To improve efficiency, save costs and elevate staff productivity and satisfaction, hospitals and health systems are automating parts of their revenue cycle using artificial intelligence (AI).
For especially complex processes, such as managing prior authorization, many are looking to generative AI (GenAI).
During a May Becker's Hospital Review webinar sponsored by AKASA, Amy Raymond, senior vice president for revenue cycle operations and deployments at AKASA, moderated a discussion with Joel Hammer, director for revenue cycle operations and innovation at Johns Hopkins Medicine in Baltimore, about practical considerations for successfully adopting GenAI within the revenue cycle.
Four key takeaways were:
1.) The rev cycle is ripe for improvements that boost accuracy and productivity. Such advances are essential, given payers' evolving AI-powered capabilities to deny pre-authorization requests and reimbursement claims and the need for providers to counter this trend by filing comprehensive, error-proof documentation on first go. Amid staffing shortages and rising labor costs, the higher administrative burden this implies for rev cycle teams requires automation that can perform tasks that require human intelligence.
2.) GenAI's underlying technology powers multiple rev cycle automation use cases. When trained on contextual healthcare data, the large language models (LLMs) that GenAI is based on can quickly scan clinical records, extract meaning from them and embed that information in predictive models that make recommendations. This output expedites work that previously required human reading, interpretation and decision-making.
Use cases in which rev cycle teams can use GenAI in their day-to-day work — with the assumption that a human still needs to validate the result before submission — include coding, clinical documentation improvement, medical necessity justification, letter drafting, claim attachments, denial appeals and producing estimates of a patient's likelihood of payment.
"By unlocking the clinical record and allowing a connection across the clinical and financial portions of the patient's record, LLMs create a very impactful space in the revenue cycle," Amy Raymond said.
3.) Johns Hopkins Medicine is leveraging AKASA to navigate this new AI space. The organization is using AKASA's GenAI solution to optimize processes related to authorization initiation and status checks on authorization requests and claim submissions.
"It eliminates the need for our staff to have to go to payer portals to submit authorizations, check on statuses and whatnot," Mr. Hammer said. "That streamlines that process and allows our team to focus on some of the more complex cases or areas that can't be automated."
4.) To identify the right vendor and solution, organizations can use these pointers. The first step is to realize that implementing GenAI is different from implementing "regular" rules-based automation. GenAI is a more advanced technology that requires having the right governance in place and being matched to the right projects. Having a lot of conversations with different vendors and asking probing questions can help organizations orient themselves to the type of solution they need, but also find the vendors that have a growth mindset.
"This is such an ever-changing landscape," Mr. Hammer concluded. "You need to find that vendor that recognizes that the products that are offered now may become obsolete within the next couple of years, is willing to grow with you and has that forward-thinking mindset.
To register for upcoming webinars, click here.