Artificial intelligence is a hot topic for healthcare stakeholders, including pharmacy benefit managers such as Abarca. But AI raises lots of questions about the kinds of applications it can and should be used for, the challenges of implementation, best practices and how its use is evolving and will continue to evolve in the future. In an interview, Abarca Chief Information Officer Serge Perras discussed how Abarca is implementing AI and where he sees the most value for it.
Abarca is leveraging different kinds of AI for its business, according to Perras.
NLP
Natural language processing (NLP) is one of the oldest forms of AI, dating back to the 1950s when world-renowned computer scientist Alan Turing wrote the paper “Computing Machinery and Intelligence.” At Abarca, NLP is used to interpret unstructured clinical and policy documents to aid in faster and more consistent benefit decisions.
“Our approach to AI isn’t about applying it indiscriminately, it’s about targeted and curated use cases that build a system that’s simpler, more intuitive, and profoundly more effective,” Perras said.
Machine Learning
Machine learning is used for Darwin RxTarget, Abarca’s adherence program giving pharmacists and payers the ability to monitor the health of their members in real-time to make interventions based on priority. Abarca harnesses predictive analytics to identify risks including high-cost claim anomalies, missed refills, and “therapy abandonment.”
Perras called attention to two areas where Abarca is not only enhancing its PBM performance but also changing the rules of engagement.
Intelligent Automation (IA)
Abarca is evolving traditional robotic process automation into intelligent automation infused with machine learning and context aware decision making capabilities. Using IA gives clinical and operational teams more time to focus on Abarca members’ care journey. One goal at Abarca for claims processing is to make benefits understandable and actionable. AI “co-pilots” also enable the development and product staff to more easily manage workflows and configurations, shortening time to market for enhancements, according to Perras.
Agentic AI
Agentic AI includes autonomous systems that can make decisions and perform tasks without human intervention. Abarca deploys Agentic AI to leverage structured and unstructured data and APIs to assess different scenarios to optimize cost and treatment opportunities.
Implementation challenges
Tech innovations always look better and easier to implement in theory than the messy reality companies take pains to address and improve over time. Perras pointed out that when it comes to implementing AI in PBMs, the hardest part is identifying the right use cases and building trust in the intelligence.
“Healthcare has long been wired for caution, and rightly so. But that also means AI must earn its place through explainability, auditability, and real-world utility,” he added.
Data fragmentation can be an obstacle. Without a full contextualized picture of the clinical history, benefit design and behavior patterns, AI can’t make responsible decisions. Abarca has invested heavily to integrate diverse data sources. These data resources include historical claims, medication data, and plan design, and the company continues to expand the universe of data its system can access via Agentic AI.
Given the concerns that even well-intentioned applications for AI could be manipulated and weaponized, ethical oversight is crucial. AI needs to be fair, bias-aware and inclusive, Perras said.
Abarca’s philosophy for AI is that it should be used to inform human decisions and be designed for augmentation. The goal isn’t to remove humans, it’s to support them with the right information at the right time.
AI offers exciting ways for healthcare organizations to grow while streamlining their operations. PBMs enlisting AI will be able to reduce or eliminate friction in the benefits management and adjudication process while also making these benefits easier for members to understand. It will also help PBMs improve communication with payers and providers. AI implementation naturally presents challenges to our heavily regulated health system. But the comfort level with using AI will improve over time. AI also offers healthcare organizations opportunities to automate administrative tasks while enabling healthcare professionals to devote more time and attention to their patients.
“Ultimately, implementing AI in PBMs isn’t about jumping on a trend. It’s about committing to a better way: smarter, faster, and more personal healthcare,” Perras said. “That’s the promise of the Darwin new era. And that’s where Abarca is heading.”
Picture: Just_Super, Getty Images