Automation Paid Off—So Why Are Denials Still Rising?

Health systems have spent two decades investing in revenue cycle automation, and it has paid off: the 2025 CAQH Index estimates the industry avoided $258 billion in administrative costs in 2024. Yet denials keep climbing, with 41 percent of providers reporting at least one in ten claims denied, a figure that has risen every year since 2022, per Experian Health’s 2025 State of Claims survey . Nowhere is the strain clearer than in denial management—and read together, those numbers point somewhere specific. The next real gain sits in how the work is designed to move from end to end, in territory automation alone can’t reach. Most revenue cycle operations have already put automation to work somewhere—eligibility checks, claim status look-ups, payment posting—and for a department that can’t add headcount, the hours saved count for something. What those projects also did was show where the hard part lives: in the work that takes context and judgment—a denial that doesn’t match any template, a payer rule that quietly changed last quarter. That is where margin slips away, in the work traditional automation was never built to handle. Why denial management still needs human judgment Give a bot a fixed rule and a process that runs the same way every time, and it executes faster and more consistently than any person. No argument there. What bots don’t do is read a situation, weigh a couple of options and decide what to do when the inputs don’t match the script—and no two denials look quite the same. That judgment work, pulling the exception out of the pile and figuring out the next action, is where the newer, more intelligent tools are designed to help. AI agents, working alongside your team, can take on that triage layer, reading the context of a denial to choose a response pathway and pulling in a specialist only when the situation calls for one. The same CAQH Index found a quarter of provider organizations using AI tools in administrative workflows. That end-to-end approach is what we at Coronis Health call Intelligent Revenue Operations: the way work is designed to move across the whole revenue lifecycle, orchestrating AI agents, people and automation around a single outcome. Better denial management starts upstream The reflex is to point at a task and ask whether it can be automated. Fair enough—it buys you some efficiency. But the better question is where the process itself needs rethinking. You can automate the denial follow-up queue and make a miserable job go faster. Or you can fix the upstream intake step that let the bad information through. The first is worth doing. The second is the one that actually shrinks the workload. That upstream focus is what revenue cycle optimization looks like in practice: half of providers in the Experian survey cite missing or inaccurate data as a top cause of their denials, and when registration, eligibility and documentation are clean from the start, there is simply less to chase later. Building revenue cycle workflows that adapt to changing payer rules Here is the practical case for redesign: the rules won’t hold still. I’ve watched a payer drop an authorization requirement with barely any notice. Teams had built whole processes around that requirement, and overnight it was gone. Every change like that becomes a manual scramble. If a major payer changed a requirement next month, how much manual rework would that set off? The sturdier approach is to design the work so it can bend: revenue cycle workflows that take on exceptions and adjust as payers and conditions shift, with agents watching how payers behave and catching emerging patterns before they harden into write-offs. The workforce math None of this is about getting people out of the revenue cycle. Most systems couldn’t hire their way out if they wanted to. The aim is to stop burning scarce expertise on work that doesn’t need it. Take the repetitive work off your team’s plate and the same people can work the messy exceptions and sit in on the conversations about performance. Organizations running orchestrated revenue operations are reporting lower denial rates and fewer days in accounts receivable, with the gains coming from how the work is designed to move rather than from added headcount. Over the next few years, the systems that pull ahead will be the ones that redesigned the work around what the revenue cycle actually needs to produce: cleaner information at the front, and people pointed at the decisions that need a human. Anitha Balasubramaniam is Vice President of Automation at Coronis Health. She has more than 19 years of leadership experience in enterprise technology, automation and digital transformation across healthcare and RCM, retail and consulting, and has built and scaled automation-driven platforms using RPA, AI/ML, OCR and advanced analytics to strengthen revenue cycle performance. For more information, visit coronishealth.com . The post Automation Paid Off—So Why Are Denials Still Rising? appeared first on Becker's Hospital Review | Healthcare News & Analysis .