Authored by: Tim Anderson, Executive Vice President of RCM, Healthcare
Medical billing requires many time-consuming and manual tasks. And, like many other industries, it’s challenged by recruiting enough qualified staff to keep up with growing workloads.
The Medical Group Management Association asked medical leaders which revenue cycle roles were the hardest to hire for during ongoing staffing shortages. The survey found the most difficult to hire were medical coders (34%), with billers coming in second (26%).
In the meantime, overburdened staff often struggle with coding errors, more denials, and billing delays — all of which adversely impact revenue cycle management (RCM). In fact, a third of hospital executives reported coding as their top concern regarding denials and denial prevention.
These challenges have created more interest in AI medical billing and whether it could help lighten the load on existing staff and improve accuracy.
How is AI medical billing used?
Medical billing tasks often involve manual processes and repetitive workflows, creating a risk of errors. As we consider AI in medical billing, a couple of interesting use cases could positively impact RCM, including:
Coding. One of the most expensive parts of RCM is denial management, that is, when the first claim isn’t successful. The reasons for denials vary, everything from not bundling a service properly to coverage issues and more. But if we can trigger AI to do more coding, clean claims rates could improve significantly, trickling down to improved RCM.
Cleaning and scrubbing data. Cleaning and scrubbing data on the front end when possible can also help improve the probability of first-claim success with payers.
However, it’s important for payers to be a part of conversations about improving RCM. As providers know, some of the reasons why payers deny claims remain a mystery, and this mystery breeds distrust. Increasing visibility into why claims are denied could allow providers the opportunity to refine internal processes, plug medical billing AI tools into the right places, and improve RCM.
AI medical billing and areas to watch
AI is a helpful tool, but not everything can be automated. For example, you might think you can set up an automatic trigger code that says, “If service X is completed, service Y must also be billed.” But there are situations where Y doesn’t need to be billed — so this, or any setup where there are outliers, is problematic for many reasons.
When using AI, you’ll also want to retain human checks and balances. A decent amount of promise exists for handheld devices with automated medical coding capabilities. A physician walks around holding the device, and the coding is done based on patient interactions. And although that device has coding suggestions, we still want the physician to be the final word on which code is used, because physicians ultimately are still very much liable for what is transmitted about claims, government claims in particular.
The future of AI medical billing
Looking into the future, I anticipate the use cases for medical billing automation will continue to grow. Currently, some RCM-related functions are completed offshore, which could be challenged as we begin to use technology differently, AI in particular. And of course, there’s always the question of whether RCM should be offshored at all, and some states have taken steps to mitigate that.
But in the meantime, I think AI will impact the need to offshore this function. Also, if we can get more efficient with RCM, hopefully, that will bring down the administrative cost of healthcare and push those dollars back to the clinical side to provide more to patient care.
Interested in staying current on revenue cycle trends? Follow us on LinkedIn and never miss a blog post.