healthcaretechoutlook

Artificial Intelligence...Applied Intelligently. Finally. An Answer to Rising Risk for Payers

By Lonny Reisman, MD, Founder & Chief Executive Officer, HealthReveal

Lonny Reisman, MD, Founder & Chief Executive Officer, HealthReveal

Chronic diseases and conditions such as heart disease, stroke, cancer, type 2 diabetes, obesity, and arthritis are among the most common, costly, and preventable of all health problems, according to the CDC. In fact, about half of all adults in the US—117 million people—had one or more chronic health conditions, while one in four adults had two or more such conditions. Further, the CDC notes, seven of the top 10 causes of death in 2014 were related to chronic disease. Two of these—heart disease and cancer—together accounted for nearly 46 percent of all US deaths that year.

"We can create a “digital replica” of each patient, using the real-time"

Health plans, perhaps more than any other stakeholders in the US healthcare system, are acutely aware of the financial burden of chronic disease. The CDC reports that 86 percent of the nation’s $2.7 trillion annual health care expenditures are attributed to people with chronic and mental health conditions. According to a report from the Partnership for Solutions, a national program whose goal is to improve care and quality of life for Americans with chronic health conditions, “People with chronic conditions, particularly those with multiple chronic conditions, are the heaviest users of health care services. Higher utilization appears in all major service categories: hospitalizations, office visits, home health care, and prescription drugs. For example, individuals with multiple chronic conditions account for two-thirds of all prescriptions filled, and those with five or more chronic conditions have an average of almost 15 physician visits and fill over 50 prescriptions in a year.”

In addition to the rising incidence of chronic disease, the number of health plan members over the age of 65 is growing, statistically placing them at a higher risk for chronic illness. As it is often these aging populations who develop chronic diseases or require more frequent care, this only adds to the financial burden that health plans are shouldering. It is therefore more important than ever for health plans to develop a strategy for maintaining the health of their members, improving patient outcomes while reducing costs.

Managing Member Health

Traditionally, payers have focused on a few key efforts to maintain member health including Utilization management, Wellness and care management, and Complex case management.

The challenge inherent in these methods of care management is not that they are ineffective, but that they are not personalized to address the complex clinical needs unique to the individual patient. On one end of the spectrum, such care management initiatives focus on population health, emphasizing efforts that are good for an entire population in general, such as the previously mentioned annual eye exams for members with diabetes. These population health efforts are of course necessary and prudent, but they’re not personal. It’s only once the preventive measures have failed for some of the members—and healthcare costs for those members are already rising substantially—that health plans typically shift their efforts from the general population to the specific member, focusing on how to control costs for the chronically ill patient who, for example, has been hospitalized three times in the last six months.

Care management naturally moves from management of the member population as a whole to management of individuals who require the most care. But what if there was a way for health plans to focus their efforts on individual patients from the start, so that fewer members ended up needing costly medical and care management services?

AI and Machine Learning for Individual Health

There are two methods in which health plans can begin to focus more on needs of individual members from the start to prevent the need for costlier downstream care: the application of guideline-directed medical therapy (GDMT) and the addition of real-world, real-time patient data.

GDMT is a series of medical protocols that follow generally accepted standards of care based on the latest research in the industry—for example, placing certain patients with diabetes and atherosclerotic cardiovascular disease on intensified antihyperglycemic therapy as an accepted standard of care to reduce the risk of cardiovascular death. Surprisingly, however, many patients do not receive these evidence-based therapies. In fact, one study in the New England Journal of Medicine (NEJM) found that almost half—46.3 percent—of chronically ill participants “did not receive recommended care.”

Another study published in the March 2017 issue of the Journal of the American Medical Association (JAMA) found that, of nearly 95,000 patients with a known history of atrial fibrillation (AF) who had acute ischemic stroke, an astonishing 83 percent “were not receiving therapeutic anticoagulation.” This was not a case of patient non-adherence, but rather of physicians either never having prescribed the medication in the first place or not having adjusted the dosage for patients receiving anticoagulants in accordance with lab results. Consider the potential value (in terms of savings to payers and patient outcomes) of avoiding these strokes, had the patients been receiving blood thinners as guidelines recommend.

The second piece of the puzzle in personalized care management is leveraging the massive amounts of patient data available today. The industry now has the ability to pull a vast array of real-time patient data from so many sources—not just from claims data and electronic health records, but from remote monitoring technology (devices that are wearable or implantable, or can be used in the home), such as blood pressure readings for heart failure patients and blood sugar tests for diabetics. It is in combining this personalized, up-to-date information with the latest advances in GDMT that physicians can intervene before acute health care events occur, resulting in substantial long-term savings for those covering the cost of care.

A Collaborative Effort

We can create a “digital replica” of each patient, using the real-time data discussed above. Through the application of AI and machine learning, that data can then be compared against the latest accepted evidence-based guidelines, and a personalized, real-time recommendation can be delivered instantaneously to the physician, patient or caregiver, alerting them to the likelihood of acute events and offering actionable interventions to help prevent those events.

We already have everything we need at our fingertips: enough information and computing power to provide an analytical bridge between medical guidelines and high-risk patients with chronic conditions. The result is prescriptive diagnostic and therapeutic solutions for physicians to incorporate into their treatment plans. The impacts include preempted strokes, heart attacks, blindness, kidney disease and amputations, and lives saved. By implementing such technology to hospitals and providers, I’m convinced that we as an industry can work together to truly change the way healthcare is delivered.

Healthcare Startup Special