treat a specific patient given her health and sociological context? What is a fair price for a new drug or device given its impact on health outcomes? And how can long-term health challenges such as cancer, obesity, heart disease, and other conditions be managed? Answering these questions is the holy grail of medicine — the path toward an entirely new system that predicts disease and delivers personalized health and wellness services to entire populations. And this change is far more important for patients and society alike than the debate now taking place in Washington.
Three key trends have converged to make AI-driven health care a necessity. First, the Affordable Care Act has fueled the industry’s focus on value over volume. We simply cannot afford to treat patients under the old model of “provide as much care as possible, regardless of the costs.” We need to get the right intervention to the right patient at the right time, while avoiding as much as possible doing things that add no value and cost a lot of money.
Second, the advent of electronic health records and the associated explosion in data has made it ever more important to learn what the data is telling us, and respond with new models of care.
And finally, the realization that treating “the whole patient” — not just isolated conditions, but attempting to improve the overall welfare of patients who often suffer from multiple health challenges — is the new definition of success, which means predictive insights are paramount.
In the emerging reimbursement environment, pharmaceutical and medical device companies will have to offer solutions instead of widgets — they will be paid based on the value their products deliver, value that is tied to a complete solution for a particular patient’s condition that improves outcomes and promotes both health and wellness. Providers, whether large hospital systems or individual doctors, will need to optimize patient pathways and drive to best outcomes; otherwise they will not get paid. And payers will make coverage decisions based on outcomes rather than clinician and patient requests.
The winners in this world will be organizations that make AI and machine learning a core aspect of what they do. It’s not enough to build analytics teams and rely on consultants for the rest. Without a core AI capability — or a long-term partnership that makes AI possible — the three key trends cannot be effectively addressed. Those who succeed in this new world will also do one other thing: They will see AI and machine learning not as a new tool, but as a whole new way of thinking about their business model.
What’s happening in healthcare is not unique — in fact, we’ve seen this transformation before, and it ends badly for those who do not adapt quickly. The automotive industry spent decades resisting government regulations about emissions standards and other issues, and it missed the start of AI and machine learning and with it, the rapidly approaching future of driverless cars and trucks. The health care industry stands at the same precipice. Let’s hope it makes wiser decisions and doesn’t end up on the scrap heap of obsolescence like Detroit.
Alex Turkeltaub is the cofounder and CEO of Roam Analytics, which created the first machine learning platform designed for health care.