In recent years, talk of what artificial intelligence could potentially bring to healthcare has come to dominate industry conversations — and those in the field haven’t been afraid to put money behind their interest.
According to recent research from Accenture, the AI healthcare market has expanded at a remarkable pace, reporting a compound annual growth rate of 40 percent. Analysts predict that the sector will top $6.6 billion by 2021 and that key clinical health AI applications could, when totaled, potentially facilitate $150 billion in annual savings for the American healthcare economy by 2026. This AI revolution is already well underway; one 2019 survey conducted by Optum found that 22 percent of surveyed health industry leaders are already in entering the final stages of their AI strategy implementation.
The health sector’s fast-track towards AI makes sense — after all, the technology stands to benefit healthcare in nearly every capacity, from administration to clinical care to facility operation. As one writer for Chilmark Research put the matter in a recent report, current artificial intelligence solutions offer a “rich array of vendor solutions in medical imaging, business operations, clinical decision support, research and drug development, patient-facing applications, and more.”
But AI’s greatest potential, arguably, is its ability to further another revolutionary force in the industry: Value-based care.
Value-based care offers a promising alternative to the troubled, fee-for-service system that dominates American healthcare today. Under this new approach, providers would be reimbursed based on their patients’ outcomes instead rather than for services rendered. This approach aligns provider incentives with patient wellness goals and — notably — aims to improve population health management, lower healthcare costs, and provide better care.
If deployed effectively, artificial intelligence could go a long way towards helping providers effectively deliver value-based care. Predictive analytics can help doctors target their interventions sooner and more effectively, allowing them to develop effective treatment plans well before the patient needs expensive hospital care.
“Generally speaking, the most powerful and effective types of AI are leveraged to power technologies that bring true clinical and financial ROI – such as digital therapeutics with AI-driven predictive analytics as well as a machine learning component,” Biofourmis CEO and founder Kuldeep Singh Rajput explained to reporters for Healthcare IT News. “Digital therapeutics powered by AI enable more informed clinical decision making and earlier interventions.”
Studies have found this to be true. In 2019, Aurora Advocate — a health system that maintains a number of value-based contracts — conducted a pilot study to gauge whether predictive analytics could prompt effective telephonic interventions for heart failure patients. The trial was remarkably successful; Aurora was able to reduce unnecessary care utilization by a remarkable 23 percent.
AI tools have also demonstrated their efficacy in oncology care studies. Last year, Cardinal Health and the prescriptive analytics firm Jvion launched another pilot study to assess how AI could benefit value-based care. In the study, three oncology treatment organizations — the Center for Cancer & Blood Disorders, Tennessee Oncology, and Northwest Medical Specialties — all incorporated one of Jvion’s AI tools into their daily workflow. Researchers then assessed how well the tools helped providers identify high-risk patients for several distinct adverse outcomes, including 30-day mortality, 6-month deterioration, 30-day avoidable admission, 30-day emergency department visit, and hospital readmission.
As with the Aurora Advocate study, the AI tool proved to be enormously useful. Researchers reported positive developments across all adverse outcomes studied and saw particular improvement in the mortality metric for palliative and hospice care. During the study period, providers noted a 225 percent uptick in hospice referrals and a 35 percent increase in palliative care referrals. The primary benefit of AI, researchers noted, was that it identifies patients who aren’t easily identifiable as high-risk and allows care providers to address and potentially prevent adverse outcomes well in advance.
“All this is still very early, but I think it definitely is a very interesting tool,” Dr. Ray Page, President & Director of Research for the Center for Cancer & Blood Disorders, reflected on the study for Obr Oncology. “It gives [doctors] pause to sit there and think, ‘Well, what’s going on with this patient? What can I do to make a difference?’”
Artificial intelligence holds tremendous potential to improve value-based care offerings — but the technology will need to clear more than a few hurdles before it can achieve its potential.
The first and perhaps most prohibitive problem is buy-in.
“There is no return on analytics or AI unless someone does something with your findings, with your models,” Aurora chief health information officer Tina Esposito declared during a speech for the HIMSS Media Machine Learning and AI Healthcare conference in 2019.
If providers don’t adopt or use the technology in their daily practice, it may as well not exist. If care organizations want to make the most of available AI tools, they will need to dedicate time and resources towards teaching providers not only what the technology does, but why they should use it. However, findings from the Cardinal Health/Jvion study suggest that convincing doctors to use the tool may not be too difficult, provided that they can see the data firsthand.
“In the beginning, we were all skeptical a little bit, until I started to manually validate [the data],” Northwest Medical Specialties director Amy Ellis told OBR Oncology. “Now [care team members] are very open to it, and they stand by me if I want to make a change.”
With all the potential that AI holds for healthcare, it’s worth hoping that the rest of the industry will be similarly receptive to the technology’s possibilities.