The 5th annual Abarca Forward Conference, hosted at the headquarters of pharmacy benefit manager (PBM) Abarca Health in San Juan, Puerto Rico, delved into important aspects of prior authorization (PA). This invitation-only event brought together payers, PBMs, specialty pharmacies, and clinicians to discuss how advances in technology and improved processes and communications can return PA to its roots in protecting patients while lowering healthcare costs.
Abarca invited five leading PA companies to showcase their innovations and insights into today’s PA and utilization management transformation. Throughout the session, the speakers returned to one idea. The challenge is not just pre-approval, but the fragmented manual experience surrounding it. Here’s what they had to say:
Akeel Williams, CoverMyMeds
Akeel Williams, senior vice president of CoverMyMeds, a drug access company owned by McKesson, says the next era of access can be defined by the answers to three questions. Which decisions should be made instantly, which decisions require human review, and how the system is experienced when it works well.
Current PA systems treat even routine decisions as if they are complex, Williams said, noting that physicians and their administrative staff spend hours each week understanding the basics of PA processes and systems. “This shows that the system doesn’t do a good job of distinguishing between what is obvious and what is uncertain, which causes wear and tear,” he says. “The impact manifests itself in delays in care, loss of capacity, loss of bandwidth, and unnecessary costs.”
Williams’ second question is which decisions deserve genuine human consideration. “Human judgment is the most expensive and valuable asset in medicine,” he said. “We should be spending our money where it changes outcomes, not where it arbitrates ambiguities created by the core design.”
Williams’ third question, about recognition and recognition when PA systems are working well, points to a shift from policy thinking to process thinking, stating that “clinicians need to experience clarity and fewer interruptions, while payers need to experience more consistency of care and an increased ability to predict and measure outcomes.”
Connecting these dots to build tomorrow’s PA systems requires shrinking the PA footprint without shrinking the purpose of the PA, Williams said. “The next generation of PA will be determined by how proportional, intelligent and intangible we can be while preserving affordability and quality,” he said. “The future of PA is not less control, but better judgment applied earlier.”
Andrew Melin, Surescripts
Next, Andrew Mellin, Chief Medical Information Officer at Surescripts, shared insights into the inefficiencies of current processes, such as manual processing of PAs, and discussed the potential of AI to improve these processes by making them more transparent and efficient.
Mr. Melin discussed the collaboration between Surescripts, a major payer, and the health system. They codified clinical policy in a three-part process. “Then, before a prescription is signed and delivered to the pharmacy, a series of real-time interactions between the EHR and PBM systems intelligently surface and apply the relevant clinical evidence already present in the EHR,” Melin said. He pointed out that this model results in decisions being made at the PBM within seconds. Providing the right information about the right medication at the right time allows patients to get to the pharmacy with fewer barriers and access medications faster.
Melin believes the lessons learned from this collaboration, particularly around AI, can be applied across the industry. “There will be a significant impact on PA, but careful vigilance is required at this time,” he said, noting that PBM leaders in the partnership with Surescripts were uncomfortable assuming risks related to AI. “It’s very important to be deterministic, at least for a while. That will change over time,” he added.
Melin also encouraged companies to simplify the questions they use in the PA process. “I have personally reviewed thousands of pre-authorization questions across many payers and many drugs, and was shocked by the vagueness of the PA questions,” he said. “When dealing with technology, ambiguity creates errors and problems.”
Similarly, Melin suggested clarifying policies regarding acceptable data for drug decisions, using the question, “Is this a chronic headache or an acute headache?” as an example. “Doctors will write on the prescription one pill for a headache and 10 pills for an acute headache, but some payers won’t accept that,” he said. “Look for sources of evidence other than strict policy.”
Kyle Kaiser, Alive Health
Kyle Kiser, CEO of Arrive Health, emphasized the importance of interoperability and AI in solving fundamental challenges in healthcare and advocated the need for modern networks that support shared decision-making. To do so, the industry needs to solve some fundamental challenges, he said. “How do we promote full participation and eligibility so that every patient who comes into our clinic can match the plan that is associated with them?” he said. “Many health plans are not fully engaged with eligibility data, and if you don’t solve that problem, you can’t solve the rest.”
According to data from Arriv Health, in up to 30% to 40% of cases, the EMR does not contain patient eligibility data. “We found that about a third of the time, we were creating a prior authorization that was unnecessary and unnecessary for that patient. The plan might need them, but the patient already met the criteria,” Kaiser said. “We want to maximize the power of interoperability and AI, so driving towards personalized insights is essential.”
Mr. Kaiser highlighted the critical role of pharmacists in real-time benefit implementation. “If you don’t incorporate pharmacist intelligence into that transaction, it’s going to be wrong,” he says. “Medication summaries set up within the EMR often speak a different language than the pharmacist’s system. Pharmacists bridge that gap by knowing what the plan pays for and knowing that’s the plan that will be implemented.”
Kaiser explained that successful implementation of real-time benefits requires an easier process. “The number of clicks has to be low and it has to be right almost every time,” he said. “And if it doesn’t meet those criteria, doctors won’t use it.”
While efforts to increase access to real-time, accurate information are occurring across many stakeholders, Kaiser emphasized the need to keep patients engaged. “You need to build the right connections so that when a patient is seeking PA status, they can ask you, the arbiter, for it, no matter where they are,” he said.
Diana Benli, Cognizant
Diana Benli, chief product officer of Cognizant’s TriZetto product group, explained the concept of “shifting left” in healthcare. This includes moving intelligence and collaboration upstream in the patient process to improve outcomes and reduce costs. She emphasized the importance of interoperable platforms and AI to achieve “10x” improvements in health systems. “It’s not about getting 10% better, it’s about being 10x better, 10x more predictive, 10x less administrative friction, 10x more resilient,” she said.
Benli believes TriZetto’s approach represents one way to bring PA processes closer to that 10x improvement. “We are combining the workflows between payers and providers and doing an integration layer between those organizations,” she said. “Friction doesn’t exist within a single organization; it exists at the seams between PBMs and health plans, for example.”
Benli described TriZetto’s approach to supporting smoother automated PA workflows and delivering value at the point of care: “We show users this is the best route, here is the fastest route to treatment, and here is an alternative. It’s not about automation, it’s about collaboration and intelligence at the point of care,” she said. “Once pre-approval becomes predictable and embedded upstream, it becomes valuable.”
Returning to the theme of his remarks, Benli explained that the shift to the left is not just about technology. “It’s really a philosophical shift from reactive to proactive, from gatekeeping to guidance, from incremental improvement to ’10x’ improvement,” she said. “Using the plumbing analogy, when you redesign the water and data flowing through pipes and infrastructure for care models, prior approval becomes a bridge, not a barrier.”
Sri Somasundaram, Potential Health
Sri Somasundaram, co-founder of Latent Health, laid out his vision of using AI to address inefficiencies in the PA process, pointing out that the common perception that the PA process is the villain of the story is dead wrong. “This system is designed with checks and balances in mind, but that’s not the problem,” he said. “We argue that this is actually a ‘computing problem.'”
What that means, Somasundaram said, is that people on both the provider and payer sides will scrutinize the data and build the case. If things go well, both sides come to the same conclusion, but it was a duplicate effort. If it doesn’t work, there will be no clinical consensus and more resources will be needed to resolve it. “I would argue that this adversarial dynamic that exists between providers and payers, which we have all just come to accept, is actually a symptom of this manual, inefficient computing problem rather than the system,” he said.
The answer, Somasundaram argued, is to deploy “clinical agents with superhuman precision” or AI “clinical reasoning systems” to handle complex clinical problems and improve the overall system. After the AI agent uses research records and applies drug criteria to answer complex clinical questions, “the idea is to communicate that information externally to all stakeholders,” he said.
He said the demand is huge, noting that the company now works with about 50 health systems in Somasundaram, up from just a few a year ago. “This is where the healthcare industry wants to move,” he said, adding that companies must resist the urge to apply AI to existing systems. “Many of the current approaches are very rudimentary when what we’re talking about is fundamentally different.”
Putting it all together
Finally, to make sense of it all, two speakers took to the stage to provide some background. Greg Ryslik, chief data and AI officer at Stellarus, and Alfredo Bird, partner at Xtillion, highlighted several unifying themes during their presentations.
Bird noted that everyone seems to feel the need to bring more humanity and transparency to the PA process. Other themes include data exchange and interoperability, and the need to build trust while reimagining the PA experience. It has become less fragmented and easier for patients and providers to navigate. “PA doesn’t have to feel like a game of ping pong; it can be a more collaborative and conversational experience, and we believe technology can play a role there,” he said.
For Ryslik, the promise of AI to improve PA is clear. “Today, right now, it’s the worst situation for these AI models. It’ll get better tomorrow and every day after that,” he said. “But they coexist with payers who process faxes manually and run older mainframes than most of the developers on my team.”
The effects of this dichotomy will continue for years to come, but will eventually fade, Bird said, noting that even organizations still running on mainframes may not want to wait years to implement AI. “That’s where very simple architecture and integration choices can make a huge difference,” he said. “Even something as simple as adding a universal data layer on top of an older application and providing a more modern way to interact with it can make a huge difference.”
