Artificial intelligence (AI) is often described as a game-changer in health care. From reducing diagnostic errors to speeding up clinical trials, its potential applications are wide-ranging.
One promising area is the use of AI to reduce time-consuming administrative tasks. While 64% of patients wish their health care providers spent more time understanding them, nearly half of a typical physician's day is spent on data entry and other administrative work.
By easing this administrative burden, AI could free up time for providers to focus more on their patients—helping to build trust and improve the quality of care. Additionally, physicians are generally more comfortable using AI for administrative support than for clinical decisions, like diagnosis or treatment, indicating an openness to use new technology in this way.
Using AI to accelerate and simplify medical necessity review
Medical necessity review is an area that shows promise when it comes to using AI to automate manual processes. Traditional approaches to medical necessity review have been criticized by providers and their staff as fragmented, time-consuming, and highly manual, which increases their administrative burden. As a result, some patients may have experienced delays or gaps in care, adding to their stress.
AI offers opportunities to address these challenges by reducing the time spent creating and processing cases, including automating submission of relevant supporting documentation, approving cases in real-time, and expediting payor-side review processes.
“As much as we can use AI to remove manual administrative processes, we should really try to do that,” said Sid Govindan, M.D., Senior Director, Healthcare Technology & Innovation, EviCore by Evernorth. “That’s going to open up more and more opportunities for humans to engage in the parts of health care where they can deliver the most value.”
A balanced approach to AI in health care
As exciting as the opportunities are for AI to dramatically improve medical necessity review, it’s important to take a balanced approach in its application. That means defining where AI won’t be used and where it’s critical to have a bonified clinical review. This approach ensures that there’s clinician expertise and understanding integrated with machine learning.
“One of the things we’ve learned over the years is that good health care AI-driven systems are very hard, if not impossible, to create without human clinical expertise,” said Dr. Govindan. “When you think about how AI gets trained and learns about clinical medicine, there’s a lot more back and forth that these systems are having with clinical experts than may be readily apparent. Training and getting health care AI systems to perform well requires close collaboration with clinical experts.”
In addition to ensuring that AI systems work effectively, human involvement is also critical in the medical necessity review process itself to ensure patients are getting the care they need. For example, EviCore intelliPath®, our electronic medical necessity review solution which leverages electronic health record (EHR) connectivity and automation, only uses AI to facilitate faster approvals of cases. For any other decision, there’s a highly skilled clinician performing an individualized clinical review. According to Evernorth’s Health Care in Focus report, 9 in 10 consumers agree that in-person interaction with providers is vital, further emphasizing the importance of combining AI with compassionate care practices.
“AI is an incredible tool to quickly and seamlessly approve cases when we have all the relevant information and there’s a high probability that the care is appropriate,” said Dr. Govindan. “However, we absolutely want a clinician review in cases when the system isn’t sure the requested test or procedure aligns with evidence-based medical guidelines. That’s where highly qualified clinicians can offer the most value in terms of making sure patients get the care that best aligns with their needs.”
AI for a more human-centered health care future
Looking ahead, one of the key enablers of AI’s future role in medical necessity review is better connectivity. As the Fast Healthcare Interoperability Resources (FHIR) standard gains more widespread adoption, it will make it easier for different entities in the health care system to share data. This may create opportunities to engage with providers through EMRs (electronic medical records) even earlier in their interactions with patients to deliver real-time information based on the data in their systems. That information can help inform care decisions in the moment rather than waiting for processes like medical necessity review to occur.
Regardless of how AI is being applied to medical necessity review, there are common goals—reduce the administrative burden for providers, eliminate concerns from patients about potential delays or gaps in their care, and facilitate more time for the humans in health care to develop relationships and deliver value where their knowledge and complex problem-solving abilities can make the most difference.
