By Dr. John Showalter, MD, MSIS, Chief Product Officer, Jvion.
COVID-19 catalyzed a rapid shift to telehealth that was years in the making. Reimbursement, once a barrier to adoption, was overcome when CMS announced that Medicare would cover telehealth to allow socially-distant care to continue. As a result, 69% of all patient encounters were done via telehealth in April, with that proportion even higher in areas with severe outbreaks of COVID-19.
Today, April feels like a lifetime ago, and telehealth accounts for only 21% of visits. But the consensus is clear: telehealth is here to stay.
A recent survey found almost 70% of providers were more motivated to continue using telehealth after the pandemic, citing better access to care (68%), more timely care (83%), improved patient health (60%), and improved financial health for their practices (57%). And now that CMS has permanently expanded telehealth coverage, any uncertainty over the long-term financial viability of telehealth can be put to rest.
Of course, not everything can be done via telehealth. Telehealth works great for chronic disease management, behavioral health, hospital/ED follow-ups and preventative care, but there are many procedures that can only be done in person. How then can providers determine which patients should be seen in person and who can be seen via telehealth?
This question is now more urgent than ever, as hospitals nationwide confront a surge in patients admitted with Covid-19. To manage capacity and keep patients safe, providers will want to see patients virtually whenever possible.
Deferred care is another concern for providers. Some 41% of US adults deferred medical care they needed this year to avoid the coronavirus. To prevent these patients from deteriorating and suffering worse health outcomes in the future, it’s critical that providers re-engage with these patients as soon as possible before it’s too late. Telehealth is often the safest way to do so.
Clinical Artificial Intelligence (AI) can be a powerful solution to triage patients for telehealth. By analyzing patient data, AI can assess which patients are at greatest risk for avoidable hospitalizations or other adverse outcomes, particularly if they’ve skipped care this year. AI can then help doctors determine the best course of action — whether that means a telehealth check-in to review the patient’s medication, or an in-person visit to take a blood test or a biopsy.
AI becomes even more powerful as a telehealth triage tool when it looks beyond the patient’s medical record to consider social determinants of health (SDOH). Socioeconomic or environmental factors can often obstruct patients’ access to the care they need. Leveraging publicly available data from the Census Bureau and other government agencies, AI can assess the likelihood that a patient is affected by these SDOH factors, based on their home address.
To understand how this works in practice, consider patients with limited access to transportation, or those who live in rural areas far from healthcare resources. These patients will have a harder time traveling to their appointments in person, and would be better served by telehealth when possible. Data from the US Department of Transportation can reveal which patients live far from public transportation or in areas with low rates of car ownership, which would suggest transportation could be a barrier to care.
Other patients may have difficulty taking time off from work to visit a doctor. These patients too would be better served by the convenience of a telehealth appointment. Here, data from the Bureau of Labor Statistics and other federal and state-level agencies can provide clues on the dominant industries and professions in any given area, and how they impact access to care.
On the other hand, telehealth may not be appropriate for patients with low digital fluency. Data from third party sources can identify these patients and help clinicians direct them to in-person appointments.
When this socioeconomic data is combined with clinical data stored in the electronic health record, that’s where AI can make a difference. Providers and health plans may have some idea of which patients are at risk based on their medical history, but rarely do they have visibility on how a patient’s lifestyle impacts their health risk. AI can connect the dots in the data to bring all these factors together for a comprehensive view of patient risk that would otherwise be invisible.
As a telehealth triage tool, AI will not only help patients at risk, but it will also support health systems in their financial recovery from the pandemic. According to the AHA, the next few years will be spent digging out of a $323 billion hole. In this time, telehealth will be a valuable source of revenue from patients that providers may not otherwise see, and AI can help pinpoint patients who would benefit from telehealth appointments.
Early interventions made via telehealth will also have the downstream impact of preventing adverse outcomes — particularly for patients who have deferred care — helping value-based care providers lower their costs. The Biden administration is expected to expand participation in value-based care models, and AI will enable health systems to thrive under the transition by delivering the insights to improve patient outcomes.
Patients stand to benefit just as much as providers. According to the CDC, nearly 75% of doctor visits are related to medication. Many of these appointments could easily be done via telehealth, saving patients time and making it easier to stay engaged with their care team. And as AI helps doctors intervene earlier on in patients’ risk trajectory, it will enable a shift to preventative care that could further reduce the number of appointments that need to be done in person.
Now that patients and providers alike have seen the value of telehealth, there’s little doubt that telehealth will continue to be an integral part of healthcare delivery beyond the pandemic. Clinical AI will help providers make the most of it.