We’ve all started to feel the impact of artificial intelligence (AI) in our lives, from entertainment to business, and in medical practice administration, AI has proven to be a time-saving blessing, especially when it comes to repetitive tasks.
But AI can also run the risk over oversimplifying medical documentation during claims analysis, leading to automatic downcoding of EM codes that can cut into practice revenue and result in unnecessary appeals and overworked staff.
In this episode of MGMA’s Ask an Advisor, senior editor Daniel Williams and senior advisor Cristy Good discussed the recent epidemic of AI downcoding, and how practices – working in concert with organizations such as MGMA – can help spot and stop the growing problem, as well as bolstering their bottom line.
Downcoding is a national issue
Good says the issue has been reported by practices across the country, as they work with payers ranging from Blue Cross Blue Shield to UnitedHealthcare. In many instances, submitted EM codes are automatically reduced to lower-paying categories, or claims are more frequently denied.
AI systems seem to be using historical data such as patient demographics, diagnosis codes and high-level billing frequency instead of directly-connected medical records, she explains.
“If a claim doesn’t match what the algorithm thinks is typical, it gets flagged,” she says. “The problem is, it doesn’t account for the real complexity of the patient encounter. This goes against CMS guidelines, which require coding to be based on either medical decision-making or total time. It’s not just frustrating, it’s also non-compliant.”
As a result, practices end up spending more staff time working on documentation for appeals, which not only creates an administrative burden, but also undermines trust in the process.
“CMS and the AMA have both made it clear that claims should not be adjusted without review of medical records. This kind of downcoding can lead to staff burnout, delay in payment and even inaccurate performance profiles for providers.”
Spotting and fixing the problem
Good says the first signs of automatic downcoding often occur in remittance advances, when disclaimers such as “service level adjusted based upon payer policy” or “billed service inconsistent with diagnosis” pop up. If higher-level EM codes begin to be routinely paid at lower levels, especially by the same payer, or several providers are more impacted than others, it might be a case of AI mismanagement.
A first course of action, she says, is documentation. Practices should track the codes and providers impacted, appeals, the timing and success in overturning those appeals. Organizations including the AMA and MGMA are working to advocate for better payer accountability, and can help practices who’ve been impacted by AI downcoding.
“Good documentation is your best defense, so be specific,” Good says. “Clearly outline your medical decision-making process. Document the total time if time-based coding is used, and make sure to include any chronic conditions or comorbidities that impact the visit.”
Take time as well to detail the claim in the EHR with original notes rather than copy-and-paste, template-styled language, which can be seen as a red flag to an AI system. Reused notes on diagnosis and reasoning for common issues may save time, but they can help feed AI denials.
Good also suggests doing annual external and internal audits, which may necessitate auditing providers who are resulting in more frequent downcoding or denials.
Using AI to spot AI issues
Ironically, Good says that practices’ growing body of AI-backed tools can be helpful in spotting and preventing AI downcoding at the payer level. If you’ve already invested in AI-driven coding assistance and audit tools, they can be helpful in spotting missing documentation or inconsistent coding, which can help ensure your claim meets payer expectations.
“It’s a great way to level the playing field when we know payers are using AI-driven tools, too,” Good says. “You should always make sure you’re checking what it’s giving you. Even if you’re sending over clean claims and getting denials, it’s still worth looking into.”
Enlist MGMA resources to take action
Good says MGMA offers a variety of tools that can help raise awareness of the downcoding epidemic. She encourages members to post their experiences on the MGMA community discussion groups, which will help elicit advice and input from other colleagues in the medical community. The more people who post, the more information that can be passed along to the AMA, she says.
“Also, if you believe your payer is violating contract terms or state regulations, you can file a complaint with your state insurance department and send us your data. We work with MGMA’s government affairs team, and they’re really pushing for transparency and fairness.”
Direct Links and Resources:
- 📧 Contact Ask MGMA: advisor@mgma.com
- 🧠 MGMA Consulting Services: consulting@mgma.com
- 💬 Join the MGMA Community Discussion
- 📢 Report Issues to MGMA Government Affairs: govaff@mgma.com
Additional Resources:
Email us at dwilliams@mgma.com if you would like to appear on an episode. If you have a question about your practice that you would like us to answer, send an email to advisor@mgma.com. Don't forget to subscribe to our network wherever you get your podcasts.
Explore Related Content
Federal Policy Resource
Implementing the No Surprises Act
Online Course
Revenue Cycle Management Certificate
